Le mercredi 08 février, Dr. David Kirkby a écrit:
> On 02/ 8/12 10:18 AM, Julien Puydt wrote:
> > Le samedi 04 février, Julien Puydt a écrit:
> >> After I compile mp.c with only the -O switch, pari and the rest of
> >> sage compiled correctly.
> >
> > Amusing remark :
> >
> > sage: float(maxima('1e+17'))
> > 1e+17
> >
> > One problem down... It's a shame others came up...
> >
> > Snark on #sagemath
> >
>
>
> Any attempt to port sage to another platform seems to bring up
> endless issues. It took me several years of my spare time to get Sage
> building on Solaris. Prior to my efforts, someone else worked on it
> full time for two years, though he never got very far with it.
>
> I've looked at building Sage on both AIX and HP-UX, but the issues
> are formidable. (In the case of HP-UX, there is probably little point
> anyway, but IBM make some very fast machines which run AIX).
>
> I think your ARM port will be less hassle than an AIX port!!!!
>
ARM is a plaftorm which is at the tip of the bleeding edge in some
sense, so gcc gets much work, and especially optimisation work.
Which also means they sometimes let bugs slip in the compiler, and
the buggy compiler slip out. I have the ubuntu gcc team and the linaro
gcc team on it, so I hope that one will go away soon. I shook quite a
few trees...
I consider portability of a piece of code a good indicator of its
quality : bad code doesn't port.
Snark on #sagemath
--
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel+unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org
