On Mon, Feb 10, 2014 at 4:32 PM, luc <[email protected]> wrote: > Le 2014-02-10 16:05, Thomas Neidhart a écrit : > > Hi Luc, >> >> it is on my queue of things to do. >> I already started to look at qhull (www.qhull.org) which is under a very >> permissive license and contains algorithms for convex hull, voronoi and >> triangulation for arbitrary dimensions. >> > > This seems nice. I did not find the license on the web page (did not > download > the source yet).
http://www.qhull.org/COPYING.txt IMO, it is similar to a Public Domain license, which only requires that the copyright notices are kept in re-distributed or derived code. > > >> Implementing the other convex hull algorithms were my first steps to get >> accustomed to geometry algorithms, and my plan was to finally port qhull >> to >> java, but this will surely take some time. >> Focusing on the 3D case for convex hulls should be feasible though, >> depending on when you need this feature, do you need it to be included in >> the 3.3 release? >> > > Well, ideally, yes, but this if you have not already started it, I'll have > an interim solution. The tests I have recently done with enclosing > spherical > on the sphere show that due to the phere topology, the Welzl algorithms > fails > to find the smallest cap in many cases. So I will remove > SphericalCapGenerator > and will implement getEnclosingCap using another trick. > > Don't hurry to do it, I have a way to compute the enclosing cap, even if in > extreme cases it will not be the smallest possible. So this mean I can > wait after 3.3. > No problem, I also do not want to give the impression that I am the only one who can work on this. In fact I would be happy to collaborate with others as certainly 3D is not my main area of expertise ;-) Thomas
