CGAL users discussion list

Peter:

The executive summary is that I am able to display
.nef3 files under OpenGL using the GLU tessellator
to tessellate all of the appropriate facets.

> > I took the basic cube.off file from Nef_3/demo and
> > converted it to cube.nef3. It is attached below
> > at the end of this message.
> >
> > At the facet[0] (marked as A), it references volume[0]
> > and shalfedge[28] (marked as B). shalfedge[28] back
> > references facet[0] (which is correct) and sface[8]
> > (marked as C). However sface[8] references volume[1]
> > which does not match the volume[0] referenced in facet[0].
> > I am confused.
> 
> Yes, that's actually correct and I agree that it is odd. I think there
> was a reason for it, but I cannot remember. But that's the only odd
> thing of the data structure.

OK.  I just wanted to make sure that I wasn't missing
anything fundamental.

> > Along similar lines. I am having some difficulties
> > trying to figure out what all of the volumes are for.
> > For a simple cube I was expecting a single volume,
> > instead there are two. Is one the inverse of the other?
> >
> > I took a large cube and subtracted a smaller cube and
> > wound up with a .nef3 file with 3 volumes in it.
> > Volume 1 strictly contained facets from the outer
> > cube. Volume 2 contained facets from the both the
> > inner and outer cubes. Volume 3 contained facets
> > from just inner cube. Is it the case that the union
> > of all three volumes would equal R^3?
> 
> The volumes, facets, edges and vertices comprise a decomposition of the
> R^3. The mark decides whether a feature is part of the polyhedron or
> not. This way, we can model open and closed polyhedra, and we can model
> the cube and its complement.

That makes perfect sense!  Why didn't I see that?
That means computing the complement is probably just
a matter of sweeping through the data structure and
flipping the appropriate mark bits.  That is sweat!

I have written down some notes on .nef3 format.  Would
it be useful for me to post them to the list?

Having a solids package that supports boolean operations
that does choke is going to be a real pleasure to work
with.

Thanks,

Peter