On 7 December 2012 13:49, Abraham Smith [Staff/Faculty [FCRH]]
<
asmi...@fordham.edu> wrote:
> Hi
Hi Abraham
> I saw that you were the author of the Sage wiki entry on Solaris/SPARC.
Yes, that is me.
> I was wondering -- do you happen to know the state of Sage 5.4 on
> Solaris? I see that the official packages are still on the 4.7
> branch, but do you know whether 5.4 will compile successfully from
> source?
I don't know for sure. All versions are checked on OpenSolaris on x86.
Last time I tried to build Sage on SPARC was about 2 months ago. It
built without any issues, but a couple of the test caused Sage to
crash. Whether the bit of code causing a problem is significant to
you, or whether it can be fixed easily I don't know. I think it was
two tests, and to be honest it is not unusual for a couple of tests to
fail on one version of some operating system.
This is the only one I can find documented
http://trac.sagemath.org/sage_trac/ticket/13151
> I ask because at Fordham, we are about to upgrade our current Sage
> server to 5.4, and it turns out that the university just freed up some
> Solaris hardware that could be re-purposed for the math department if
> it is compatible.
It should be compatible.
Sage has tended to get less testing on Solaris than Linux, but
upgrades to Solaris don't seem to cause any problems, whereas on Linux
it seems every time a new release comes out, another issue arrises. I
have the same version of Sage on versions of Solaris as old as 2005
and as recent as 2011, and it built and passed on tests. But I think
someone has recently committed a bit of code which has a bug, which is
showing up on Solaris.
> Thanks,
> Abraham Smith
> Visiting Assistant Professor
> Fordham University Mathematics
> Bronx, NY
Dave