Revision 86c2549 by Campbell Barton February 6, 2009, 16:38 (GMT) |
[#18257] Workaround for bug in freealut 1.1.0 by Jörg Müller (nexyon) Since this area is not maintained, I can only say it works on my system (linux/32bit) And nexyon assured me he checked over this patch well. Blender using the deprecated function is not great, but nexyon's bugfix would not be available until the next version of freealut so better to work around it by using the new function. Please test game sounds play back on win32 and Mac, in cases that it worked in the first place. |
Revision 14f180f by Ton Roosendaal February 6, 2009, 16:34 (GMT) |
2.5 Node editor: - Linking sockets back. - Removed obsolete code - Add node gives correct pointers for Render stuff |
Revision 61fdeb1 by Ton Roosendaal February 6, 2009, 16:31 (GMT) |
2.5 View2D bug: zooming failed in node editor due to check for wrong variable (keepzoom instead of keepofs :) |
Revision 8f70b93 by Campbell Barton February 6, 2009, 14:03 (GMT) |
2.4x Sequencer Python API - add metastrips now possible - access to speed options - access to wipe options - added seq.startDisp seq.endDisp seq.update() |
Revision 1646f05 by Ton Roosendaal February 6, 2009, 13:07 (GMT) |
2.5 Compo node fixes: - adding new composite in scene crashed, wrong pointers passed on to addnode function - scene render-layer node now gets default scene ID pointer, which can be viewed/edited with menu in node - on using headerbuttons in node window, display refreshes now |
Revision 9d5958f by Ton Roosendaal February 6, 2009, 12:33 (GMT) |
2.5 Bugfix for Bugfix for 2.5 - Bugfix #18251! - the call to GPU_default_lights has to be done after windows opened, in case you have no .b.blend saved it crashed - further this function is only called on .B.blend reading, so it's correct to reset the lighting to user-def. |
Revision 586d660 by Ton Roosendaal February 6, 2009, 11:11 (GMT) |
2.5 Bugfix: on reading multiple pre-2.5 files, the keymaps were generated again, causing items to be added double. |
Revision 0355dcc by Joshua Leung February 6, 2009, 10:48 (GMT) |
2.5 - Armatures Code * Brought back recalculate bone roll (Ctrl N). This should be quite a straightforward example of a cleaned-up + ported armature tool. * Cleaned up a few warnings |
Revision 2756060 by Joshua Leung February 6, 2009, 10:04 (GMT) |
2.5 Converted all countall() calls in armature code to armature_sync_selection() |
Revision 4b0313b by Joshua Leung February 6, 2009, 01:39 (GMT) |
2.5 Armature Editing - Restored 'Align Bones' (Ctrl Alt A) * I've had to remap 'Duplicate window' to Ctrl Alt W instead... * Fixed some bugs with armature context iterators for X-Axis Mirroring. Now, the code there checks for mirrored bones in the right way. For details on how to write tools that need to cope with this option, refer to the 'Align Bones' operator. I'll port another simple operator as another good example soon. Additional notes: Currently, armature_sync_selection() is really buggy (not part of this commit), and needs further attention. |
Revision 6a669d0 by Robin Allen February 6, 2009, 01:21 (GMT) |
Patch #18015, adds "Value to Normal" node |
Revision fbdd75a by Robin Allen February 6, 2009, 00:55 (GMT) |
Added compose/decompose, fixed bugs. |
Revision e1b92bc by Ton Roosendaal February 5, 2009, 19:28 (GMT) |
2.5 Safe method to move render results to the displayed image. It now allocates a single image for display, and on each refresh callback from render, it copies the refreshed section over to this image, in 32 bits. While rendering that image then only shows progress updates, as usual. This also now works for scenes in composte and results for composite. This should solve reported crashes for MBlur or SSS. |
Revision 0a3697c by Joshua Leung February 5, 2009, 05:51 (GMT) |
2.5 - Bugfixes for Pose 'Clear' Operators * Fixed context iterators for PoseChannels. The actual selection test was missing (but layer visibility was getting checked). Also added check for hidden bones here too. * Removed unnecessary defines added by mfoxdogg in BKE_context.h * Removed the old hacks in 'Clear ...' operators for ensuring that object animation (ipos/actions) don't overwrite poses. These are no longer necessary, since normal animation calculation is only done in a separate step on frame-changes now. * Removed the manual checks for layer visibility from 'Clear ...' operators, since that's taken care of by context iterators. |
Revision 6947624 by Michael Fox February 5, 2009, 05:13 (GMT) |
2.5 ******* -ported clear loc,rot,scale operators to pose mode operators - for some reason when animated and you clear anything, the whole aramture clears. somthing to do with the depsgraph stuff that was directly ported from clear armature() |
Revision f89e3d2 by Joshua Leung February 5, 2009, 03:28 (GMT) |
2.5 - Context API access for Bones (EditMode and PoseMode) * Added selected, selected+editable, and active to access EditBones for Armature Operators to use. These take into account X-Axis Mirror too, so there is really no need to check that sort of thing anymore in tools. * Added a quick testing operator for verifying that these loops filter the data correctly. I've dumped this in armature_ops.c for now. It can be activated using the TKEY hotkey in Armature EditMode only. This should be removed once we have a few more functional tools. * Ported over cleaned up roll-calculation tools from AnimSys2 * Removed a few ugly stubs from posemode code |
Revision 61a6e19 by Shaul Kedem February 5, 2009, 01:32 (GMT) |
flip_editnormals is back, alt+3key |
Revision 66f90c4 by Nicholas Bishop February 5, 2009, 01:12 (GMT) |
* Added range function for Multires's level RNA * Removed RNA for old multires in Mesh, the old mesh->mr struct should never be used (except for loading old files) * Dealt with sculpt + render (just needed to force multires update) * sculpt memory leaks fixed * Entering editmode now always exits paint modes (previously, only happened if editmode was called from the header button) |
Revision 5fc657c by Joshua Leung February 4, 2009, 22:33 (GMT) |
Bugfix for 2.5 - Bugfix #18251 The patch was pasted in the wrong place, thus destroying all custom OpenGL solid light setups too. |
Revision e25318e by Martin Poirier February 4, 2009, 21:38 (GMT) |
Rename correlation subdividing method into adaptative subdividing method, since that's what it's trying to do and much more understandable from a user's pov. |
|