[Bf-committers] State of the affairs on Os X

Jean-Luc Peurière jlp at nerim.net
Wed Jan 18 00:32:53 CET 2006

Le 17 janv. 06 à 01:14, Ben a écrit :

> lukep,
> I get the same error on windows, its related to the graphics  
> hardware. That line is where it calls the fragment shader. Unless  
> you have a card that supports opengl 2.0, the fragment shader isnt  
> supported by the hardware, so the script will fail with that message

Hmm. An ATI 9650 is normally able to do that, which is confirmed by  

Detected GL_ARB_texture_env_combine
Detected GL_ARB_texture_cube_map
Enabled GL_ARB_multitexture
Enabled GL_ARB_shader_objects
Enabled GL_ARB_vertex_shader
Detected GL_ARB_fragment_shader
Detected GL_EXT_separate_specular_color
GL_VERSION: 1.5 (1.5 ATI-1.4.18)

> Jean-Luc Peurière wrote:
>> After much tinkering (mostly useless, the solution was rather  
>> simpler than expected) i finally managed to get a working build
>> for OS X (posted in test build forum),
>> but there is still quite some problems :
>> - what works is make build with python 2.4, OpenAL works here
>> - make build with python 2.3 fails at launch in  
>> lib_link_scriptlink function (readfile.c), line 1241 in the  
>> newlibadr call
>> -scons build succeed but the new ARB stuff and GLSL dont work (i  
>> may know why, even if it is strange), openAL dont work

test build with python 2.3 posted.
doing scons tomorrow

>> In the working build, even if the GLSL stuff works (as far i can  
>> see) it spew the following error at each iteration of the main loop :
>> ValueError: GLSL Error
>> Program failed
>> Traceback (most recent call last):
>>   File "ToonShader.py", line 123, in ?
>>   File "ToonShader.py", line 115, in MainLoop
>> it can crash on exit of GE.
>> Same for Sebulba
>> also, there is a lot of warnings for the GE, especially many about  
>> C++ standard and forbidden cast
>> All tests with gcc3.3
>> will continue work to sort things.

Erwin can you make at least that the error is reported on console  
only once ?


More information about the Bf-committers mailing list