On May 14, 1:18 pm, "Nicolas M. Thiery" <nicolas.thi...@u-psud.fr>
wrote:
> On Thu, May 14, 2009 at 08:34:56PM +0100, Martin Albrecht wrote:
Hi,
> > > * Solaris: I finally *fixed* the symmetrica issues and all it took
> > > was 6 hours of staring at disgusting code.
>
> > Congratulations!
Thanks.
> And thanks so much for the investment you did into this.
Well, there are so many things fundamentally wrong with that code I
would rather get rid of it sooner than later, i.e. I *really* don't
want to port this mess to MSVC. And here the main problem is not so
much compiling the code (as usual), but to have it eventually link
into Sage and actually work.
My main gripe is simple that every common design choice was ignored in
that code. It is old, i.e. it goes back to at least 1991 from comments
I read in the code and it seems to do things well in certain cases.
Most function names and a lot of comments are in German (which is not
a problem for me, but many other potential coders in the Sage project)
and upstream is plain and simply not interested in cooperation it
appears.
> I very much hope Alex will get your fixes into Symmetrica.
Since I have never gotten a reply from him I doubt I will try to
contact him again any time soon. I have so many other things to do
where I have a positive impact and given that it was 4 years between
releases this isn't exactly one I have to keep an eye on. The patch I
wrote trivially rebases and even if it doesn't recreating it is an
hours work at most.
The main question is whether this code base is fixable, i.e. the
avoidance of doing headers and just declaring *everything* as extern
in def.h is the main gripe I have at the moment. The missing test
suite (I can now take all the examples from the doc files and write a
'make check' target easily if I wanted to) is something trivial to
fix. There is also a giant leak in Jack polynomials, i.e. about 170KB
for the example in the Sage docstring. That is the second largest leak
in Sage after one in padic L functions that I will loudly complain
about at SD16 unless I find the time to fix it myself :).
In the end I know that there are bits in Symmetrica that are useful to
the combinat project and not available anywhere else. But we have to
weigh short term gain with long term pain and I doubt you will find
any sucker willing to work on the Symmetrica codebase and for
symmetric functions there is a much better maintained alternative
codebase that is licensed the right way and takes relatively little
work to get the same functionality we currently have. And maybe it is
even faster :)
> Cheers,
> Nicolas
> --
Cheers,
Michael
> Nicolas M. Thiéry "Isil" <nthi...@users.sf.net>http://Nicolas.Thiery.name/
--~--~---------~--~----~------------~-------~--~----~
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
-~----------~----~----~----~------~----~------~--~---
