this is mainly because adding pose keyframes recalculates every handle so importing became increasingly slow.
added a 'fast' argument to insertkey that python api's insertPoseKey can make use of since it alredy accepts a 'fast' option.
The ~4450 frame, 31 bone BVH imports in ~108sec now
Seperated editmode switch statement in space.c's event handling,
if editmode is disabled, or the images is a render or composite, UV editing operations are ignored.
In previous releases it has given an annoying warning if selecting or scaling is attempted when out of UV/Face mode.
This is implimented by defining macro's that take the mode into account when get/set/checking the selection and that a face can be drawn.
When this is enabled, things work a bit differently
* all faces are drawn in the UV view (except hidden faces).
* selecting faces/verts in the UV window selects them on the mesh also and vice vercer.
* when EditMode (Face only) is enabled, selecting faces in the UV view is not sticky.
* hiding/(de)select all and invert selection simle call the editmesh functions.
* there is no way to select 1 uv corner of 1 face (because a mesh cannot have this as a valid selection) either all UV's that use a vert (sticky), or the whole face.
Notes
* you cant edit UV's in the image window in "UV Face Select" mode. (removed UV from the name)
* going into Face Select mode no longer adds UV's and does not need UV's to work.
* The UV Calculation menu is now in editmode (Alt+W)
Todo..
* Image replace - partly broken in stable also.
* Rotate/Mirror UV/VCol are still only in Face Select mode.
* Hide/Reveal is not quite right, (issue with editmode flushing)
- Removing evil InitTransModeFlags (I would always forget to update that when adding new stuff). It's properly done in the per transform init now.
- Add toggleable mode for Shear (with MMB) for horizontal or vertical shear. Toggle shear mode also changes the input mode from vertical to horizontal and vice versa (it's intuitiver that way). (This was suggested in a really old patch but never implemented because I disliked adding transform specific handling in the main event switch).
- Add custom event handlers per transform (currently only used for Shear)
- Add custom data pointer and freeing flag (only (a)bused by Shear).
NOTE to Levi Schooley: Some stuff in there might be useful for you.
===============================
An improved CrazySpace correction is now used for Armature modifiers that use
vertex groups, and that are the first enabled modifiers in the stack. This is
a a specific case, but also a common one.
http://www.blender.org/development/current-projects/changes-since-244/skinning/
Implementation Notes:
- The quaternion crazyspace correction is still used for modifiers other than
the armature modifier.
- Modifiers can now provide a deform matrix per vertex to be used for
crazyspace correction, only the armature modifier implements this now.
Tracking is now tempolarily disabled while an object matrix used for space conversions stuff is calculated. This used to be done before the constraints recode.
Now PoseBones should transform normally again with constraints applied and in other cases. I've gone back to the old code that used to be there before the recode, and added the constraint inverse (as appropriate) into that.
* Reverting a previous commit where I wrongly assumed that the code was not doing things the right way. This makes a few cases work better normally again.
* Object-level Transforms should now perform normally again. Now, transforms are only get inverse-corrected for constraints if certain constraints are the first 'active' constraint.
* PoseBone-level Transforms still need to be fixed, although I might have done a few tweaks here.
Here are some attempts at fixing bugs being encountered with the constraint system after the recode. I've just fixed a few typos (where relevant), and added some extra checks for things.
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
I've moved the Auto-Keyframing functionality out of the special_aftertrans_update function into two separate functions, which can be called independent of the Transform system. One is for Bones, while the other is for Objects.
This now means that the Shift-S Snapping Tools will now work with auto-keyframing.
using mface->flag for both.
Also found that the cdDM_drawMappedFaces and cdDM_drawFacesTex_common could get normals mixed up when rendering hidden faces. because hidden/invisible faces used continue without advancing to the next normal.
Now, the 'Only Needed' Auto-Keyframing tool only adds keyframes for the relevant transforms when working with bones in PoseMode. Previously, all transform channels were keyed in such cases, as Auto-IK used to make things difficult.
removed most custom add_*data* wrappers from Main.c
removed makeCurrent() from Text.c (was never in a release), use "bpy.texts.active = text" now
clamp new image sizes
made add_empty_action accept a string rather then a blocktype since the blocktype was only being used to choose one of 3 strings anyway.
This commit upgrades the modifier stack to only calculate the data which is
needed, either by modifiers further down the stack or by other functions at
the end of the stack (e.g. drawing functions).
This speeds up modifier stack recalculation, especially where vertex
groups and UV coordinates are concerned. For example, a mesh with an Armature
modifier followed by a Subsurf modifier would previously have required the
Subsurf modifier to interpolate all the vertex groups in the mesh, slowing
down modifier calculations considerably. With this update, vertex group data
is not propagated beyond the Armature modifier, so calculations are faster.
Note that this depends on the order of modifiers in the stack. If the Armature
and Subsurf modifiers were swapped in the above example, the Subsurf modifier
would have to interpolate vertex groups, as they are needed by the Armature
modifier.
Now, when auto-keyframing, there is the option to have 'cleaner'
ipo curves. This is a user-pref option, and is based on the same
code responsible for the 'Only Needed' keyframing mode.
* Bugfix for Nicholas Bishop's recent sculptmode hotkeys:
When entering text in a text object, hotkeys were not disabled,
eg. type n = floating panel appears press 2 = layer 2 becomes visable.
Was 'if' instead of 'else if'
Reported by mfoxdogg on cvs.
Modified version of patch #5281 by Joshua Leung.
It's a two part patch:
Add a Roll button in the transform properties floating panel with bones selected in edit mode
Add a Roll transform (Ctrl-R) that can be used to modify the roll of selected bones in edit mode
The transformation modifies the roll of all selected bones incrementally (like a rotation does to the bones' rotation) [that is the part that differs from the patch. The patch would set the same roll value to all bones]
Also, this commit includes some shuffling around of the functions, to keep the previously logical order. :)
PET Connectivity calculation code had a bug that only occured on really large scale meshes which caused a hang (infinite loop).
Fixed by raising the threshold to prevent float errors.
UV edit option 'constrain to rectange' didn't restore correctly on ESC.
(note, on ESC it doesn't undo the constrainting, in case the edited face
was not a quad).
Also:
- The UVs pulldown in Image window was far too large, moved the scripts to
an own sublevel menu
- Previous commit in vpaint.c had warnings
All data layers, including MVert/MEdge/MFace, are now managed as custom
data layers. The pointers like Mesh.mvert, Mesh.dvert or Mesh.mcol are
still used of course, but allocating, copying or freeing these arrays
should be done through the CustomData API.
Work in progress documentation on this is here:
http://mediawiki.blender.org/index.php/BlenderDev/BlenderArchitecture/CustomData
Replaced TFace by MTFace:
This is the same struct, except that it does not contain color, that now
always stays separated in MCol. This was not a good design decision to
begin with, and it is needed for adding multiple color layers later. Note
that this does mean older Blender versions will not be able to read UV
coordinates from the next release, due to an SDNA limitation.
Removed DispListMesh:
This now fully replaced by DerivedMesh. To provide access to arrays of
vertices, edges and faces, like DispListMesh does. The semantics of the
DerivedMesh.getVertArray() and similar functions were changed to return
a pointer to an array if one exists, or otherwise allocate a temporary
one. On releasing the DerivedMesh, this temporary array will be removed
automatically.
Removed ssDM and meshDM DerivedMesh backends:
The ssDM backend was for DispListMesh, so that became obsolete automatically.
The meshDM backend was replaced by the custom data backend, that now figures
out which layers need to be modified, and only duplicates those.
This changes code in many places, and overall removes 2514 lines of code.
So, there's a good chance this might break some stuff, although I've been
testing it for a few days now. The good news is, adding multiple color and
uv layers should now become easy.
Small tweak in depsgraph usage of transform: to make sure recalc tags get
set in invisible objects (in other layers) as well. That way switching
layers won't give errors with child-objects not being in the correct place.
- CTRL+Click bones in editmode gave too small envelope distances (was 0.1
instead of 0.25*bone->length)
- PoseMode: ALT+S in Envelope drawmode only scaled the Parent Bone(s).
This is a much faster and easier way to give a bevelled curve a taper, without
using taper curves. Each point on a curve now has a 'radius' value that you can
shrink and fatten using Alt S, which will influence the taper when the curve is
bevelled (either with a bevob, or with front/back turned off and a bevel dept
set). Alt S shrinks and fattens the selected points in an interactive transform,
and you can set an absolute radius for selected points with 'Set Radius' in the
curve specials menu.
See demo: http://mke3.net/blender/etc/curve_shrinkfatten-h264.mov
This can be a quick way to create revolved surfaces (eg.
http://mke3.net/blender/etc/wineglass-h264.mov ) and it would be very
interesting to use this radius value in other tools, such as a 'freehand curve'
tool that would let you draw a curve freehand, with the radius affected by pen
pressure, or even using the radius at each point to control curve guides for
particles more precisely, rather than the continous maxdist.
Bug #3657
Texture Space transform
* grab/size: did not support local constraints properly
* size: constraint support did not respect object orientation
- Found several places, where people explicitly casted the frame number
to short.
- Fixed the crash in BPY_interface by adding an empty line (to make it
recompile everywhere, make clean doesn't help...)
For the build system maintainers:
Problem was: The change in makesdna changed the position of the
scriptlink structure. BPY_interface.c somehow didn't get recompiled
(not even after a make clean!!!) which triggered crashes on adding
scriptlinks.
only in Face-Select mode. It then uses for rotate and scaling the face
center itself as reference.
Code uses a loop-in-loop to find the face that belongs to the vertex...
means it will be slow with operations on 10k or more faces. Acceptable
for now, will make it nicer later. :)
- Set local sticky in the uv editor as default.
- Don't do live unwrap on fully selected charts or charts with no pins
selected.
- Fixed bug with live unwrap not respecting transform cancel in some cases.
- "View Home" didn't work without an image.
- Move UV Calculation settings (cube size, cylinder radius, ..) into the scene
toolsettings, instead of global variables
- Remove the name LSCM from the UI (and python docs on seams), and replace it
with 'Unwrap', with upcoming ABF this didn't make sense anymore.
- Move the Old/New LSCM switch into the UV Calculation panel. New LSCM is the
default now. Also renamed LSCM there to "Conformal".
- Made some room in the UV Calculation panel by removing the buttons to execute
the UV calculation, only leaving the settings.
Fill Holes:
- LSCM now has an option to fill holes in the chart before unwrapping. This on
by default, and enables two things:
- Prevent internal overlaps (e.g. eyes, mouth) for LSCM unwrapping.
- Allow the internal boundaries to move freely during stretch minimize.
- The possibility to switch it off is there because it is not always possible
to define which the outer boundary is. For example with an open cylinder
where there are two identical holes.
A full detailed description of this will be done later... is several days
of work. Here's a summary:
Render:
- Full cleanup of render code, removing *all* globals and bad level calls
all over blender. Render module is now not called abusive anymore
- API-fied calls to rendering
- Full recode of internal render pipeline. Is now rendering tiles by
default, prepared for much smarter 'bucket' render later.
- Each thread now can render a full part
- Renders were tested with 4 threads, goes fine, apart from some lookup
tables in softshadow and AO still
- Rendering is prepared to do multiple layers and passes
- No single 32 bits trick in render code anymore, all 100% floats now.
Writing images/movies
- moved writing images to blender kernel (bye bye 'schrijfplaatje'!)
- made a new Movie handle system, also in kernel. This will enable much
easier use of movies in Blender
PreviewRender:
- Using new render API, previewrender (in buttons) now uses regular render
code to generate images.
- new datafile 'preview.blend.c' has the preview scenes in it
- previews get rendered in exact displayed size (1 pixel = 1 pixel)
3D Preview render
- new; press Pkey in 3d window, for a panel that continuously renders
(pkey is for games, i know... but we dont do that in orange now!)
- this render works nearly identical to buttons-preview render, so it stops
rendering on any event (mouse, keyboard, etc)
- on moving/scaling the panel, the render code doesn't recreate all geometry
- same for shifting/panning view
- all other operations (now) regenerate the full render database still.
- this is WIP... but big fun, especially for simple scenes!
Compositor
- Using same node system as now in use for shaders, you can composit images
- works pretty straightforward... needs much more options/tools and integration
with rendering still
- is not threaded yet, nor is so smart to only recalculate changes... will be
done soon!
- the "Render Result" node will get all layers/passes as output sockets
- The "Output" node renders to a builtin image, which you can view in the Image
window. (yes, output nodes to render-result, and to files, is on the list!)
The Bad News
- "Unified Render" is removed. It might come back in some stage, but this
system should be built from scratch. I can't really understand this code...
I expect it is not much needed, especially with advanced layer/passes
control
- Panorama render, Field render, Motion blur, is not coded yet... (I had to
recode every single feature in render, so...!)
- Lens Flare is also not back... needs total revision, might become composit
effect though (using zbuffer for visibility)
- Part render is gone! (well, thats obvious, its default now).
- The render window is only restored with limited functionality... I am going
to check first the option to render to a Image window, so Blender can become
a true single-window application. :)
For example, the 'Spare render buffer' (jkey) doesnt work.
- Render with border, now default creates a smaller image
- No zbuffers are written yet... on the todo!
- Scons files and MSVC will need work to get compiling again
OK... thats what I can quickly recall. Now go compiling!
movie friends :)
There were two issues with it, which both have been tackled as follows:
- the correction transformations (quaternions) were calculated per face,
and then averaged over the vertices. This gave annoying inaccuracies,
especially when the geometry is irregular.
The new code first calculates two tangent vectors in a vertex, based on
the associated edges it has in a face. These tangents then are used to
define the transform. Tangents are 20% of the length of an edge now.
- When a SubSurf modifier was in the chain, the deformation caused by the
subsurf was also included in CrazySpace correction, giving even larger
errors.
New code temporally disables Subsurf, recalculates vertices, and then
does the crazy tricks. :)
All in all, quite well working!
guitargeek), this commit enhances the support for temporary storage
for the structs EditVert, EditEdge, and EditFace. The field
"EditVert *vn" has been removed and replaced by a union called
"tmp" that can hold:
v, an EditVert pointer;
e, an EditEdge pointer;
f, an EditFace pointer;
fp, a float pointer;
p, a void pointer;
l, a long;
Please see the mailing list post here for more information about
this:
http://projects.blender.org/pipermail/bf-committers/2005-December/012877.html
Works like for Object layers, but local within Armature itself. Each Bone
can be in (16 now) any layer, and the Armature layer defines what is
visible or not. Also note that hiding will still work too.
Since the Blender code is *stuffed* with Bone options now, this commit
requires a good test if all tools we got now comply to layers...
(I counted 130 cases for checking for selected Bones in code!)
In PoseMode; hotkey M will show 'movetolayer' menu. Not in editmode...
then its the mirror menu.
Todo: make action/nla drawing comply to Armature layer settings.
- There is now a (temporary) dropdown box in the image window header for
switching between the old an new unwrapper code. So to test the changes
described below you need to enable the new unwrapper code.
- Pinning is now more predictable, if one uv is pinned, the others belonging
to the same vertex are pinned also.
- Live LSCM is much faster, since the LU factorization, the most expensive
part of the computation, is now stored and reused (was Jens' idea).
- Packing multiple uv charts is slightly improved, by doing a binary search
over the texture width. This fixes the case where all the charts are
packed at the bottom of the image.
- LSCM now uses an angle based formulation, and the results seem somewhat
different (maybe slightly better?), didn't find out why yet.
Set the option in EditMode, in 2nd mesh tools panel. It only works on
transform options now (grab/rot/scale), and of course assumes a near-
perfect symmetrical mesh. Mesh Object itself can be on any location
though (and rotated etc).
