* Loading old files didn't initialise the new rotation variables properly
* Fixed some errors with the newly added operator for copying RNA-paths for properties
* Auto-keyframing now correctly refreshes animation editors after adding keyframes. Made the keyingsets code send notifiers again, but now using the newly added WM_main_event_add() (thanks Brecht)
* A few UI tweaks again for animation stuff (timeline, keyingsets UI)
- editing properties from python wasnt running their update function.
- missing commas made dir(context) give joined strings.
- added __undo__ as an operator class attribute so python ops can be set as undoable. (like existing __register__)
* Added a new option for Auto-Keyframing which makes it only insert keyframes for the items included in the active Keying Set.
This only works for Transform Auto-Keyframing so far (other tools will get it added later). The option is disabled by default.
* Fixed bug where adding an 'entire' array to some KeyingSet would only start from the index of the button that the mouse was over at the time
* Made some UI tweaks for Keying Sets buttons (still heaps of missing options there).
The hardcoded paths for rotation keyframes on objects got broken by my commits to rename the rotation properties. I've taken this opportunity to recode the auto-keyframing code here to use the builtin keyingsets instead of going through and manually calling insert_keyframe(), thus preventing this problem in future.
* Removed the hack-functions to set euler rotations from the values set in the other rotation representations, even when euler rotations weren't being used. I pressume that this was added for being able to represent quats in terms of eulers for the UI, but really it would break animation evaluation (i.e. euler curves after quaternion curves would always block the quaternion curves).
* Object rotation values in the transform properties panel now take into account rotation modes, so the appropriate rotations will get converted to quaternions before being drawn.
* Fixed a few bugs with some of the conversion code (minor stuff left out).
Animating rotations using different rotation modes should now work more often than before.
Previously, quaternion and axis-angle values were stored in the same variable in DNA, but that was causing problems with other animation curves overwriting the values and causing the rotations to not work as expected.
There are still some issues, but I'll track those down later tonight
* Compo node backdrop works again.
* Compo node previews and backdrop now get correct color management
float to byte conversion.
* Compo nodes got unecessarily recalculated while moving nodes.
* Fix compo node viewer nodes not getting activated correctly.
* Main compo node preview render computations are now outside of
mutex lock, so better for multithreading.
* Tex node outputs did not work in some files loaded from 2.4.
* Change RNA updates to take into account groups that may be shared
between multiple node trees.
of the main reasons for passing along context, while actually
they don't need much context at all. Might be removed again,
but would like to have this especially for RNA API functions.
* Make Ctrl+M key work for mirror in 3D view.
* Fix mirror along global axis, was mirroring around all axes when
the object was rotated, due to wrong matrix order, was also not
working in 2.4.
* Pressing e.g. X twice still doesn't go to local mode, would fix
but don't know how the code is intended to work.
- added TexMesh access ([#19505] Missing option : TexMesh)
- Ctrl+Tab works again, not-so-nice workaround, disallow switching to paint modes from editmode, but would be nicer to manage this with keymaps.
- Useful for dragging buttons to the far right when theyd otherwise hit the screen edge.
- Useful for transform though probably NOT what you want when using the transform manipulator (should make an option).
- When enabled, number buttons use this as well as a different conversion of mouse movement
float numbuts: mouse 1px == 1-clickstep
int numbuts: 2px == 1 (tried 1:1 but its too jitter prone)
details...
- access as an option to GHOST_SetCursorGrab(grab, warp)
- Currently all operators that grab use this, could be made an operator flag
- only Ghost/X11 supported currently
The makefile is a copy of the source one, not tuned, just to unbreak build.
This raises some questions: why separate dirs? why each build system takes
a different approach (different libs vs all source files into one)?
(Mostly for very early testers)
Cocoa uses coordinates with y=0 at bottom : updated wm_window.c and wm_event_system.c for COCOA build to avoid double conversions in response to mouse move events and GHOST_getCursorPosition
Known limitations:
No fullscreen support
Font issue in preference panel
libSDL uses some Carbon functions
Selecting a material in the node tree sets this as the active material and the buttons view redraws.
Added rna prop material.active_node_material
Currently its not clear what settings are used by the node material and the base material (needs some tedious research) so I made most panels use the node material with the exceptions of volumetrics, physics and halo settings.
We'll probably need to split the panels up to do this properly.
This calls a registered menu as a popup so we can reuse header menus , currently used for Node editor and Sequencer add menu (Shift+A), can be used for toolbox too.
In the Outliner, it is now possible to toggle per bone the selectability of the bone in the viewport, as for Objects using the restriction columns. This can also be set using the RNA-api.
I've tested all commonly used tools IMO, but there may still be a few which I've missed. Please report those cases.
PS. For some reason, the define was already there, but not connected up to anything. Can't remember why anymore, but here it is...
*Disabled ray counter (can be enabled on render/extern/include/RE_raytrace.h by commenting out the define)
*marked bvh_node_merge() as static inline (hopping it now compiles on gcc and mingw)
* Made bvh_node_merge() in svbvh.h static (fix suggested by jaguarandi). This makes mingw link again.
* Also, patched my previous fix for ... = {}; since mingw didn't like the other fix (which was for msvc).
* Replaced ... = {}; with ... = {0};
* Solved problem with logf(), where msvc couldn't figure out which version of log() to call (solved by casting the int argument to a float, but could also have been to double)...
* The cflags and cxxflags for scons when compiling the rendering module were only valid for gcc compiles. These will still need to get added for msvc sometime, but for now, there are no more warnings about unknown options...
* mingw almost compiles again cleanly, except for a linking error when linking blender http://www.pasteall.org/8297
* win64 should compile again too to a similar degree?
* silenced warnings about no newlines...
* Code for generating 'Object' summary of Keyframes for DopeSheet (which is also used by the TimeLine for getting keyframes to draw) now considers materials, object data, and particles too.
* Rearranged the way that keyframing-related settings were presented in the User Preferences. The way the settings were grouped was plain confusing, and based on biased views from the old system. For the record, 'needed'+'visual' are always considered when inserting keyframes, 'always' is for autokeyframing, and default interpolation is only used for newly created F-Curves.
* Fixed bug #19472 - Scroll wheel scrolls in the wrong direction for enum-menus that were flipped (i.e. window type menu and 3d-view mode selector).
