CGAL users discussion list

["Wayne C. Gramlich" <>]

All:

I think I figured out what the problem is.  When I feed
the original data into nef_3_homogeneous, I use .off
format.  For one of my structures, it is a bunch of
quadralaterals in a loop with top and bottom capping
polygon.  I think that there were some small errors
in the quadralateral points that caused all four points
to not be in a plane.  When I switched from quadralaterals
to triangles, the segmentation faults went away.

As I say, I have found Nef_3 to remarkably solid code.

-Wayne