Andy Novocin and I have been working on flint 1.6 for about a year or maybe a year and a half. This has involved numerous visits of him to the UK and me to France. The big improvement... factoring polynomials over Z with a new algorithm due to him and Mark van Hoeij.
This is a massive project to implement. When we started I think FLINT had 80,000 lines of code. Now it has about 107,000, and still there is very little test code written for the new stuff. But the result *should* be soundly better than any other algorithm out there for any kind of polynomial over Z (there's a nice collection on the NTL website, and Mark van Hoeij and Andy Novocin have been collecting pathological examples too). The algorithm works, we've implemented it (prototype implementation really), Andy will have finished texing the (long) proof by tomorrow, and even the prototype implementation performs well, modulo the huge number of things in flint which need optimising to make this work fast. We are going to be working very hard on it at the coming Sage Days in Leiden. Shortly thereafter, we hope to release flint 1.6 which will finally have factorisation in Z[x]!! This will be the last flint 1.x.y release evah!! FLINT 1.6 will also include a nice new module written by Sebastian Pancratz ages ago (he's been waiting and waiting for us to release this, but I think his code is used directly by Sage anyway by now). That's all ready to go. Some minor documentation issues remain, everything else is ready. There'll also be some minor bug fixes for issues people have reported, including stuff to do with the makefile, which needs to be incrementally improved to properly support recent Sage build targets. So, if flint 1.x.y is going to finally be put to rest, what is to replace it? That will be the subject of my third post to sage-devel..... Bill. -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
