MSVC is not C99-compliant, but with the macro as it was, the "args" part was consuming and dropping one of the params, so the format string wanted N extra args, and there were only N-1... boom, crash! Without "args" this works correctly (checked on MSVC10 release build)
Crash was caused by non-initialized original index passing to compareDrawOptions
callback. Due to in some cases it's enough to know indices of faces in final DM
(like for comparing if the same texture used for mesh drawing) assume this
callback receives index in final DM and if it's needed it will make conversion
to original index itself.
This should help reaching extra speed if textured object is affected by
bevel modifiers, i.e.
* Compile all of cycles with -ffast-math again
* Add scons compilation of cuda binaries, tested on mac/linux.
* Add UI option for supported/experimental features, to make it
more clear what is supported, opencl/subdivision is experimental.
* Remove cycles xml exporter, was just for testing.
Added compareDrawSettings callback to driver mesh's callbacks which are
drawing textured faces (mapped and not mapped). This new callback checks
if two faces are drawing with the same settings (testures, shading etc)
and if they not, flush of faces happens into ogl using glDrawArrays and
next face would be drawn with it's own settings.
Currently implemented compareDrawSettings is used to resolve issue from
bug report only, probably there are extra places where this callback is
needed, but haven't seen configuration where current logic will fail,
so it should be ok.
Also reordered arguments passing to drawMappedFaces DM's callbacks,
so now all drawing callback are accepting list of callbacks and then
userData, instead of using mixed order of callbacks and userData which
was a bit confusing to work with.
This commit fixes:
- #26410: VBO & multitexture doesnt work
- #29464: VBO enabled causes UV coruption
It was error in CMakeLists.txt caused by automatic bundling script which
expanded variables instead of substituting them as-is.
Fixed both of bundling script and CMakeLists.txt
bit so that it's more symmetric and resulting float ops are compiled the same
way. Otherwise extended float precision being used in one place and not the
other can make comparisons fail.
This also revealed an issue where the opengl render float buffer was not linear,
and toggling back to a render slot would show wrong colors. Now it converts the
float buffer to linear so that this goes ok, disadvantage is that it's slower.
*** use-case:
"I have 10 reference images that overlap each other and every time I want to see one, I need to change the transparency of all the others.
therefore it would be nice to have a little button by each menu to allow enable/disabling individual background images"
To avoid subversioning bump I created a define that is negative (DISABLED) and of course a rna that is a boolean_negative.
Talked with Campbell and he actually prefers this way over do_version, so there it goes.
There has been a comment on the bug tracker about similar issues with drawing in logic buttons, this should be checked as well (Note on the bf-committers ML).
"The Blender Foundation also sells licenses for use in proprietary software under the Blender Licens"
also remove NaN references from files that have been added since blender went opensource.
from Andrew Hale (trumanblending)
Tracker description
*******************
The current python noise module included with Blender has yet to be updated to the new Py API. This patch does so, with the following major points:
- The noise module has now been moved to a submodule of mathutils, it can be accessed by mathutils.noise. It was moved from it's own module as it will now return mathutils types and also have greater visibility to the user.
- All functions which return vectors will now return mathutils.Vector types to be consistent with the rest of the API. Previously (x, y, z) tuples were returned.
- A different implementation of random_unit_vector is now used, this allows 2D, 3D and 4D vectors to be returned. Previously only 3D was possible.
- Some function names have been changed to remove ambiguities and make naming consistent within the module. noise.vector is now noise.noise_vector and noise.vl_vector is now noise.variable_lacunarity
- Doc strings have been updated to be compatible with auto docs.
- Code style and internal naming has been changed to match the conventions in other mathutils code.
Thanks,
Andrew
The ratios were only being maintained between the altered weights when they were redistributed,
and the resulting weights did not fully respect the locks--now they do.
