CGAL users discussion list

["Fernando Cacciola" <>]

Hi Pablo,

> I'm trying some other solution here, I suspect that this model is already
> simple enough and no triangle can be removed somehow without changing the
> shape....
Right, that's exactly the case here.
I tested it and every edge in your example is "non-collapsable".

One way to notice this is via a visitor, as shown in the second example in the 
user manual.

In fact, if you use that example you can see how it reports that none of the 
191316 edges can be collapsed due to topological constriants.
Furthermore, if you use the LindstromTurk policies with that example, you can 
see that more than 190000 edges cannot be collapsed due to cost computation 
constrians, which in this case means that any new vertex placement would incurr 
a very large error, so it discarded.

> I need some topological structure where I can query for adjacent triangles
> of a vertice / triangle or adjecent edges of a vertive / egde.. is there
> anything like this in cgal.. (I think there is)..
> any one can give-me an
> example or a doc about some structure like this? I'm currently using a CGAL
> Polyhedron to store my triangles...

You can query all adjacencies of a given element using the Polyhedron, but when 
you build a polyhedron you do the opposite: define those adjacencies rather than 
query them.
I did not investigate why none of the edges are non-collapsable, but from your 
question above: is this a real mesh or just a collection of isolated triangles? 
that is, are edges shared between triangles? If not, then you probem is how to 
build a proper mesh from the whatever input you have.
If that's the case, please top-post a question, with a proper subject, asking 
for that specifically, providing details of the input you have and what you know 
about the adjacencies of the triangles (if anything)

Btw: I couldn't use your viewer because it depends on a "osgViewer.dll" which I 
don't have.

HTH

Fernando Cacciola
GeometryFactory