http://wiki.blender.org/index.php/Rewriting_the_2.4x_BPython_API
Some areas are not yet finished, the 2.4 api can remain "import Blender" and the bpy api "import bpy" gives access to the new api.
Todo
* event, link, etc are not assigned to Blender.*
* Scripts need to be tested and updated.
* Scripts cannot mix bpy and Blender modules anymore.
* Makefiles need to be updated (only works with scons now)
* Make the 2.4 api optional with IFDEF's
* Lots of testing and small fixes
=================
Big commit, but little user visible changes.
- Dupliverts and duplifaces are now rendered as instances, instead
of storing all of the geometry for each dupli, now an instance is
created with a matrix transform refering to the source object.
This should allow us to render tree leaves more memory efficient.
- Radiosity and to some degree raytracing of such objects is not
really efficient still. For radiosity this is fundamentally hard
to solve, but raytracing an octree could be created for each object,
but the current octree code with it's fixed size doesn't allow this
efficiently.
- The regression tests survived, but with I expect that some bugs will
pop up .. hopefully not too many :).
Implementation Notes
====================
- Dupligroups and linked meshes are not rendered as instances yet,
since they can in fact be different due to various reasons,
instancing of these types of duplis that are the same can be added
for them at a later point.
- Each ObjectRen now stores it's own database, instead of there being
one big databases of faces, verts, .. . Which objects that are actually
rendered are defined by the list of ObjectRenInstances, which all refer
to an ObjectRen.
- Homogeneous coordinatess and clipping is now not stored in vertices
anymore, but instead computed on the fly. This couldn't work for
instances. That does mean some extra computation has to be done, but
memory lookups can be slow too, and this saves some memory. Overall
I didn't find a significant speed impact.
- OSA rendering for solid and ztransp now is different. Instead of e.g.
going 8 times over the databases times and rendering the z-buffer, it
now goes over the database once and renders each polygon 8 times. That
was necessary to keep instances efficient, and can also give some
performance improvement without instances.
- There was already instancing support in the yafray export code, now it
uses Blender's render instances for export.
- UV and color layer storage in the render was a bit messy before, now
should be easier to understand.
- convertblender.c was reorganized somewhat. Regular render, speedvector
and baking now use a single function to create the database, previously
there was code duplicated for it.
- Some of these changes were done with future multithreading of scene
and shadow buffer creation in mind, though especially for scene creation
much work remains to be done to make it threadsafe, since it also involves
a lot of code from blenkernel, and there is an ugly conflict with the way
dupli groups work here .. though in the render code itself it's almost there.
not mirrored exactly, though the problem is not completely solved.
The way local frames are computed for particles is still not fully
symmetric, which shows especially on long hairs...
Also made the shift+o subsurf switch work recursively into
dupli-groups, did only the first level before.
the depsgraph relations for constraints with a local ipo were being
skipped, but I have no idea why this code was added? Uncommenting
it seems to work fine.
* Old files now get initialised with the correct scale. The wrong calculation was being used
* Added a new tool to Alt-S menu: "Apply Scale"
This tool causes all the keyframes in the active strip to be moved to their NLA-scaled times, the scale to be set to 1.0, and the frame ranges recalculated accordingly (to remove any nasty weird errors)
* Scale field now draws red when the action-range is < 1, and the tooltip in this case instructs the user how to fix this (by using "Apply Scale").
NLA-Strips now have a new setting: Scale.
It determines how much the action-range is scaled for each repeat, instead of the scaling being implicitly determined based on repeats + strip-length.
One of the instant benefits of this, is that when increasing the number of repeats, the strip length increases by the right amount. Thus, increasing the number of repeats retains a constant speed.
Hopefully we can prevent weirdly scaled actions this way. (i.e. 0.00001 frames long action * 10000 or so)
Todo:
- Transform code needs to be able to set the scale setting (it doesn't yet)
- Add a new option to "apply scaling", to fix up problems with old files that have really bad scaling. Situations when this is needed could get indicated in the interface too... (red background for "Scale" field?)
=========
- Normalize strand coordinates over the length of the strand, not
cut off when with e.g. length texture control.
- Weight paint and particle mode x-mirror now works with deformation
and shape keys active.
- X-ray is now disabled in particle mode, messed up depth tests.
IK-constraint "flush_targets" function was causing segfaults on debug builds from certain MSVC compilers.
The cause of this, is that ct is freed in the SINGLETARGET_FLUSH_TARS macro already, but ct is accessed in the following line to get the next target (ct= ct->next). Caused by brecht's commit for the pole-target stuff for IK-constraint.
Now the SINGLETARGET_FLUSH_TARS macro, and the SINGLETARGETNS_FLUSH_TARS macro will correctly advance the location that ct points to.
used for clipping were being stored in the mesh, but modifiers
can result in two objects with the same mesh having a different
bounding box. Solution is to store bounding box in the object.
=========
- Texture orco coordinates for particles are now actual orcos instead
of just the vertex positions, which means they are the same under
deformations and the same as the ones on the mesh.
- Particle distribution now uses these orcos to get consistent
distributions independent of deformation.
- This required changing the way orco's are computed for meshes. Now
instead of generating an orco derivedmesh separately, the derivedmesh
is generated alongside the regular one and stored in an orco custom
data layer.
* [Peach Request] AutoIK now respects axis locking (rotation locks).
- Temporary DOF-Locks are turned on for those bones that are part of an AutoIK chain while transforming. These locks get cleared after transforming.
- This works for all bones except the root bone of the chain, which doesn't seem to be able to be locked.
* Limit Location constraint can now optionally affect Translations too (i.e. NKEY panel values won't change).
- LimitRot,LimitScale support (for their respective transforms) will be done at a later date
- This only works if the constraint is using World/Local space (the other spaces are not supported yet).
- I've added a temporary button in the LimitLoc panel to enable this option (it is disabled by default). This button will be properly assigned a place in that panel sometime.
=========
- Fix crash in particle transform with the particle system not editable.
- Particle child distribution and caching is now multithreaded.
- Child particles now have a separate Render Amount next to the existing
Amount. The render amount particles are now only distributed and cached
at render time, which should make editing with child particles faster.
- Two new options for diffuse strand shading:
- Surface Diffuse: computes the strand normal taking the normal at
the surface into account.
- Blending Distance: the distance in Blender units over which to
blend in the normal at the surface.
- Special strand rendering for more memory efficient and faster hair and
grass. This is a work in progress, and has a number of known issues,
don't report bugs to me for this feature yet.
More info:
http://www.blender.org/development/current-projects/changes-since-244/particles/
====================
Bugfix for mdef + groups. The mdef modifier treated the cage object as
a sort of parent, which meant that transforming the cage object also
transformed the mesh. However, this behaviour is different from e.g.
armatures, where transforming the armature object has no effect for
the modifier, but instead this transform is taken into account by the
fact that the armature is usually the parent of the mesh.
For groups and duplis, this basically lead to the parent transform
being applied twice. So, now transforming the cage object will have no
effect, except if you make the cage object the parent of the object.
-----
Bugfix #7831: fix image filename extensions on saved files. Make FFMPEG save JPEG images by default, allow JPEG images to have either ".jpg" or ".jpeg" extension.
- Library referencing didn't add the new local Ipo for constraints
- Proxy command didn't set the driver object in constraint ipo to be
the new proxy object.
=========
Bugfix for crash with disabling the particles system and saving/loading
files, the way it detected if the mesh had changed could not work correct.
- new option for Local Constraint Ipos did not set user counter in
Ipo at file reading, causing data to get lost (not saved).
- Driver feature: the channels "Loc X Y Z" now also use the result
of constraints, but transformed back into local space, as if it
was action X Y Z. Nice stuff for those who understand this...
it means you can drive something with a bone that has constraints.
modifier.c, BKE_modifier.h - flag for modifiers to say they use pointcache, also new func modifiers_usesPointCache
renamed //pointcache to //blendcache_blendfilename so blendfiles in the same dir dont conflict, and other to show this dir isnt limited to pointcache only (nodes way want to use this)
wizard_curve2tree.py - better defaults for pretty tree's
Fixes for bugs in #7833
- Hair softbody cache was reset on rendering.
- When Alt-a playback is synced to audio frames are skipped. This caused dynamics caching to go crazy. Now the first loop of playback goes through all frames if there are non cached dynamic systems.
Invalid point caches weren't cleared when new systems were created. This caused messed up shapes when softbody was added.
==========
- Added a Mute button for shape keys, useful for debugging the influence
of driven shape keys.
- Bugfix for the shape keys relative to others keys, was hanging in
an eternal loop when deleting the other key.
=============
A new "Selected to Active" option in the Bake panel, to (typically) bake
a high poly object onto a low poly object. Code based on patch #7339 by
Frank Richter (Crystal Space developer), thanks!.
Normal Mapping
==============
Camera, World, Object and Tangent space is now supported for baking, and
for material textures. The "NMap TS" setting is replaced with a dropdown
of the four choices in the image texture buttons.
http://www.blender.org/development/current-projects/changes-since-244/render-baking/
in EditButtons, panel "Links and Materials", there's now a browse button
to directly assign a material to selected faces. It does:
- check if material was already in one of the 'slots' of the object
- if so, then use this as index to assign
- if not, then add a new slot, and assign the new index
-> Constraint Influence Ipo now can be local, linked to constraint itself
You enable this in the IpoWindow header, with the Action icon to the left
of the Ipo Type menu. The button tooltips give the clue as well.
Tech note: the Ipo now can get directly linked to a constraint, and is
being called during regular pose constraint solving.
Actions (and drivers in actions) are being calculated *before* pose
constraint solving. Result of actions then is written in bones, which
then solves the entire pose.
This means you can have a driver on both the constraint, as on the action
channel for the constraint! Not that I'm going to debug that easily :)
Additional fix: Joshua added a copy/paste IpoCurve feature, but he broke
the functionality to be able to paste in an empty ipo channel. That now
works again
