> Don't you have until the scholl starts? Yup, which is in less than a month. This summer has sure been going fast. :( > He he, you know I want that too. Unfortunately we won't be able to > take > advantages of 3D hardware until > we have support from the drivers - and read here: made by NVidia & > ATi. > Rudolf may be able to provide us with *some* 3D futures, > but they will be few. Until we don't have support from at least > one of the > 2 manufaturers for advanced features we have > to cope with what we have. So, what do we have? We have 2D > acceleration > (blitting), acess to video memory and the ability to use it, and > AGP > FastWrites support. If you ask me, that is enough for double > buffering. > Later, when new 3D functions are implemented in drivers we can > start to > use OpenGL, AND, this would be easy because we *already* would > have > experience and a great deal of written code. One of the things I want to change for R2 is the video driver model itself, if for no other reason than to expand the number of hook functions that we can use. > Let me say where I see this going: > I see a drawing/composition engine (ala Quatz) with 4 (ATM) > interfaces: > 1) DirectInterface - pretty much what we have now, only better > 2) 2DInterface - uses 2D capabilities(blit only) BUT adds double > buffering > support by placing bitmaps in videoMem, ready for blitting. > 3) 3DInterface - uses 3D capabilities(blit, scalling, rotations, > 3D > transformations(surfaces)). Of course, double buffering is there! > 4) ShaderInterface - uses the newest 3D capabilities (pixel > shaders - for > drawing instructions) (vertex shaders - for transformations) Sounds interesting. :) > :-) This was exactly what I was thinking. It's easier to just > provide a > surface to ServerWindow. R1 will be just like R5. Not that I like > that in > this area... :-)). Me neither. --DW