CGAL users discussion list

Hi,
thanks for your prompt reply!
1)Why unique function applied to the list doesn't remove all multiple
intersections? It seems not linked with the choice of the geometrical
kernel.
If you increase the precision of the output you will see that
the points are different.
Actually I had tried to increase the precision but not sufficiently :). So
I was wondering about a correct use of the unique function (as about the
criteria I used to select all triangles, surface/volume, which can
originate multiple intersections from double counting).

2) I obtain on average only 30 intersection points for about 30000
triangles. How to establish whether the number of intersections is
"reasonable"?
What do you mean by "reasonable"?

I mean few or too many intersections. For example I could guess naively an
increase in the intersection number when decreasing the radius bound of
the mesh. I'm probably not taking into account again the orientation
effects that you mention in your answer to the question 3 which make the
output more complex.

3) I'm wrong expecting roughly uniform spacing for the intersection points
when tuning radius bound of cells and boundary triangles to obtain a
uniform mesh?
I think you are wrong. The fact that the mesh is uniform does not
imply that any line intersecting it has intersection points with
tetrahedron uniformly spaced (the mesh is not made of axis aligned cube
with the line being collinear to an axis).

Thanks again