2002-10-12 11:37:38 +00:00
|
|
|
/**
|
|
|
|
* $Id$
|
|
|
|
*
|
|
|
|
* ***** BEGIN GPL/BL DUAL LICENSE BLOCK *****
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU General Public License
|
|
|
|
* as published by the Free Software Foundation; either version 2
|
|
|
|
* of the License, or (at your option) any later version. The Blender
|
|
|
|
* Foundation also sells licenses for use in proprietary software under
|
|
|
|
* the Blender License. See http://www.blender.org/BL/ for information
|
|
|
|
* about this.
|
|
|
|
*
|
|
|
|
* This program is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
* GNU General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU General Public License
|
|
|
|
* along with this program; if not, write to the Free Software Foundation,
|
|
|
|
* Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
|
|
|
|
*
|
|
|
|
* The Original Code is Copyright (C) 2001-2002 by NaN Holding BV.
|
|
|
|
* All rights reserved.
|
|
|
|
*
|
|
|
|
* The Original Code is: all of this file.
|
|
|
|
*
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
* Contributor(s): 2007, Joshua Leung, major recode
|
2002-10-12 11:37:38 +00:00
|
|
|
*
|
|
|
|
* ***** END GPL/BL DUAL LICENSE BLOCK *****
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <stdio.h>
|
2008-01-20 23:53:13 +00:00
|
|
|
#include <stddef.h>
|
2002-10-12 11:37:38 +00:00
|
|
|
#include <string.h>
|
2003-12-24 11:44:57 +00:00
|
|
|
#include <math.h>
|
2002-10-12 11:37:38 +00:00
|
|
|
|
|
|
|
#include "MEM_guardedalloc.h"
|
|
|
|
#include "nla.h"
|
|
|
|
|
|
|
|
#include "BLI_blenlib.h"
|
|
|
|
#include "BLI_arithb.h"
|
|
|
|
|
|
|
|
#include "DNA_armature_types.h"
|
|
|
|
#include "DNA_constraint_types.h"
|
|
|
|
#include "DNA_object_types.h"
|
|
|
|
#include "DNA_action_types.h"
|
|
|
|
#include "DNA_curve_types.h"
|
2007-08-17 11:23:48 +00:00
|
|
|
#include "DNA_meshdata_types.h"
|
|
|
|
#include "DNA_lattice_types.h"
|
2004-01-27 06:08:37 +00:00
|
|
|
#include "DNA_scene_types.h"
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
#include "DNA_text_types.h"
|
2002-10-12 11:37:38 +00:00
|
|
|
|
|
|
|
#include "BKE_utildefines.h"
|
|
|
|
#include "BKE_action.h"
|
2007-08-28 08:43:38 +00:00
|
|
|
#include "BKE_anim.h" /* for the curve calculation part */
|
2002-10-12 11:37:38 +00:00
|
|
|
#include "BKE_armature.h"
|
|
|
|
#include "BKE_blender.h"
|
|
|
|
#include "BKE_constraint.h"
|
2005-12-21 17:49:43 +00:00
|
|
|
#include "BKE_displist.h"
|
2007-08-17 11:23:48 +00:00
|
|
|
#include "BKE_deform.h"
|
2007-08-28 08:43:38 +00:00
|
|
|
#include "BKE_DerivedMesh.h" /* for geometry targets */
|
|
|
|
#include "BKE_cdderivedmesh.h" /* for geometry targets */
|
2002-10-12 11:37:38 +00:00
|
|
|
#include "BKE_object.h"
|
|
|
|
#include "BKE_ipo.h"
|
|
|
|
#include "BKE_global.h"
|
|
|
|
#include "BKE_library.h"
|
2007-06-18 07:41:21 +00:00
|
|
|
#include "BKE_idprop.h"
|
|
|
|
|
2007-08-17 11:23:48 +00:00
|
|
|
|
2007-06-18 07:41:21 +00:00
|
|
|
#include "BPY_extern.h"
|
2002-10-12 11:37:38 +00:00
|
|
|
|
|
|
|
#include "blendef.h"
|
|
|
|
|
2002-11-25 12:02:15 +00:00
|
|
|
#ifdef HAVE_CONFIG_H
|
|
|
|
#include <config.h>
|
|
|
|
#endif
|
|
|
|
|
2002-10-12 11:37:38 +00:00
|
|
|
#ifndef M_PI
|
|
|
|
#define M_PI 3.14159265358979323846
|
|
|
|
#endif
|
2005-03-09 19:45:59 +00:00
|
|
|
|
2002-10-12 11:37:38 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* ******************* Constraint Channels ********************** */
|
|
|
|
/* Constraint Channels exist in one of two places:
|
|
|
|
* - Under Action Channels in an Action (act->chanbase->achan->constraintChannels)
|
2008-03-25 10:36:36 +00:00
|
|
|
* - Under Object without Object-level Action yet (ob->constraintChannels)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
*
|
2008-03-25 10:36:36 +00:00
|
|
|
* The main purpose that Constraint Channels serve is to act as a link
|
|
|
|
* between an IPO-block (which provides values to interpolate between for some settings)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
*/
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* ------------ Data Management ----------- */
|
|
|
|
|
|
|
|
/* Free constraint channels, and reduce the number of users of the related ipo-blocks */
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
void free_constraint_channels (ListBase *chanbase)
|
|
|
|
{
|
|
|
|
bConstraintChannel *chan;
|
|
|
|
|
2007-05-17 06:45:16 +00:00
|
|
|
for (chan=chanbase->first; chan; chan=chan->next) {
|
|
|
|
if (chan->ipo) {
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
chan->ipo->id.us--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
BLI_freelistN(chanbase);
|
|
|
|
}
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* Make a copy of the constraint channels from dst to src, and also give the
|
|
|
|
* new constraint channels their own copy of the original's IPO.
|
|
|
|
*/
|
2005-08-24 20:18:03 +00:00
|
|
|
void copy_constraint_channels (ListBase *dst, ListBase *src)
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
{
|
|
|
|
bConstraintChannel *dchan, *schan;
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
dst->first = dst->last = NULL;
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
duplicatelist(dst, src);
|
|
|
|
|
2007-05-17 06:45:16 +00:00
|
|
|
for (dchan=dst->first, schan=src->first; dchan; dchan=dchan->next, schan=schan->next) {
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
dchan->ipo = copy_ipo(schan->ipo);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* Make a copy of the constraint channels from dst to src, but make the
|
|
|
|
* new constraint channels use the same IPO-data as their twin.
|
|
|
|
*/
|
2005-08-24 20:18:03 +00:00
|
|
|
void clone_constraint_channels (ListBase *dst, ListBase *src)
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
{
|
|
|
|
bConstraintChannel *dchan, *schan;
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
dst->first = dst->last = NULL;
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
duplicatelist(dst, src);
|
|
|
|
|
2007-05-17 06:45:16 +00:00
|
|
|
for (dchan=dst->first, schan=src->first; dchan; dchan=dchan->next, schan=schan->next) {
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
id_us_plus((ID *)dchan->ipo);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* ------------- Constraint Channel Tools ------------ */
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* Find the constraint channel with a given name */
|
|
|
|
bConstraintChannel *get_constraint_channel (ListBase *list, const char name[])
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
bConstraintChannel *chan;
|
|
|
|
|
2008-03-25 10:36:36 +00:00
|
|
|
if (list) {
|
2007-11-28 12:11:06 +00:00
|
|
|
for (chan = list->first; chan; chan=chan->next) {
|
|
|
|
if (!strcmp(name, chan->name)) {
|
|
|
|
return chan;
|
|
|
|
}
|
2007-07-21 07:26:15 +00:00
|
|
|
}
|
2003-11-23 20:28:35 +00:00
|
|
|
}
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
return NULL;
|
2003-11-23 20:28:35 +00:00
|
|
|
}
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* Find or create a new constraint channel */
|
|
|
|
bConstraintChannel *verify_constraint_channel (ListBase *list, const char name[])
|
2004-08-16 01:14:18 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
bConstraintChannel *chan;
|
|
|
|
|
|
|
|
chan= get_constraint_channel(list, name);
|
|
|
|
|
|
|
|
if (chan == NULL) {
|
|
|
|
chan= MEM_callocN(sizeof(bConstraintChannel), "new constraint channel");
|
|
|
|
BLI_addtail(list, chan);
|
|
|
|
strcpy(chan->name, name);
|
2004-08-16 01:14:18 +00:00
|
|
|
}
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
return chan;
|
2004-08-16 01:14:18 +00:00
|
|
|
}
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* --------- Constraint Channel Evaluation/Execution --------- */
|
|
|
|
|
|
|
|
/* IPO-system call: calculate IPO-block for constraint channels, and flush that
|
|
|
|
* info onto the corresponding constraint.
|
|
|
|
*/
|
|
|
|
void do_constraint_channels (ListBase *conbase, ListBase *chanbase, float ctime, short onlydrivers)
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
bConstraint *con;
|
|
|
|
|
|
|
|
/* for each Constraint, calculate its Influence from the corresponding ConstraintChannel */
|
|
|
|
for (con=conbase->first; con; con=con->next) {
|
2007-11-28 12:11:06 +00:00
|
|
|
Ipo *ipo= NULL;
|
|
|
|
|
2008-03-25 10:36:36 +00:00
|
|
|
if (con->flag & CONSTRAINT_OWN_IPO)
|
2007-11-28 12:11:06 +00:00
|
|
|
ipo= con->ipo;
|
|
|
|
else {
|
|
|
|
bConstraintChannel *chan = get_constraint_channel(chanbase, con->name);
|
2008-03-25 10:36:36 +00:00
|
|
|
if (chan) ipo= chan->ipo;
|
2007-11-28 12:11:06 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
2007-11-28 12:11:06 +00:00
|
|
|
if (ipo) {
|
|
|
|
IpoCurve *icu;
|
|
|
|
|
|
|
|
calc_ipo(ipo, ctime);
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
2007-11-28 12:11:06 +00:00
|
|
|
for (icu=ipo->curve.first; icu; icu=icu->next) {
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (!onlydrivers || icu->driver) {
|
|
|
|
switch (icu->adrcode) {
|
|
|
|
case CO_ENFORCE:
|
|
|
|
{
|
|
|
|
/* Influence is clamped to 0.0f -> 1.0f range */
|
|
|
|
con->enforce = CLAMPIS(icu->curval, 0.0f, 1.0f);
|
|
|
|
}
|
|
|
|
break;
|
Patch #7767: Constraint Subtargets can now target anywhere on a bone, not just the head or tail
Patch by: Roland Hess (harkyman)
For example, a constraint can be sub-targeted at the 50% (or 31.2% or 85% etc.) point of its target bone, giving you enormous rigging flexibility and removing the need for complex contraptions to do such things as:
- A bone whose base slides only between to points on a rig (CopyLoc with a variable, animated subtarget point)
- Bones that attach to multiple points along another bone (CopyLocs, each with a different head/tail percentage)
- Bones that need to stretch to a point midway between specific spots on two other bones (old way: too crazy to mention; new way: stretch bone between points on end bones, then another stretch to the midpoint of the first stretch)
It is only used for the constraint types for which it is relevant: CopyLoc, TrackTo, StretchTo and MinMax, TrackTo, and Floor.
Notes:
- This is accessed by the Head/Tail number-slider.
- This value can be animated per constraint
- The old "Copy Bone Tail" option for the CopyLoc constraint has been automatically converted to 1.0 Head/Bone values for the affected constraints
- In the code, this value is in the bConstraint struct, so it is available for all constraints, even though only a few implement it.
2007-11-12 04:17:03 +00:00
|
|
|
case CO_HEADTAIL:
|
|
|
|
{
|
2008-03-25 10:36:36 +00:00
|
|
|
/* we need to check types of constraints that can get this here, as user
|
|
|
|
* may have created an IPO-curve for this from IPO-editor but for a constraint
|
|
|
|
* that cannot support this
|
|
|
|
*/
|
|
|
|
switch (con->type) {
|
|
|
|
/* supported constraints go here... */
|
|
|
|
case CONSTRAINT_TYPE_LOCLIKE:
|
|
|
|
case CONSTRAINT_TYPE_TRACKTO:
|
|
|
|
case CONSTRAINT_TYPE_MINMAX:
|
|
|
|
case CONSTRAINT_TYPE_STRETCHTO:
|
|
|
|
case CONSTRAINT_TYPE_DISTLIMIT:
|
|
|
|
con->headtail = icu->curval;
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
/* not supported */
|
|
|
|
break;
|
|
|
|
}
|
Patch #7767: Constraint Subtargets can now target anywhere on a bone, not just the head or tail
Patch by: Roland Hess (harkyman)
For example, a constraint can be sub-targeted at the 50% (or 31.2% or 85% etc.) point of its target bone, giving you enormous rigging flexibility and removing the need for complex contraptions to do such things as:
- A bone whose base slides only between to points on a rig (CopyLoc with a variable, animated subtarget point)
- Bones that attach to multiple points along another bone (CopyLocs, each with a different head/tail percentage)
- Bones that need to stretch to a point midway between specific spots on two other bones (old way: too crazy to mention; new way: stretch bone between points on end bones, then another stretch to the midpoint of the first stretch)
It is only used for the constraint types for which it is relevant: CopyLoc, TrackTo, StretchTo and MinMax, TrackTo, and Floor.
Notes:
- This is accessed by the Head/Tail number-slider.
- This value can be animated per constraint
- The old "Copy Bone Tail" option for the CopyLoc constraint has been automatically converted to 1.0 Head/Bone values for the affected constraints
- In the code, this value is in the bConstraint struct, so it is available for all constraints, even though only a few implement it.
2007-11-12 04:17:03 +00:00
|
|
|
}
|
|
|
|
break;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2007-07-21 07:26:15 +00:00
|
|
|
}
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* ************************ Constraints - General Utilities *************************** */
|
|
|
|
/* These functions here don't act on any specific constraints, and are therefore should/will
|
|
|
|
* not require any of the special function-pointers afforded by the relevant constraint
|
|
|
|
* type-info structs.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* -------------- Naming -------------- */
|
|
|
|
|
|
|
|
/* Find the first available, non-duplicate name for a given constraint */
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
void unique_constraint_name (bConstraint *con, ListBase *list)
|
|
|
|
{
|
2008-01-20 23:53:13 +00:00
|
|
|
BLI_uniquename(list, con, "Const", offsetof(bConstraint, name), 32);
|
2003-10-21 13:22:07 +00:00
|
|
|
}
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* ----------------- Evaluation Loop Preparation --------------- */
|
|
|
|
|
|
|
|
/* package an object/bone for use in constraint evaluation */
|
|
|
|
/* This function MEM_calloc's a bConstraintOb struct, that will need to be freed after evaluation */
|
|
|
|
bConstraintOb *constraints_make_evalob (Object *ob, void *subdata, short datatype)
|
2003-10-21 13:22:07 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
bConstraintOb *cob;
|
|
|
|
|
|
|
|
/* create regardless of whether we have any data! */
|
|
|
|
cob= MEM_callocN(sizeof(bConstraintOb), "bConstraintOb");
|
2003-10-21 13:22:07 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* based on type of available data */
|
|
|
|
switch (datatype) {
|
|
|
|
case CONSTRAINT_OBTYPE_OBJECT:
|
2007-06-18 07:41:21 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* disregard subdata... calloc should set other values right */
|
|
|
|
if (ob) {
|
|
|
|
cob->ob = ob;
|
|
|
|
cob->type = datatype;
|
|
|
|
Mat4CpyMat4(cob->matrix, ob->obmat);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
Mat4One(cob->matrix);
|
2007-06-18 07:41:21 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
Mat4CpyMat4(cob->startmat, cob->matrix);
|
2007-06-18 07:41:21 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
break;
|
|
|
|
case CONSTRAINT_OBTYPE_BONE:
|
2003-10-21 13:22:07 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* only set if we have valid bone, otherwise default */
|
|
|
|
if (ob && subdata) {
|
|
|
|
cob->ob = ob;
|
|
|
|
cob->pchan = (bPoseChannel *)subdata;
|
|
|
|
cob->type = datatype;
|
|
|
|
|
|
|
|
/* matrix in world-space */
|
|
|
|
Mat4MulMat4(cob->matrix, cob->pchan->pose_mat, ob->obmat);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
Mat4One(cob->matrix);
|
|
|
|
|
|
|
|
Mat4CpyMat4(cob->startmat, cob->matrix);
|
2003-10-21 13:22:07 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
break;
|
2007-05-17 06:45:16 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
default: /* other types not yet handled */
|
|
|
|
Mat4One(cob->matrix);
|
|
|
|
Mat4One(cob->startmat);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
return cob;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* cleanup after constraint evaluation */
|
|
|
|
void constraints_clear_evalob (bConstraintOb *cob)
|
|
|
|
{
|
|
|
|
float delta[4][4], imat[4][4];
|
|
|
|
|
|
|
|
/* prevent crashes */
|
|
|
|
if (cob == NULL)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* calculate delta of constraints evaluation */
|
|
|
|
Mat4Invert(imat, cob->startmat);
|
2008-01-22 20:59:45 +00:00
|
|
|
Mat4MulMat4(delta, imat, cob->matrix);
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
|
|
|
/* copy matrices back to source */
|
|
|
|
switch (cob->type) {
|
|
|
|
case CONSTRAINT_OBTYPE_OBJECT:
|
2005-09-07 00:11:39 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* cob->ob might not exist! */
|
|
|
|
if (cob->ob) {
|
|
|
|
/* copy new ob-matrix back to owner */
|
|
|
|
Mat4CpyMat4(cob->ob->obmat, cob->matrix);
|
|
|
|
|
|
|
|
/* copy inverse of delta back to owner */
|
|
|
|
Mat4Invert(cob->ob->constinv, delta);
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
case CONSTRAINT_OBTYPE_BONE:
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* cob->ob or cob->pchan might not exist */
|
|
|
|
if (cob->ob && cob->pchan) {
|
|
|
|
/* copy new pose-matrix back to owner */
|
|
|
|
Mat4MulMat4(cob->pchan->pose_mat, cob->matrix, cob->ob->imat);
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* copy inverse of delta back to owner */
|
|
|
|
Mat4Invert(cob->pchan->constinv, delta);
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* free tempolary struct */
|
|
|
|
MEM_freeN(cob);
|
|
|
|
}
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* -------------- Space-Conversion API -------------- */
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
|
|
|
/* This function is responsible for the correct transformations/conversions
|
|
|
|
* of a matrix from one space to another for constraint evaluation.
|
|
|
|
* For now, this is only implemented for Objects and PoseChannels.
|
|
|
|
*/
|
2007-08-14 11:19:35 +00:00
|
|
|
void constraint_mat_convertspace (Object *ob, bPoseChannel *pchan, float mat[][4], short from, short to)
|
2002-10-12 11:37:38 +00:00
|
|
|
{
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
float tempmat[4][4];
|
|
|
|
float diff_mat[4][4];
|
|
|
|
float imat[4][4];
|
|
|
|
|
|
|
|
/* prevent crashes in these unlikely events */
|
|
|
|
if (ob==NULL || mat==NULL) return;
|
|
|
|
/* optimise trick - check if need to do anything */
|
|
|
|
if (from == to) return;
|
|
|
|
|
|
|
|
/* are we dealing with pose-channels or objects */
|
|
|
|
if (pchan) {
|
|
|
|
/* pose channels */
|
|
|
|
switch (from) {
|
|
|
|
case CONSTRAINT_SPACE_WORLD: /* ---------- FROM WORLDSPACE ---------- */
|
|
|
|
{
|
|
|
|
/* world to pose */
|
|
|
|
if (to==CONSTRAINT_SPACE_POSE || to==CONSTRAINT_SPACE_LOCAL || to==CONSTRAINT_SPACE_PARLOCAL) {
|
|
|
|
Mat4Invert(imat, ob->obmat);
|
|
|
|
Mat4CpyMat4(tempmat, mat);
|
|
|
|
Mat4MulMat4(mat, tempmat, imat);
|
|
|
|
}
|
2007-07-28 09:00:15 +00:00
|
|
|
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
/* pose to local */
|
|
|
|
if (to == CONSTRAINT_SPACE_LOCAL) {
|
|
|
|
/* call self with slightly different values */
|
|
|
|
constraint_mat_convertspace(ob, pchan, mat, CONSTRAINT_SPACE_POSE, to);
|
|
|
|
}
|
|
|
|
/* pose to local + parent */
|
|
|
|
else if (to == CONSTRAINT_SPACE_PARLOCAL) {
|
|
|
|
/* call self with slightly different values */
|
|
|
|
constraint_mat_convertspace(ob, pchan, mat, CONSTRAINT_SPACE_POSE, to);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case CONSTRAINT_SPACE_POSE: /* ---------- FROM POSESPACE ---------- */
|
|
|
|
{
|
|
|
|
/* pose to world */
|
|
|
|
if (to == CONSTRAINT_SPACE_WORLD) {
|
|
|
|
Mat4CpyMat4(tempmat, mat);
|
|
|
|
Mat4MulMat4(mat, tempmat, ob->obmat);
|
|
|
|
}
|
|
|
|
/* pose to local */
|
|
|
|
else if (to == CONSTRAINT_SPACE_LOCAL) {
|
|
|
|
if (pchan->bone) {
|
2007-07-28 10:44:03 +00:00
|
|
|
if (pchan->parent) {
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
float offs_bone[4][4];
|
|
|
|
|
|
|
|
/* construct offs_bone the same way it is done in armature.c */
|
|
|
|
Mat4CpyMat3(offs_bone, pchan->bone->bone_mat);
|
|
|
|
VECCOPY(offs_bone[3], pchan->bone->head);
|
2007-07-28 10:44:03 +00:00
|
|
|
offs_bone[3][1]+= pchan->bone->parent->length;
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
|
|
|
if (pchan->bone->flag & BONE_HINGE) {
|
|
|
|
/* pose_mat = par_pose-space_location * chan_mat */
|
|
|
|
float tmat[4][4];
|
|
|
|
|
|
|
|
/* the rotation of the parent restposition */
|
2007-07-28 10:44:03 +00:00
|
|
|
Mat4CpyMat4(tmat, pchan->bone->parent->arm_mat);
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
|
|
|
/* the location of actual parent transform */
|
|
|
|
VECCOPY(tmat[3], offs_bone[3]);
|
|
|
|
offs_bone[3][0]= offs_bone[3][1]= offs_bone[3][2]= 0.0f;
|
|
|
|
Mat4MulVecfl(pchan->parent->pose_mat, tmat[3]);
|
|
|
|
|
2007-07-28 10:44:03 +00:00
|
|
|
Mat4MulMat4(diff_mat, offs_bone, tmat);
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
Mat4Invert(imat, diff_mat);
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* pose_mat = par_pose_mat * bone_mat * chan_mat */
|
2007-07-28 10:44:03 +00:00
|
|
|
Mat4MulMat4(diff_mat, offs_bone, pchan->parent->pose_mat);
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
Mat4Invert(imat, diff_mat);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* pose_mat = chan_mat * arm_mat */
|
|
|
|
Mat4Invert(imat, pchan->bone->arm_mat);
|
|
|
|
}
|
|
|
|
|
|
|
|
Mat4CpyMat4(tempmat, mat);
|
|
|
|
Mat4MulMat4(mat, tempmat, imat);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/* pose to local with parent */
|
|
|
|
else if (to == CONSTRAINT_SPACE_PARLOCAL) {
|
|
|
|
if (pchan->bone) {
|
|
|
|
Mat4Invert(imat, pchan->bone->arm_mat);
|
|
|
|
Mat4CpyMat4(tempmat, mat);
|
|
|
|
Mat4MulMat4(mat, tempmat, imat);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case CONSTRAINT_SPACE_LOCAL: /* ------------ FROM LOCALSPACE --------- */
|
|
|
|
{
|
|
|
|
/* local to pose */
|
|
|
|
if (to==CONSTRAINT_SPACE_POSE || to==CONSTRAINT_SPACE_WORLD) {
|
|
|
|
/* do inverse procedure that was done for pose to local */
|
|
|
|
if (pchan->bone) {
|
|
|
|
/* we need the posespace_matrix = local_matrix + (parent_posespace_matrix + restpos) */
|
|
|
|
if (pchan->parent) {
|
|
|
|
float offs_bone[4][4];
|
2008-03-25 10:36:36 +00:00
|
|
|
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
/* construct offs_bone the same way it is done in armature.c */
|
|
|
|
Mat4CpyMat3(offs_bone, pchan->bone->bone_mat);
|
|
|
|
VECCOPY(offs_bone[3], pchan->bone->head);
|
2007-07-28 10:44:03 +00:00
|
|
|
offs_bone[3][1]+= pchan->bone->parent->length;
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
2007-07-24 05:08:55 +00:00
|
|
|
if (pchan->bone->flag & BONE_HINGE) {
|
|
|
|
/* pose_mat = par_pose-space_location * chan_mat */
|
|
|
|
float tmat[4][4];
|
|
|
|
|
|
|
|
/* the rotation of the parent restposition */
|
2007-07-28 10:44:03 +00:00
|
|
|
Mat4CpyMat4(tmat, pchan->bone->parent->arm_mat);
|
2007-07-24 05:08:55 +00:00
|
|
|
|
|
|
|
/* the location of actual parent transform */
|
|
|
|
VECCOPY(tmat[3], offs_bone[3]);
|
|
|
|
offs_bone[3][0]= offs_bone[3][1]= offs_bone[3][2]= 0.0f;
|
|
|
|
Mat4MulVecfl(pchan->parent->pose_mat, tmat[3]);
|
|
|
|
|
2007-07-28 10:44:03 +00:00
|
|
|
Mat4MulMat4(diff_mat, offs_bone, tmat);
|
2007-07-24 05:08:55 +00:00
|
|
|
Mat4CpyMat4(tempmat, mat);
|
|
|
|
Mat4MulMat4(mat, tempmat, diff_mat);
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* pose_mat = par_pose_mat * bone_mat * chan_mat */
|
2007-07-28 10:44:03 +00:00
|
|
|
Mat4MulMat4(diff_mat, offs_bone, pchan->parent->pose_mat);
|
2007-07-24 05:08:55 +00:00
|
|
|
Mat4CpyMat4(tempmat, mat);
|
|
|
|
Mat4MulMat4(mat, tempmat, diff_mat);
|
|
|
|
}
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
}
|
|
|
|
else {
|
|
|
|
Mat4CpyMat4(diff_mat, pchan->bone->arm_mat);
|
|
|
|
|
|
|
|
Mat4CpyMat4(tempmat, mat);
|
|
|
|
Mat4MulMat4(mat, tempmat, diff_mat);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/* local to world */
|
|
|
|
if (to == CONSTRAINT_SPACE_WORLD) {
|
|
|
|
/* call self with slightly different values */
|
|
|
|
constraint_mat_convertspace(ob, pchan, mat, CONSTRAINT_SPACE_POSE, to);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case CONSTRAINT_SPACE_PARLOCAL: /* -------------- FROM LOCAL WITH PARENT ---------- */
|
|
|
|
{
|
|
|
|
/* local to pose */
|
|
|
|
if (to==CONSTRAINT_SPACE_POSE || to==CONSTRAINT_SPACE_WORLD) {
|
|
|
|
if (pchan->bone) {
|
|
|
|
Mat4CpyMat4(diff_mat, pchan->bone->arm_mat);
|
|
|
|
Mat4CpyMat4(tempmat, mat);
|
|
|
|
Mat4MulMat4(mat, diff_mat, tempmat);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/* local to world */
|
|
|
|
if (to == CONSTRAINT_SPACE_WORLD) {
|
|
|
|
/* call self with slightly different values */
|
|
|
|
constraint_mat_convertspace(ob, pchan, mat, CONSTRAINT_SPACE_POSE, to);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* objects */
|
|
|
|
if (from==CONSTRAINT_SPACE_WORLD && to==CONSTRAINT_SPACE_LOCAL) {
|
|
|
|
/* check if object has a parent - otherwise this won't work */
|
|
|
|
if (ob->parent) {
|
|
|
|
/* 'subtract' parent's effects from owner */
|
|
|
|
Mat4MulMat4(diff_mat, ob->parentinv, ob->parent->obmat);
|
|
|
|
Mat4Invert(imat, diff_mat);
|
|
|
|
Mat4CpyMat4(tempmat, mat);
|
|
|
|
Mat4MulMat4(mat, tempmat, imat);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else if (from==CONSTRAINT_SPACE_LOCAL && to==CONSTRAINT_SPACE_WORLD) {
|
|
|
|
/* check that object has a parent - otherwise this won't work */
|
|
|
|
if (ob->parent) {
|
|
|
|
/* 'add' parent's effect back to owner */
|
|
|
|
Mat4CpyMat4(tempmat, mat);
|
|
|
|
Mat4MulMat4(diff_mat, ob->parentinv, ob->parent->obmat);
|
|
|
|
Mat4MulMat4(mat, tempmat, diff_mat);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* ------------ General Target Matrix Tools ---------- */
|
2002-10-12 11:37:38 +00:00
|
|
|
|
2007-08-17 11:23:48 +00:00
|
|
|
/* function that sets the given matrix based on given vertex group in mesh */
|
|
|
|
static void contarget_get_mesh_mat (Object *ob, char *substring, float mat[][4])
|
|
|
|
{
|
2007-08-28 08:43:38 +00:00
|
|
|
DerivedMesh *dm;
|
2007-08-17 11:23:48 +00:00
|
|
|
float vec[3] = {0.0f, 0.0f, 0.0f}, tvec[3];
|
2007-08-18 04:03:03 +00:00
|
|
|
float normal[3] = {0.0f, 0.0f, 0.0f}, plane[3];
|
|
|
|
float imat[3][3], tmat[3][3];
|
2007-08-17 11:23:48 +00:00
|
|
|
int dgroup;
|
|
|
|
|
|
|
|
/* initialize target matrix using target matrix */
|
|
|
|
Mat4CpyMat4(mat, ob->obmat);
|
|
|
|
|
|
|
|
/* get index of vertex group */
|
|
|
|
dgroup = get_named_vertexgroup_num(ob, substring);
|
|
|
|
if (dgroup < 0) return;
|
|
|
|
|
2007-08-28 08:43:38 +00:00
|
|
|
/* get DerivedMesh */
|
|
|
|
if (G.obedit && G.editMesh) {
|
|
|
|
/* we are in editmode, so get a special derived mesh */
|
|
|
|
dm = CDDM_from_editmesh(G.editMesh, ob->data);
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* when not in EditMode, this should exist */
|
|
|
|
dm = (DerivedMesh *)ob->derivedFinal;
|
|
|
|
}
|
|
|
|
|
2007-08-17 11:23:48 +00:00
|
|
|
/* only continue if there's a valid DerivedMesh */
|
|
|
|
if (dm) {
|
|
|
|
MDeformVert *dvert = dm->getVertDataArray(dm, CD_MDEFORMVERT);
|
|
|
|
int *index = (int *)dm->getVertDataArray(dm, CD_ORIGINDEX);
|
|
|
|
int numVerts = dm->getNumVerts(dm);
|
|
|
|
int i, j, count = 0;
|
2007-08-18 04:03:03 +00:00
|
|
|
float co[3], nor[3];
|
|
|
|
|
2007-09-27 23:36:17 +00:00
|
|
|
/* check that dvert and index are valid pointers (just in case) */
|
|
|
|
if (dvert && index) {
|
|
|
|
/* get the average of all verts with that are in the vertex-group */
|
|
|
|
for (i = 0; i < numVerts; i++, index++) {
|
|
|
|
for (j = 0; j < dvert[i].totweight; j++) {
|
|
|
|
/* does this vertex belong to nominated vertex group? */
|
|
|
|
if (dvert[i].dw[j].def_nr == dgroup) {
|
|
|
|
dm->getVertCo(dm, i, co);
|
|
|
|
dm->getVertNo(dm, i, nor);
|
|
|
|
VecAddf(vec, vec, co);
|
|
|
|
VecAddf(normal, normal, nor);
|
|
|
|
count++;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2007-08-17 11:23:48 +00:00
|
|
|
}
|
|
|
|
}
|
2007-09-27 23:36:17 +00:00
|
|
|
|
|
|
|
|
|
|
|
/* calculate averages of normal and coordinates */
|
|
|
|
if (count > 0) {
|
|
|
|
VecMulf(vec, 1.0f / count);
|
|
|
|
VecMulf(normal, 1.0f / count);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/* derive the rotation from the average normal:
|
|
|
|
* - code taken from transform_manipulator.c,
|
|
|
|
* calc_manipulator_stats, V3D_MANIP_NORMAL case
|
|
|
|
*/
|
|
|
|
/* we need the transpose of the inverse for a normal... */
|
|
|
|
Mat3CpyMat4(imat, ob->obmat);
|
|
|
|
|
|
|
|
Mat3Inv(tmat, imat);
|
|
|
|
Mat3Transp(tmat);
|
|
|
|
Mat3MulVecfl(tmat, normal);
|
|
|
|
|
|
|
|
Normalize(normal);
|
|
|
|
VECCOPY(plane, tmat[1]);
|
|
|
|
|
|
|
|
VECCOPY(tmat[2], normal);
|
|
|
|
Crossf(tmat[0], normal, plane);
|
|
|
|
Crossf(tmat[1], tmat[2], tmat[0]);
|
|
|
|
|
|
|
|
Mat4CpyMat3(mat, tmat);
|
|
|
|
Mat4Ortho(mat);
|
|
|
|
|
|
|
|
|
|
|
|
/* apply the average coordinate as the new location */
|
|
|
|
VecMat4MulVecfl(tvec, ob->obmat, vec);
|
|
|
|
VECCOPY(mat[3], tvec);
|
2007-08-17 11:23:48 +00:00
|
|
|
}
|
|
|
|
}
|
2007-08-28 08:43:38 +00:00
|
|
|
|
|
|
|
/* free temporary DerivedMesh created (in EditMode case) */
|
|
|
|
if (G.editMesh) {
|
|
|
|
if (dm) dm->release(dm);
|
|
|
|
}
|
2007-08-17 11:23:48 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/* function that sets the given matrix based on given vertex group in lattice */
|
|
|
|
static void contarget_get_lattice_mat (Object *ob, char *substring, float mat[][4])
|
|
|
|
{
|
|
|
|
Lattice *lt= (Lattice *)ob->data;
|
|
|
|
|
|
|
|
DispList *dl = find_displist(&ob->disp, DL_VERTS);
|
|
|
|
float *co = dl?dl->verts:NULL;
|
|
|
|
BPoint *bp = lt->def;
|
|
|
|
|
|
|
|
MDeformVert *dvert = lt->dvert;
|
|
|
|
int tot_verts= lt->pntsu*lt->pntsv*lt->pntsw;
|
|
|
|
float vec[3]= {0.0f, 0.0f, 0.0f}, tvec[3];
|
|
|
|
int dgroup=0, grouped=0;
|
|
|
|
int i, n;
|
|
|
|
|
|
|
|
/* initialize target matrix using target matrix */
|
|
|
|
Mat4CpyMat4(mat, ob->obmat);
|
|
|
|
|
|
|
|
/* get index of vertex group */
|
|
|
|
dgroup = get_named_vertexgroup_num(ob, substring);
|
|
|
|
if (dgroup < 0) return;
|
2007-09-27 23:36:17 +00:00
|
|
|
if (dvert == NULL) return;
|
2007-08-17 11:23:48 +00:00
|
|
|
|
|
|
|
/* 1. Loop through control-points checking if in nominated vertex-group.
|
|
|
|
* 2. If it is, add it to vec to find the average point.
|
|
|
|
*/
|
|
|
|
for (i=0; i < tot_verts; i++, dvert++) {
|
|
|
|
for (n= 0; n < dvert->totweight; n++) {
|
|
|
|
/* found match - vert is in vgroup */
|
|
|
|
if (dvert->dw[n].def_nr == dgroup) {
|
|
|
|
/* copy coordinates of point to temporary vector, then add to find average */
|
|
|
|
if (co)
|
|
|
|
memcpy(tvec, co, 3*sizeof(float));
|
|
|
|
else
|
|
|
|
memcpy(tvec, bp->vec, 3*sizeof(float));
|
|
|
|
|
|
|
|
VecAddf(vec, vec, tvec);
|
|
|
|
grouped++;
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* advance pointer to coordinate data */
|
|
|
|
if (co) co+= 3;
|
|
|
|
else bp++;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* find average location, then multiply by ob->obmat to find world-space location */
|
|
|
|
if (grouped)
|
|
|
|
VecMulf(vec, 1.0f / grouped);
|
|
|
|
VecMat4MulVecfl(tvec, ob->obmat, vec);
|
|
|
|
|
|
|
|
/* copy new location to matrix */
|
|
|
|
VECCOPY(mat[3], tvec);
|
|
|
|
}
|
|
|
|
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
/* generic function to get the appropriate matrix for most target cases */
|
|
|
|
/* The cases where the target can be object data have not been implemented */
|
Patch #7767: Constraint Subtargets can now target anywhere on a bone, not just the head or tail
Patch by: Roland Hess (harkyman)
For example, a constraint can be sub-targeted at the 50% (or 31.2% or 85% etc.) point of its target bone, giving you enormous rigging flexibility and removing the need for complex contraptions to do such things as:
- A bone whose base slides only between to points on a rig (CopyLoc with a variable, animated subtarget point)
- Bones that attach to multiple points along another bone (CopyLocs, each with a different head/tail percentage)
- Bones that need to stretch to a point midway between specific spots on two other bones (old way: too crazy to mention; new way: stretch bone between points on end bones, then another stretch to the midpoint of the first stretch)
It is only used for the constraint types for which it is relevant: CopyLoc, TrackTo, StretchTo and MinMax, TrackTo, and Floor.
Notes:
- This is accessed by the Head/Tail number-slider.
- This value can be animated per constraint
- The old "Copy Bone Tail" option for the CopyLoc constraint has been automatically converted to 1.0 Head/Bone values for the affected constraints
- In the code, this value is in the bConstraint struct, so it is available for all constraints, even though only a few implement it.
2007-11-12 04:17:03 +00:00
|
|
|
static void constraint_target_to_mat4 (Object *ob, char *substring, float mat[][4], short from, short to, float headtail)
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
{
|
2002-10-12 11:37:38 +00:00
|
|
|
/* Case OBJECT */
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
if (!strlen(substring)) {
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
Mat4CpyMat4(mat, ob->obmat);
|
|
|
|
constraint_mat_convertspace(ob, NULL, mat, from, to);
|
|
|
|
}
|
|
|
|
/* Case VERTEXGROUP */
|
2007-08-17 11:23:48 +00:00
|
|
|
/* Current method just takes the average location of all the points in the
|
2007-08-18 04:03:03 +00:00
|
|
|
* VertexGroup, and uses that as the location value of the targets. Where
|
|
|
|
* possible, the orientation will also be calculated, by calculating an
|
|
|
|
* 'average' vertex normal, and deriving the rotaation from that.
|
2007-08-17 11:23:48 +00:00
|
|
|
*
|
2007-08-18 04:03:03 +00:00
|
|
|
* NOTE: EditMode is not currently supported, and will most likely remain that
|
|
|
|
* way as constraints can only really affect things on object/bone level.
|
2007-08-17 11:23:48 +00:00
|
|
|
*/
|
|
|
|
else if (ob->type == OB_MESH) {
|
|
|
|
contarget_get_mesh_mat(ob, substring, mat);
|
|
|
|
constraint_mat_convertspace(ob, NULL, mat, from, to);
|
|
|
|
}
|
|
|
|
else if (ob->type == OB_LATTICE) {
|
|
|
|
contarget_get_lattice_mat(ob, substring, mat);
|
|
|
|
constraint_mat_convertspace(ob, NULL, mat, from, to);
|
2002-10-12 11:37:38 +00:00
|
|
|
}
|
|
|
|
/* Case BONE */
|
|
|
|
else {
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
bPoseChannel *pchan;
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
pchan = get_pose_channel(ob->pose, substring);
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
if (pchan) {
|
|
|
|
/* Multiply the PoseSpace accumulation/final matrix for this
|
|
|
|
* PoseChannel by the Armature Object's Matrix to get a worldspace
|
|
|
|
* matrix.
|
Result of 2 weeks of quiet coding work in Greece :)
Aim was to get a total refresh of the animation system. This
is needed because;
- we need to upgrade it with 21st century features
- current code is spaghetti/hack combo, and hides good design
- it should become lag-free with using dependency graphs
A full log, with complete code API/structure/design explanation
will follow, that's a load of work... so here below the list with
hot changes;
- The entire object update system (matrices, geometry) is now
centralized. Calls to where_is_object and makeDispList are
forbidden, instead we tag objects 'changed' and let the
depgraph code sort it out
- Removed all old "Ika" code
- Depgraph is aware of all relationships, including meta balls,
constraints, bevelcurve, and so on.
- Made depgraph aware of relation types and layers, to do smart
flushing of 'changed' events. Nothing gets calculated too often!
- Transform uses depgraph to detect changes
- On frame-advance, depgraph flushes animated changes
Armatures;
Almost all armature related code has been fully built from scratch.
It now reveils the original design much better, with a very clean
implementation, lag free without even calculating each Bone more than
once. Result is quite a speedup yes!
Important to note is;
1) Armature is data containing the 'rest position'
2) Pose is the changes of rest position, and always on object level.
That way more Objects can use same Pose. Also constraints are in Pose
3) Actions only contain the Ipos to change values in Poses.
- Bones draw unrotated now
- Drawing bones speedup enormously (10-20 times)
- Bone selecting in EditMode, selection state is saved for PoseMode,
and vice-versa
- Undo in editmode
- Bone renaming does vertexgroups, constraints, posechannels, actions,
for all users of Armature in entire file
- Added Bone renaming in NKey panel
- Nkey PoseMode shows eulers now
- EditMode and PoseMode now have 'active' bone too (last clicked)
- Parenting in EditMode' CTRL+P, ALT+P, with nice options!
- Pose is added in Outliner now, with showing that constraints are in
the Pose, not Armature
- Disconnected IK solving from constraints. It's a separate phase now,
on top of the full Pose calculations
- Pose itself has a dependency graph too, so evaluation order is lag free.
TODO NOW;
- Rotating in Posemode has incorrect inverse transform (Martin will fix)
- Python Bone/Armature/Pose API disabled... needs full recode too
(wait for my doc!)
- Game engine will need upgrade too
- Depgraph code needs revision, cleanup, can be much faster!
(But, compliments for Jean-Luc, it works like a charm!)
- IK changed, it now doesnt use previous position to advance to next
position anymore. That system looks nice (no flips) but is not well
suited for NLA and background render.
TODO LATER;
We now can do loadsa new nifty features as well; like:
- Kill PoseMode (can be option for armatures itself)
- Make B-Bones (Bezier, Bspline, like for spines)
- Move all silly button level edit to 3d window (like CTRL+I = add
IK)
- Much better & informative drawing
- Fix action/nla editors
- Put all ipos in Actions (object, mesh key, lamp color)
- Add hooks
- Null bones
- Much more advanced constraints...
Bugfixes;
- OGL render (view3d header) had wrong first frame on anim render
- Ipo 'recording' mode had wrong playback speed
- Vertex-key mode now sticks to show 'active key', until frame change
-Ton-
2005-07-03 17:35:38 +00:00
|
|
|
*/
|
Patch #7767: Constraint Subtargets can now target anywhere on a bone, not just the head or tail
Patch by: Roland Hess (harkyman)
For example, a constraint can be sub-targeted at the 50% (or 31.2% or 85% etc.) point of its target bone, giving you enormous rigging flexibility and removing the need for complex contraptions to do such things as:
- A bone whose base slides only between to points on a rig (CopyLoc with a variable, animated subtarget point)
- Bones that attach to multiple points along another bone (CopyLocs, each with a different head/tail percentage)
- Bones that need to stretch to a point midway between specific spots on two other bones (old way: too crazy to mention; new way: stretch bone between points on end bones, then another stretch to the midpoint of the first stretch)
It is only used for the constraint types for which it is relevant: CopyLoc, TrackTo, StretchTo and MinMax, TrackTo, and Floor.
Notes:
- This is accessed by the Head/Tail number-slider.
- This value can be animated per constraint
- The old "Copy Bone Tail" option for the CopyLoc constraint has been automatically converted to 1.0 Head/Bone values for the affected constraints
- In the code, this value is in the bConstraint struct, so it is available for all constraints, even though only a few implement it.
2007-11-12 04:17:03 +00:00
|
|
|
if (headtail < 0.000001) {
|
|
|
|
/* skip length interpolation if set to head */
|
|
|
|
Mat4MulMat4(mat, pchan->pose_mat, ob->obmat);
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
float tempmat[4][4], loc[3];
|
|
|
|
|
|
|
|
/* interpolate along length of bone */
|
|
|
|
VecLerpf(loc, pchan->pose_head, pchan->pose_tail, headtail);
|
|
|
|
|
|
|
|
/* use interpolated distance for subtarget */
|
|
|
|
Mat4CpyMat4(tempmat, pchan->pose_mat);
|
|
|
|
VecCopyf(tempmat[3], loc);
|
|
|
|
|
|
|
|
Mat4MulMat4(mat, tempmat, ob->obmat);
|
|
|
|
}
|
2002-10-12 11:37:38 +00:00
|
|
|
}
|
|
|
|
else
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
Mat4CpyMat4(mat, ob->obmat);
|
|
|
|
|
|
|
|
/* convert matrix space as required */
|
|
|
|
constraint_mat_convertspace(ob, pchan, mat, from, to);
|
2004-01-27 06:08:37 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* ************************* Specific Constraints ***************************** */
|
|
|
|
/* Each constraint defines a set of functions, which will be called at the appropriate
|
|
|
|
* times. In addition to this, each constraint should have a type-info struct, where
|
|
|
|
* its functions are attached for use.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* Template for type-info data:
|
|
|
|
* - make a copy of this when creating new constraints, and just change the functions
|
|
|
|
* pointed to as necessary
|
|
|
|
* - although the naming of functions doesn't matter, it would help for code
|
|
|
|
* readability, to follow the same naming convention as is presented here
|
|
|
|
* - any functions that a constraint doesn't need to define, don't define
|
|
|
|
* for such cases, just use NULL
|
|
|
|
* - these should be defined after all the functions have been defined, so that
|
|
|
|
* forward-definitions/prototypes don't need to be used!
|
|
|
|
* - keep this copy #if-def'd so that future constraints can get based off this
|
|
|
|
*/
|
|
|
|
#if 0
|
|
|
|
static bConstraintTypeInfo CTI_CONSTRNAME = {
|
|
|
|
CONSTRAINT_TYPE_CONSTRNAME, /* type */
|
|
|
|
sizeof(bConstrNameConstraint), /* size */
|
|
|
|
"ConstrName", /* name */
|
|
|
|
"bConstrNameConstraint", /* struct name */
|
|
|
|
constrname_free, /* free data */
|
|
|
|
constrname_relink, /* relink data */
|
|
|
|
constrname_copy, /* copy data */
|
|
|
|
constrname_new_data, /* new data */
|
|
|
|
constrname_get_tars, /* get constraint targets */
|
|
|
|
constrname_flush_tars, /* flush constraint targets */
|
|
|
|
constrname_get_tarmat, /* get target matrix */
|
|
|
|
constrname_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
#endif
|
|
|
|
|
|
|
|
/* This function should be used for the get_target_matrix member of all
|
|
|
|
* constraints that are not picky about what happens to their target matrix.
|
|
|
|
*/
|
|
|
|
static void default_get_tarmat (bConstraint *con, bConstraintOb *cob, bConstraintTarget *ct, float ctime)
|
|
|
|
{
|
|
|
|
if (VALID_CONS_TARGET(ct))
|
Patch #7767: Constraint Subtargets can now target anywhere on a bone, not just the head or tail
Patch by: Roland Hess (harkyman)
For example, a constraint can be sub-targeted at the 50% (or 31.2% or 85% etc.) point of its target bone, giving you enormous rigging flexibility and removing the need for complex contraptions to do such things as:
- A bone whose base slides only between to points on a rig (CopyLoc with a variable, animated subtarget point)
- Bones that attach to multiple points along another bone (CopyLocs, each with a different head/tail percentage)
- Bones that need to stretch to a point midway between specific spots on two other bones (old way: too crazy to mention; new way: stretch bone between points on end bones, then another stretch to the midpoint of the first stretch)
It is only used for the constraint types for which it is relevant: CopyLoc, TrackTo, StretchTo and MinMax, TrackTo, and Floor.
Notes:
- This is accessed by the Head/Tail number-slider.
- This value can be animated per constraint
- The old "Copy Bone Tail" option for the CopyLoc constraint has been automatically converted to 1.0 Head/Bone values for the affected constraints
- In the code, this value is in the bConstraint struct, so it is available for all constraints, even though only a few implement it.
2007-11-12 04:17:03 +00:00
|
|
|
constraint_target_to_mat4(ct->tar, ct->subtarget, ct->matrix, CONSTRAINT_SPACE_WORLD, ct->space, con->headtail);
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
else if (ct)
|
|
|
|
Mat4One(ct->matrix);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This following macro should be used for all standard single-target *_get_tars functions
|
|
|
|
* to save typing and reduce maintainance woes.
|
|
|
|
* (Hopefully all compilers will be happy with the lines with just a space on them. Those are
|
|
|
|
* really just to help this code easier to read)
|
|
|
|
*/
|
2007-10-24 14:58:31 +00:00
|
|
|
#define SINGLETARGET_GET_TARS(con, datatar, datasubtarget, ct, list) \
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{ \
|
|
|
|
ct= MEM_callocN(sizeof(bConstraintTarget), "tempConstraintTarget"); \
|
|
|
|
\
|
2007-10-24 14:58:31 +00:00
|
|
|
ct->tar= datatar; \
|
|
|
|
strcpy(ct->subtarget, datasubtarget); \
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
ct->space= con->tarspace; \
|
|
|
|
ct->flag= CONSTRAINT_TAR_TEMP; \
|
|
|
|
\
|
|
|
|
if (ct->tar) { \
|
|
|
|
if ((ct->tar->type==OB_ARMATURE) && (ct->subtarget[0])) ct->type = CONSTRAINT_OBTYPE_BONE; \
|
|
|
|
else if (ELEM(ct->tar->type, OB_MESH, OB_LATTICE) && (ct->subtarget[0])) ct->type = CONSTRAINT_OBTYPE_VERT; \
|
|
|
|
else ct->type = CONSTRAINT_OBTYPE_OBJECT; \
|
|
|
|
} \
|
|
|
|
\
|
|
|
|
BLI_addtail(list, ct); \
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This following macro should be used for all standard single-target *_get_tars functions
|
|
|
|
* to save typing and reduce maintainance woes. It does not do the subtarget related operations
|
|
|
|
* (Hopefully all compilers will be happy with the lines with just a space on them. Those are
|
|
|
|
* really just to help this code easier to read)
|
|
|
|
*/
|
2007-10-24 14:58:31 +00:00
|
|
|
#define SINGLETARGETNS_GET_TARS(con, datatar, ct, list) \
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{ \
|
|
|
|
ct= MEM_callocN(sizeof(bConstraintTarget), "tempConstraintTarget"); \
|
|
|
|
\
|
2007-10-24 14:58:31 +00:00
|
|
|
ct->tar= datatar; \
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
ct->space= con->tarspace; \
|
|
|
|
ct->flag= CONSTRAINT_TAR_TEMP; \
|
|
|
|
\
|
|
|
|
if (ct->tar) ct->type = CONSTRAINT_OBTYPE_OBJECT; \
|
|
|
|
\
|
|
|
|
BLI_addtail(list, ct); \
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This following macro should be used for all standard single-target *_flush_tars functions
|
|
|
|
* to save typing and reduce maintainance woes.
|
2007-12-06 10:20:03 +00:00
|
|
|
* Note: the pointer to ct will be changed to point to the next in the list (as it gets removed)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
* (Hopefully all compilers will be happy with the lines with just a space on them. Those are
|
|
|
|
* really just to help this code easier to read)
|
|
|
|
*/
|
2007-10-24 14:58:31 +00:00
|
|
|
#define SINGLETARGET_FLUSH_TARS(con, datatar, datasubtarget, ct, list, nocopy) \
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{ \
|
|
|
|
if (ct) { \
|
2007-12-06 10:20:03 +00:00
|
|
|
bConstraintTarget *ctn = ct->next; \
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (nocopy == 0) { \
|
2007-10-24 14:58:31 +00:00
|
|
|
datatar= ct->tar; \
|
|
|
|
strcpy(datasubtarget, ct->subtarget); \
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
con->tarspace= ct->space; \
|
|
|
|
} \
|
|
|
|
\
|
2007-10-24 14:58:31 +00:00
|
|
|
BLI_freelinkN(list, ct); \
|
2007-12-06 10:20:03 +00:00
|
|
|
ct= ctn; \
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
} \
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This following macro should be used for all standard single-target *_flush_tars functions
|
2007-12-06 10:20:03 +00:00
|
|
|
* to save typing and reduce maintainance woes. It does not do the subtarget related operations.
|
|
|
|
* Note: the pointer to ct will be changed to point to the next in the list (as it gets removed)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
* (Hopefully all compilers will be happy with the lines with just a space on them. Those are
|
|
|
|
* really just to help this code easier to read)
|
|
|
|
*/
|
2007-10-24 14:58:31 +00:00
|
|
|
#define SINGLETARGETNS_FLUSH_TARS(con, datatar, ct, list, nocopy) \
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{ \
|
|
|
|
if (ct) { \
|
2007-12-06 10:20:03 +00:00
|
|
|
bConstraintTarget *ctn = ct->next; \
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (nocopy == 0) { \
|
2007-10-24 14:58:31 +00:00
|
|
|
datatar= ct->tar; \
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
con->tarspace= ct->space; \
|
|
|
|
} \
|
|
|
|
\
|
2007-10-24 14:58:31 +00:00
|
|
|
BLI_freelinkN(list, ct); \
|
2007-12-06 10:20:03 +00:00
|
|
|
ct= ctn; \
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
} \
|
|
|
|
}
|
|
|
|
|
|
|
|
/* --------- ChildOf Constraint ------------ */
|
|
|
|
|
|
|
|
static void childof_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bChildOfConstraint *data= (bChildOfConstraint *)cdata;
|
|
|
|
|
|
|
|
data->flag = (CHILDOF_LOCX | CHILDOF_LOCY | CHILDOF_LOCZ |
|
|
|
|
CHILDOF_ROTX |CHILDOF_ROTY | CHILDOF_ROTZ |
|
|
|
|
CHILDOF_SIZEX | CHILDOF_SIZEY | CHILDOF_SIZEZ);
|
|
|
|
Mat4One(data->invmat);
|
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int childof_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bChildOfConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* standard target-getting macro for single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void childof_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bChildOfConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void childof_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bChildOfConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= targets->first;
|
|
|
|
|
|
|
|
/* only evaluate if there is a target */
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
float parmat[4][4], invmat[4][4], tempmat[4][4];
|
|
|
|
float loc[3], eul[3], size[3];
|
|
|
|
float loco[3], eulo[3], sizo[3];
|
|
|
|
|
|
|
|
/* get offset (parent-inverse) matrix */
|
|
|
|
Mat4CpyMat4(invmat, data->invmat);
|
|
|
|
|
|
|
|
/* extract components of both matrices */
|
|
|
|
VECCOPY(loc, ct->matrix[3]);
|
|
|
|
Mat4ToEul(ct->matrix, eul);
|
|
|
|
Mat4ToSize(ct->matrix, size);
|
|
|
|
|
|
|
|
VECCOPY(loco, invmat[3]);
|
|
|
|
Mat4ToEul(invmat, eulo);
|
|
|
|
Mat4ToSize(invmat, sizo);
|
|
|
|
|
|
|
|
/* disable channels not enabled */
|
|
|
|
if (!(data->flag & CHILDOF_LOCX)) loc[0]= loco[0]= 0.0f;
|
|
|
|
if (!(data->flag & CHILDOF_LOCY)) loc[1]= loco[1]= 0.0f;
|
|
|
|
if (!(data->flag & CHILDOF_LOCZ)) loc[2]= loco[2]= 0.0f;
|
|
|
|
if (!(data->flag & CHILDOF_ROTX)) eul[0]= eulo[0]= 0.0f;
|
|
|
|
if (!(data->flag & CHILDOF_ROTY)) eul[1]= eulo[1]= 0.0f;
|
|
|
|
if (!(data->flag & CHILDOF_ROTZ)) eul[2]= eulo[2]= 0.0f;
|
|
|
|
if (!(data->flag & CHILDOF_SIZEX)) size[0]= sizo[0]= 1.0f;
|
|
|
|
if (!(data->flag & CHILDOF_SIZEY)) size[1]= sizo[1]= 1.0f;
|
|
|
|
if (!(data->flag & CHILDOF_SIZEZ)) size[2]= sizo[2]= 1.0f;
|
|
|
|
|
|
|
|
/* make new target mat and offset mat */
|
|
|
|
LocEulSizeToMat4(ct->matrix, loc, eul, size);
|
|
|
|
LocEulSizeToMat4(invmat, loco, eulo, sizo);
|
|
|
|
|
|
|
|
/* multiply target (parent matrix) by offset (parent inverse) to get
|
|
|
|
* the effect of the parent that will be exherted on the owner
|
|
|
|
*/
|
|
|
|
Mat4MulMat4(parmat, invmat, ct->matrix);
|
|
|
|
|
|
|
|
/* now multiply the parent matrix by the owner matrix to get the
|
|
|
|
* the effect of this constraint (i.e. owner is 'parented' to parent)
|
|
|
|
*/
|
|
|
|
Mat4CpyMat4(tempmat, cob->matrix);
|
|
|
|
Mat4MulMat4(cob->matrix, tempmat, parmat);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_CHILDOF = {
|
|
|
|
CONSTRAINT_TYPE_CHILDOF, /* type */
|
|
|
|
sizeof(bChildOfConstraint), /* size */
|
|
|
|
"ChildOf", /* name */
|
|
|
|
"bChildOfConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
childof_new_data, /* new data */
|
|
|
|
childof_get_tars, /* get constraint targets */
|
|
|
|
childof_flush_tars, /* flush constraint targets */
|
|
|
|
default_get_tarmat, /* get a target matrix */
|
|
|
|
childof_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* -------- TrackTo Constraint ------- */
|
|
|
|
|
|
|
|
static void trackto_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bTrackToConstraint *data= (bTrackToConstraint *)cdata;
|
|
|
|
|
|
|
|
data->reserved1 = TRACK_Y;
|
|
|
|
data->reserved2 = UP_Z;
|
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int trackto_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bTrackToConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* standard target-getting macro for single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void trackto_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bTrackToConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2006-12-03 18:38:51 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
static int basis_cross (int n, int m)
|
2006-12-03 18:38:51 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
switch (n-m) {
|
|
|
|
case 1:
|
|
|
|
case -2:
|
|
|
|
return 1;
|
|
|
|
|
|
|
|
case -1:
|
|
|
|
case 2:
|
|
|
|
return -1;
|
|
|
|
|
|
|
|
default:
|
|
|
|
return 0;
|
|
|
|
}
|
2006-12-03 18:38:51 +00:00
|
|
|
}
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
static void vectomat (float *vec, float *target_up, short axis, short upflag, short flags, float m[][3])
|
2006-12-03 18:38:51 +00:00
|
|
|
{
|
|
|
|
float n[3];
|
|
|
|
float u[3]; /* vector specifying the up axis */
|
|
|
|
float proj[3];
|
|
|
|
float right[3];
|
|
|
|
float neg = -1;
|
2006-12-05 02:06:51 +00:00
|
|
|
int right_index;
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
2006-12-03 18:38:51 +00:00
|
|
|
VecCopyf(n, vec);
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (Normalize(n) == 0.0) {
|
2006-12-03 18:38:51 +00:00
|
|
|
n[0] = 0.0;
|
|
|
|
n[1] = 0.0;
|
|
|
|
n[2] = 1.0;
|
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (axis > 2) axis -= 3;
|
2006-12-03 18:38:51 +00:00
|
|
|
else VecMulf(n,-1);
|
|
|
|
|
|
|
|
/* n specifies the transformation of the track axis */
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (flags & TARGET_Z_UP) {
|
2006-12-03 18:38:51 +00:00
|
|
|
/* target Z axis is the global up axis */
|
|
|
|
u[0] = target_up[0];
|
|
|
|
u[1] = target_up[1];
|
|
|
|
u[2] = target_up[2];
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* world Z axis is the global up axis */
|
|
|
|
u[0] = 0;
|
|
|
|
u[1] = 0;
|
|
|
|
u[2] = 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* project the up vector onto the plane specified by n */
|
|
|
|
Projf(proj, u, n); /* first u onto n... */
|
|
|
|
VecSubf(proj, u, proj); /* then onto the plane */
|
|
|
|
/* proj specifies the transformation of the up axis */
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (Normalize(proj) == 0.0) { /* degenerate projection */
|
2006-12-03 18:38:51 +00:00
|
|
|
proj[0] = 0.0;
|
|
|
|
proj[1] = 1.0;
|
|
|
|
proj[2] = 0.0;
|
|
|
|
}
|
|
|
|
|
2007-04-04 13:18:41 +00:00
|
|
|
/* Normalized cross product of n and proj specifies transformation of the right axis */
|
2006-12-03 18:38:51 +00:00
|
|
|
Crossf(right, proj, n);
|
2007-04-04 13:18:41 +00:00
|
|
|
Normalize(right);
|
2006-12-03 18:38:51 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (axis != upflag) {
|
2006-12-03 18:38:51 +00:00
|
|
|
right_index = 3 - axis - upflag;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
neg = (float)basis_cross(axis, upflag);
|
|
|
|
|
2006-12-03 18:38:51 +00:00
|
|
|
/* account for up direction, track direction */
|
|
|
|
m[right_index][0] = neg * right[0];
|
|
|
|
m[right_index][1] = neg * right[1];
|
|
|
|
m[right_index][2] = neg * right[2];
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
2006-12-03 18:38:51 +00:00
|
|
|
m[upflag][0] = proj[0];
|
|
|
|
m[upflag][1] = proj[1];
|
|
|
|
m[upflag][2] = proj[2];
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
2006-12-03 18:38:51 +00:00
|
|
|
m[axis][0] = n[0];
|
|
|
|
m[axis][1] = n[1];
|
|
|
|
m[axis][2] = n[2];
|
|
|
|
}
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
/* identity matrix - don't do anything if the two axes are the same */
|
2006-12-03 18:38:51 +00:00
|
|
|
else {
|
|
|
|
m[0][0]= m[1][1]= m[2][2]= 1.0;
|
|
|
|
m[0][1]= m[0][2]= m[0][3]= 0.0;
|
|
|
|
m[1][0]= m[1][2]= m[1][3]= 0.0;
|
|
|
|
m[2][0]= m[2][1]= m[2][3]= 0.0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
static void trackto_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
2002-10-12 11:37:38 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
bTrackToConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= targets->first;
|
|
|
|
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
float size[3], vec[3];
|
|
|
|
float totmat[3][3];
|
|
|
|
float tmat[4][4];
|
|
|
|
|
|
|
|
/* Get size property, since ob->size is only the object's own relative size, not its global one */
|
|
|
|
Mat4ToSize(cob->matrix, size);
|
|
|
|
|
|
|
|
/* Clear the object's rotation */
|
|
|
|
cob->matrix[0][0]=size[0];
|
|
|
|
cob->matrix[0][1]=0;
|
|
|
|
cob->matrix[0][2]=0;
|
|
|
|
cob->matrix[1][0]=0;
|
|
|
|
cob->matrix[1][1]=size[1];
|
|
|
|
cob->matrix[1][2]=0;
|
|
|
|
cob->matrix[2][0]=0;
|
|
|
|
cob->matrix[2][1]=0;
|
|
|
|
cob->matrix[2][2]=size[2];
|
|
|
|
|
|
|
|
/* targetmat[2] instead of ownermat[2] is passed to vectomat
|
|
|
|
* for backwards compatability it seems... (Aligorith)
|
|
|
|
*/
|
|
|
|
VecSubf(vec, cob->matrix[3], ct->matrix[3]);
|
|
|
|
vectomat(vec, ct->matrix[2],
|
|
|
|
(short)data->reserved1, (short)data->reserved2,
|
|
|
|
data->flags, totmat);
|
|
|
|
|
|
|
|
Mat4CpyMat4(tmat, cob->matrix);
|
|
|
|
Mat4MulMat34(cob->matrix, totmat, tmat);
|
|
|
|
}
|
|
|
|
}
|
2002-10-12 11:37:38 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
static bConstraintTypeInfo CTI_TRACKTO = {
|
|
|
|
CONSTRAINT_TYPE_TRACKTO, /* type */
|
|
|
|
sizeof(bTrackToConstraint), /* size */
|
|
|
|
"TrackTo", /* name */
|
|
|
|
"bTrackToConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
trackto_new_data, /* new data */
|
|
|
|
trackto_get_tars, /* get constraint targets */
|
|
|
|
trackto_flush_tars, /* flush constraint targets */
|
|
|
|
default_get_tarmat, /* get target matrix */
|
|
|
|
trackto_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* --------- Inverse-Kinemetics --------- */
|
|
|
|
|
|
|
|
static void kinematic_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bKinematicConstraint *data= (bKinematicConstraint *)cdata;
|
|
|
|
|
|
|
|
data->weight= (float)1.0;
|
|
|
|
data->orientweight= (float)1.0;
|
|
|
|
data->iterations = 500;
|
|
|
|
data->flag= CONSTRAINT_IK_TIP|CONSTRAINT_IK_STRETCH|CONSTRAINT_IK_POS;
|
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int kinematic_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bKinematicConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
/* standard target-getting macro for single-target constraints is used twice here */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
|
|
|
|
SINGLETARGET_GET_TARS(con, data->poletar, data->polesubtarget, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 2;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void kinematic_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bKinematicConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
|
|
|
|
SINGLETARGET_FLUSH_TARS(con, data->poletar, data->polesubtarget, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void kinematic_get_tarmat (bConstraint *con, bConstraintOb *cob, bConstraintTarget *ct, float ctime)
|
|
|
|
{
|
|
|
|
bKinematicConstraint *data= con->data;
|
|
|
|
|
|
|
|
if (VALID_CONS_TARGET(ct))
|
Patch #7767: Constraint Subtargets can now target anywhere on a bone, not just the head or tail
Patch by: Roland Hess (harkyman)
For example, a constraint can be sub-targeted at the 50% (or 31.2% or 85% etc.) point of its target bone, giving you enormous rigging flexibility and removing the need for complex contraptions to do such things as:
- A bone whose base slides only between to points on a rig (CopyLoc with a variable, animated subtarget point)
- Bones that attach to multiple points along another bone (CopyLocs, each with a different head/tail percentage)
- Bones that need to stretch to a point midway between specific spots on two other bones (old way: too crazy to mention; new way: stretch bone between points on end bones, then another stretch to the midpoint of the first stretch)
It is only used for the constraint types for which it is relevant: CopyLoc, TrackTo, StretchTo and MinMax, TrackTo, and Floor.
Notes:
- This is accessed by the Head/Tail number-slider.
- This value can be animated per constraint
- The old "Copy Bone Tail" option for the CopyLoc constraint has been automatically converted to 1.0 Head/Bone values for the affected constraints
- In the code, this value is in the bConstraint struct, so it is available for all constraints, even though only a few implement it.
2007-11-12 04:17:03 +00:00
|
|
|
constraint_target_to_mat4(ct->tar, ct->subtarget, ct->matrix, CONSTRAINT_SPACE_WORLD, ct->space, con->headtail);
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
else if (ct) {
|
|
|
|
if (data->flag & CONSTRAINT_IK_AUTO) {
|
|
|
|
Object *ob= cob->ob;
|
|
|
|
|
|
|
|
if (ob == NULL) {
|
|
|
|
Mat4One(ct->matrix);
|
2006-02-19 17:42:20 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
else {
|
|
|
|
float vec[3];
|
|
|
|
/* move grabtarget into world space */
|
|
|
|
VECCOPY(vec, data->grabtarget);
|
|
|
|
Mat4MulVecfl(ob->obmat, vec);
|
|
|
|
Mat4CpyMat4(ct->matrix, ob->obmat);
|
|
|
|
VECCOPY(ct->matrix[3], vec);
|
2002-10-12 11:37:38 +00:00
|
|
|
}
|
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
else
|
|
|
|
Mat4One(ct->matrix);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_KINEMATIC = {
|
|
|
|
CONSTRAINT_TYPE_KINEMATIC, /* type */
|
|
|
|
sizeof(bKinematicConstraint), /* size */
|
|
|
|
"IK", /* name */
|
|
|
|
"bKinematicConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
kinematic_new_data, /* new data */
|
|
|
|
kinematic_get_tars, /* get constraint targets */
|
|
|
|
kinematic_flush_tars, /* flush constraint targets */
|
|
|
|
kinematic_get_tarmat, /* get target matrix */
|
|
|
|
NULL /* evaluate - solved as separate loop */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* -------- Follow-Path Constraint ---------- */
|
|
|
|
|
|
|
|
static void followpath_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bFollowPathConstraint *data= (bFollowPathConstraint *)cdata;
|
|
|
|
|
|
|
|
data->trackflag = TRACK_Y;
|
|
|
|
data->upflag = UP_Z;
|
|
|
|
data->offset = 0;
|
|
|
|
data->followflag = 0;
|
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int followpath_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bFollowPathConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* standard target-getting macro for single-target constraints without subtargets */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGETNS_GET_TARS(con, data->tar, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void followpath_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bFollowPathConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGETNS_FLUSH_TARS(con, data->tar, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void followpath_get_tarmat (bConstraint *con, bConstraintOb *cob, bConstraintTarget *ct, float ctime)
|
|
|
|
{
|
|
|
|
bFollowPathConstraint *data= con->data;
|
|
|
|
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
Curve *cu= ct->tar->data;
|
2008-02-27 14:17:32 +00:00
|
|
|
float q[4], vec[4], dir[3], quat[4], x1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
float totmat[4][4];
|
|
|
|
float curvetime;
|
|
|
|
|
|
|
|
Mat4One(totmat);
|
|
|
|
Mat4One(ct->matrix);
|
|
|
|
|
|
|
|
/* note: when creating constraints that follow path, the curve gets the CU_PATH set now,
|
|
|
|
* currently for paths to work it needs to go through the bevlist/displist system (ton)
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* only happens on reload file, but violates depsgraph still... fix! */
|
|
|
|
if (cu->path==NULL || cu->path->data==NULL)
|
|
|
|
makeDispListCurveTypes(ct->tar, 0);
|
|
|
|
|
|
|
|
if (cu->path && cu->path->data) {
|
|
|
|
curvetime= bsystem_time(ct->tar, (float)ctime, 0.0) - data->offset;
|
2007-10-26 11:37:29 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (calc_ipo_spec(cu->ipo, CU_SPEED, &curvetime)==0) {
|
|
|
|
curvetime /= cu->pathlen;
|
|
|
|
CLAMP(curvetime, 0.0, 1.0);
|
2003-10-21 13:22:07 +00:00
|
|
|
}
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if ( where_on_path(ct->tar, curvetime, vec, dir) ) {
|
|
|
|
if (data->followflag) {
|
2008-02-27 14:17:32 +00:00
|
|
|
vectoquat(dir, (short) data->trackflag, (short) data->upflag, quat);
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
Normalize(dir);
|
|
|
|
q[0]= (float)cos(0.5*vec[3]);
|
|
|
|
x1= (float)sin(0.5*vec[3]);
|
|
|
|
q[1]= -x1*dir[0];
|
|
|
|
q[2]= -x1*dir[1];
|
|
|
|
q[3]= -x1*dir[2];
|
|
|
|
QuatMul(quat, q, quat);
|
2007-07-03 00:58:38 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
QuatToMat4(quat, totmat);
|
2003-10-21 13:22:07 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
VECCOPY(totmat[3], vec);
|
2007-06-18 07:41:21 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
Mat4MulSerie(ct->matrix, ct->tar->obmat, totmat, NULL, NULL, NULL, NULL, NULL, NULL);
|
2007-06-18 07:41:21 +00:00
|
|
|
}
|
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
else if (ct)
|
|
|
|
Mat4One(ct->matrix);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void followpath_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bConstraintTarget *ct= targets->first;
|
|
|
|
|
|
|
|
/* only evaluate if there is a target */
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
float obmat[4][4];
|
|
|
|
float size[3], obsize[3];
|
|
|
|
|
|
|
|
/* get Object local transform (loc/rot/size) to determine transformation from path */
|
|
|
|
//object_to_mat4(ob, obmat);
|
|
|
|
Mat4CpyMat4(obmat, cob->matrix); // FIXME!!!
|
|
|
|
|
|
|
|
/* get scaling of object before applying constraint */
|
|
|
|
Mat4ToSize(cob->matrix, size);
|
|
|
|
|
|
|
|
/* apply targetmat - containing location on path, and rotation */
|
|
|
|
Mat4MulSerie(cob->matrix, ct->matrix, obmat, NULL, NULL, NULL, NULL, NULL, NULL);
|
|
|
|
|
|
|
|
/* un-apply scaling caused by path */
|
|
|
|
Mat4ToSize(cob->matrix, obsize);
|
|
|
|
if (obsize[0])
|
|
|
|
VecMulf(cob->matrix[0], size[0] / obsize[0]);
|
|
|
|
if (obsize[1])
|
|
|
|
VecMulf(cob->matrix[1], size[1] / obsize[1]);
|
|
|
|
if (obsize[2])
|
|
|
|
VecMulf(cob->matrix[2], size[2] / obsize[2]);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_FOLLOWPATH = {
|
|
|
|
CONSTRAINT_TYPE_FOLLOWPATH, /* type */
|
|
|
|
sizeof(bFollowPathConstraint), /* size */
|
|
|
|
"Follow Path", /* name */
|
|
|
|
"bFollowPathConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
followpath_new_data, /* new data */
|
|
|
|
followpath_get_tars, /* get constraint targets */
|
|
|
|
followpath_flush_tars, /* flush constraint targets */
|
|
|
|
followpath_get_tarmat, /* get target matrix */
|
|
|
|
followpath_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* --------- Limit Location --------- */
|
|
|
|
|
|
|
|
|
|
|
|
static void loclimit_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bLocLimitConstraint *data = con->data;
|
|
|
|
|
|
|
|
if (data->flag & LIMIT_XMIN) {
|
|
|
|
if (cob->matrix[3][0] < data->xmin)
|
|
|
|
cob->matrix[3][0] = data->xmin;
|
|
|
|
}
|
|
|
|
if (data->flag & LIMIT_XMAX) {
|
|
|
|
if (cob->matrix[3][0] > data->xmax)
|
|
|
|
cob->matrix[3][0] = data->xmax;
|
|
|
|
}
|
|
|
|
if (data->flag & LIMIT_YMIN) {
|
|
|
|
if (cob->matrix[3][1] < data->ymin)
|
|
|
|
cob->matrix[3][1] = data->ymin;
|
|
|
|
}
|
|
|
|
if (data->flag & LIMIT_YMAX) {
|
|
|
|
if (cob->matrix[3][1] > data->ymax)
|
|
|
|
cob->matrix[3][1] = data->ymax;
|
|
|
|
}
|
|
|
|
if (data->flag & LIMIT_ZMIN) {
|
|
|
|
if (cob->matrix[3][2] < data->zmin)
|
|
|
|
cob->matrix[3][2] = data->zmin;
|
|
|
|
}
|
|
|
|
if (data->flag & LIMIT_ZMAX) {
|
|
|
|
if (cob->matrix[3][2] > data->zmax)
|
|
|
|
cob->matrix[3][2] = data->zmax;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_LOCLIMIT = {
|
|
|
|
CONSTRAINT_TYPE_LOCLIMIT, /* type */
|
|
|
|
sizeof(bLocLimitConstraint), /* size */
|
|
|
|
"Limit Location", /* name */
|
|
|
|
"bLocLimitConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
NULL, /* new data */
|
|
|
|
NULL, /* get constraint targets */
|
|
|
|
NULL, /* flush constraint targets */
|
|
|
|
NULL, /* get target matrix */
|
|
|
|
loclimit_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* -------- Limit Rotation --------- */
|
|
|
|
|
|
|
|
static void rotlimit_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bRotLimitConstraint *data = con->data;
|
|
|
|
float loc[3];
|
|
|
|
float eul[3];
|
|
|
|
float size[3];
|
|
|
|
|
|
|
|
VECCOPY(loc, cob->matrix[3]);
|
|
|
|
Mat4ToSize(cob->matrix, size);
|
|
|
|
|
|
|
|
Mat4ToEul(cob->matrix, eul);
|
|
|
|
|
|
|
|
/* eulers: radians to degrees! */
|
|
|
|
eul[0] = (eul[0] / M_PI * 180);
|
|
|
|
eul[1] = (eul[1] / M_PI * 180);
|
|
|
|
eul[2] = (eul[2] / M_PI * 180);
|
|
|
|
|
|
|
|
/* limiting of euler values... */
|
|
|
|
if (data->flag & LIMIT_XROT) {
|
|
|
|
if (eul[0] < data->xmin)
|
|
|
|
eul[0] = data->xmin;
|
2007-04-07 03:32:57 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (eul[0] > data->xmax)
|
|
|
|
eul[0] = data->xmax;
|
|
|
|
}
|
|
|
|
if (data->flag & LIMIT_YROT) {
|
|
|
|
if (eul[1] < data->ymin)
|
|
|
|
eul[1] = data->ymin;
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (eul[1] > data->ymax)
|
|
|
|
eul[1] = data->ymax;
|
|
|
|
}
|
|
|
|
if (data->flag & LIMIT_ZROT) {
|
|
|
|
if (eul[2] < data->zmin)
|
|
|
|
eul[2] = data->zmin;
|
2007-07-21 07:26:15 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (eul[2] > data->zmax)
|
|
|
|
eul[2] = data->zmax;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* eulers: degrees to radians ! */
|
|
|
|
eul[0] = (eul[0] / 180 * M_PI);
|
|
|
|
eul[1] = (eul[1] / 180 * M_PI);
|
|
|
|
eul[2] = (eul[2] / 180 * M_PI);
|
|
|
|
|
|
|
|
LocEulSizeToMat4(cob->matrix, loc, eul, size);
|
|
|
|
}
|
2004-09-05 20:21:16 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
static bConstraintTypeInfo CTI_ROTLIMIT = {
|
|
|
|
CONSTRAINT_TYPE_ROTLIMIT, /* type */
|
|
|
|
sizeof(bRotLimitConstraint), /* size */
|
|
|
|
"Limit Rotation", /* name */
|
|
|
|
"bRotLimitConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
NULL, /* new data */
|
|
|
|
NULL, /* get constraint targets */
|
|
|
|
NULL, /* flush constraint targets */
|
|
|
|
NULL, /* get target matrix */
|
|
|
|
rotlimit_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* --------- Limit Scaling --------- */
|
|
|
|
|
|
|
|
|
|
|
|
static void sizelimit_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bSizeLimitConstraint *data = con->data;
|
|
|
|
float obsize[3], size[3];
|
|
|
|
|
|
|
|
Mat4ToSize(cob->matrix, size);
|
|
|
|
Mat4ToSize(cob->matrix, obsize);
|
|
|
|
|
|
|
|
if (data->flag & LIMIT_XMIN) {
|
|
|
|
if (size[0] < data->xmin)
|
|
|
|
size[0] = data->xmin;
|
|
|
|
}
|
|
|
|
if (data->flag & LIMIT_XMAX) {
|
|
|
|
if (size[0] > data->xmax)
|
|
|
|
size[0] = data->xmax;
|
2002-10-12 11:37:38 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (data->flag & LIMIT_YMIN) {
|
|
|
|
if (size[1] < data->ymin)
|
|
|
|
size[1] = data->ymin;
|
|
|
|
}
|
|
|
|
if (data->flag & LIMIT_YMAX) {
|
|
|
|
if (size[1] > data->ymax)
|
|
|
|
size[1] = data->ymax;
|
|
|
|
}
|
|
|
|
if (data->flag & LIMIT_ZMIN) {
|
|
|
|
if (size[2] < data->zmin)
|
|
|
|
size[2] = data->zmin;
|
|
|
|
}
|
|
|
|
if (data->flag & LIMIT_ZMAX) {
|
|
|
|
if (size[2] > data->zmax)
|
|
|
|
size[2] = data->zmax;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (obsize[0])
|
|
|
|
VecMulf(cob->matrix[0], size[0]/obsize[0]);
|
|
|
|
if (obsize[1])
|
|
|
|
VecMulf(cob->matrix[1], size[1]/obsize[1]);
|
|
|
|
if (obsize[2])
|
|
|
|
VecMulf(cob->matrix[2], size[2]/obsize[2]);
|
|
|
|
}
|
2002-10-12 11:37:38 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
static bConstraintTypeInfo CTI_SIZELIMIT = {
|
|
|
|
CONSTRAINT_TYPE_SIZELIMIT, /* type */
|
|
|
|
sizeof(bSizeLimitConstraint), /* size */
|
|
|
|
"Limit Scaling", /* name */
|
|
|
|
"bSizeLimitConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
NULL, /* new data */
|
|
|
|
NULL, /* get constraint targets */
|
|
|
|
NULL, /* flush constraint targets */
|
|
|
|
NULL, /* get target matrix */
|
|
|
|
sizelimit_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* ----------- Copy Location ------------- */
|
|
|
|
|
|
|
|
static void loclike_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bLocateLikeConstraint *data= (bLocateLikeConstraint *)cdata;
|
|
|
|
|
|
|
|
data->flag = LOCLIKE_X|LOCLIKE_Y|LOCLIKE_Z;
|
2002-10-12 11:37:38 +00:00
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int loclike_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bLocateLikeConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* standard target-getting macro for single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
static void loclike_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
2002-10-12 11:37:38 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (con && list) {
|
|
|
|
bLocateLikeConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
2002-10-12 11:37:38 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
static void loclike_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bLocateLikeConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= targets->first;
|
|
|
|
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
float offset[3] = {0.0f, 0.0f, 0.0f};
|
|
|
|
|
|
|
|
if (data->flag & LOCLIKE_OFFSET)
|
|
|
|
VECCOPY(offset, cob->matrix[3]);
|
2006-09-22 15:57:58 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (data->flag & LOCLIKE_X) {
|
|
|
|
cob->matrix[3][0] = ct->matrix[3][0];
|
2006-09-22 15:57:58 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (data->flag & LOCLIKE_X_INVERT) cob->matrix[3][0] *= -1;
|
|
|
|
cob->matrix[3][0] += offset[0];
|
|
|
|
}
|
|
|
|
if (data->flag & LOCLIKE_Y) {
|
|
|
|
cob->matrix[3][1] = ct->matrix[3][1];
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (data->flag & LOCLIKE_Y_INVERT) cob->matrix[3][1] *= -1;
|
|
|
|
cob->matrix[3][1] += offset[1];
|
|
|
|
}
|
|
|
|
if (data->flag & LOCLIKE_Z) {
|
|
|
|
cob->matrix[3][2] = ct->matrix[3][2];
|
2005-09-07 00:11:39 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (data->flag & LOCLIKE_Z_INVERT) cob->matrix[3][2] *= -1;
|
|
|
|
cob->matrix[3][2] += offset[2];
|
2005-09-07 00:11:39 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
2002-10-12 11:37:38 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
static bConstraintTypeInfo CTI_LOCLIKE = {
|
|
|
|
CONSTRAINT_TYPE_LOCLIKE, /* type */
|
|
|
|
sizeof(bLocateLikeConstraint), /* size */
|
|
|
|
"Copy Location", /* name */
|
|
|
|
"bLocateLikeConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
loclike_new_data, /* new data */
|
|
|
|
loclike_get_tars, /* get constraint targets */
|
|
|
|
loclike_flush_tars, /* flush constraint targets */
|
Patch #7767: Constraint Subtargets can now target anywhere on a bone, not just the head or tail
Patch by: Roland Hess (harkyman)
For example, a constraint can be sub-targeted at the 50% (or 31.2% or 85% etc.) point of its target bone, giving you enormous rigging flexibility and removing the need for complex contraptions to do such things as:
- A bone whose base slides only between to points on a rig (CopyLoc with a variable, animated subtarget point)
- Bones that attach to multiple points along another bone (CopyLocs, each with a different head/tail percentage)
- Bones that need to stretch to a point midway between specific spots on two other bones (old way: too crazy to mention; new way: stretch bone between points on end bones, then another stretch to the midpoint of the first stretch)
It is only used for the constraint types for which it is relevant: CopyLoc, TrackTo, StretchTo and MinMax, TrackTo, and Floor.
Notes:
- This is accessed by the Head/Tail number-slider.
- This value can be animated per constraint
- The old "Copy Bone Tail" option for the CopyLoc constraint has been automatically converted to 1.0 Head/Bone values for the affected constraints
- In the code, this value is in the bConstraint struct, so it is available for all constraints, even though only a few implement it.
2007-11-12 04:17:03 +00:00
|
|
|
default_get_tarmat, /* get target matrix */
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
loclike_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* ----------- Copy Rotation ------------- */
|
|
|
|
|
|
|
|
static void rotlike_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bRotateLikeConstraint *data= (bRotateLikeConstraint *)cdata;
|
|
|
|
|
|
|
|
data->flag = ROTLIKE_X|ROTLIKE_Y|ROTLIKE_Z;
|
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int rotlike_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bRotateLikeConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* standard target-getting macro for single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void rotlike_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bRotateLikeConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void rotlike_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bRotateLikeConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= targets->first;
|
|
|
|
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
float loc[3];
|
|
|
|
float eul[3], obeul[3];
|
|
|
|
float size[3];
|
|
|
|
|
|
|
|
VECCOPY(loc, cob->matrix[3]);
|
|
|
|
Mat4ToSize(cob->matrix, size);
|
|
|
|
|
|
|
|
Mat4ToEul(ct->matrix, eul);
|
|
|
|
Mat4ToEul(cob->matrix, obeul);
|
|
|
|
|
2007-10-30 11:18:50 +00:00
|
|
|
if ((data->flag & ROTLIKE_X)==0)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
eul[0] = obeul[0];
|
2007-10-30 11:18:50 +00:00
|
|
|
else {
|
|
|
|
if (data->flag & ROTLIKE_OFFSET)
|
|
|
|
euler_rot(eul, obeul[0], 'x');
|
|
|
|
|
|
|
|
if (data->flag & ROTLIKE_X_INVERT)
|
|
|
|
eul[0] *= -1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
2007-10-30 11:18:50 +00:00
|
|
|
if ((data->flag & ROTLIKE_Y)==0)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
eul[1] = obeul[1];
|
2007-10-30 11:18:50 +00:00
|
|
|
else {
|
|
|
|
if (data->flag & ROTLIKE_OFFSET)
|
|
|
|
euler_rot(eul, obeul[1], 'y');
|
|
|
|
|
|
|
|
if (data->flag & ROTLIKE_Y_INVERT)
|
|
|
|
eul[1] *= -1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
2007-10-30 11:18:50 +00:00
|
|
|
if ((data->flag & ROTLIKE_Z)==0)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
eul[2] = obeul[2];
|
2007-10-30 11:18:50 +00:00
|
|
|
else {
|
|
|
|
if (data->flag & ROTLIKE_OFFSET)
|
|
|
|
euler_rot(eul, obeul[2], 'z');
|
|
|
|
|
|
|
|
if (data->flag & ROTLIKE_Z_INVERT)
|
|
|
|
eul[2] *= -1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
compatible_eul(eul, obeul);
|
|
|
|
LocEulSizeToMat4(cob->matrix, loc, eul, size);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_ROTLIKE = {
|
|
|
|
CONSTRAINT_TYPE_ROTLIKE, /* type */
|
|
|
|
sizeof(bRotateLikeConstraint), /* size */
|
|
|
|
"Copy Rotation", /* name */
|
|
|
|
"bRotateLikeConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
rotlike_new_data, /* new data */
|
|
|
|
rotlike_get_tars, /* get constraint targets */
|
|
|
|
rotlike_flush_tars, /* flush constraint targets */
|
|
|
|
default_get_tarmat, /* get target matrix */
|
|
|
|
rotlike_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* ---------- Copy Scaling ---------- */
|
|
|
|
|
|
|
|
static void sizelike_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bSizeLikeConstraint *data= (bSizeLikeConstraint *)cdata;
|
|
|
|
|
|
|
|
data->flag = SIZELIKE_X|SIZELIKE_Y|SIZELIKE_Z;
|
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int sizelike_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bSizeLikeConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* standard target-getting macro for single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void sizelike_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bSizeLikeConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void sizelike_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bSizeLikeConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= targets->first;
|
|
|
|
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
float obsize[3], size[3];
|
|
|
|
|
|
|
|
Mat4ToSize(ct->matrix, size);
|
|
|
|
Mat4ToSize(cob->matrix, obsize);
|
|
|
|
|
2007-10-31 10:04:57 +00:00
|
|
|
if ((data->flag & SIZELIKE_X) && (obsize[0] != 0)) {
|
|
|
|
if (data->flag & SIZELIKE_OFFSET) {
|
|
|
|
size[0] += (obsize[0] - 1.0f);
|
|
|
|
VecMulf(cob->matrix[0], size[0] / obsize[0]);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
VecMulf(cob->matrix[0], size[0] / obsize[0]);
|
|
|
|
}
|
|
|
|
if ((data->flag & SIZELIKE_Y) && (obsize[1] != 0)) {
|
|
|
|
if (data->flag & SIZELIKE_OFFSET) {
|
|
|
|
size[1] += (obsize[1] - 1.0f);
|
|
|
|
VecMulf(cob->matrix[1], size[1] / obsize[1]);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
VecMulf(cob->matrix[1], size[1] / obsize[1]);
|
|
|
|
}
|
|
|
|
if ((data->flag & SIZELIKE_Z) && (obsize[2] != 0)) {
|
|
|
|
if (data->flag & SIZELIKE_OFFSET) {
|
|
|
|
size[2] += (obsize[2] - 1.0f);
|
|
|
|
VecMulf(cob->matrix[2], size[2] / obsize[2]);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
VecMulf(cob->matrix[2], size[2] / obsize[2]);
|
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_SIZELIKE = {
|
|
|
|
CONSTRAINT_TYPE_SIZELIKE, /* type */
|
|
|
|
sizeof(bSizeLikeConstraint), /* size */
|
|
|
|
"Copy Scale", /* name */
|
|
|
|
"bSizeLikeConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
sizelike_new_data, /* new data */
|
|
|
|
sizelike_get_tars, /* get constraint targets */
|
|
|
|
sizelike_flush_tars, /* flush constraint targets */
|
|
|
|
default_get_tarmat, /* get target matrix */
|
|
|
|
sizelike_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* ----------- Python Constraint -------------- */
|
|
|
|
|
|
|
|
static void pycon_free (bConstraint *con)
|
|
|
|
{
|
|
|
|
bPythonConstraint *data= con->data;
|
|
|
|
|
|
|
|
/* id-properties */
|
|
|
|
IDP_FreeProperty(data->prop);
|
|
|
|
MEM_freeN(data->prop);
|
|
|
|
|
|
|
|
/* multiple targets */
|
2007-10-23 08:23:57 +00:00
|
|
|
BLI_freelistN(&data->targets);
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void pycon_relink (bConstraint *con)
|
|
|
|
{
|
|
|
|
bPythonConstraint *data= con->data;
|
|
|
|
|
|
|
|
ID_NEW(data->text);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void pycon_copy (bConstraint *con, bConstraint *srccon)
|
|
|
|
{
|
|
|
|
bPythonConstraint *pycon = (bPythonConstraint *)con->data;
|
|
|
|
bPythonConstraint *opycon = (bPythonConstraint *)srccon->data;
|
|
|
|
|
|
|
|
pycon->prop = IDP_CopyProperty(opycon->prop);
|
|
|
|
duplicatelist(&pycon->targets, &opycon->targets);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void pycon_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bPythonConstraint *data= (bPythonConstraint *)cdata;
|
|
|
|
|
|
|
|
/* everything should be set correctly by calloc, except for the prop->type constant.*/
|
|
|
|
data->prop = MEM_callocN(sizeof(IDProperty), "PyConstraintProps");
|
|
|
|
data->prop->type = IDP_GROUP;
|
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int pycon_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bPythonConstraint *data= con->data;
|
|
|
|
|
|
|
|
list->first = data->targets.first;
|
|
|
|
list->last = data->targets.last;
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return data->tarnum;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Whether this approach is maintained remains to be seen (aligorith) */
|
|
|
|
static void pycon_get_tarmat (bConstraint *con, bConstraintOb *cob, bConstraintTarget *ct, float ctime)
|
|
|
|
{
|
|
|
|
bPythonConstraint *data= con->data;
|
|
|
|
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
/* special exception for curves - depsgraph issues */
|
|
|
|
if (ct->tar->type == OB_CURVE) {
|
|
|
|
Curve *cu= ct->tar->data;
|
2002-10-12 11:37:38 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* this check is to make sure curve objects get updated on file load correctly.*/
|
|
|
|
if (cu->path==NULL || cu->path->data==NULL) /* only happens on reload file, but violates depsgraph still... fix! */
|
|
|
|
makeDispListCurveTypes(ct->tar, 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* firstly calculate the matrix the normal way, then let the py-function override
|
|
|
|
* this matrix if it needs to do so
|
|
|
|
*/
|
Patch #7767: Constraint Subtargets can now target anywhere on a bone, not just the head or tail
Patch by: Roland Hess (harkyman)
For example, a constraint can be sub-targeted at the 50% (or 31.2% or 85% etc.) point of its target bone, giving you enormous rigging flexibility and removing the need for complex contraptions to do such things as:
- A bone whose base slides only between to points on a rig (CopyLoc with a variable, animated subtarget point)
- Bones that attach to multiple points along another bone (CopyLocs, each with a different head/tail percentage)
- Bones that need to stretch to a point midway between specific spots on two other bones (old way: too crazy to mention; new way: stretch bone between points on end bones, then another stretch to the midpoint of the first stretch)
It is only used for the constraint types for which it is relevant: CopyLoc, TrackTo, StretchTo and MinMax, TrackTo, and Floor.
Notes:
- This is accessed by the Head/Tail number-slider.
- This value can be animated per constraint
- The old "Copy Bone Tail" option for the CopyLoc constraint has been automatically converted to 1.0 Head/Bone values for the affected constraints
- In the code, this value is in the bConstraint struct, so it is available for all constraints, even though only a few implement it.
2007-11-12 04:17:03 +00:00
|
|
|
constraint_target_to_mat4(ct->tar, ct->subtarget, ct->matrix, CONSTRAINT_SPACE_WORLD, ct->space, con->headtail);
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
BPY_pyconstraint_target(data, ct);
|
|
|
|
}
|
|
|
|
else if (ct)
|
|
|
|
Mat4One(ct->matrix);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void pycon_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bPythonConstraint *data= con->data;
|
|
|
|
|
|
|
|
/* currently removed, until I this can be re-implemented for multiple targets */
|
|
|
|
#if 0
|
|
|
|
/* Firstly, run the 'driver' function which has direct access to the objects involved
|
|
|
|
* Technically, this is potentially dangerous as users may abuse this and cause dependency-problems,
|
|
|
|
* but it also allows certain 'clever' rigging hacks to work.
|
|
|
|
*/
|
|
|
|
BPY_pyconstraint_driver(data, cob, targets);
|
|
|
|
#endif
|
|
|
|
|
|
|
|
/* Now, run the actual 'constraint' function, which should only access the matrices */
|
|
|
|
BPY_pyconstraint_eval(data, cob, targets);
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_PYTHON = {
|
|
|
|
CONSTRAINT_TYPE_PYTHON, /* type */
|
|
|
|
sizeof(bPythonConstraint), /* size */
|
|
|
|
"Script", /* name */
|
|
|
|
"bPythonConstraint", /* struct name */
|
|
|
|
pycon_free, /* free data */
|
|
|
|
pycon_relink, /* relink data */
|
|
|
|
pycon_copy, /* copy data */
|
|
|
|
pycon_new_data, /* new data */
|
|
|
|
pycon_get_tars, /* get constraint targets */
|
|
|
|
NULL, /* flush constraint targets */
|
|
|
|
pycon_get_tarmat, /* get target matrix */
|
|
|
|
pycon_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* -------- Action Constraint ----------- */
|
|
|
|
|
|
|
|
static void actcon_relink (bConstraint *con)
|
|
|
|
{
|
|
|
|
bActionConstraint *data= con->data;
|
|
|
|
ID_NEW(data->act);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void actcon_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bActionConstraint *data= (bActionConstraint *)cdata;
|
|
|
|
|
|
|
|
/* set type to 20 (Loc X), as 0 is Rot X for backwards compatability */
|
|
|
|
data->type = 20;
|
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int actcon_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bActionConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* standard target-getting macro for single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void actcon_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bActionConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void actcon_get_tarmat (bConstraint *con, bConstraintOb *cob, bConstraintTarget *ct, float ctime)
|
|
|
|
{
|
|
|
|
extern void chan_calc_mat(bPoseChannel *chan);
|
|
|
|
bActionConstraint *data = con->data;
|
|
|
|
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
float tempmat[4][4], vec[3];
|
|
|
|
float s, t;
|
|
|
|
short axis;
|
|
|
|
|
|
|
|
/* initialise return matrix */
|
|
|
|
Mat4One(ct->matrix);
|
|
|
|
|
|
|
|
/* get the transform matrix of the target */
|
Patch #7767: Constraint Subtargets can now target anywhere on a bone, not just the head or tail
Patch by: Roland Hess (harkyman)
For example, a constraint can be sub-targeted at the 50% (or 31.2% or 85% etc.) point of its target bone, giving you enormous rigging flexibility and removing the need for complex contraptions to do such things as:
- A bone whose base slides only between to points on a rig (CopyLoc with a variable, animated subtarget point)
- Bones that attach to multiple points along another bone (CopyLocs, each with a different head/tail percentage)
- Bones that need to stretch to a point midway between specific spots on two other bones (old way: too crazy to mention; new way: stretch bone between points on end bones, then another stretch to the midpoint of the first stretch)
It is only used for the constraint types for which it is relevant: CopyLoc, TrackTo, StretchTo and MinMax, TrackTo, and Floor.
Notes:
- This is accessed by the Head/Tail number-slider.
- This value can be animated per constraint
- The old "Copy Bone Tail" option for the CopyLoc constraint has been automatically converted to 1.0 Head/Bone values for the affected constraints
- In the code, this value is in the bConstraint struct, so it is available for all constraints, even though only a few implement it.
2007-11-12 04:17:03 +00:00
|
|
|
constraint_target_to_mat4(ct->tar, ct->subtarget, tempmat, CONSTRAINT_SPACE_WORLD, ct->space, con->headtail);
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
|
|
|
/* determine where in transform range target is */
|
|
|
|
/* data->type is mapped as follows for backwards compatability:
|
|
|
|
* 00,01,02 - rotation (it used to be like this)
|
|
|
|
* 10,11,12 - scaling
|
|
|
|
* 20,21,22 - location
|
|
|
|
*/
|
|
|
|
if (data->type < 10) {
|
|
|
|
/* extract rotation (is in whatever space target should be in) */
|
|
|
|
Mat4ToEul(tempmat, vec);
|
|
|
|
vec[0] *= (float)(180.0/M_PI);
|
|
|
|
vec[1] *= (float)(180.0/M_PI);
|
|
|
|
vec[2] *= (float)(180.0/M_PI);
|
|
|
|
axis= data->type;
|
|
|
|
}
|
|
|
|
else if (data->type < 20) {
|
|
|
|
/* extract scaling (is in whatever space target should be in) */
|
|
|
|
Mat4ToSize(tempmat, vec);
|
|
|
|
axis= data->type - 10;
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* extract location */
|
|
|
|
VECCOPY(vec, tempmat[3]);
|
|
|
|
axis= data->type - 20;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Target defines the animation */
|
|
|
|
s = (vec[axis]-data->min) / (data->max-data->min);
|
|
|
|
CLAMP(s, 0, 1);
|
|
|
|
t = ( s * (data->end-data->start)) + data->start;
|
|
|
|
|
|
|
|
/* Get the appropriate information from the action */
|
2007-10-22 02:43:07 +00:00
|
|
|
if (cob->type == CONSTRAINT_OBTYPE_BONE) {
|
|
|
|
bPose *pose;
|
|
|
|
bPoseChannel *pchan, *tchan;
|
|
|
|
|
|
|
|
/* make a temporary pose and evaluate using that */
|
|
|
|
pose = MEM_callocN(sizeof(bPose), "pose");
|
|
|
|
|
|
|
|
pchan = cob->pchan;
|
|
|
|
tchan= verify_pose_channel(pose, pchan->name);
|
|
|
|
extract_pose_from_action(pose, data->act, t);
|
|
|
|
|
|
|
|
chan_calc_mat(tchan);
|
|
|
|
|
|
|
|
Mat4CpyMat4(ct->matrix, tchan->chan_mat);
|
|
|
|
|
|
|
|
/* Clean up */
|
== Bone Groups ==
I'm committing some work-in-progress code for "bone groups" now, as I there have been are some major bugs caused by the timeoffset stuff (some of my test files were not loading, and other files were showing all sorts of weird problems).
Anyway, in this commit, the following things for "bone groups" have been done:
* Bone groups are stored per armature (internally, this is per bPose block)
* Added controls for editing bone-groups per armature - "add", "remove", "rename". These can be found in the "Links and Materials" panel in PoseMode, beside the settings for PoseLib.
* Reorganised buttons for editing selected bones in PoseMode. I've replaced the "dist" and "weight" buttons (they existed in EditMode anyway) with a menu to choose the bone-group and the custom-shape-ob field. In the place of the old custom-shape-ob field, I've restored the "Hide" button. This might break muscle-memory a bit, but there isn't a lot of space to play with there.
Some stuff I'd been originally planning to do before committing:
* When adding keyframes for bones, an action-group with the same name as the bone's group will be added to the action, and the action-channel will be made a member of that.
* New action/bone groups have unique names (renaming/adding new should check if name exists before assigning it)
* There's a setting under Bone-Groups stuff which sets which custom-colour set is used to colour that group's bones. Currently, this is non-functional, as the necessary drawing code for armatures is not in place yet.
2008-01-20 02:55:35 +00:00
|
|
|
free_pose(pose);
|
2007-10-22 02:43:07 +00:00
|
|
|
}
|
|
|
|
else if (cob->type == CONSTRAINT_OBTYPE_OBJECT) {
|
|
|
|
/* evaluate using workob */
|
|
|
|
what_does_obaction(cob->ob, data->act, t);
|
|
|
|
object_to_mat4(&workob, ct->matrix);
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* behaviour undefined... */
|
|
|
|
puts("Error: unknown owner type for Action Constraint");
|
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void actcon_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bConstraintTarget *ct= targets->first;
|
|
|
|
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
float temp[4][4];
|
|
|
|
|
|
|
|
/* Nice and simple... we just need to multiply the matrices, as the get_target_matrix
|
|
|
|
* function has already taken care of everything else.
|
|
|
|
*/
|
|
|
|
Mat4CpyMat4(temp, cob->matrix);
|
|
|
|
Mat4MulMat4(cob->matrix, ct->matrix, temp);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_ACTION = {
|
|
|
|
CONSTRAINT_TYPE_ACTION, /* type */
|
|
|
|
sizeof(bActionConstraint), /* size */
|
|
|
|
"Action", /* name */
|
|
|
|
"bActionConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
actcon_relink, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
actcon_new_data, /* new data */
|
|
|
|
actcon_get_tars, /* get constraint targets */
|
|
|
|
actcon_flush_tars, /* flush constraint targets */
|
|
|
|
actcon_get_tarmat, /* get target matrix */
|
|
|
|
actcon_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* --------- Locked Track ---------- */
|
|
|
|
|
|
|
|
static void locktrack_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bLockTrackConstraint *data= (bLockTrackConstraint *)cdata;
|
|
|
|
|
|
|
|
data->trackflag = TRACK_Y;
|
2007-10-26 11:37:29 +00:00
|
|
|
data->lockflag = LOCK_Z;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int locktrack_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bLockTrackConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void locktrack_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bLockTrackConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void locktrack_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bLockTrackConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= targets->first;
|
|
|
|
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
float vec[3],vec2[3];
|
|
|
|
float totmat[3][3];
|
|
|
|
float tmpmat[3][3];
|
|
|
|
float invmat[3][3];
|
|
|
|
float tmat[4][4];
|
|
|
|
float mdet;
|
|
|
|
|
|
|
|
/* Vector object -> target */
|
|
|
|
VecSubf(vec, ct->matrix[3], cob->matrix[3]);
|
|
|
|
switch (data->lockflag){
|
|
|
|
case LOCK_X: /* LOCK X */
|
|
|
|
{
|
|
|
|
switch (data->trackflag) {
|
|
|
|
case TRACK_Y: /* LOCK X TRACK Y */
|
|
|
|
{
|
|
|
|
/* Projection of Vector on the plane */
|
|
|
|
Projf(vec2, vec, cob->matrix[0]);
|
|
|
|
VecSubf(totmat[1], vec, vec2);
|
|
|
|
Normalize(totmat[1]);
|
|
|
|
|
|
|
|
/* the x axis is fixed */
|
|
|
|
totmat[0][0] = cob->matrix[0][0];
|
|
|
|
totmat[0][1] = cob->matrix[0][1];
|
|
|
|
totmat[0][2] = cob->matrix[0][2];
|
|
|
|
Normalize(totmat[0]);
|
|
|
|
|
|
|
|
/* the z axis gets mapped onto a third orthogonal vector */
|
|
|
|
Crossf(totmat[2], totmat[0], totmat[1]);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case TRACK_Z: /* LOCK X TRACK Z */
|
|
|
|
{
|
|
|
|
/* Projection of Vector on the plane */
|
|
|
|
Projf(vec2, vec, cob->matrix[0]);
|
|
|
|
VecSubf(totmat[2], vec, vec2);
|
|
|
|
Normalize(totmat[2]);
|
2007-10-26 11:37:29 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* the x axis is fixed */
|
|
|
|
totmat[0][0] = cob->matrix[0][0];
|
|
|
|
totmat[0][1] = cob->matrix[0][1];
|
|
|
|
totmat[0][2] = cob->matrix[0][2];
|
|
|
|
Normalize(totmat[0]);
|
2007-10-26 11:37:29 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* the z axis gets mapped onto a third orthogonal vector */
|
|
|
|
Crossf(totmat[1], totmat[2], totmat[0]);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case TRACK_nY: /* LOCK X TRACK -Y */
|
|
|
|
{
|
|
|
|
/* Projection of Vector on the plane */
|
|
|
|
Projf(vec2, vec, cob->matrix[0]);
|
|
|
|
VecSubf(totmat[1], vec, vec2);
|
|
|
|
Normalize(totmat[1]);
|
|
|
|
VecMulf(totmat[1],-1);
|
|
|
|
|
|
|
|
/* the x axis is fixed */
|
|
|
|
totmat[0][0] = cob->matrix[0][0];
|
|
|
|
totmat[0][1] = cob->matrix[0][1];
|
|
|
|
totmat[0][2] = cob->matrix[0][2];
|
|
|
|
Normalize(totmat[0]);
|
|
|
|
|
|
|
|
/* the z axis gets mapped onto a third orthogonal vector */
|
|
|
|
Crossf(totmat[2], totmat[0], totmat[1]);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case TRACK_nZ: /* LOCK X TRACK -Z */
|
|
|
|
{
|
|
|
|
/* Projection of Vector on the plane */
|
|
|
|
Projf(vec2, vec, cob->matrix[0]);
|
|
|
|
VecSubf(totmat[2], vec, vec2);
|
|
|
|
Normalize(totmat[2]);
|
|
|
|
VecMulf(totmat[2],-1);
|
|
|
|
|
|
|
|
/* the x axis is fixed */
|
|
|
|
totmat[0][0] = cob->matrix[0][0];
|
|
|
|
totmat[0][1] = cob->matrix[0][1];
|
|
|
|
totmat[0][2] = cob->matrix[0][2];
|
|
|
|
Normalize(totmat[0]);
|
|
|
|
|
|
|
|
/* the z axis gets mapped onto a third orthogonal vector */
|
|
|
|
Crossf(totmat[1], totmat[2], totmat[0]);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
{
|
|
|
|
totmat[0][0] = 1;totmat[0][1] = 0;totmat[0][2] = 0;
|
|
|
|
totmat[1][0] = 0;totmat[1][1] = 1;totmat[1][2] = 0;
|
|
|
|
totmat[2][0] = 0;totmat[2][1] = 0;totmat[2][2] = 1;
|
|
|
|
}
|
|
|
|
break;
|
2002-10-12 11:37:38 +00:00
|
|
|
}
|
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
break;
|
|
|
|
case LOCK_Y: /* LOCK Y */
|
2003-10-21 13:22:07 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
switch (data->trackflag) {
|
|
|
|
case TRACK_X: /* LOCK Y TRACK X */
|
2007-07-03 00:58:38 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* Projection of Vector on the plane */
|
|
|
|
Projf(vec2, vec, cob->matrix[1]);
|
|
|
|
VecSubf(totmat[0], vec, vec2);
|
|
|
|
Normalize(totmat[0]);
|
2007-07-03 00:58:38 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* the y axis is fixed */
|
|
|
|
totmat[1][0] = cob->matrix[1][0];
|
|
|
|
totmat[1][1] = cob->matrix[1][1];
|
|
|
|
totmat[1][2] = cob->matrix[1][2];
|
|
|
|
Normalize(totmat[1]);
|
|
|
|
|
|
|
|
/* the z axis gets mapped onto a third orthogonal vector */
|
|
|
|
Crossf(totmat[2], totmat[0], totmat[1]);
|
2007-07-03 00:58:38 +00:00
|
|
|
}
|
2003-10-21 13:22:07 +00:00
|
|
|
break;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
case TRACK_Z: /* LOCK Y TRACK Z */
|
2007-07-03 00:58:38 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* Projection of Vector on the plane */
|
|
|
|
Projf(vec2, vec, cob->matrix[1]);
|
|
|
|
VecSubf(totmat[2], vec, vec2);
|
|
|
|
Normalize(totmat[2]);
|
|
|
|
|
|
|
|
/* the y axis is fixed */
|
|
|
|
totmat[1][0] = cob->matrix[1][0];
|
|
|
|
totmat[1][1] = cob->matrix[1][1];
|
|
|
|
totmat[1][2] = cob->matrix[1][2];
|
|
|
|
Normalize(totmat[1]);
|
|
|
|
|
|
|
|
/* the z axis gets mapped onto a third orthogonal vector */
|
|
|
|
Crossf(totmat[0], totmat[1], totmat[2]);
|
2007-07-03 00:58:38 +00:00
|
|
|
}
|
2003-10-21 13:22:07 +00:00
|
|
|
break;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
case TRACK_nX: /* LOCK Y TRACK -X */
|
2007-07-03 00:58:38 +00:00
|
|
|
{
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* Projection of Vector on the plane */
|
|
|
|
Projf(vec2, vec, cob->matrix[1]);
|
|
|
|
VecSubf(totmat[0], vec, vec2);
|
|
|
|
Normalize(totmat[0]);
|
|
|
|
VecMulf(totmat[0],-1);
|
|
|
|
|
|
|
|
/* the y axis is fixed */
|
|
|
|
totmat[1][0] = cob->matrix[1][0];
|
|
|
|
totmat[1][1] = cob->matrix[1][1];
|
|
|
|
totmat[1][2] = cob->matrix[1][2];
|
|
|
|
Normalize(totmat[1]);
|
|
|
|
|
|
|
|
/* the z axis gets mapped onto a third orthogonal vector */
|
|
|
|
Crossf(totmat[2], totmat[0], totmat[1]);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case TRACK_nZ: /* LOCK Y TRACK -Z */
|
|
|
|
{
|
|
|
|
/* Projection of Vector on the plane */
|
|
|
|
Projf(vec2, vec, cob->matrix[1]);
|
|
|
|
VecSubf(totmat[2], vec, vec2);
|
|
|
|
Normalize(totmat[2]);
|
|
|
|
VecMulf(totmat[2],-1);
|
|
|
|
|
|
|
|
/* the y axis is fixed */
|
|
|
|
totmat[1][0] = cob->matrix[1][0];
|
|
|
|
totmat[1][1] = cob->matrix[1][1];
|
|
|
|
totmat[1][2] = cob->matrix[1][2];
|
|
|
|
Normalize(totmat[1]);
|
|
|
|
|
|
|
|
/* the z axis gets mapped onto a third orthogonal vector */
|
|
|
|
Crossf(totmat[0], totmat[1], totmat[2]);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
{
|
|
|
|
totmat[0][0] = 1;totmat[0][1] = 0;totmat[0][2] = 0;
|
|
|
|
totmat[1][0] = 0;totmat[1][1] = 1;totmat[1][2] = 0;
|
|
|
|
totmat[2][0] = 0;totmat[2][1] = 0;totmat[2][2] = 1;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case LOCK_Z: /* LOCK Z */
|
|
|
|
{
|
|
|
|
switch (data->trackflag) {
|
|
|
|
case TRACK_X: /* LOCK Z TRACK X */
|
|
|
|
{
|
|
|
|
/* Projection of Vector on the plane */
|
|
|
|
Projf(vec2, vec, cob->matrix[2]);
|
|
|
|
VecSubf(totmat[0], vec, vec2);
|
|
|
|
Normalize(totmat[0]);
|
|
|
|
|
|
|
|
/* the z axis is fixed */
|
|
|
|
totmat[2][0] = cob->matrix[2][0];
|
|
|
|
totmat[2][1] = cob->matrix[2][1];
|
|
|
|
totmat[2][2] = cob->matrix[2][2];
|
|
|
|
Normalize(totmat[2]);
|
|
|
|
|
|
|
|
/* the x axis gets mapped onto a third orthogonal vector */
|
|
|
|
Crossf(totmat[1], totmat[2], totmat[0]);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case TRACK_Y: /* LOCK Z TRACK Y */
|
|
|
|
{
|
|
|
|
/* Projection of Vector on the plane */
|
|
|
|
Projf(vec2, vec, cob->matrix[2]);
|
|
|
|
VecSubf(totmat[1], vec, vec2);
|
|
|
|
Normalize(totmat[1]);
|
|
|
|
|
|
|
|
/* the z axis is fixed */
|
|
|
|
totmat[2][0] = cob->matrix[2][0];
|
|
|
|
totmat[2][1] = cob->matrix[2][1];
|
|
|
|
totmat[2][2] = cob->matrix[2][2];
|
|
|
|
Normalize(totmat[2]);
|
|
|
|
|
|
|
|
/* the x axis gets mapped onto a third orthogonal vector */
|
|
|
|
Crossf(totmat[0], totmat[1], totmat[2]);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case TRACK_nX: /* LOCK Z TRACK -X */
|
|
|
|
{
|
|
|
|
/* Projection of Vector on the plane */
|
|
|
|
Projf(vec2, vec, cob->matrix[2]);
|
|
|
|
VecSubf(totmat[0], vec, vec2);
|
|
|
|
Normalize(totmat[0]);
|
|
|
|
VecMulf(totmat[0],-1);
|
|
|
|
|
|
|
|
/* the z axis is fixed */
|
|
|
|
totmat[2][0] = cob->matrix[2][0];
|
|
|
|
totmat[2][1] = cob->matrix[2][1];
|
|
|
|
totmat[2][2] = cob->matrix[2][2];
|
|
|
|
Normalize(totmat[2]);
|
|
|
|
|
|
|
|
/* the x axis gets mapped onto a third orthogonal vector */
|
|
|
|
Crossf(totmat[1], totmat[2], totmat[0]);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case TRACK_nY: /* LOCK Z TRACK -Y */
|
|
|
|
{
|
|
|
|
/* Projection of Vector on the plane */
|
|
|
|
Projf(vec2, vec, cob->matrix[2]);
|
|
|
|
VecSubf(totmat[1], vec, vec2);
|
|
|
|
Normalize(totmat[1]);
|
|
|
|
VecMulf(totmat[1],-1);
|
|
|
|
|
|
|
|
/* the z axis is fixed */
|
|
|
|
totmat[2][0] = cob->matrix[2][0];
|
|
|
|
totmat[2][1] = cob->matrix[2][1];
|
|
|
|
totmat[2][2] = cob->matrix[2][2];
|
|
|
|
Normalize(totmat[2]);
|
|
|
|
|
|
|
|
/* the x axis gets mapped onto a third orthogonal vector */
|
|
|
|
Crossf(totmat[0], totmat[1], totmat[2]);
|
2007-07-03 00:58:38 +00:00
|
|
|
}
|
2003-10-21 13:22:07 +00:00
|
|
|
break;
|
|
|
|
default:
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
2003-10-21 13:22:07 +00:00
|
|
|
totmat[0][0] = 1;totmat[0][1] = 0;totmat[0][2] = 0;
|
|
|
|
totmat[1][0] = 0;totmat[1][1] = 1;totmat[1][2] = 0;
|
|
|
|
totmat[2][0] = 0;totmat[2][1] = 0;totmat[2][2] = 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
{
|
|
|
|
totmat[0][0] = 1;totmat[0][1] = 0;totmat[0][2] = 0;
|
|
|
|
totmat[1][0] = 0;totmat[1][1] = 1;totmat[1][2] = 0;
|
|
|
|
totmat[2][0] = 0;totmat[2][1] = 0;totmat[2][2] = 1;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
/* Block to keep matrix heading */
|
|
|
|
tmpmat[0][0] = cob->matrix[0][0];tmpmat[0][1] = cob->matrix[0][1];tmpmat[0][2] = cob->matrix[0][2];
|
|
|
|
tmpmat[1][0] = cob->matrix[1][0];tmpmat[1][1] = cob->matrix[1][1];tmpmat[1][2] = cob->matrix[1][2];
|
|
|
|
tmpmat[2][0] = cob->matrix[2][0];tmpmat[2][1] = cob->matrix[2][1];tmpmat[2][2] = cob->matrix[2][2];
|
|
|
|
Normalize(tmpmat[0]);
|
|
|
|
Normalize(tmpmat[1]);
|
|
|
|
Normalize(tmpmat[2]);
|
|
|
|
Mat3Inv(invmat, tmpmat);
|
|
|
|
Mat3MulMat3(tmpmat, totmat, invmat);
|
|
|
|
totmat[0][0] = tmpmat[0][0];totmat[0][1] = tmpmat[0][1];totmat[0][2] = tmpmat[0][2];
|
|
|
|
totmat[1][0] = tmpmat[1][0];totmat[1][1] = tmpmat[1][1];totmat[1][2] = tmpmat[1][2];
|
|
|
|
totmat[2][0] = tmpmat[2][0];totmat[2][1] = tmpmat[2][1];totmat[2][2] = tmpmat[2][2];
|
|
|
|
|
|
|
|
Mat4CpyMat4(tmat, cob->matrix);
|
|
|
|
|
|
|
|
mdet = Det3x3( totmat[0][0],totmat[0][1],totmat[0][2],
|
|
|
|
totmat[1][0],totmat[1][1],totmat[1][2],
|
|
|
|
totmat[2][0],totmat[2][1],totmat[2][2]);
|
|
|
|
if (mdet==0) {
|
|
|
|
totmat[0][0] = 1;totmat[0][1] = 0;totmat[0][2] = 0;
|
|
|
|
totmat[1][0] = 0;totmat[1][1] = 1;totmat[1][2] = 0;
|
|
|
|
totmat[2][0] = 0;totmat[2][1] = 0;totmat[2][2] = 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* apply out transformaton to the object */
|
|
|
|
Mat4MulMat34(cob->matrix, totmat, tmat);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_LOCKTRACK = {
|
|
|
|
CONSTRAINT_TYPE_LOCKTRACK, /* type */
|
|
|
|
sizeof(bLockTrackConstraint), /* size */
|
|
|
|
"Locked Track", /* name */
|
|
|
|
"bLockTrackConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
locktrack_new_data, /* new data */
|
|
|
|
locktrack_get_tars, /* get constraint targets */
|
|
|
|
locktrack_flush_tars, /* flush constraint targets */
|
|
|
|
default_get_tarmat, /* get target matrix */
|
|
|
|
locktrack_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
2008-01-25 11:23:36 +00:00
|
|
|
/* ---------- Limit Distance Constraint ----------- */
|
|
|
|
|
|
|
|
static void distlimit_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bDistLimitConstraint *data= (bDistLimitConstraint *)cdata;
|
|
|
|
|
|
|
|
data->dist= 0.0;
|
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int distlimit_get_tars (bConstraint *con, ListBase *list)
|
2008-01-25 11:23:36 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bDistLimitConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* standard target-getting macro for single-target constraints */
|
|
|
|
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
2008-01-25 11:23:36 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
2008-01-25 11:23:36 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void distlimit_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bDistLimitConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
|
|
|
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void distlimit_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bDistLimitConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= targets->first;
|
|
|
|
|
|
|
|
/* only evaluate if there is a target */
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
float dvec[3], dist=0.0f, sfac=1.0f;
|
|
|
|
short clamp_surf= 0;
|
|
|
|
|
|
|
|
/* calculate our current distance from the target */
|
|
|
|
dist= VecLenf(cob->matrix[3], ct->matrix[3]);
|
|
|
|
|
|
|
|
/* set distance (flag is only set when user demands it) */
|
|
|
|
if (data->dist == 0)
|
|
|
|
data->dist= dist;
|
|
|
|
|
|
|
|
/* check if we're which way to clamp from, and calculate interpolation factor (if needed) */
|
|
|
|
if (data->mode == LIMITDIST_OUTSIDE) {
|
|
|
|
/* if inside, then move to surface */
|
|
|
|
if (dist <= data->dist) {
|
|
|
|
clamp_surf= 1;
|
|
|
|
sfac= data->dist / dist;
|
|
|
|
}
|
|
|
|
/* if soft-distance is enabled, start fading once owner is dist+softdist from the target */
|
|
|
|
else if (data->flag & LIMITDIST_USESOFT) {
|
|
|
|
if (dist <= (data->dist + data->soft)) {
|
|
|
|
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else if (data->mode == LIMITDIST_INSIDE) {
|
|
|
|
/* if outside, then move to surface */
|
|
|
|
if (dist >= data->dist) {
|
|
|
|
clamp_surf= 1;
|
|
|
|
sfac= data->dist / dist;
|
|
|
|
}
|
|
|
|
/* if soft-distance is enabled, start fading once owner is dist-soft from the target */
|
|
|
|
else if (data->flag & LIMITDIST_USESOFT) {
|
|
|
|
// FIXME: there's a problem with "jumping" when this kicks in
|
|
|
|
if (dist >= (data->dist - data->soft)) {
|
|
|
|
sfac = data->soft*(1.0 - exp(-(dist - data->dist)/data->soft)) + data->dist;
|
|
|
|
sfac /= dist;
|
|
|
|
|
|
|
|
clamp_surf= 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
if (IS_EQ(dist, data->dist)==0) {
|
|
|
|
clamp_surf= 1;
|
|
|
|
sfac= data->dist / dist;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* clamp to 'surface' (i.e. move owner so that dist == data->dist) */
|
|
|
|
if (clamp_surf) {
|
|
|
|
/* simply interpolate along line formed by target -> owner */
|
|
|
|
VecLerpf(dvec, ct->matrix[3], cob->matrix[3], sfac);
|
|
|
|
|
|
|
|
/* copy new vector onto owner */
|
|
|
|
VECCOPY(cob->matrix[3], dvec);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_DISTLIMIT = {
|
|
|
|
CONSTRAINT_TYPE_DISTLIMIT, /* type */
|
|
|
|
sizeof(bDistLimitConstraint), /* size */
|
|
|
|
"Limit Distance", /* name */
|
|
|
|
"bDistLimitConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
distlimit_new_data, /* new data */
|
|
|
|
distlimit_get_tars, /* get constraint targets */
|
|
|
|
distlimit_flush_tars, /* flush constraint targets */
|
|
|
|
default_get_tarmat, /* get a target matrix */
|
|
|
|
distlimit_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* ---------- Stretch To ------------ */
|
|
|
|
|
|
|
|
static void stretchto_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bStretchToConstraint *data= (bStretchToConstraint *)cdata;
|
|
|
|
|
|
|
|
data->volmode = 0;
|
|
|
|
data->plane = 0;
|
|
|
|
data->orglength = 0.0;
|
|
|
|
data->bulge = 1.0;
|
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int stretchto_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bStretchToConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* standard target-getting macro for single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void stretchto_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bStretchToConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void stretchto_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bStretchToConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= targets->first;
|
|
|
|
|
|
|
|
/* only evaluate if there is a target */
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
float size[3], scale[3], vec[3], xx[3], zz[3], orth[3];
|
|
|
|
float totmat[3][3];
|
|
|
|
float tmat[4][4];
|
|
|
|
float dist;
|
|
|
|
|
|
|
|
/* store scaling before destroying obmat */
|
|
|
|
Mat4ToSize(cob->matrix, size);
|
|
|
|
|
|
|
|
/* store X orientation before destroying obmat */
|
|
|
|
xx[0] = cob->matrix[0][0];
|
|
|
|
xx[1] = cob->matrix[0][1];
|
|
|
|
xx[2] = cob->matrix[0][2];
|
|
|
|
Normalize(xx);
|
|
|
|
|
|
|
|
/* store Z orientation before destroying obmat */
|
|
|
|
zz[0] = cob->matrix[2][0];
|
|
|
|
zz[1] = cob->matrix[2][1];
|
|
|
|
zz[2] = cob->matrix[2][2];
|
|
|
|
Normalize(zz);
|
|
|
|
|
|
|
|
VecSubf(vec, cob->matrix[3], ct->matrix[3]);
|
|
|
|
vec[0] /= size[0];
|
|
|
|
vec[1] /= size[1];
|
|
|
|
vec[2] /= size[2];
|
|
|
|
|
|
|
|
dist = Normalize(vec);
|
|
|
|
//dist = VecLenf( ob->obmat[3], targetmat[3]);
|
|
|
|
|
|
|
|
/* data->orglength==0 occurs on first run, and after 'R' button is clicked */
|
|
|
|
if (data->orglength == 0)
|
|
|
|
data->orglength = dist;
|
|
|
|
if (data->bulge == 0)
|
|
|
|
data->bulge = 1.0;
|
|
|
|
|
|
|
|
scale[1] = dist/data->orglength;
|
|
|
|
switch (data->volmode) {
|
|
|
|
/* volume preserving scaling */
|
|
|
|
case VOLUME_XZ :
|
|
|
|
scale[0] = 1.0f - (float)sqrt(data->bulge) + (float)sqrt(data->bulge*(data->orglength/dist));
|
|
|
|
scale[2] = scale[0];
|
|
|
|
break;
|
|
|
|
case VOLUME_X:
|
|
|
|
scale[0] = 1.0f + data->bulge * (data->orglength /dist - 1);
|
|
|
|
scale[2] = 1.0;
|
|
|
|
break;
|
|
|
|
case VOLUME_Z:
|
|
|
|
scale[0] = 1.0;
|
|
|
|
scale[2] = 1.0f + data->bulge * (data->orglength /dist - 1);
|
|
|
|
break;
|
|
|
|
/* don't care for volume */
|
|
|
|
case NO_VOLUME:
|
|
|
|
scale[0] = 1.0;
|
|
|
|
scale[2] = 1.0;
|
|
|
|
break;
|
|
|
|
default: /* should not happen, but in case*/
|
|
|
|
return;
|
|
|
|
} /* switch (data->volmode) */
|
|
|
|
|
|
|
|
/* Clear the object's rotation and scale */
|
|
|
|
cob->matrix[0][0]=size[0]*scale[0];
|
|
|
|
cob->matrix[0][1]=0;
|
|
|
|
cob->matrix[0][2]=0;
|
|
|
|
cob->matrix[1][0]=0;
|
|
|
|
cob->matrix[1][1]=size[1]*scale[1];
|
|
|
|
cob->matrix[1][2]=0;
|
|
|
|
cob->matrix[2][0]=0;
|
|
|
|
cob->matrix[2][1]=0;
|
|
|
|
cob->matrix[2][2]=size[2]*scale[2];
|
|
|
|
|
|
|
|
VecSubf(vec, cob->matrix[3], ct->matrix[3]);
|
|
|
|
Normalize(vec);
|
|
|
|
|
|
|
|
/* new Y aligns object target connection*/
|
|
|
|
totmat[1][0] = -vec[0];
|
|
|
|
totmat[1][1] = -vec[1];
|
|
|
|
totmat[1][2] = -vec[2];
|
|
|
|
switch (data->plane) {
|
|
|
|
case PLANE_X:
|
|
|
|
/* build new Z vector */
|
|
|
|
/* othogonal to "new Y" "old X! plane */
|
|
|
|
Crossf(orth, vec, xx);
|
|
|
|
Normalize(orth);
|
|
|
|
|
|
|
|
/* new Z*/
|
|
|
|
totmat[2][0] = orth[0];
|
|
|
|
totmat[2][1] = orth[1];
|
|
|
|
totmat[2][2] = orth[2];
|
|
|
|
|
|
|
|
/* we decided to keep X plane*/
|
|
|
|
Crossf(xx, orth, vec);
|
|
|
|
Normalize(xx);
|
|
|
|
totmat[0][0] = xx[0];
|
|
|
|
totmat[0][1] = xx[1];
|
|
|
|
totmat[0][2] = xx[2];
|
|
|
|
break;
|
|
|
|
case PLANE_Z:
|
|
|
|
/* build new X vector */
|
|
|
|
/* othogonal to "new Y" "old Z! plane */
|
|
|
|
Crossf(orth, vec, zz);
|
|
|
|
Normalize(orth);
|
|
|
|
|
|
|
|
/* new X */
|
|
|
|
totmat[0][0] = -orth[0];
|
|
|
|
totmat[0][1] = -orth[1];
|
|
|
|
totmat[0][2] = -orth[2];
|
|
|
|
|
|
|
|
/* we decided to keep Z */
|
|
|
|
Crossf(zz, orth, vec);
|
|
|
|
Normalize(zz);
|
|
|
|
totmat[2][0] = zz[0];
|
|
|
|
totmat[2][1] = zz[1];
|
|
|
|
totmat[2][2] = zz[2];
|
|
|
|
break;
|
|
|
|
} /* switch (data->plane) */
|
|
|
|
|
|
|
|
Mat4CpyMat4(tmat, cob->matrix);
|
|
|
|
Mat4MulMat34(cob->matrix, totmat, tmat);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_STRETCHTO = {
|
|
|
|
CONSTRAINT_TYPE_STRETCHTO, /* type */
|
|
|
|
sizeof(bStretchToConstraint), /* size */
|
|
|
|
"Stretch To", /* name */
|
|
|
|
"bStretchToConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
stretchto_new_data, /* new data */
|
|
|
|
stretchto_get_tars, /* get constraint targets */
|
|
|
|
stretchto_flush_tars, /* flush constraint targets */
|
|
|
|
default_get_tarmat, /* get target matrix */
|
|
|
|
stretchto_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* ---------- Floor ------------ */
|
|
|
|
|
|
|
|
static void minmax_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bMinMaxConstraint *data= (bMinMaxConstraint *)cdata;
|
|
|
|
|
|
|
|
data->minmaxflag = TRACK_Z;
|
|
|
|
data->offset = 0.0f;
|
|
|
|
data->cache[0] = data->cache[1] = data->cache[2] = 0.0f;
|
|
|
|
data->flag = 0;
|
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int minmax_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bMinMaxConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* standard target-getting macro for single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void minmax_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bMinMaxConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void minmax_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bMinMaxConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= targets->first;
|
|
|
|
|
|
|
|
/* only evaluate if there is a target */
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
float obmat[4][4], imat[4][4], tarmat[4][4], tmat[4][4];
|
|
|
|
float val1, val2;
|
|
|
|
int index;
|
|
|
|
|
|
|
|
Mat4CpyMat4(obmat, cob->matrix);
|
|
|
|
Mat4CpyMat4(tarmat, ct->matrix);
|
|
|
|
|
|
|
|
if (data->flag & MINMAX_USEROT) {
|
|
|
|
/* take rotation of target into account by doing the transaction in target's localspace */
|
|
|
|
Mat4Invert(imat, tarmat);
|
|
|
|
Mat4MulMat4(tmat, obmat, imat);
|
|
|
|
Mat4CpyMat4(obmat, tmat);
|
|
|
|
Mat4One(tarmat);
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (data->minmaxflag) {
|
|
|
|
case TRACK_Z:
|
|
|
|
val1 = tarmat[3][2];
|
|
|
|
val2 = obmat[3][2]-data->offset;
|
|
|
|
index = 2;
|
|
|
|
break;
|
|
|
|
case TRACK_Y:
|
|
|
|
val1 = tarmat[3][1];
|
|
|
|
val2 = obmat[3][1]-data->offset;
|
|
|
|
index = 1;
|
|
|
|
break;
|
|
|
|
case TRACK_X:
|
|
|
|
val1 = tarmat[3][0];
|
|
|
|
val2 = obmat[3][0]-data->offset;
|
|
|
|
index = 0;
|
|
|
|
break;
|
|
|
|
case TRACK_nZ:
|
|
|
|
val2 = tarmat[3][2];
|
|
|
|
val1 = obmat[3][2]-data->offset;
|
|
|
|
index = 2;
|
|
|
|
break;
|
|
|
|
case TRACK_nY:
|
|
|
|
val2 = tarmat[3][1];
|
|
|
|
val1 = obmat[3][1]-data->offset;
|
|
|
|
index = 1;
|
|
|
|
break;
|
|
|
|
case TRACK_nX:
|
|
|
|
val2 = tarmat[3][0];
|
|
|
|
val1 = obmat[3][0]-data->offset;
|
|
|
|
index = 0;
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (val1 > val2) {
|
|
|
|
obmat[3][index] = tarmat[3][index] + data->offset;
|
|
|
|
if (data->flag & MINMAX_STICKY) {
|
|
|
|
if (data->flag & MINMAX_STUCK) {
|
|
|
|
VECCOPY(obmat[3], data->cache);
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
VECCOPY(data->cache, obmat[3]);
|
|
|
|
data->flag |= MINMAX_STUCK;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (data->flag & MINMAX_USEROT) {
|
|
|
|
/* get out of localspace */
|
|
|
|
Mat4MulMat4(tmat, obmat, ct->matrix);
|
|
|
|
Mat4CpyMat4(cob->matrix, tmat);
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
VECCOPY(cob->matrix[3], obmat[3]);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
data->flag &= ~MINMAX_STUCK;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_MINMAX = {
|
|
|
|
CONSTRAINT_TYPE_MINMAX, /* type */
|
|
|
|
sizeof(bMinMaxConstraint), /* size */
|
|
|
|
"Floor", /* name */
|
|
|
|
"bMinMaxConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
minmax_new_data, /* new data */
|
|
|
|
minmax_get_tars, /* get constraint targets */
|
|
|
|
minmax_flush_tars, /* flush constraint targets */
|
|
|
|
default_get_tarmat, /* get target matrix */
|
|
|
|
minmax_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* ------- RigidBody Joint ---------- */
|
|
|
|
|
|
|
|
static void rbj_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bRigidBodyJointConstraint *data= (bRigidBodyJointConstraint *)cdata;
|
|
|
|
|
|
|
|
// removed code which set target of this constraint
|
|
|
|
data->type=1;
|
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int rbj_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bRigidBodyJointConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* standard target-getting macro for single-target constraints without subtargets */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGETNS_GET_TARS(con, data->tar, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void rbj_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bRigidBodyJointConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGETNS_FLUSH_TARS(con, data->tar, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_RIGIDBODYJOINT = {
|
|
|
|
CONSTRAINT_TYPE_RIGIDBODYJOINT, /* type */
|
|
|
|
sizeof(bRigidBodyJointConstraint), /* size */
|
|
|
|
"RigidBody Joint", /* name */
|
|
|
|
"bRigidBodyJointConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
rbj_new_data, /* new data */
|
|
|
|
rbj_get_tars, /* get constraint targets */
|
|
|
|
rbj_flush_tars, /* flush constraint targets */
|
|
|
|
default_get_tarmat, /* get target matrix */
|
|
|
|
NULL /* evaluate - this is not solved here... is just an interface for game-engine */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* -------- Clamp To ---------- */
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int clampto_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bClampToConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* standard target-getting macro for single-target constraints without subtargets */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGETNS_GET_TARS(con, data->tar, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void clampto_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bClampToConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGETNS_FLUSH_TARS(con, data->tar, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void clampto_get_tarmat (bConstraint *con, bConstraintOb *cob, bConstraintTarget *ct, float ctime)
|
|
|
|
{
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
Curve *cu= ct->tar->data;
|
|
|
|
|
|
|
|
/* note: when creating constraints that follow path, the curve gets the CU_PATH set now,
|
|
|
|
* currently for paths to work it needs to go through the bevlist/displist system (ton)
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* only happens on reload file, but violates depsgraph still... fix! */
|
|
|
|
if (cu->path==NULL || cu->path->data==NULL)
|
|
|
|
makeDispListCurveTypes(ct->tar, 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* technically, this isn't really needed for evaluation, but we don't know what else
|
|
|
|
* might end up calling this...
|
|
|
|
*/
|
|
|
|
if (ct)
|
|
|
|
Mat4One(ct->matrix);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void clampto_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bClampToConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= targets->first;
|
|
|
|
|
|
|
|
/* only evaluate if there is a target and it is a curve */
|
|
|
|
if (VALID_CONS_TARGET(ct) && (ct->tar->type == OB_CURVE)) {
|
|
|
|
Curve *cu= data->tar->data;
|
|
|
|
float obmat[4][4], targetMatrix[4][4], ownLoc[3];
|
|
|
|
float curveMin[3], curveMax[3];
|
|
|
|
|
|
|
|
Mat4CpyMat4(obmat, cob->matrix);
|
|
|
|
Mat4One(targetMatrix);
|
|
|
|
VECCOPY(ownLoc, obmat[3]);
|
|
|
|
|
|
|
|
INIT_MINMAX(curveMin, curveMax)
|
|
|
|
minmax_object(ct->tar, curveMin, curveMax);
|
|
|
|
|
|
|
|
/* get targetmatrix */
|
|
|
|
if (cu->path && cu->path->data) {
|
|
|
|
float vec[4], dir[3], totmat[4][4];
|
|
|
|
float curvetime;
|
|
|
|
short clamp_axis;
|
|
|
|
|
|
|
|
/* find best position on curve */
|
|
|
|
/* 1. determine which axis to sample on? */
|
|
|
|
if (data->flag == CLAMPTO_AUTO) {
|
|
|
|
float size[3];
|
|
|
|
VecSubf(size, curveMax, curveMin);
|
|
|
|
|
|
|
|
/* find axis along which the bounding box has the greatest
|
|
|
|
* extent. Otherwise, default to the x-axis, as that is quite
|
|
|
|
* frequently used.
|
|
|
|
*/
|
|
|
|
if ((size[2]>size[0]) && (size[2]>size[1]))
|
|
|
|
clamp_axis= CLAMPTO_Z - 1;
|
|
|
|
else if ((size[1]>size[0]) && (size[1]>size[2]))
|
|
|
|
clamp_axis= CLAMPTO_Y - 1;
|
|
|
|
else
|
|
|
|
clamp_axis = CLAMPTO_X - 1;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
clamp_axis= data->flag - 1;
|
|
|
|
|
|
|
|
/* 2. determine position relative to curve on a 0-1 scale based on bounding box */
|
|
|
|
if (data->flag2 & CLAMPTO_CYCLIC) {
|
|
|
|
/* cyclic, so offset within relative bounding box is used */
|
|
|
|
float len= (curveMax[clamp_axis] - curveMin[clamp_axis]);
|
|
|
|
float offset;
|
|
|
|
|
|
|
|
/* find bounding-box range where target is located */
|
|
|
|
if (ownLoc[clamp_axis] < curveMin[clamp_axis]) {
|
|
|
|
/* bounding-box range is before */
|
|
|
|
offset= curveMin[clamp_axis];
|
|
|
|
|
|
|
|
while (ownLoc[clamp_axis] < offset)
|
|
|
|
offset -= len;
|
|
|
|
|
|
|
|
/* now, we calculate as per normal, except using offset instead of curveMin[clamp_axis] */
|
|
|
|
curvetime = (ownLoc[clamp_axis] - offset) / (len);
|
|
|
|
}
|
|
|
|
else if (ownLoc[clamp_axis] > curveMax[clamp_axis]) {
|
|
|
|
/* bounding-box range is after */
|
|
|
|
offset= curveMax[clamp_axis];
|
|
|
|
|
|
|
|
while (ownLoc[clamp_axis] > offset) {
|
|
|
|
if ((offset + len) > ownLoc[clamp_axis])
|
|
|
|
break;
|
|
|
|
else
|
|
|
|
offset += len;
|
2003-10-21 13:22:07 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
|
|
|
/* now, we calculate as per normal, except using offset instead of curveMax[clamp_axis] */
|
|
|
|
curvetime = (ownLoc[clamp_axis] - offset) / (len);
|
2003-10-21 13:22:07 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
else {
|
|
|
|
/* as the location falls within bounds, just calculate */
|
|
|
|
curvetime = (ownLoc[clamp_axis] - curveMin[clamp_axis]) / (len);
|
2003-10-21 13:22:07 +00:00
|
|
|
}
|
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
else {
|
|
|
|
/* no cyclic, so position is clamped to within the bounding box */
|
|
|
|
if (ownLoc[clamp_axis] <= curveMin[clamp_axis])
|
|
|
|
curvetime = 0.0;
|
|
|
|
else if (ownLoc[clamp_axis] >= curveMax[clamp_axis])
|
|
|
|
curvetime = 1.0;
|
|
|
|
else
|
|
|
|
curvetime = (ownLoc[clamp_axis] - curveMin[clamp_axis]) / (curveMax[clamp_axis] - curveMin[clamp_axis]);
|
|
|
|
}
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* 3. position on curve */
|
|
|
|
if (where_on_path(ct->tar, curvetime, vec, dir) ) {
|
|
|
|
Mat4One(totmat);
|
|
|
|
VECCOPY(totmat[3], vec);
|
2007-01-14 23:42:06 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
Mat4MulSerie(targetMatrix, ct->tar->obmat, totmat, NULL, NULL, NULL, NULL, NULL, NULL);
|
2003-10-21 13:22:07 +00:00
|
|
|
}
|
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
|
|
|
/* obtain final object position */
|
|
|
|
VECCOPY(cob->matrix[3], targetMatrix[3]);
|
|
|
|
}
|
|
|
|
}
|
2006-08-13 14:18:15 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
static bConstraintTypeInfo CTI_CLAMPTO = {
|
|
|
|
CONSTRAINT_TYPE_CLAMPTO, /* type */
|
|
|
|
sizeof(bClampToConstraint), /* size */
|
|
|
|
"Clamp To", /* name */
|
|
|
|
"bClampToConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
NULL, /* new data */
|
|
|
|
clampto_get_tars, /* get constraint targets */
|
|
|
|
clampto_flush_tars, /* flush constraint targets */
|
|
|
|
clampto_get_tarmat, /* get target matrix */
|
|
|
|
clampto_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* ---------- Transform Constraint ----------- */
|
|
|
|
|
|
|
|
static void transform_new_data (void *cdata)
|
|
|
|
{
|
|
|
|
bTransformConstraint *data= (bTransformConstraint *)cdata;
|
|
|
|
|
|
|
|
data->map[0]= 0;
|
|
|
|
data->map[1]= 1;
|
|
|
|
data->map[2]= 2;
|
|
|
|
}
|
|
|
|
|
2008-03-08 02:16:37 +00:00
|
|
|
static int transform_get_tars (bConstraint *con, ListBase *list)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bTransformConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* standard target-getting macro for single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_GET_TARS(con, data->tar, data->subtarget, ct, list)
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 1;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
2008-03-08 02:16:37 +00:00
|
|
|
|
|
|
|
return 0;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void transform_flush_tars (bConstraint *con, ListBase *list, short nocopy)
|
|
|
|
{
|
|
|
|
if (con && list) {
|
|
|
|
bTransformConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= list->first;
|
|
|
|
|
|
|
|
/* the following macro is used for all standard single-target constraints */
|
2007-10-24 14:58:31 +00:00
|
|
|
SINGLETARGET_FLUSH_TARS(con, data->tar, data->subtarget, ct, list, nocopy)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void transform_evaluate (bConstraint *con, bConstraintOb *cob, ListBase *targets)
|
|
|
|
{
|
|
|
|
bTransformConstraint *data= con->data;
|
|
|
|
bConstraintTarget *ct= targets->first;
|
|
|
|
|
|
|
|
/* only evaluate if there is a target */
|
|
|
|
if (VALID_CONS_TARGET(ct)) {
|
|
|
|
float loc[3], eul[3], size[3];
|
|
|
|
float dvec[3], sval[3];
|
|
|
|
short i;
|
|
|
|
|
|
|
|
/* obtain target effect */
|
|
|
|
switch (data->from) {
|
2008-02-10 10:01:57 +00:00
|
|
|
case 2: /* scale */
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
Mat4ToSize(ct->matrix, dvec);
|
|
|
|
break;
|
2008-02-15 04:42:48 +00:00
|
|
|
case 1: /* rotation (convert to degrees first) */
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
Mat4ToEul(ct->matrix, dvec);
|
2008-02-15 04:42:48 +00:00
|
|
|
for (i=0; i<3; i++)
|
|
|
|
dvec[i] = dvec[i] / M_PI * 180;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
break;
|
|
|
|
default: /* location */
|
|
|
|
VecCopyf(dvec, ct->matrix[3]);
|
|
|
|
break;
|
2006-08-13 14:18:15 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
|
|
|
/* extract components of owner's matrix */
|
|
|
|
VECCOPY(loc, cob->matrix[3]);
|
|
|
|
Mat4ToEul(cob->matrix, eul);
|
2008-02-10 10:01:57 +00:00
|
|
|
Mat4ToSize(cob->matrix, size);
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
|
|
|
/* determine where in range current transforms lie */
|
|
|
|
if (data->expo) {
|
|
|
|
for (i=0; i<3; i++) {
|
|
|
|
if (data->from_max[i] - data->from_min[i])
|
|
|
|
sval[i]= (dvec[i] - data->from_min[i]) / (data->from_max[i] - data->from_min[i]);
|
|
|
|
else
|
|
|
|
sval[i]= 0.0f;
|
2006-08-13 14:18:15 +00:00
|
|
|
}
|
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
else {
|
|
|
|
/* clamp transforms out of range */
|
|
|
|
for (i=0; i<3; i++) {
|
|
|
|
CLAMP(dvec[i], data->from_min[i], data->from_max[i]);
|
|
|
|
if (data->from_max[i] - data->from_min[i])
|
|
|
|
sval[i]= (dvec[i] - data->from_min[i]) / (data->from_max[i] - data->from_min[i]);
|
|
|
|
else
|
|
|
|
sval[i]= 0.0f;
|
2006-08-13 14:18:15 +00:00
|
|
|
}
|
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
2008-02-15 04:42:48 +00:00
|
|
|
/* convert radians<->degrees */
|
|
|
|
if (data->to == 1) {
|
|
|
|
/* if output is rotation, convert to radians from degrees */
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
for (i=0; i<3; i++)
|
|
|
|
sval[i] = sval[i] / 180 * M_PI;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* apply transforms */
|
|
|
|
switch (data->to) {
|
|
|
|
case 2: /* scaling */
|
|
|
|
for (i=0; i<3; i++)
|
|
|
|
size[i]= data->to_min[i] + (sval[data->map[i]] * (data->to_max[i] - data->to_min[i]));
|
|
|
|
break;
|
|
|
|
case 1: /* rotation */
|
|
|
|
for (i=0; i<3; i++) {
|
|
|
|
float tmin, tmax;
|
2007-04-07 03:32:57 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* convert destination min/max ranges from degrees to radians */
|
|
|
|
tmin= data->to_min[i] / M_PI * 180;
|
|
|
|
tmax= data->to_max[i] / M_PI * 180;
|
2007-09-26 07:33:31 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
eul[i]= tmin + (sval[data->map[i]] * (tmax - tmin));
|
2007-09-26 07:33:31 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
break;
|
|
|
|
default: /* location */
|
|
|
|
/* get new location */
|
|
|
|
for (i=0; i<3; i++)
|
|
|
|
loc[i]= (data->to_min[i] + (sval[data->map[i]] * (data->to_max[i] - data->to_min[i])));
|
2007-04-07 03:32:57 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* add original location back on (so that it can still be moved) */
|
|
|
|
VecAddf(loc, cob->matrix[3], loc);
|
|
|
|
break;
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
|
|
|
/* apply to matrix */
|
|
|
|
LocEulSizeToMat4(cob->matrix, loc, eul, size);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static bConstraintTypeInfo CTI_TRANSFORM = {
|
|
|
|
CONSTRAINT_TYPE_TRANSFORM, /* type */
|
|
|
|
sizeof(bTransformConstraint), /* size */
|
|
|
|
"Transform", /* name */
|
|
|
|
"bTransformConstraint", /* struct name */
|
|
|
|
NULL, /* free data */
|
|
|
|
NULL, /* relink data */
|
|
|
|
NULL, /* copy data */
|
|
|
|
transform_new_data, /* new data */
|
|
|
|
transform_get_tars, /* get constraint targets */
|
|
|
|
transform_flush_tars, /* flush constraint targets */
|
|
|
|
default_get_tarmat, /* get a target matrix */
|
|
|
|
transform_evaluate /* evaluate */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* ************************* Constraints Type-Info *************************** */
|
|
|
|
/* All of the constraints api functions use bConstraintTypeInfo structs to carry out
|
|
|
|
* and operations that involve constraint specifc code.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* These globals only ever get directly accessed in this file */
|
2007-11-14 21:19:49 +00:00
|
|
|
static bConstraintTypeInfo *constraintsTypeInfo[NUM_CONSTRAINT_TYPES+1];
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
static short CTI_INIT= 1; /* when non-zero, the list needs to be updated */
|
|
|
|
|
|
|
|
/* This function only gets called when CTI_INIT is non-zero */
|
|
|
|
static void constraints_init_typeinfo () {
|
|
|
|
constraintsTypeInfo[0]= NULL; /* 'Null' Constraint */
|
|
|
|
constraintsTypeInfo[1]= &CTI_CHILDOF; /* ChildOf Constraint */
|
|
|
|
constraintsTypeInfo[2]= &CTI_TRACKTO; /* TrackTo Constraint */
|
|
|
|
constraintsTypeInfo[3]= &CTI_KINEMATIC; /* IK Constraint */
|
|
|
|
constraintsTypeInfo[4]= &CTI_FOLLOWPATH; /* Follow-Path Constraint */
|
|
|
|
constraintsTypeInfo[5]= &CTI_ROTLIMIT; /* Limit Rotation Constraint */
|
|
|
|
constraintsTypeInfo[6]= &CTI_LOCLIMIT; /* Limit Location Constraint */
|
|
|
|
constraintsTypeInfo[7]= &CTI_SIZELIMIT; /* Limit Scaling Constraint */
|
|
|
|
constraintsTypeInfo[8]= &CTI_ROTLIKE; /* Copy Rotation Constraint */
|
|
|
|
constraintsTypeInfo[9]= &CTI_LOCLIKE; /* Copy Location Constraint */
|
|
|
|
constraintsTypeInfo[10]= &CTI_SIZELIKE; /* Copy Scaling Constraint */
|
|
|
|
constraintsTypeInfo[11]= &CTI_PYTHON; /* Python/Script Constraint */
|
|
|
|
constraintsTypeInfo[12]= &CTI_ACTION; /* Action Constraint */
|
|
|
|
constraintsTypeInfo[13]= &CTI_LOCKTRACK; /* Locked-Track Constraint */
|
2008-01-25 11:23:36 +00:00
|
|
|
constraintsTypeInfo[14]= &CTI_DISTLIMIT; /* Limit Distance Constraint */
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
constraintsTypeInfo[15]= &CTI_STRETCHTO; /* StretchTo Constaint */
|
|
|
|
constraintsTypeInfo[16]= &CTI_MINMAX; /* Floor Constraint */
|
|
|
|
constraintsTypeInfo[17]= &CTI_RIGIDBODYJOINT; /* RigidBody Constraint */
|
|
|
|
constraintsTypeInfo[18]= &CTI_CLAMPTO; /* ClampTo Constraint */
|
2008-01-25 11:23:36 +00:00
|
|
|
constraintsTypeInfo[19]= &CTI_TRANSFORM; /* Transformation Constraint */
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/* This function should be used for getting the appropriate type-info when only
|
|
|
|
* a constraint type is known
|
|
|
|
*/
|
|
|
|
bConstraintTypeInfo *get_constraint_typeinfo (int type)
|
|
|
|
{
|
|
|
|
/* initialise the type-info list? */
|
|
|
|
if (CTI_INIT) {
|
|
|
|
constraints_init_typeinfo();
|
|
|
|
CTI_INIT = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* only return for valid types */
|
|
|
|
if ( (type >= CONSTRAINT_TYPE_NULL) &&
|
|
|
|
(type <= NUM_CONSTRAINT_TYPES ) )
|
|
|
|
{
|
|
|
|
/* there shouldn't be any segfaults here... */
|
|
|
|
return constraintsTypeInfo[type];
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
printf("No valid constraint type-info data available. Type = %i \n", type);
|
|
|
|
}
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This function should always be used to get the appropriate type-info, as it
|
|
|
|
* has checks which prevent segfaults in some weird cases.
|
|
|
|
*/
|
|
|
|
bConstraintTypeInfo *constraint_get_typeinfo (bConstraint *con)
|
|
|
|
{
|
|
|
|
/* only return typeinfo for valid constraints */
|
|
|
|
if (con)
|
|
|
|
return get_constraint_typeinfo(con->type);
|
|
|
|
else
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* ************************* General Constraints API ************************** */
|
|
|
|
/* The functions here are called by various parts of Blender. Very few (should be none if possible)
|
|
|
|
* constraint-specific code should occur here.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* ---------- Data Management ------- */
|
|
|
|
|
|
|
|
/* Free data of a specific constraint if it has any info */
|
|
|
|
void free_constraint_data (bConstraint *con)
|
|
|
|
{
|
|
|
|
if (con->data) {
|
|
|
|
bConstraintTypeInfo *cti= constraint_get_typeinfo(con);
|
|
|
|
|
|
|
|
/* perform any special freeing constraint may have */
|
|
|
|
if (cti && cti->free_data)
|
|
|
|
cti->free_data(con);
|
|
|
|
|
|
|
|
/* free constraint data now */
|
|
|
|
MEM_freeN(con->data);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Free all constraints from a constraint-stack */
|
|
|
|
void free_constraints (ListBase *conlist)
|
|
|
|
{
|
|
|
|
bConstraint *con;
|
|
|
|
|
|
|
|
/* Free constraint data and also any extra data */
|
|
|
|
for (con= conlist->first; con; con= con->next) {
|
|
|
|
free_constraint_data(con);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Free the whole list */
|
|
|
|
BLI_freelistN(conlist);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Reassign links that constraints have to other data (called during file loading?) */
|
|
|
|
void relink_constraints (ListBase *conlist)
|
|
|
|
{
|
|
|
|
bConstraint *con;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
for (con= conlist->first; con; con= con->next) {
|
|
|
|
bConstraintTypeInfo *cti= constraint_get_typeinfo(con);
|
|
|
|
|
|
|
|
if (cti) {
|
|
|
|
/* relink any targets */
|
|
|
|
if (cti->get_constraint_targets) {
|
|
|
|
ListBase targets = {NULL, NULL};
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
cti->get_constraint_targets(con, &targets);
|
|
|
|
for (ct= targets.first; ct; ct= ct->next) {
|
|
|
|
ID_NEW(ct->tar);
|
|
|
|
}
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (cti->flush_constraint_targets)
|
|
|
|
cti->flush_constraint_targets(con, &targets, 0);
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
}
|
2007-07-21 07:26:15 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* relink any other special data */
|
|
|
|
if (cti->relink_data)
|
|
|
|
cti->relink_data(con);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* duplicate all of the constraints in a constraint stack */
|
|
|
|
void copy_constraints (ListBase *dst, ListBase *src)
|
|
|
|
{
|
|
|
|
bConstraint *con, *srccon;
|
|
|
|
|
|
|
|
dst->first= dst->last= NULL;
|
|
|
|
duplicatelist(dst, src);
|
|
|
|
|
|
|
|
for (con=dst->first, srccon=src->first; con; srccon=srccon->next, con=con->next) {
|
|
|
|
bConstraintTypeInfo *cti= constraint_get_typeinfo(con);
|
|
|
|
|
|
|
|
/* make a new copy of the constraint's data */
|
|
|
|
con->data = MEM_dupallocN(con->data);
|
|
|
|
|
|
|
|
/* only do specific constraints if required */
|
|
|
|
if (cti && cti->copy_data)
|
|
|
|
cti->copy_data(con, srccon);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-01-04 11:21:50 +00:00
|
|
|
/* -------- Constraints and Proxies ------- */
|
|
|
|
|
|
|
|
/* Rescue all constraints tagged as being CONSTRAINT_PROXY_LOCAL (i.e. added to bone that's proxy-synced in this file) */
|
|
|
|
void extract_proxylocal_constraints (ListBase *dst, ListBase *src)
|
|
|
|
{
|
|
|
|
bConstraint *con, *next;
|
|
|
|
|
|
|
|
/* for each tagged constraint, remove from src and move to dst */
|
|
|
|
for (con= src->first; con; con= next) {
|
|
|
|
next= con->next;
|
|
|
|
|
|
|
|
/* check if tagged */
|
|
|
|
if (con->flag & CONSTRAINT_PROXY_LOCAL) {
|
|
|
|
BLI_remlink(src, con);
|
|
|
|
BLI_addtail(dst, con);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Returns if the owner of the constraint is proxy-protected */
|
|
|
|
short proxylocked_constraints_owner (Object *ob, bPoseChannel *pchan)
|
|
|
|
{
|
|
|
|
/* Currently, constraints can only be on object or bone level */
|
|
|
|
if (ob && ob->proxy) {
|
|
|
|
if (ob->pose && pchan) {
|
|
|
|
bArmature *arm= ob->data;
|
|
|
|
|
|
|
|
/* On bone-level, check if bone is on proxy-protected layer */
|
|
|
|
if ((pchan->bone) && (pchan->bone->layer & arm->layer_protected))
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* FIXME: constraints on object-level are not handled well yet */
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* -------- Target-Matrix Stuff ------- */
|
|
|
|
|
|
|
|
/* This function is a relic from the prior implementations of the constraints system, when all
|
|
|
|
* constraints either had one or no targets. It used to be called during the main constraint solving
|
|
|
|
* loop, but is now only used for the remaining cases for a few constraints.
|
|
|
|
*
|
|
|
|
* None of the actual calculations of the matricies should be done here! Also, this function is
|
|
|
|
* not to be used by any new constraints, particularly any that have multiple targets.
|
|
|
|
*/
|
2007-10-24 14:58:31 +00:00
|
|
|
void get_constraint_target_matrix (bConstraint *con, int n, short ownertype, void *ownerdata, float mat[][4], float ctime)
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
{
|
|
|
|
bConstraintTypeInfo *cti= constraint_get_typeinfo(con);
|
|
|
|
ListBase targets = {NULL, NULL};
|
|
|
|
bConstraintOb *cob;
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
if (cti && cti->get_constraint_targets) {
|
|
|
|
/* make 'constraint-ob' */
|
|
|
|
cob= MEM_callocN(sizeof(bConstraintOb), "tempConstraintOb");
|
|
|
|
cob->type= ownertype;
|
|
|
|
switch (ownertype) {
|
|
|
|
case CONSTRAINT_OBTYPE_OBJECT: /* it is usually this case */
|
|
|
|
{
|
|
|
|
cob->ob= (Object *)ownerdata;
|
|
|
|
cob->pchan= NULL;
|
|
|
|
if (cob->ob) {
|
|
|
|
Mat4CpyMat4(cob->matrix, cob->ob->obmat);
|
|
|
|
Mat4CpyMat4(cob->startmat, cob->matrix);
|
2007-07-21 07:26:15 +00:00
|
|
|
}
|
|
|
|
else {
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
Mat4One(cob->matrix);
|
|
|
|
Mat4One(cob->startmat);
|
2007-07-21 07:26:15 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
}
|
|
|
|
break;
|
|
|
|
case CONSTRAINT_OBTYPE_BONE: /* this may occur in some cases */
|
|
|
|
{
|
|
|
|
cob->ob= NULL; /* this might not work at all :/ */
|
|
|
|
cob->pchan= (bPoseChannel *)ownerdata;
|
|
|
|
if (cob->pchan) {
|
|
|
|
Mat4CpyMat4(cob->matrix, cob->pchan->pose_mat);
|
|
|
|
Mat4CpyMat4(cob->startmat, cob->matrix);
|
2007-07-21 07:26:15 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
else {
|
|
|
|
Mat4One(cob->matrix);
|
|
|
|
Mat4One(cob->startmat);
|
2007-07-21 07:26:15 +00:00
|
|
|
}
|
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
break;
|
2007-07-21 07:26:15 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
|
|
|
/* get targets - we only need the first one though (and there should only be one) */
|
|
|
|
cti->get_constraint_targets(con, &targets);
|
|
|
|
|
|
|
|
/* only calculate the target matrix on the first target */
|
|
|
|
ct= (bConstraintTarget *)targets.first;
|
2007-10-24 14:58:31 +00:00
|
|
|
while(ct && n-- > 0)
|
|
|
|
ct= ct->next;
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
if (ct) {
|
|
|
|
if (cti->get_target_matrix)
|
|
|
|
cti->get_target_matrix(con, cob, ct, ctime);
|
|
|
|
Mat4CpyMat4(mat, ct->matrix);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* free targets + 'constraint-ob' */
|
|
|
|
if (cti->flush_constraint_targets)
|
|
|
|
cti->flush_constraint_targets(con, &targets, 1);
|
|
|
|
MEM_freeN(cob);
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
/* invalid constraint - perhaps... */
|
|
|
|
Mat4One(mat);
|
2002-10-12 11:37:38 +00:00
|
|
|
}
|
2005-09-12 04:46:34 +00:00
|
|
|
}
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
|
|
|
|
/* ---------- Evaluation ----------- */
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* This function is called whenever constraints need to be evaluated. Currently, all
|
|
|
|
* constraints that can be evaluated are everytime this gets run.
|
|
|
|
*
|
|
|
|
* constraints_make_evalob and constraints_clear_evalob should be called before and
|
|
|
|
* after running this function, to sort out cob
|
|
|
|
*/
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
void solve_constraints (ListBase *conlist, bConstraintOb *cob, float ctime)
|
|
|
|
{
|
|
|
|
bConstraint *con;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
float solution[4][4], delta[4][4];
|
|
|
|
float oldmat[4][4], imat[4][4];
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
float enf;
|
|
|
|
|
|
|
|
/* check that there is a valid constraint object to evaluate */
|
|
|
|
if (cob == NULL)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* loop over available constraints, solving and blending them */
|
|
|
|
for (con= conlist->first; con; con= con->next) {
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
bConstraintTypeInfo *cti= constraint_get_typeinfo(con);
|
|
|
|
ListBase targets = {NULL, NULL};
|
|
|
|
|
|
|
|
/* these we can skip completely (invalid constraints...) */
|
|
|
|
if (cti == NULL) continue;
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
if (con->flag & CONSTRAINT_DISABLE) continue;
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* these constraints can't be evaluated anyway */
|
|
|
|
if (cti->evaluate_constraint == NULL) continue;
|
2007-07-23 11:28:43 +00:00
|
|
|
/* influence == 0 should be ignored */
|
|
|
|
if (con->enforce == 0.0f) continue;
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* influence of constraint
|
|
|
|
* - value should have been set from IPO's/Constraint Channels already
|
|
|
|
*/
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
enf = con->enforce;
|
|
|
|
|
2007-08-18 06:17:50 +00:00
|
|
|
/* move owner matrix into right space */
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
constraint_mat_convertspace(cob->ob, cob->pchan, cob->matrix, CONSTRAINT_SPACE_WORLD, con->ownspace);
|
2007-08-18 06:17:50 +00:00
|
|
|
Mat4CpyMat4(oldmat, cob->matrix);
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* prepare targets for constraint solving */
|
|
|
|
if (cti->get_constraint_targets) {
|
|
|
|
bConstraintTarget *ct;
|
|
|
|
|
|
|
|
/* get targets
|
|
|
|
* - constraints should use ct->matrix, not directly accessing values
|
|
|
|
* - ct->matrix members have not yet been calculated here!
|
|
|
|
*/
|
|
|
|
cti->get_constraint_targets(con, &targets);
|
2007-08-18 06:17:50 +00:00
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* set matrices
|
|
|
|
* - calculate if possible, otherwise just initialise as identity matrix
|
|
|
|
*/
|
|
|
|
if (cti->get_target_matrix) {
|
|
|
|
for (ct= targets.first; ct; ct= ct->next)
|
|
|
|
cti->get_target_matrix(con, cob, ct, ctime);
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
for (ct= targets.first; ct; ct= ct->next)
|
|
|
|
Mat4One(ct->matrix);
|
|
|
|
}
|
2007-08-18 06:17:50 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Solve the constraint */
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
cti->evaluate_constraint(con, cob, &targets);
|
|
|
|
|
|
|
|
/* clear targets after use
|
|
|
|
* - this should free temp targets but no data should be copied back
|
|
|
|
* as constraints may have done some nasty things to it...
|
|
|
|
*/
|
|
|
|
if (cti->flush_constraint_targets) {
|
|
|
|
cti->flush_constraint_targets(con, &targets, 1);
|
|
|
|
}
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
|
|
|
|
/* Interpolate the enforcement, to blend result of constraint into final owner transform */
|
|
|
|
/* 1. Remove effects of original matrix from constraint solution ==> delta */
|
|
|
|
Mat4Invert(imat, oldmat);
|
|
|
|
Mat4CpyMat4(solution, cob->matrix);
|
|
|
|
Mat4MulMat4(delta, solution, imat);
|
|
|
|
|
|
|
|
/* 2. If constraint influence is not full strength, then interpolate
|
|
|
|
* identity_matrix --> delta_matrix to get the effect the constraint actually exerts
|
|
|
|
*/
|
|
|
|
if (enf < 1.0) {
|
|
|
|
float identity[4][4];
|
|
|
|
Mat4One(identity);
|
|
|
|
Mat4BlendMat4(delta, identity, delta, enf);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* 3. Now multiply the delta by the matrix in use before the evaluation */
|
|
|
|
Mat4MulMat4(cob->matrix, delta, oldmat);
|
|
|
|
|
== Constraints System - Recode 2 ==
Once again, I've recoded the constraints system. This time, the goals were:
* To make it more future-proof by 'modernising' the coding style. The long functions filled with switch statements, have given way to function-pointers with smaller functions for specific purposes.
* To make it support constraints which use multiple targets more readily that it did. In the past, it was assumed that constraints could only have at most one target.
As a result, a lot of code has been shuffled around, and modified. Also, the subversion number has been bumped up.
Known issues:
* PyConstraints, which were the main motivation for supporting multiple-targets, are currently broken. There are some bimport() error that keeps causing problems. I've also temporarily removed the doDriver support, although it may return in another form soon.
* Constraints BPy-API is currently has a few features which currently don't work yet
* Outliner currently only displays the names of the constraints instead of the fancy subtarget/target/constraint-name display it used to do. What gets displayed here needs further investigation, as the old way was certainly not that great (and is not compatible with the new system too)
2007-10-21 23:00:29 +00:00
|
|
|
/* move owner back into worldspace for next constraint/other business */
|
== Constraints System ==
After just over a week of coding, I've finished doing a major refactor/cleanup of the constraints code. In the process, quite a few old kludges and ugly hacks have been removed. Also, some new features which will greatly benefit riggers have been implemented.
=== What's New ===
* The long-awaited ``ChildOf Constraint'':
This allows you to animate parent influences, and choose which transformation channels the parent affects the child on (i.e. no translation/rotation/scaling). It should be noted that disabling some combinations may not totally work as expected. Also, the 'Set Inverse' and 'Clear Inverse' buttons at the bottom of this constraint's panel set/clear the inverse correction for the parent's effects. Use these to make the owner not stick/be glued to the parent.
* Constraint/Target Evaluation Spaces:
In some constraints, there are now 1-2 combo boxes at the bottom of their panel, which allows you to pick which `co-ordinate space' they are evaluated in. This is much more flexible than the old 'local' options for bones only were.
* Action Constraint - Loc/Rot/Size Inputs
The Action Constraint can finally use the target's location/rotation/scaling transforms as input, to control the owner of the constraint. This should work much more reliably than it used to. The target evaluation should now also be more accurate due to the new space conversion stuff.
* Transform - No longer in Crazy Space (TM)
Transforming objects/bones with constraints applied should no longer occur in Crazy Space. They are now correctly inverse-corrected. This also applies to old-style object tracking.
=== General Code Changes ===
* solve_constraints is now in constraints.c. I've removed the old `blend consecutive constraints of same type' junk, which made the code more complex than it needed to be.
* evaluate_constraint is now only passed the constraint, and two matrices. A few unused variables have been removed from here.
* A tempolary struct, bConstraintOb, is now passed to solve_constraints instead of relying on an ugly, static workobject in some cases. This works much better.
* Made the formatting of constraint code consistent
* There's a version patch for older files so that constraint settings are correctly converted to the new system. This is currently done for MajorVersion <= 244, and SubVersion < 3. I've bumped up the subversion to 3 for this purpose. However, with the imminent 2.45 release, this may need to be adjusted accordingly.
* LocEulSizeToMat4 and LocQuatSizeToMat4 now work in the order Size, Rot, Location. I've also added a few other math functions.
* Mat4BlendMat4 is now in arithb. I've modified it's method slightly, to use other arithb functions, instead of its crazy blending scheme.
* Moved some of the RigidBodyJoint constraint's code out of blenkernel, and into src. It shouldn't be setting its target in its data initialisation function based + accessing scene stuff where it was doing so.
=== Future Work ===
* Geometry to act as targets for constraints. A space has been reserved for this already.
* Tidy up UI buttons of constraints
2007-07-15 03:35:37 +00:00
|
|
|
if ((con->flag & CONSTRAINT_SPACEONCE) == 0)
|
|
|
|
constraint_mat_convertspace(cob->ob, cob->pchan, cob->matrix, con->ownspace, CONSTRAINT_SPACE_WORLD);
|
|
|
|
}
|
|
|
|
}
|