| Age | Commit message (Collapse) | Author |
|---|
|
[#7113] GE crash pressing as soon as P on 64 bit
Note: glext.h has been removed from the source
If you get errors compiling with it you have 2 options
download/install glext.h (preferred method)
or set WITH_BF_GLEXT=false
If your a user and having problems with game engine try
setting the env var: WITHOUT_GLEXT 1
Kent
|
|
things default on for linux. ideasman helped me get scons working.
Cmake still needs some love...
Kent
|
|
systems Mesa now.
|
|
|
|
i386-64 (was just i386)
default python to 2.5
|
|
Kent
|
|
|
|
[#7660] Solaris 10 x86 support (Makefiles)
Hopefully it will not mess up anything for anyone else. I removed
some hardcoded static libs and made NAN_*_LIB definitions so they could be
overridden, to allow greater flexability.
Let me know if there are any problems/questions.
Kent
|
|
was py 2.2 which probably wont even compile anymore)
pose.c - left a print here
export obj would ask about file overwriting twice.
|
|
Updated FFMPEG to latest version. (And fixed some Tab-codes, that Emacs
didn't like in OPENEXR-related lines)
|
|
It adds read only dds support. (Writing will come later)
Kent
|
|
|
|
targets to build, the definitions file had different names from the
Makefiles, so these two were broken.
WITH_BF_BLENDERGAMEENGINE
WITH_BF_BLENDERPLAYER
Fixed now.
Kent
|
|
|
|
|
|
OpenEXR, and I switched all of the other platforms back to
no ThreadLib. I really was trying to make this easier for people not
harder, and I'm afraid I failed. Hopefully this clears things up for
people and everyone will switch to using the Makefiles. ;)
If your having problems you should be able to just add the IlmThread lib
back into your NAN_OPENEXR_LIBS inside of your user-def.mk and be good to
go. let me know if you have problems with it.
Kent
|
|
The first is it adds libIlmThread.a to the OpenEXR libs.
The second thing it does, is make it possible to define what TARGETS
you want to build inside of your user-def.mk file. This simplifies
source/Makefile quite a bit. I made each platform have the same
defaults (build dynamic blender, gameengine and gameplayer) I think
most platforms should be building this combo anyway and if you need to
change it its trivial so no need to have different defaults for each platform.
The new defines are as follows with their default settings in:
source/nan_definitions.mk
export WITH_BF_DYNAMICOPENGL ?= true
export WITH_BF_STATICOPENGL ?= false
export WITH_BF_GAMEENGINE ?= true
export WITH_BF_GAMEPLAYER ?= true
export WITH_BF_WEBPLUGIN ?= false
Let me know if you have any problems with it.
Kent
|
|
|
|
Temporarily disabled game engine for scons and Makefiles (nan_definitions?), this to upgrade to Bullet 2.x
scons for Darwin, Linux, BSD3, SunOS5, Win32, and Makefiles should not compile game engine right now.
CMake and MSVC needs still disabling of game engine.
|
|
into an icons folder in the release package on linux, freebsd and openbsd
when doing 'make release'.
|
|
Kent
|
|
- All code is in #ifdef ... #endif
- Only make build system is supported and you have to add:
export WITH_VERSE=true
to user-def.mk file
- Blender can share only mesh objects and bitmaps now
- More informations can be found at wiki:
http://mediawiki.blender.org/index.php/BlenderDev/VerseIntegrationToBlender
http://mediawiki.blender.org/index.php/BlenderDev/VerseIntegrationToBlenderUserDoc
I hope, that I didn't forget at anything
|
|
Everyone else was using gzip. On the latest downloads many of them
are using bzip2, (I think everyone using scons)
So I decided to update make release so that everyone is using
gzip2. I don't mean to step on anyones toes, I know at least some of
the platform maintainers do not use make though so feel free to change
it back and or to something different for your platform.
While I was at it I also changed default path of OPENEXR to /usr on linux.
Was suggested on mailing list by Matt.
(I'm pretty sure the linux maintainer is using scons)
Kent
|
|
- Added a new Brush datablock, only used by image paint, but intended
to be used in texture paint, vertex paint, weight paint and sculpt
mode also.
- Being a datablock, these brushes can be saved, appended and linked.
They have a fake user by default, to make sure they are saved even if
not selected.
Image Painting:
- Replaced the img module with C code in imagepaint.c
- Airbrush is no longer a separate tool, but rather an option that can
be used for soften, smear and clone also.
- Blend modes mix, add, subtract, multiply, darken and lighten have been
added, code taken directly from vertex paint.
Note to project files maintainers:
- The img module was removed from SCons and Makefiles, and this should
be done in other build systems also. I'll wait to remove the module
from cvs, to not break compilation.
|
|
|
|
- patch from Douglas with endian fixes
- Makefile adds static libiconv.a from lib/
- this version will use OpenEXR libs from lib/ too, and is latest release
with threading support
- openAL is missing, added it as default to not include it
|
|
Made internal ffmpeg-libraries link by explicitly specifying the archive-files.
(Thanks GSR for pointing this out)
Otherwise, blender always links to the system-libs.
|
|
forgot to commit nan_definition changes
|
|
|
|
export WITH_FFMPEG=true
(and be sure to update lib/windows). Test, test, test.
|
|
with NAN_FFMPEG_LIBS more general (i.e., not just for linux, and
moving some of the linux linking stuff to source/nan_definitions.mk).
|
|
If "export NAN_USE_FFMPEG_CONFIG=true" is added to user-def.mk,
the system executes the ffmeg-config program to set values
for NAN_FFMPEG (--prefix), NAN_FFMPEGLIBS (--libs avcodec avformat),
and NAN_FFMPEGCFLAGS (--cflags). Only one used so far is the
NAN_FFMPEGLIBS for linking on linux (if requested to do so).
Current default is not to do this.
|
|
|
|
(the one in doc)
|
|
always statically (you have to force it to build a dynamic library) the
resulting binary is redistributable.
The code is made ffmpeg-version independent using #ifdef's.
|
|
Here are my notes on things to look out for as potential problem
spots:
source/blender/blenkernel/intern/displist.c:
+ is initfastshade(void) supposed to be empty? I had
to make it empty to get the merged tree to compile.
source/blender/python/api2_2x/Armature.c:
+ went with the version that had Armature_getLayers()
source/blender/python/api2_2x/Object.c
+ went with the version of Object_getPose() from bf-blender.
(#ifdef 0-ed the other version)
source/blender/python/api2_2x/Pose.[ch]
+ had problems linking due to no Pose_Init() ... copied these
two files straight from bf-blender.
source/blender/src/drawview.c:
+ view3d_panel_properties() had things shifted a few things shifted
a few pixels, otherwise, things were painless
source/blender/src/splash.jpg.c:
+ went with bf-blender version (orange is dead)
source/gameengine:
+ went with bf-blender version -- does not compile due to IMB_rect* stuff,
Ton should look into this.
|
|
(with make, need to confirm with scons)
after cleaning the changes are in fact minimal, but the situation
is still quite a bit hackish.
Game engine coders, there is also quite a number of warnings that
need to be fixed.
current situation is that everything seems to work, but GLSL shaders
spew a lot of errors on console and blender may crash on exit when
a GLSL shader was used. ARB stuff works fine.
|
|
|
|
|
|
some strange merging). Unfortunately, Bullet doesn't build correctly
for gcc under cygwin (some private vs public stuff), so it is
important to be able to disable the game engine in these circumstances.
|
|
|
|
|
|
These should make it so that other people can compile with OpenEXR support.
(I also added the OPENAL fix erwin commited to bf-blender since I
need it for my machine, and this syncs up the file)
Kent
|
|
|
|
Then after cvs is unfrozen, development can continue.
|
|
python.def and upped Python version to 2.4. This means that Cygwin/GCC will
now be building Python 2.4 binaries. Thanks all!
|
|
|
|
To enable dynamic tiff support.
I had to fix some of the logic in the fileselect box for icons,
I also expanded the patch to look in various default locations for
a dynamic libtiff.so/libtiff.dll
and look at the env variable BF_TIFF_LIB if it can't find it automatically.
If unable to load the library it prints a message about setting BF_TIFF_LIB
to the console.
I haven't been able to test it on a lot of platforms but hopefully it
will just work ;) I added the files to scons but have not had a chance to
test that as well.
Kent
|
|
(which uses SDL_mixer), one thing in nan_definitions.mk slipped
through.
My apologies :)
|
|
so there will be more files following.
Anyway: NEW BOOLEANS from Google Summer of Code (Courtesy of Marc Freixas)
Known problems:
- Random freezes while using them as a modifier. This may not be directly
related to modifiers though - it's maybe just the huge number of
operations that leads to a higher probability of triggering a bug
- Static booleans (the first 3 entries in the WKEY menu) are borked
anyway, this is not due to this commit.
- Errors when exiting Blender (dupli_alloc stuff), is not related to this
commit, either.
Please test if everything works, and check the other build systems, I only
know that make works.
Also, compare the results of, say, cube-cylinder, in old and new booleans
:)
|
