- ghost data wasn't being freed (added wm_ghost_exit() call to wm_init_exit.c)
- GHOST_EventManager wasn't freeing GHOST_IEventConsumer's
- ghost/X11 wasnt calling XCloseDisplay(), some junk from X11 wasnt being freed
- ghost/X11 XAllocNamedColor wasn't freeing the colors when done making a custom cursor.
When there were no strips but some empty tracks, the active action should be evaluated normally. This is not an optimal solution (more user testing + suggestions regarding the best way to do this is welcome).
The main contribution of this commit is the possibility to save Freestyle configuration information inside .blend files. It required an extensive refactoring of the previous interface code.
The code has been tested and does not crash on my machine. If you encounter issues, please let me know (and if possible, sending me an sample .blend file).
DETAILS:
- refactored code to notify Freestyle when adding/removing render layers (FRS_freestyle_config.{h,cpp})
- corrected the freeing of style modules when files are read from / written to disk
- moved Freestyle configuration information into scene renderlayers, to allow loading / saving .blend files (DNA_scene_types.h DNA_freestyle_types.h FRS_freestyle.cpp)
- inserted temporary trick to prevent crashes when orthographic camera is used (SilhouetteGeomEngine.cpp), but outputting incorrect feature line calculations
* That also means you can't really change them and save it before we actually do a do_version.
* Marking the old variables in DNA_world_types.h to be removed.
* meshes are now exported without modifiers applied (for skinning to work), this should become a configurable option later
* had to edit OpenCollada code to make armature/skinning import work correctly
* code layout improvements
* Added callback function for some metaball properties:
When some properties (wiresize, threshold, update flags) of metaball are changed, then these properties are copied to all metaballs in the group (all metaballs with same base name). This is important to "share" some properties between metaballs, because polygonisation of metaball is influenced only by properties of base metaball and base metaball can be changed.
* Improved drawing of selected Metaball objects
There was a known issue for a long time that we occasionally encountered
strange "TypeError: an integer is required" and "RuntimeWarning: tp_compare
didn't return -1 or -2 for exception", as shown in the following unit test
log. The source of the former error was PyInt_AsLong(obj) being used by
IntegrationType_from_BPy_IntegrationType(obj), where "obj" was not properly
initialized in "__init__" before the converter was called. The TypeError
was left unattended for a while and showed up when a comparison occurred
and the TypeError was detected, which resulted in the latter warning.
> runTest (__main__.UnaryFunction1DDoubleInitTestCase) ...
> .\blender:211: RuntimeWarning: tp_compare didn't return -1 or -2 for exception
> ERROR
>
> ======================================================================
> ERROR: runTest (__main__.UnaryFunction1DDoubleInitTestCase)
> ----------------------------------------------------------------------
> Traceback (most recent call last):
> File "init_tests.py", line 211, in runTest
> TypeError: an integer is required
>
> ----------------------------------------------------------------------
* Also removed unnecessary error messages in "__init__" methods of
UnaryFunction1D types.
Multires > Multiresolution
The term 'Subsurf' was really confusing, because of the similar term Subsurface Scattering, meaning something completely different. Additionally, the prefix 'sub' (meaning 'lower' or 'under'), as well as the word 'surf' (meaning water foam / a water sport) really mean completely different things on their own, making 'subsurf' a terrible abbreviation.
files but only manipulates paths. It produces both ablsolute and relative paths. COLLADA exporter can now use it.
- re-wrote unit test for it, this is now more compact and readable
- RNA API Image.get_export_path takes a boolean arg to indicate whether a relative or absolute path should be returned. Python scripts
can now use it.
