I'll give it some more thought. Q1.1: I dunno. 1mil triangle = 3 mil points. If we on the begining, after transformating them on the screen coordinates (*1+1/Zw) roud them all, and then check if intX=intX, intY=intY, now we have at least 1/3 of points or even less for calculating because for picture to be homogenous at least 3 triangles have to have same point. And also this eliminates some of accumulated round-off errors. I'm not fanaticaly pursuing this just brain storming :).
Q2.1: Why do we add 0.5 to everything (A,R,G,B,LOD...). Isn't 0.5 supposed just to be a offset for screen coordinates (Xsreal and Ysreal) so we could check if they're in trapezoid or not? Q4(stupid one): Which way is the code supposed to work on the chip? In sequential mode like this: (Number stands for processed frame, 1st,2nd...) 1-0-0-0-.. 0-1-0-0-... 0-0-1-0-... 0-0-0-1-... . . . 2-0-0-0-... or in like this(seq+parralel): 0 / \ 1-0-0-0... \ / 0 1 / \ 2-1-0-0... \ / 1 2 / \ 3-2-1-0... \ / 2 . . . Sorry for drawing but i don't know enough english to explain what i mean. TIA Dushan _______________________________________________ Open-graphics mailing list [email protected] http://lists.duskglow.com/mailman/listinfo/open-graphics List service provided by Duskglow Consulting, LLC (www.duskglow.com)
