Blender Git Commits

Cycles Hair: refactoring to support generic attributes for hair curves. There
should be no functional changes yet. UV, tangent and intercept are now stored
as attributes, with the intention to add more like multiple uv's, vertex
colors, generated coordinates and motion vectors later.

Things got a bit messy due to having both triangle and curve data in the same
mesh data structure, which also gives us two sets of attributes. This will get
cleaned up when we split the mesh class.

change limited dissolve angle limit default from 15 -> 5 deg. since 15deg is quite high for 2 surfaces to be considered co-planar

code cleanup: BM_face_create_ngon()

Change some of the paranoid checks into an assert since they should never happen.

fix [#33689] Crash with decimate modifier

in fact a more general bug in BM_face_create_ngon() where edges containing the same vertex multiple times were not accounted for.

fix [#33715] Dirty Vertex Colors display problem since 2.65a

fix [#33682] Animation not updated when the active shape index isnt set

CMake: Looks like a period sneaked into r53515, which caused CMake to fail.

CMake: Setting vc10 to use Python 3.3 (it was still set to 3.2). For the time being, I'm having vc10 use the vc9 Python libs, which SCons is currently doing.

add option WITH_SYSTEM_BULLET to link against the bullet installation found on the system.

Note: this doesn't work yet for everything with latest stable bullet (2.81), need to look into why and likely apply some patches upstream.
However I managed to link blender by disabling some features, likely it can be made to work without too much trouble.

add id property clear function (matching the same python function for dicts/lists)

r53351 removed a NULL check for name in defgroup_name_index(). name can be NULL when deforming verts in the BGE, so I'm adding a NULL check back in.

revert 53478, it looks this is still useful for a couple of use cases in non-projective texture painting such as stamping a static texture and using image space size in pixels.

Some UI message fixes...

changes needed for EDL import to work again.
- add sequence.update(data=False) function.
- made some sequence vars editable.
- correct some comments.

also rename rna function sequence.getStripElem() --> strip_elem_from_frame()

Bug fix - post 2.65a

New option "Keep Session" was not using "untitled.blend" as default save name,
in case a startup.blend was kept as session. Could give bad accidents.

Fix [#33189] AFTER 2.65 - Units in doc strings in ocean modifier.

Also:
* Fixes a (op prop) bug which prevented, once you had baked and freed ocean once, to bake again.
* Fixed infinite values of acumulated foam when baking with foam_fade values above 1.0, now simply clipping accumulated foam value to 1.0, as already done for the "instantaneaous" foam value returned by BKE_ocean_jminus_to_foam().
* Added missing RNA descriptions.
* Made foam_fade unanimatable!
* Added in UI some missing properties that are imho useful: random seed, size (kindof 'surface scaling'), and foam_fade (baking only).
* Removed custom lerp() func from bke's ocean.c, BLI's interpf does exactly the same thing (the first two args are just in reversed order). Note: this could most certainly be done in other parts of the code, bpy's mathutils for e.g. has its own linear interpolation code for vectors and matrices :/).
* Did some general code cleanup (mostly line length and no C++ -> C comments)...

Fix playercompile with stub for BKE_brush_gen_texture_cache

fix [#33726] keyframe_insert() and keyframe_delete() with index=-1 always fail

The bug was introduced in r53297, was calling pyrna_struct_anim_args_parse() twice for no reason, perhaps baddly applied patch?.

add assert to BLI_buffer_at if a type is given different to 'elem_size'

extend BLI_buffer
- add option to calloc or not, existing code wasnt consistent here, would calloc on first alloc but not on realloc, also - the static memory was never zero'd.
use flag BLI_BUFFER_USE_CALLOC to ensure all new memory is zero'd (static/alloc/realloc's).

- add BLI_buffer_declare_static / BLI_buffer_declare so its possible to have a buffer that never uses static memory.