[SciPy-User] Central File Exchange for SciPy

[william ratcliff]

I've been thinking about this a bit more, so hopefully two last questions
before starting the actual prototype:

1)  There are a number of open-id packages which integrate with Django.  Has
anyone used one and if so are there any preferences?  I assume that we don't
need to support twitter, openauth, facebook, etc.

2)  I took a look at gist and it's really interesting.  If we run git on the
server, then we can also make a repository on the server.   The interesting
challenge is how to pick out a particular revision of a given repository.
 For example, suppose somebody posts an svm code snippet and we have a
series of edits and forks like:

A->B->C->D->E
        \
         F->G->H->I
                \
                  J->K

Is there an easy way to get to say K, or H directly through command line?  I
imagine that every post can be treated then as a git repository (as with
gist).  Comments will be attached to a given step (say for example J) and in
our django database, we will store the repository and the identifier (for
example J) to display that currently has the most votes for relevance.

So, the workflow would be that someone would submit a code snippet (A).
 Anyone who wants to can edit the code snippet, which will create a new view
(B) with it's own comments.  If someone things they want to work on
something that's related, but a bit further afield, they can fork (F).
Because the code can change each time, comments will follow the particular
instance of the code.  The score on forking or editing will decrease by some
fraction from the original score.  That way, it will still pop up in
searches, but if it's deemed more relevant, it will be the entry point that
people will see first.   There will be an interface so people can either go
forward or backward along the commits, or can explore the branches.  I don't
want to deal with displaying the whole structure initially.  I also don't
want to deal with merging.

Since we're primarily interested in code snippets, then only one file will
be initially supported.

Does this seem too complicated?

Thanks,
William


On Mon, Nov 1, 2010 at 11:21 PM, <josef.pktd at gmail.com> wrote:

> On Mon, Nov 1, 2010 at 10:51 PM, Robert Kern <robert.kern at gmail.com>
> wrote:
> > On Mon, Nov 1, 2010 at 19:31, Vincent Davis <vincent at vincentdavis.net>
> wrote:
> >
> >> I would prefer the discussion stay on the topic of central file exchange
> and
> >> not on licensing, but maybe this is more important to the planing than I
> >> realize.
> >
> > The license flamewar is the eigenstate of all threads concerning open
> > source software distribution.
>
> Which is still much better than specifying no license. As I discovered
> again, license statements are very scarce when you look at
> econometrics or signal processing code published on the web (outside
> of the matlab fileexchange !).
>
> Josef
> >
> > --
> > Robert Kern
> >
> > "I have come to believe that the whole world is an enigma, a harmless
> > enigma that is made terrible by our own mad attempt to interpret it as
> > though it had an underlying truth."
> >   -- Umberto Eco
> > _______________________________________________
> > SciPy-User mailing list
> > SciPy-User at scipy.org
> > http://mail.scipy.org/mailman/listinfo/scipy-user
> >
> _______________________________________________
> SciPy-User mailing list
> SciPy-User at scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20101102/ceb8458f/attachment.html>