[geos-devel] How robust is GEOS?

Norman Vine nhv at cape.com
Fri May 6 12:26:53 EDT 2005


Martin Davis writes:

> This still doesn't produce the ultimate desired result of computing intersections to the full 56 bits of double precision,
however.
> We're working on a solution for that - will post it if and when it's done.

I am sure you are familiar with this technique

 http://www-2.cs.cmu.edu/~quake/robust.html

Cheers

Norman




More information about the geos-devel mailing list