I wrote:
>  f=.normalize center-eye 
>  s=.normalize f X normalize up 
>  u=.normalize s X f 

Why did I put in that extraneous normalize?
 
The normalize on the right exists in the documentation for gluLookAt:
http://www.opengl.org/documentation/specs/man_pages/hardcopy/GL/html/glu/lookat.html
 
But the resulting values for s and u are not normalized when eye/center/up
were not orthogonal.
 
So, to make -: work in my previous message, I normalized s and u.
 
Is this ever a significant issue, visually?  (Probably not for typical 
arguments, but
it makes a difference for pathological cases.)
 
Are glu implementations are bug-compatible with the openGl doc on this issue?
(Since glu is almost always implemented in software, I imagine implementations
always closely follow the docs.)
 
If I was slightly less lazy, I'd report this as a bug to the opengl standards 
body.
(But there's a chance they're already aware of the issue.)
 
-- 
Raul
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to