#18929: Include igraph library
-------------------------------------+-------------------------------------
Reporter: borassi | Owner:
Type: enhancement | Status: needs_info
Priority: major | Milestone: sage-6.8
Component: packages: | Resolution:
optional | Merged in:
Keywords: igraph library | Reviewers:
Authors: Michele Borassi | Work issues:
Report Upstream: N/A | Commit:
Branch: | ac2c5c39a65c1ef8e003219bf59fc9f6e35f61e1
u/borassi/include_igraph_library | Stopgaps:
Dependencies: |
-------------------------------------+-------------------------------------
Comment (by borassi):
Hello!
Replying to [comment:12 ncohen]:
> Hello again Michele,
>
> Don't worry about the many steps it takes to get this through. I also
worry that I may be a bit boring, with all these things I add one by one
to your stack because I cannot seem to figure out them all at the same
time.
>
> I installed the two packages, and things went well. I still have a
question about the igraph package: upstream is apparently a 'fake' .tar.gz
archive (it is not gzipped) -> is this why your igraph package is 'not
exactly' (up to md5sum hash) equal to upstream?
What do you mean by 'fake' .tar.gz archive? Anyway, I have downloaded
again the file igraph.tar.gz, and I have run again the `sage -sh sage-fix-
pkg-checksums` script, hoping it will solve the problem. Does it?
> While looking at the (Di)Graph constructor, I thought that something
unlikely may happen:
> 1) Somebody has a igraph object in a Sage session
> 2) igraph is not installed in Sage
> I thought that it could happen if by some chance such an object was
obtained through pickling.
>
> Perhaps it sounds stupid. This being said, I also 'fixed' something
else: importing modules is not totally free, and it was even the
bottleneck in some design code I met in the past. For this reason, it is
better to not import things in potentially critical parts of the code (the
graph constructor can be one of those, for instance when you enumerate
graphs up to isomorphism) and so I tried to make it less frequent. This
commit is located at `public/18929`, you can take/remove what you like/do
not like in it.
I have included the commit.
> Have you figured out this GPL2 vs GPLv2+ issue I mentionned in an email?
The igraph file COPYING that comes with both packages starts with:
GNU GENERAL PUBLIC LICENSE
Version 2, June 1991
So I believe it is GPLv2. However, it does not seem to be a problem:
http://www.gnu.org/licenses/gpl-faq.en.html#WhatDoesCompatMean. Moreover,
also package latte_int is distributed with Sage with GPLv2.
> Byeeeeeeeeeee,
>
> Nathann
>
> P.S.: I had some issue when installing the package, which I reported
upstream (https://github.com/igraph/igraph/issues/842)
Ouch, yes, the problem is libxml2. I have uninstalled it from my computer,
and I have re-installed igraph (without Sage): the problem still appears,
even if ./configure realizes that libxml2 is missing, because it writes
"GraphML format support -- no". Probably it is an igraph bug.
In order to solve this problem, I have added a check in the spkg-install
script, and I had to modify the sage-spkg file to allow an error which
depends on the user. Do you agree with this modification? Or do you think
I should do something less "invasive"? Because I do not like that we tell
the user to contact sage-devel if (s)he only has to install libxml2.
--
Ticket URL: <http://trac.sagemath.org/ticket/18929#comment:14>
Sage <http://www.sagemath.org>
Sage: Creating a Viable Open Source Alternative to Magma, Maple, Mathematica,
and MATLAB
--
You received this message because you are subscribed to the Google Groups
"sage-trac" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sage-trac.
For more options, visit https://groups.google.com/d/optout.
