Age | Commit message (Collapse) | Author |
|
|
|
billboards.
|
|
object center button was duplicated
|
|
Forgot REDRAW in obscure cases...
|
|
Reworked seq_cut completely. Now it is equivalent to: dup and set startofs /
endofs, which is what most people probably expected...
Fixes:
[#8388] Sequencer Cut on empty meta
[#8546] Sequencer: 2 different problems cutting some metastrips
|
|
* Attempted fix for bug #8599c (inserting new keyframes mucked up existing IPO-Keys). Now it calls make_ipokey() regardless of whether they exist already (if ipokeys are being shown in the IPO-Editor). This might be a bit slower on some files.
* 'Auto-Sync PoseLib' button is now only visible when there's a PoseLib
|
|
edges
Removed an incorrect check for original vertices on whether any faces used them.
|
|
Undoing in sculpt mode correctly returns to previous pivots now.
|
|
zooming behind an object to one with xray now works, transp objects work like normal objects.
|
|
|
|
large maps)
|
|
cloth settings in old files but I had to do it before release because naming convention was really bad in cloth (e.g. using SimulationSettings instead of ClothSimSettings in DNA). Same for some structure in CollisionModifier but with no sideeffects.
|
|
dynamic update and catches therefore also faster movements - reported by nudelZ (#blendercoders)
|
|
Not time for documention and polishing before release, hiding the panel behind (rt == 42) check.
|
|
|
|
|
|
Insert-Key in Action Editor didn't work correctly for Constraint Channels yet. This was because a pointer to constraint data was not being obtained (it's now hacked to assume that 'Influence' is being keyed all the time).
|
|
Deleting the first Action Channel in an Action Group acted odd. In some cases it crashed, other times it corrupted stuff (removing all subsequent Action Channels). Was missing a call to remove the Action Channel from group(s).
* Fixed whitespace in many places...
|
|
----------
Bugfix #6543: Blender.Library.Load() could segfault if called more than once
with a library of different endian type. In the process discovered an
invalid memory reference in other another function calling library_append().
Thanks to Brecht for the pointers on tracking this old bug down.
|
|
|
|
|
|
* sanitise IKEY handling block for view3d (really weird if-structure)
|
|
|
|
A new tool for for fixing colors on meshes imported from other files with
conflicting color palettes has been added to the FLT toolbar script. This
will find the closest match for all the colors in the faces of selected
meshes and update their baked vertex color representation.
Also added a small tweak to the 'update all' tool which will fix black faces
|
|
FLT importer had problems with external references that used
relative paths. Fixed.
|
|
|
|
|
|
|
|
docs and an exception if called.
|
|
|
|
Several transform bugs in the flt importer have been fixed. This included
parent/child relationships and matrices not importing correctly along with
some problems with the import scale option.
|
|
Default FLT palette had all colors with an alpha of zero
which caused problems with some versions of OpenSceneGraph.
|
|
Though I could not reproduce the crash, the example file would always reference an invalid pointer.
|
|
|
|
Should look into why XGetWindowProperty isnt working also.
|
|
|
|
[#7124] knot vector access for SurfNurb
|
|
|
|
possible crash if not enough memory there.
|
|
ray mirror and transparency, was using the same stack for all
recursion depths, now creates new stacks as needed.
Fix for bug #8592: crash saving non-float image as 16bit tiff.
|
|
|
|
down some values unnecessarily, which made it not entirely exact.
|
|
case of a deforming modifier after subsurf.
|
|
Fix for another bug I introduced with persistent python windows.
also removed unused var.
|
|
|
|
Manipulator event handler was out of sync with transform event handler.
The real hotkeys to increase/decrease the area of influence are Alt PAD+ / PAD- and PageUp / PageDown (has been that way in non-manipulator transform for years now).
|
|
|
|
drawing in the group row.
|
|
setting image user correct. Also found that this texface rendering was
not threadsafe for > 2 threads.
|
|
|