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)

Reply via email to