Blender Git Commits

Blender Git "master" branch commits.

Page: 5325 / 5574

October 10, 2005, 18:57 (GMT)
Ack! Recently the get_ipocurve was added in the game engine, for baking
physics. This creates very bad dependencies... tsk!

comments indicate it though;

//quick hack
//todo, before 2.38/2.40 release, Erwin

Hrms... :)
October 10, 2005, 18:21 (GMT)
added editipo_lib.c editipo_mods.c to project
October 10, 2005, 18:05 (GMT)
Stupid me! Committed in wrong console with wrong dir... here's the rest of
all files for the Ipo/Action/NLA makeover...
October 10, 2005, 17:57 (GMT)
-mathutils fix
* cos takes -1 to +1
October 10, 2005, 17:55 (GMT)
Missing include file made file not compilable...
ALso fixed midsing bracket.
October 10, 2005, 17:42 (GMT)
Stage two of the giant animation recode project; Ipo/Action/NLA cleanup

-> Note; added 2 new c files (split editipo.c). MSVC needs upgrade.

Impatient people can check the goodies in CMS:
http://www.blender3d.org/cms/Action_and_NLA_editor.706.0.html

Most work was on trying to unwind the spaghetti for editing ipos. Too much
history and bad design got added here. Most evident changes:
- made generic 'context' for detecting which Ipo is being edited, or to
assign ipos or to retrieve ipo curves.
- made generic insertkey() for all ipo types, including actions
- shuffled a lot of code around to make things more logical. Also made
sure local functions are not exported

It is far from ready... when action/nla was added in Blender, a lot of
duplicate code was generated. That's for another time.

Now the goodies;
- made Actions to allow any Ipo type
- made NLA to define active actions, for Action window too
- corrected timing for active action, so it shows the 'real time', as
defined in NLA editor.

I did update python code, but that would require testing. Testing is
needed for this commit in general, too many changes happened on all
levels of the animation system. :)
Will keep track of all reports this evening, hopefully it doesnt break
the pre-release schedule!
October 10, 2005, 17:27 (GMT)
msvc6 project file suite complete again
October 10, 2005, 15:34 (GMT)
Projectfile update
* added new file to elbeem project
October 10, 2005, 15:28 (GMT)
Patch for mathutils
* fixed another issue with anglebetweenvecs
Revision 7095d8a by Nils Thuerey
October 10, 2005, 11:31 (GMT)
- minor solver fix for msvc compilation
- added pop up menues for fluid sim bake with no
fluid objects or without valid output dir settings
- fixed 3d view redraw for fluidsim buttons
Revision d3d8441 by Nils Thuerey
October 10, 2005, 08:52 (GMT)
- fixed fluidsim inflow speed init
- fixed intersecting fluid objects init
- added MSVC6 fixes from bjornmose (thanks!)
Revision 328eeaf by Nils Thuerey
October 10, 2005, 06:59 (GMT)
- added option to switch off compiling elbeem to scons files (set USE_FLUIDSIM=true)
in this case only the new blenderdummy.cpp and utilities.cpp have to be compiled
- restructured gui:
* domain options split up into 2 sections
* added compressibility and refinement settings
* added inflow/outflow object types
- increased progress bar by 1
October 9, 2005, 21:01 (GMT)
- silly test inversion, thanks to Stephan Kassemeyer for noticing
October 9, 2005, 18:47 (GMT)
- bug fix, paste clipboard on windows didn't free buffer
October 9, 2005, 18:38 (GMT)
- bug fix #2940, at some point assumption that mesh_get_derived_...
returned NULL in editmode was invalidated which caused a crash in
particles (was freeing subsurf cache). I can't remember but hope
that I didn't intentionally break this assumption for some
reason. :/
October 9, 2005, 18:18 (GMT)
- bug fix for #2948, made ccgSubSurf bail (completely) when it encounters
an inconsistent mesh. I would really like this to be corrected at some
other stage, probably on loading, but at least things shouldn't crash
now.
October 9, 2005, 16:59 (GMT)
o another warning fix for msvc
October 9, 2005, 16:57 (GMT)
o some warning fixes (compare of int to NULL, missing prototype)
October 8, 2005, 22:40 (GMT)
make compiling with free ms vc7 toolkit easier
hum .. do we need that ODBC .. looks like no
October 7, 2005, 15:24 (GMT)
-fix for AngleBetweenVecs
* correctly finds zero length vectors now
By: Miika HämäläinenLast update: Nov-07-2014 14:18MiikaHweb | 2003-2021