Re: [boost] Formal Review: Boost.Polygon starts today August 24, 2009

Subject: Re: [boost] Formal Review: Boost.Polygon starts today August 24, 2009
From: Simonson, Lucanus J (lucanus.j.simonson_at_[hidden])
Date: 2009-08-25 12:27:50

Andreas Fabri wrote:
> Hello,
> 
> In the BoostCon paper you make a benchmark with, among others, the
> Boolean operations 
> of CGAL, the Computational Geometry Algorithms Library.  Which of the
> following 
> two packages did you use:
> -
> http://www.cgal.org/Manual/beta/doc_html/cgal_manual/Boolean_set_operations_2/Chapter_main.html

This one.

My friend Barend Gehrels [Barend.Gehrels_at_[hidden]] also has several countires for the geographic database that CGAL fails on.  You might be able to get them from him quite easily.

> I would be glad to check your claim that CGAL failed on one of your
> data sets. 
> Are your benchmark testdrivers for the various libs under the boost
> svn, or can you 
> post them, or send  them in a private mail.

I can add them.  My colleague Gyuszi implemented the test driver and I can get it from him again since I may have deleted my copy.

> Also I am wondering if I get it right, that you always operate on
> integer coordinates of 
> bounded size? Do you give any guarantees on topolgical correctness
> when you snap 
> round intersection points to grid points.

In the benchmarks the integer coordinates are of bounded size.  The library operates on integer coordinates bounded only by the integer data type itself.

I'm not sure what you mean by topological correctness.  I guarantee to consistently follow my snap rounding policy, but if snapping changes the topology so be it.

Thanks,
Luke