[ODE] Triangle Meshes
pseudosig at yahoo.com
Mon Sep 10 10:38:40 MST 2007
As soon as you suggested this, I had a flashback to
the model creation in Wings3D, when I used the bevelor
tool. When you do extreme beveling, it seems
sometimes that polygons faces can merge to a single
point and are ALSO included in the object file export.
So I checked my model against this hypothesis and it
indeed has degenerate triangles where extreme beveling
I haven't tried this in the ODE demo example, but I'm
sure this is a step in the right direction.
--- Flavien Brebion <f.brebion at vrcontext.com> wrote:
> Jeremy wrote:
> > I compiled ode with both and the behavior was
> > much the same for either. Basically my trimesh is
> > topologically detailed rounded die (your usualy 6
> > sided die). There are regions in the mesh that
> > closely located vertices. Could there be issues
> > precision with closely grouped vertices so as to
> > the demo app?
> One thing you could check are degenerated triangles;
> at many places in
> the code, a division is made by the area of a
> triangle. If the triangle
> is degenerated, it generates a division-by-zero
> error which creates an
> avalanche of problems. Just an idea.
> F. Brebion
> ODE mailing list
> ODE at ode.org
Fussy? Opinionated? Impossible to please? Perfect. Join Yahoo!'s user panel and lay it on us. http://surveylink.yahoo.com/gmrs/yahoo_panel_invite.asp?a=7
More information about the ODE