Fixing UV Quad Constraint when cancelling transform. Also, transdata conversion is a bit faster for all UV transforms.
Patch by Cambo, reviewed and revised.
Fixed by only conditionally resetting the start of the triangle used to calculate the angle to when an angle is actually measured (so really small angle don't result in continuous deltas of 0 degree).
However, the smallest rotation angle is still limited by the precision of the acos function (here: 0.02 degrees, 0.02 / 30 with Shift pressed).
Added (disabled) MKV support. Might come handy later, if someone finds
out, why it instantly crashes after 2 frames.
Made it impossible to set min_rate higher than max_rate in render buttons
panel.
Add a function GameLogic.expandPath() that works like Blender.sys.expandpath() and is also available in the BlenderPlayer.
Fix the game actuator in the BlenderPlayer to work like in Blender:
- try first to load the .blend from the current working directory
- if not found, try to load from the startup .blend or runtime base directory
* Scaling NLA-strips now resets their scale setting correctly at all times
* Added safe-guards against negative scaling being created through the use of the transform tools.
Note: at scale 0 or thereabouts, there will still be a little blip, when keyframes are scaled as if scale were 1.0f. It's quite harmless.
for running these scripts outside blender.
The error looked like this.
SyntaxError: Non-ASCII character '\xe4' in file ./3ds_export.py on line 10, but no encoding declared; see http://www.python.org/peps/pep-0263.html for details
* Move armatures out of posemode before joining/separating, so that bones don't mysteriously disappear after the operation
* Added an undo push and fixed up hotkey access for armature separation
Found a little bug in the Armature Joining code due to the 'imat' var being defined twice and used for two different things. This was brought to light by Valgrind output (thanks Campbell)
1) Bug #6692 is actually a feature request. While I do think that
Michalis' patch is a worthy change, we don't have time to test
this at this stage in the release cycle (and svn is supposedly
closed to new features anyways). I'm sorry Michalis, this will
have to wait until after release.
2) I have looked at bug #8814 a while ago, but I have not yet found
a good solution (and perhaps the solution is to *not* export faces
without materials assigned). I don't have time to assess whether
revision 14774 presents a worthwhile change or not, and I have no
time to test before release. Also I don't like that the 'fix' for
#8814 is mixed with a fix for a previous faulty commit.
3) Again, another commit to a script I maintain without consulting
me first. Not cool!
Chris
