On Feb 1, 4:23 pm, Tim Abbott <tabb...@mit.edu> wrote:
> Hello all,
Hi Tim,
> As of today, Sage 3.0.5 is available in Debian sid, so you can now run
> "apt-get install sagemath" and get Sage installed on your Debian
> system[1].
Excellent.
> However, the sage package has bitrotted substantially during the 7 months
> it was hanging out in the NEW queue, so I'd not count on it to work. In
> fact, Ondrej Certik has already reported it segfaulting at startup on
> i386. Bugs are being tracked (as usual) at
> <http://bugs.debian.org/sagemath> and we're currently using
> sage-deb...@googlegroups.com for related discussion. I plan send mail
> here again once I upload a version that is believed working on x86.
Ok.
> I'm planning to begin work on upgrading to 3.3 this weekend, since I
> believe that will help with some of the problems.
>
> It looks like there are a couple of likely complications:
>
> (1) Sage 3.3 alphas seem to be using an svn revision of MPIR. Is it
> possible to use GMP instead? Or alternatively, is there a stable MPIR
> release yet?
It is possible to use GMP, but you will see doctest failures with
mathematically equivalent, but correct results. A stable MPIR release
should be out very soon, right now the only bugs fixed are the ones
that pop up when using --enable-fat on 64 bit Intel and AMD CPUs
(which never worked in GMP).
> (2) Sage seems to contain an spkg for "pynac", which is apparently a
> python-integrated gynac, where Sage is the upstream source? Does it have
> anything to do with this pynac: <http://sourceforge.net/projects/pynac/>?
No, that is different. The upstream code is in the spkg. It is also
about to be updated to 1.1.2 IIRC.
> -Tim Abbott
>
> [1] Well, if your system is running one of the architectures on which Sage
> compiles. Apparently, Sage builds on s390 but not MIPS; who'd have known?
> The apparent problem with MIPS is the use of long double functions like
> sqrtl in the Sage library (in sage/combinat/partitions_c.cc, and maybe
> nowhere else). MIPS support is quite low on my priority list, but if
> you're interested, feel free to send me patches.
Strange, when I build Sage 3.1.2 or so on a MIPS64 SiCortex box I do
not recall running into that problem. But it is no problem to use MPFR
there and so that problem will go away.
Cheers,
Michael
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to
sage-devel-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
