Re: [PyOpenGL-Users] GLUError and gluNurbsCurve

[Ivo U. <iv...@iu...>]

Mike C. Fletcher wrote:
> Hmm, are you saying those ints are supposed to be returned as values
> from gluNurbsCurve?  My manual says that it is supposed to have a void
> return type.  Or is this supposed to show up in a glGetError return
> value?  Basically I'm wondering whether these are supposed to be part of
> GLU or part of the wrapper?  If they are part of GLU I don't yet know
> how to retrieve them.

sorry for not responding earlier Mike
(but better ever than never :D)

No! gluNurbsCurve does not return int (it returns void).
I've tracked error through (sgi) source and it goes like this:
1) *gluNurbsCurve* calls *nurbscurve* on GLUnurbs object
2) *nurbscurve* method creates knotvector object and
calls *do_check_knots* function on that object
3) *do_check_knots* calls *knotvector::validate* method
(which is the one I sent you) and puts the result
in the status variable. If there is any error (status!=0)
*do_nurbserror(status)* is called
4) *do_nurbs_error* calls *errorHandler*
5) *errorHandler* calls *postError* with appropriate GluError

after this I'm not sure what comes next :(

> 
> For now, I've added them to the wrapper and have them controlled by
> OpenGL.ERROR_CHECKING, so you can turn them off when your code is no
> longer producing invalid values and want to reclaim the performance. 
> You should be able to check out current head via:
> 
>    bzr branch lp:pyopengl
> 
> and check that the code now catches your errors.
> 
> Thanks for the bug report,
> Mike
> 

Ok. I'll test it this weekend

have fun,
-- 
Ivo Ugrina | Jab: gro.rebbajTAanirgui
www.iugrina.com | ICQ: 47508335
------------------------------------
baza matematickih pojmova
http://baza.iugrina.com
------------------------------------
anime, manga, Japan fanzin
http://yoshi.iugrina.com