Fernando Cacciola wrote:
5) There is a reason why CGAL doesn't provide a distance function at all but only a squared_distance, and I agree with that reason: sqrt() kills robustness and distances are usually used for comparisons, so you seldom need that.
This is one of the things I think proto can help with. I'm playing around with a simple vector/point/matrix library just to get a feel for it. proto should be able to transform expressions of the sort "length(a) > length(b)" into "dot(a,a) > dot(b,b)". I think proper use of proto can lead to high levels of expressivity combined with good optimization. -- Noah No virus found in this outgoing message. Checked by AVG. Version: 7.5.519 / Virus Database: 269.22.1/1346 - Release Date: 3/27/2008 10:03 AM