Re: Sponsoring science packages
On Wed, Aug 17, 2005 at 05:15:09PM +0200, Daniel Leidert wrote:
> Am Mittwoch, den 17.08.2005, 16:51 +0200 schrieb Frederic Lehobey:
> > On Wed, Aug 17, 2005 at 04:29:26PM +0200, Daniel Leidert wrote:
> > > Am Mittwoch, den 17.08.2005, 15:39 +0200 schrieb Frederic Lehobey:
> > > > There are several projects (inside Debian) I know of for collective
> > > > packaging practice and review:
> > > [..]
> > > > Could not we settle something similar for science packages?
> > > > (like pkg-debian-science).
> > >
> > > On alioth? This could become a really big project or do you only want to
> > > keep the debian files there (as e.g. the pkg-java  guys do)?
> > Yes, this is what I meant. I was thinking of source+debian files as
> > ocaml maintainers do.
> Ok, now I understand. But be aware, that even the compressed source can
> be large (e.g. Jmol, gopenmol). Would this still allow a packaging
> practice, where the upstream is put in compressed form in orig-tarball
> (e.g. used by the apache maintainers team). Sorry, if this is a dumb
> question, but I am not familiar with svn and e.g. the Chemistry
> Development Kit needs such a packaging practice because of the circular
> dependency with Joelib (the tarball then contains compressed joelib, cdk
> and cdk-plugin archives).
I don't know anything about any of the version tracking systems, but
it might be possible, in a large number of cases, to not include the
.orig.tar.gz itself, but only a hyperlink to it, and the .diff.gz. In
fact, this makes some sense, because then there is no bother worrying
about a trojaned orig.