On Tue, Sep 8, 2009 at 2:05 PM, osman <[email protected]> wrote: > > On Tue, 2009-09-08 at 09:34 -0400, Jonathan Guyer wrote: > > > > On Sep 7, 2009, at 9:48 PM, Mike Atkins wrote: > > > > > I get a segfault when I run this test as well. It seems to happen > > > when solve/sweep is called with certain solvers, so my guess is > > > there is something wrong with pysparse we have. I installed the > > > pysparse package that was in the Debian Lenny repository, and since > > > ubuntu is based on Debian, I'll bet we are using the same > > > package(1.01-5) for pysparse. That's my idea, but I've no idea on > > > fixing it. > > > > PySparse 1.0.1 is pretty old (mid 2007). The current release (from > > about a week ago) is 1.1. I don't know what's involved in replacing or > > overriding your package manager's version, though. > > > > > > Thanks Jonathan. On my 32 bit laptop, apt-get remove python-sparse , and > building one from an svn copy, installing, worked. No failures, all 476 > tests passed. > Maybe a note to ubuntu users should be in the FAQ? >
I think this bug (http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=526527) is the same bug we've discovered. It appears that it's just a matter of time until the package with the bug fix is put into the stable repository. > Thanks again > -osman > > > >
