CGAL users discussion list

[Ruud Vlaming <>]

BTW,

I just want to add i was able to generate two almost
identical example stl files. Both can be meshed
in the absence of feature_detection (but with Delaunay refinement)
but only one of them survives meshing with feature detection.
The samples are identical up to one edge. (I removed the
offending edge, and replaced it by one 'perpendicular' to
it.) If you are interested i can send these files (ascii-stl,
400KB) directly to you, together with the meshing criteria).
Unfortunately, i cannot post these files here.

Regards,
Ruud

On 08/10/14 10:42, Laurent Rineau (CGAL/GeometryFactory) wrote:
> Le Tuesday 07 October 2014 17:12:25 Ruud Vlaming a écrit :
> > Hey Guys,
> >
> > I get the error:
> >
> > Point p[-9.91211 -13.5246 -10.8856], dim=1 and q[-9.9571 -13.5666 -10.9184],
> > dim=1 form an edge but do not intersect ! CGAL error: assertion violation!
> > Expression : minimal_size_ > 0 || c3t3_.is_valid()
> > File       : /opt/local/include/CGAL/Mesh_3/Protect_edges_sizing_field.h
> > Line       : 370
> >
> > and stumbled on this link:
> >
> >
> > http://cgal-discuss.949826.n4.nabble.com/Problem-when-meshing-polyhedron-wi
> > th-sharp-features-td4659217.html
> >
> > Are those issues solved in CGAL 4.4_0?
> >
> > If so, what might be causing this? The stl used
> > was checked with MeshLab to be singly connected,
> > but has same very sharp features and "tunnels".
> >
> > Thank you for any assistance.
>
> There is a bug in the file <CGAL/Mesh_3/Protect_edges_sizing_field.h>, in
> CGAL-4.5-beta1. I assume that you are using that version of CGAL. The bug is
> not present in the previous released version CGAL-4.4, and it will be fixed in
> next version CGAL-4.5, coming soon.