CGAL users discussion list

["Wesley Smith" <>]

Given the kighly parametric design of CGAL, I was wondering how
trivial/nontrivial it would be to reparametrize the 2D cartesian
kernal as a toroidal space in both dimensions.  The best possible
answer would be that I would only have to change the equality measure
of Point_2, but somehow I doubt it's that simple.  My ultimate goal is
to implement 2D geometry constructions such as voronoi on the surface
of a sphere where the 2 axes are angles for lattitude and longitude.
Any thoughts from people who have worked in the various kernels as to
the difficulty and steps required?

thanks,
wes