CGAL users discussion list

Angelina Novacheva wrote:
> Many thanks! The v.set_point ( Point p) was what I was
> looking for.
> What problems should I expect with arithmetics? The
> rest of the TDS stays the same, right - all the
> neighbourhood information etc.?

The data structure stays the same but the coordinates of the points will 
be rounded (because of sines and cosines that are not rational).

So, when you check whether the triangulation is still valid after the 
rotation, with the new coordinates, it can very well be that the 
orientation tests will return wrong values.

(also, the perturbation technique used internally to solve degenerate 
situations is not invariant through rotations, and this might cause 
other problems)