> From: Carsten Neumann [mailto:[email protected]] 
> Sent: Wednesday, July 01, 2009 09:42

[snip]

> hm, you have not really mentioned what problem the degenerate triangles 
> are causing for you, can you give us an idea what you are trying to do?
> In the case of the stitched strips degenerate tris can not be avoided, 
> in the other case they are not nice, but the hardware should just 
> discard them and since their number is low compared to real triangles.
>
> @Johannes: do you have a description of the striper (other than the code 
> ;) ), e.g. a paper about it? do you recall if it always puts a 
> degenerate tri in a strip to support stitching and where in the strip it 
> is placed (I'm wondering if the problem can be fixed simply by 
> decreasing the length of the strip by one?). Thanks!


Hi Carsten.

Actually, the only problem that the degenerate triangles have caused me is
that I thought I must be doing something wrong.  Although the on-screen
result looked correct, I suspected it wasn't robust, and I spent about a
day trying to figure out what I was doing wrong, until I finally searched
the mail archives and discovered that it was "a feature, not a bug."  I
can live with that.

Personally, I would rather see the triangle strips divided to remove the
degenerate triangles: more strips, no degenerate triangles.  From my user
perspective, that's the expected behavior.

(My current task is to marry OpenSG to a physics simulation of tearing
 cloth while preserving colors and textures across the tear.  Tears
 produce new vertices and change triangle position indices, but should not
 change color or texture indices.  The physics engine works with discrete
 triangles, so I need to decompose triangle-strips and triangle-fans into
 discrete triangles, reconfigure the index mapping to join positions and
 normals in one index and isolate colors and textures in other indices,
 and mate these triangles to the physics triangles.  Ultimately, the
 degenerate triangles disappear after they fail to mate with any physics
 triangles.)

-- 

Ted


------------------------------------------------------------------------------
_______________________________________________
Opensg-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/opensg-users

Reply via email to