I’m considering using bezier surfaces in my terrain visualization application. I also need collision detection for the terrain. My question is now if I should use opengl’s splines and surfaces, or use my own software implementation(which would produce a triangulation out of control points), since I need a software implementation anyway to do collision detection. Are there any pros and contras speaking for each side?
The tesselated primitives are all glu, not
gl functions, and thus not hardware rendered
anyway.
Thus, you’ll do well writing your own
implementation (assuming you know how to
optimize it).
Well I dont mean glu tesselation functions, but true splines(glMap, glEvalCoord, glEvalMesh, …), which can be hardware accelerated (as far as I know NV20 will do them in hardware(on the consumer market)).