On Jul 15, 7:29 am, "[EMAIL PROTECTED]" <[EMAIL PROTECTED]>
wrote:
> Hi,
Hi Vincent
> Tried to answer earlier, but for some reason my answer didn't go
> through ...
Yeah, some times Google groups is a little funny in that regard.
> Anyway, if it's not too much trouble to you, I am greatly
> interested in a solaris9-sparc version, and I'm sure I'm not the only
> one (even though potentially interested people might not subscribe to
> this particular list).
I have had Solaris 9/Sparc builds in the past. Since clisp does not
compile properly on any Sparc box I ever tried I have moved to Solaris
10/Intel for now as primary port platform. Once we switch from clisp
to ecl a Solaris 9/Sparc port should be doable. There was some trouble
in Givaro due to them using fenv.h which is not present on Solaris
before and including Solaris 10, but IIRC that code has been removed.
So in short: It is likely that by mid to late August I can provide you
with such a binary, but not at the moment.
> I tried to build the thing myself, and was able to patch things up all
> the way to ATLAS, where I had to give up.
Which gcc did you use? I used a gcc 4.3.0 with GNU binutils 2.18 and
did not hit any trouble.
> Along the way, I was
> wondering: exactly how many of the packages are needed in order to
> reach the sage: prompt ? Even if 'make' does not go smoothly, some
> functionality is better than nothing I guess ... and sage as a package
> manager for its packages would still be extremely useful even if some
> of them don't build out of the box.
This is tricky and you do not to get past the Sage library stage. Then
you need Maxima, GAP and pyprocessing and Sage should start up.
> BTW, do you have building instructions written somewhere ? I am
> willing to experiment, but I would need a starting point ...
I have a bunch of rather cryptic notes which has fixes that should
mostly go into Sage 3.0.6.alpha0 which should happen in the next 24 to
48 hours. My main focus in the next two to three months are the
Solaris, native Windows and OSX 64 bit port (in order of priority :)).
As I mentioned the 3.0.5 binary works better than any other Sage
release on Solaris in the last 15 months and I tracked down a number
of issues to numpy doing bad things. At the moment 41 doctests fail of
which about 20 are trivial failures that are caused by maybe 2, 3
bugs. So things are looking up.
> Thanks for all your work in any case !
No problem. The more people play with the Solaris binary the better. I
can also offer an SSE2 Opteron build, but my access to that box will
cease while I am on the road starting in two days until I get back.
> /vincent
Cheers,
Michael
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
