On 10/23/2012 08:02 PM, Lars Buitinck wrote: > 2012/10/23 Gael Varoquaux <[email protected]>: >> the current master segfaults at import, which I find a very serious problem. >> It segfaulted the buildbot, and I can reproduce the problem under a computer >> with Python 2.7 (interstingly not Python 2.6). I spent 3 hours on this today >> but I must really switch to other urgent matters. > I got segfaults with 2.6, so that's not it. They don't occur > consistently though. As I already posted in the issue tracker, I got > them at IPython shutdown as well as in nosetests. > >> I could really use some help here: having the master is this state is very >> bad for the health of the project. Reverting the corresponding commits is >> not something easy, as we would have to revert a long list of commits in a >> pull request. > Some proper Git magic takes care of that. The commit at > https://github.com/larsmans/scikit-learn/tree/revert-euclidean-speedup > reverts the entire merge in one commit, that can later be un-reverted > easily once we've tackled the problem. Feel free to pull it into > master if a solution cannot be found quickly. > +1 for reversing. It seems I was to quick with the merge - Sorry about that.
Another issue came up for me: I could again reproduce my linker error on my other box. The problem is that the blas3 that is used in the euclidean distance computation is not included in the cblas we ship. I guess that means we have to complete the cblas we ship. ------------------------------------------------------------------------------ Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_sfd2d_oct _______________________________________________ Scikit-learn-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
