[sage-devel] Re: Polymake/ Are the Sage polytope constructors (too) slow?

2010-12-03 Thread Volker Braun
For the record, Polymake switched to jReality (http://www3.math.tu- berlin.de/jreality) for its 3d graphics recently. So, apart from nauty, which probably isn't that crucial to Polymake, I think its all GPL now. I saw a few glimpses of the jReality GUI and it looks really nice! Volker -- To post

[sage-devel] Re: Polymake/ Are the Sage polytope constructors (too) slow?

2010-12-03 Thread mhampton
On Dec 2, 8:52 am, jplab wrote: > I'm a student currently at Techniche Universität Berlin, so Polymake > is quite the most "popular" software when dealing with polytopes as it > was created here some years ago. I went through the threads about > Polymake in Sage-devel... if I may summarize what

[sage-devel] Re: Polymake/ Are the Sage polytope constructors (too) slow?

2010-12-03 Thread Volker Braun
On Dec 3, 7:47 am, Dima Pasechnik wrote: > You can also develop a Cython interface to cddlib, to get Sage on par > with Polymake in this regard. I don't think this is a particularly useful project. The cddlib API is, by today's standards, not very well designed. No automatic documentation. PPL is

[sage-devel] Re: Polymake/ Are the Sage polytope constructors (too) slow?

2010-12-02 Thread Dima Pasechnik
Jean-Philippe, if you feel adventurous, you might try using PPL with Sage, as Volker suggests. I understand that the basic functionality is already there. You can also develop a Cython interface to cddlib, to get Sage on par with Polymake in this regard. Dmitrii -- To post to this group, send a

[sage-devel] Re: Polymake/ Are the Sage polytope constructors (too) slow?

2010-12-02 Thread Volker Braun
My personal roadmap for Sage's Polyhedron class is: 1) wait until PPL is a standard spkg, see http://trac.sagemath.org/sage_trac/ticket/10039 Right now Sage is communicating with cddlib via virtual terminals by printing/parsing ascii text. This is obviously slow. The PPL Cython interface, which