home *** CD-ROM | disk | FTP | other *** search
- In article <shfD0938M.4vu@netcom.com>, Stuart Ferguson <shf@netcom.com> wrote:
- >a 10,000 point model, that might be the bug. If you render a hundred
- >frames and find a single pixel the wrong color, that might be the bug.
-
- Almost everyone who has talked about the bug and graphics has said that
- one wrong pixel is not a fatal error. I always look at it with the
- question: What if the bug occurs during the computation of a rotation
- matrix or some computation that would affect more than a pixel? Or if
- it occurs while doing a write to the screen or some other array, under
- the right circumstances, the program will crash. These would all be
- very bad :-) and I would consider them to be fatal.
- As to whether any of the above will actually occur with any frequency
- (although I would consider just once enough cause to give Intel a call),
- I guess we will have to wait and see.
-
- J.J.
-
- --
- ___________________________________________________________________________
- J.J. Hoesing | hoesing@cs.tulane.edu | http://www.cs.tulane.edu/www/Hoesing
-
- The story drives everything.
- -John Lasseter
-
-
-
-