Yes bridgman, that was what I was talking... Shader-based *decode* acceleration.
Personally I think its cleaner to use it rather than something that had proprietary origins and you need to learn (some/most things) from scratch...
Deanjo, thanks for you graphs, I read them, but they've not convinced me...
I don't forget that current OpenGL support in Mesa drivers isn't as developed as it's on nVidia/ATI proprietary drivers.
Cheers!![]()


Reply With Quote

