Hi, I'm having a weird(?) idea. As we can now program the vertex processing and pixel processing pipeline, the only fixed stage is the linear interpolation between them. I can see some kind of tendency that a large part of pixel shaders do nothing else, but something per pixel that was available per vertex only. Don't you think that if we could modify the way the interpolation works, then we could do the almost the same with a vertex shader and an interpolator? Maybe it's already specified in sharer version X, and I'm the only one who does not know about this :) What do you think? kp