VBGamer

Right now all my D3D stuff uses arrays of D3DVERTEX structures along with arrays of indices to store object geometry. Lately I've been taking a look at vertex buffers as an alternative. While there is ample documentation on MSDN for applying both methods, it doesn't really say what each is particularily good at, and doesn't do much comparison. So I'd like to open up this thread to discuss the 2, and compare/contrast their strengths/weaknesses in different situations (mostly I'm interested in raw rendering speed here). For example: Static versus dynamic objects Hardware TnL support issues The use of untransformed, unlit vertices So guys, who's been through this before? With all the D3DIM engines in development right now, there's gotta be at least one of you out there who knows what's up with geometry representation! :-)