Bug reported by Dominique Lorre
Fix submitted by Jeroen Bakker
When importing COLLADA files, the name of a custom data layer can be longer than 32 bytes. Make sure only 32 bytes are copied.
Sculpt data shouldn't be lost when making topology changes without
quads<->tris face converison.
General idea:
- Go through all grid points of each corner and convert per-corner
coordiante to per-face cooredinate
- Apply weights and convert new point to per-corner coordinate
- Use bilinear interpolation to get needed displacement vector
Some additional work was necessery:
- Two neighbour corners could have different displacements along common
boundary. multires_mdisp_smooth_bounds() makes displacement "symmetrical"
- Point could change it's corner, so displacement vector should be flipped
in some way. In some cases it's not only flipping, because corner could
be mapped with some rotation. It's not solved for triangular faces yet,
so only z-axis displacement would be interpolated for tris.
More limitations:
- Interpolation will give incorrect result after quad<->triangle
face conversion.
- When face normal was fillped displacement would change it's direction too.
Also use const char in many other parts of blenders code.
Currently this gives warnings for setting operator id, label and description since these are an exception and allocated beforehand.
- fixed this error 7 different functions (deform groups, uv layers & similar).
- support for numbers over 999.
- renamed splitIDname() to BLI_split_name_num(), moved to BLI_path_utils
Both stored the filename of the blend file, but G.sce stored the last opened file.
This will make blender act differently in some cases since a relative path to the last opened file will no longer resolve (which is correct IMHO since that file isnt open and the path might not even be valid anymore).
Tested linking with durian files and rendering to relative paths when no files is loaded however we may need to have some operators give an error if they are used on the default startup.blend.
- omit render code from this warning (cmake only), until render branch is merged.
- moved -Wunused-parameter warning to apply to all C code in blender (not just ./source/blender), (cmake only).
- mdisp_corners used to return incorrect number of verts in some cases
- fixed memory corruption when face changed vertex count in
edit mode (forgot displacement for such faces atm, could be changed
in the future)
- removed deprecated bitmap arg from IMB_allocImBuf (plugins will need updating).
- mostly tagged UNUSED() since some of these functions look like they may need to have the arguments used later.
scene.active_keying_set --> scene.keying_sets.active
...same for active_uv_texture. active_vertex_color, active_keyconfig,
- move mesh.add_uv_layer() and mesh.add_vertex_color() into their collections
also have them return the newly created layer and dont set the layer active.
uvtex = mesh.uv_layers.new(name)
vcol = mesh.vertex_colors.new(name)
- BGE Shader.setSampler(name, index): index range check was wrong.
- Compositor check for an invalid channel was incorrect.
- getting the center of selected verts used an uninitalized z axis.
- do_init_render_material() used && rather then & when testing for MA_TRANSP.
- weight paint activate flipped bone used && rather then & for flag checking.
- vertex normals were not being flipped (though faces are)
- rim faces didnt influence edge vertex normals
apply solidify on top of solidify modifier now works correctly
mask, also avoid making vertex groups if they will not be filled.
* Add back image pin option in image editor header.
* Fix deep shadow not respecting Cast Buffer Shadows option.
* Tangent space normal map baking should work again now.
* Fix a problem with particle duplis, due to own bugfix for #20350,
the problem for that seems to be in dupliverts, not particles.
* Fix external multires data link getting lost on exiting editmode.
(commits 27776,27777,27830,27840,27841,27862 by Brecht from render25 branch)
to specify different spring lengths.
Implementation is quite ugly because the shape key has to be pulled
through the modifier stack somehow, need a more flexible data mask
system to solve this properly.
(commits 27773,27775,27778 by Brecht from render25 branch)
replaces another so it can do updates (e.g. dopesheet editor can
sync channel selection).
Also coded a simple optimization for allocating small objects,
based on mempools. It's #ifdef'd out, you can enabled it by
defining OPTIMIZE_SMALL_BLOCKS (e.g. adding -DDOPTIMIZE_SMALL_BLOCKS to
your compiler flags).
We suffer from a great deal of performance loss from the system allocator
(vgroups, ghash, edgehash, the singly-linked list implementation in blenlib,
editmesh, and likely a great many areas I'm forgetting), and this is the
common solution for handling the many-small-objects problem. It's not
really production-ready yet (it's long-term memory consequencers need to
be profiled first, and the implementation tweaked as necassary), but for
people on systems with slow system allocators it's worth trying.
Note that since this creates a guardedalloc<->blenlib link, the build systems
need to be updated accordingly (I've already done this for scons, though I'm
not sure if the player builds).
Revised external multires file saving. Now it is more manual in that you
have to specify where to save it, like an image file, but still saved at
the same time as the .blend. It would ideally be automatic, but this is
difficult to implement, so for now this should at least be more reliable.
* Don't allow adding/removing multires levels in editmode.
* Customdata code for swapping mdisps restored.
* Fix inflate brush crashing with multires.
* Smooth and layer brush don't work yet with multires, but at
least avoids crashing now.
* Fix threading issue with flatten brush.
* This is experimental, the file format may change still!
* Helps reduce memory usage, keeps .blend files smaller, and makes
saving quicker when not editing multires.
* This is implemented at the customdata level, currently only the
multires displacements can be stored externally.
ToDo
* Better integration with object duplication/removal/..
* Memory is not yet freed when exiting sculpt mode.
* Loading only lower levels is not supported yet.
* Displacement coordinates are now stored differently, as a grid per
face corner. This means there is duplication of coordinates, especially
at low subdivision levels, but the simpler implementation justifies it
I think.
* ToDo: conversion of existing multires files (2.4x or 2.5x), loading them
may even crash now.
* Editmode preservation/interpolation code also has not been updated yet.
* Multires now works on the CCGDerivedMesh grids instead of CDDerivedMesh,
which should be more memory efficient.
* There are still bad memory peaks (if you're using 32bit) when subdividing
or propagating displacements. Though at least there should be no huge
memory blocks allocated, which windows is now to have trouble with.
* Still found some weird spike artifacts at lower multires levels, some also
happening before this commit. Perhaps computation of tangents needs to be
tweaked more.
* Multires modifier now has viewport, sculpt and render levels. Also the
levels have been made consistent with subsurf, previously the same level
of subdivision was one less for multires.
* Both multires and subsurf modifier now can have their subdivision level
set to 0 for no subdivision.
* Convert all code to use new functions.
* Branch maintainers may want to skip this commit, and run this
conversion script instead, if they use a lot of math functions
in new code:
http://www.pasteall.org/9052/python
svn merge https://svn.blender.org/svnroot/bf-blender/trunk/blender -r19820:HEAD
Notes:
* Game and sequencer RNA, and sequencer header are now out of date
a bit after changes in trunk.
* I didn't know how to port these bugfixes, most likely they are
not needed anymore.
* Fix "duplicate strip" always increase the user count for ipo.
* IPO pinning on sequencer strips was lost during Undo.
soc-2008-nicholasbishop branch.
Note: any old code with multires_test() or multires_level1_test() can
just be deleted, not needed by the multires modifier.
* pixels with <= the current opacity are not painted onto, speeds up painting especially with low spacing. (only when airbrush is disabled)
* use qsort rather then own crufty sorting function
* grouped projection painting functions together.
* subsurf code had a lot of unused variables, removed these where they are obviously not needed. commented if they could be useful later.
* some variables declorations hide existing variables (many of these left), but fixed some that could cause confusion.
* removed unused vars
* obscure python memory leak with colorband.
* make_sample_tables had a loop running wasnt used.
* if 0'd functions in arithb.c that are not used yet.
* made many functions static
=======================================
Alpha blending + sorting was revised, to fix bugs and get it
to work more predictable.
* A new per texture face "Sort" setting defines if the face
is alpha sorted or not, instead of abusing the "ZTransp"
setting as it did before.
* Existing files are converted to hopefully match the old
behavior as much as possible with a version patch.
* On new meshes the Sort flag is disabled by the default, to
avoid unexpected and hard to find slowdowns.
* Alpha sorting for faces was incredibly slow. Sorting faces
in a mesh with 600 faces lowered the framerate from 200 to
70 fps in my test.. the sorting there case goes about 15x
faster now, but it is still advised to use Clip Alpha if
possible instead of regular Alpha.
* There still various limitations in the alpha sorting code,
I've added some comments to the code about this.
Some docs at the bottom of the page:
http://www.blender.org/development/current-projects/changes-since-246/realtime-glsl-materials/
Merged some fixes from the apricot branch, most important
change is that tangents are now exactly the same as the rest
of Blender, instead of being computed in the game engine with a
different algorithm.
Also, the subversion was bumped to 1.
