Since I haven't seen any updates on the work to build a Debian package
for Sage, here is a summary of the past discussions that I believe
is up-to-date:
- I was mistaken about gdmodule: it is included in the Debian package
python-gd. So no work is needed there. Good.
- The Cephes Mathematical Library is only needed for Cygwin (Windows)
and so is not required for Debian (William Stein).
- f2c is only needed for Scipy and therefore should not be required
for a Debian package because Debian already has Scipy (François Bissey,
William Stein).
- flintqs / SIMPQS issue. FlintQS should probably be removed as a SPKG
in Sage, but it may require work (qsieve). Bill Hart may be able to
comment further (William Stein).
- opencdk / gnutls. William Stein thinks it is only needed to "provide
a secure SSL mode for the Sage notebook". So we can probably use
the Debian version of gnutls, but that will require assessment.
- Issues Identified by François Bissey:
* Scipy 0.6.7, which is in sid (Debian's development branch), will
break things, at least because Sage's scipy requires this patch
http://projects.scipy.org/scipy/changeset/5790
* Sage does not yet support scipy 0.8 because it depends on numpy-1.4
and "Sage is not going there yet". Strangely, Debian sid currently
has scipy 0.7.2 (good), but includes numpy 1.4.1 (ut-oh). So that
will probably create issues.
* This pickles patch is still not integrated into Python upstream,
but it is needed for Sage: http://bugs.python.org/issue7689
* This python2.6 issue probably affects sid:
http://bugs.python.org/issue7491 which breaks stuff, witness:
http://github.com/cschwan/sage-on-gentoo/issues#issue/1
* pexpect issue (Debian pkg: python-pexpect): Debian ships version 2.3,
but Sage requires version 2.0 (though one of the necessary patches
is already in 2.4). Plotting in the notebook is likely to break with
newer versions that are in Debian; and >2.0 is reported to be slow.
* ecl is OK unless compiled with Unicode
support (http://github.com/cschwan/sage-on-gentoo/issues/closed#issue/2).
Debian may do that, so it could be an issue.
- There are still two unpackaged dependencies: ratpoints and cliquer.
- There are at least five Debian packages that should be updated to newer
upstream versions: libflint, libfplll0, gfan, lcalc, m4ri.
There are some general issues mentioned on this page:
http://wiki.debconf.org/wiki/Debconf10/Unofficial/Talks/MathematicalSoftware
There was a 10-15 minute discussion on packaging
Sage at DebConf10. Video is here:
http://penta.debconf.org/dc10_schedule/events/557.en.html
At a later talk, David Bremner gave an eloquent
summary of the challenges in packaging Sage in Debian:
http://penta.debconf.org/dc10_schedule/events/604.en.html
Giovanni Mascellani kindly put the old, buggy Debian Sage source code
into git: git://git.debian.org/git/debian-science/packages/sagemath.git
So we can build on Tim Abbott's prior work.
Did I miss any other issues?
So we need volunteers to help on at least the seven packages that are
dependencies of Sage.
It might be helpful to maintain this "list of issues for Sage in Debian"
so that it can be edited collectively. Where? How?
We also need volunteers to coordinate between the Sage, Upstream, and
Debian package maintainers to resolve each of the documented issues.
Finally, it is still not clear to me who is on the team working to
build a Debian Sage package. Is everyone interested on the debian-sage
google group? Is there another place where the team should congregate
to coordinate the work?
On Mon, Aug 09, 2010 at 08:17:43PM -0400, CJ Fearnley wrote:
> On Fri, Aug 06, 2010 at 05:44:32PM -0400, kamaraju kusumanchi wrote:
> > But now the situation is a bit different. Are we sure that we have all
> > the deps of sagemath packaged into Debian? If the answer is yes, then
> > I am happy to start with 4.5 right away.
>
> I spent some time analyzing the components of sage-4.5.2 which was released
> recently. I looked at http://www.sagemath.org/packages/standard/ and
> http://www.sagemath.org/links-components.html and used packages.debian.org to
> see what we have available. Here is my first cut assessment:
>
> * eclib is Sage's packaging of mwrank. Neither eclib nor mwrank are
> in Debian. Upstream has not changed in ages. I do not know of any
> other software that depends on eclib besides sage. So we /might/
> be able to use sage's code/packaging and skip packaging it for Debian.
> o http://www.sagemath.org/packages/standard/eclib-20080310.p10.txt
> * Cephes Mathematical Library. Not in Debian. I cannot find any licensing
> statement upstream nor in Sage's SPKG. I think it is free, but
> debian-{policy,legal} will complain if we try to package this.
> Maybe we should lobby for its removal from sage? Or maybe the license
> issues can be resolved? Or Debian's sage may just need to skip it?
> o http://www.sagemath.org/packages/standard/cephes-2.8.txt
> * Cliquer. Needs a Debian package.
> o http://www.sagemath.org/packages/standard/cliquer-1.2.p5.txt
> * Data files. I think we can ship the following SPKG's as part of Sage and
> do not need to package them for Debian, since they appear to be data files
> only:
> o http://www.sagemath.org/packages/standard/conway_polynomials-0.2.txt
> o http://www.sagemath.org/packages/standard/elliptic_curves-0.1.txt
> o http://www.sagemath.org/packages/standard/examples-4.5.2.txt
> o http://www.sagemath.org/packages/standard/extcode-4.5.2.txt
> o http://www.sagemath.org/packages/standard/graphs-20070722.p1.txt
> o http://www.sagemath.org/packages/standard/polytopes_db-20100210.txt
> * extcode: a miscellany. Appears to have Tim's debian subdirectory for
> building the package, jsMath, and tons of other stuff. Most (all?) of
> it is not a problem (jsMath is already in Debian, but which version
> is included in Sage is not clear to me), but some things may require
> finding upstream and building a Debian package. More work needed.
> o http://www.sagemath.org/packages/standard/extcode-4.5.2.txt
> * f2c: Debian (20090411-1+b1) is way newer than Sage (20070816.p2).
> We might need to push Sage to upgrade.
> o http://www.sagemath.org/packages/standard/f2c-20070816.p2.txt
> * flintqs (SIMPQS): this seems to be part of flint (in Debian but old
> version, see below) but Sage distributes it as a separate SPKG.
> Uggh, my brain hurts.
> o http://www.sagemath.org/packages/standard/flintqs-20070817.p5.txt
> * gdmodule. Needs a Debian package.
> o http://www.sagemath.org/packages/standard/gdmodule-0.56.p7.txt
> * opencdk. This ships as part of gnutls. However, sage's fork differs
> quite a bit from the code in gnutls. Do we need a package or is it
> part of gnutls with no work needed? I don't know.
> o http://www.sagemath.org/packages/standard/opencdk-0.6.6.p5.txt
> * ratpoints. Needs a Debian package.
> o http://www.sagemath.org/packages/standard/ratpoints-2.1.3.p1.txt
> * rubiks. I think these three upstream "packages" are too small to
> package for Debian. I recommend using sage's source for them.
> o http://www.sagemath.org/packages/standard/rubiks-20070912.p12.txt
> * scipy_sandbox. Seems to include a few optional/experimental scipy
> packages. They do not appear to be in Debian, but are probably too small
> to warrant separate packages.
> o http://www.sagemath.org/packages/standard/scipy_sandbox-20071020.p5.txt
>
> New versions needed in Debian:
> * I submitted http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=592349
> since libflint is out of date
> * I submitted http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=592349
> since libfplll0 is out of date
> * I submitted http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=592425
> since gfan is out of date
> * I submitted http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=592426
> since lcalc is out of date
> * I submitted http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=592429
> since m4ri is out of date
> * I did not go through all dependencies comprehensively looking for
> Sage/Debian version mismatch issues. More work needed.
>
> I believe everything else is either already in Debian or is really just
> a Sage SPKG that we can and should treat as upstream source. So the
> library situation is not too bad: most of the upstreams are actively
> maintained in Debian already!
>
> I might have enough bandwidth to package gdmodule. Anyone want ratpoints
> or Cliquer? Anyone see how to handle the headaches that I identified?
--
We are on a spaceship; a beautiful one. It took billions of years to develop.
We're not going to get another. Now, how do we make this spaceship work?
-- Buckminster Fuller
CJ Fearnley | Explorer in Universe
c...@cjfearnley.com | "Dare to be Naive" -- Bucky Fuller
http://www.CJFearnley.com | http://blog.remoteresponder.net/
--
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
