The dependency on MTL is really absurd. The only function you use from it is mtl::mult. That is an insane amount of bloaty template baggage to be carrying around just to get one 4x4 M * V multiply that takes 3 lines of code to write.
Normally I'd say fine ok whatever at least it works. However, it actually doesn't work. MTL doesn't compile on any microsoft compiler other than MSVC6 because of a bunch of workarounds someone stuck in the code that are blindly applied to all versions of MSVC. Anyway, MTL is crap in my opinion. Some neat ideas, but typical academic abandonware. They got something sort working, tried to get people to use it, then the students working on it all graduated or got busy with other projects. MTL is not worth the hassle for ANY purpose in my opinion. Use ublas++ or anything if you really need to do some heavy linear algebra. Use simple Vec4/Mat44 classes if that's all you need.
Anyway, end of rant.
I was able to replace the MTL code in about an hour using a Vec4/Mat44 class I have lying around.
I'll post a patch over in "patches". I think the diff files get wrapped incorrectly here.
thanks for the critique on this and on the QUAD STRIPS thing. half the point of the project was for me to reteach myself how OpenGL works, albeit in an ass backwards manner. none of the software I was testing against used QUAD STRIPS so I never caught the bug (duh!).
if I put the code into sourceforge CVS and then added you as a developer, would you be down to roll in your changes to fix the QUAD strips and MTL problems?
I have in fact largely abandoned this project since I am working on totally other stuff these days, but since i guess there are still users out there it's maybe worth fixing the problems you mentioned.
Log in to post a comment.
Sign up for the SourceForge newsletter:
You seem to have CSS turned off.
Please don't fill out this field.