[Debichem-devel] [Pkg-scicomp-devel] Package categories

[Andreas Tille]

On Sat, 2 Aug 2008, Michael Banck wrote:

> We have the following, as far as I can tell:
>
> biology -> debian-med
> chemistry -> debichem, with some in debian-med and debian-science

ACK (for both, even if the scope of debian-med contains more than only
Biology).

> physics -> (?, right now pkg-scicomp and debian-science, AIUI)

According to my perception physics is covered not separately
but is just under the Debian Science cover.  It needs to be picked
up by a (not yet existing) Debian Physics team.  It might be born
out of the existing projects if there are people with a certain amount
of time to spend on it and who are willing to do some grunt work
to make some things happen ...

> That is for the applications.  Then there is the underlying
> numerical/mathematical support packages like blas, lapack, etc. which go
> into pkg-scicomp mostly.

Well, the underlying tools might be used in several sciences (comparable
to the typesetting tools that are used in different sciences).  A CDD
does not actually mean packaging ralated applications but rather providing
applications to the user which makes a complete working environment.
That definitely means to provide applications from other fields - you
do not necessarily package it yourself in the CDD - just tell YOUR users
what is there and were you have at least an eye uppon for them.

> My personal opinion would be to package end-user applications either in
> a specialized team/repo like debian-med or debichem, or, if no such team
> exists for that field of science (like physics, as I understand it) in
> debian-science.

Makes sense.  But there is finally no harm if some maintainer does
the job alone in a perfect quality.  Just keep an eye on it (and tools
for Debian Med to do so will be enhanced soon to enable others with
nice QA pages) and insist if something happens that harms your QA
expectations.

> Packages (mostly libraries), which are of general use
> to scientific packages would be packaged by pkg-scicomp.

Well, as I said in another thread, I did never completeyl understand
the scope of pkg-scicomp.  Perhaps it is just an effort which intended
to fix the weakness of the Debian Science team form about one year ago
when it was rather just a simple mailing list and intended to become
the technical arm.  I would be in favour of joining pkg-scicomp back
into Debian Science team.  But this is just my personal opinion ...

> At this point I assume that it is no problem in all of the above
> mentioned (alioth) teams that members from one team get added to the
> other team once we established that a particular package that person
> maintains should get moved over.

Yes.

> One issue is the DVCS the team uses, some seem to use git and some svn,
> so we should see how to collaborate here.

Yes.  I think Vcs divergence in Debian is a burden for several people -
but we will not solve this issue.  Debian Med team decided decided to stick
to SVN - but we just lost Git adictives who were unwilling to cope with
this.  So if those Git people will find a home in Debian Science I'm
fine with this.  This group is actually not particularly wrong and
the package is somehow group maintained.  So I alsways try to go the
the pragmatic way which puts a package under a reasonably working group
maintenance and leaves the person who is mainly working on it enough
freedom.  We than collect the packages in question in our tasks files
to present them for the user.  This is a clear implementation of the
fact that there is the user view onto Debian which has to be focussed
onto the packages of his very interest and the Debian developer who just
makes sure that there are high quality packages.  A CDD is just the
implementation of the missing link between both views which is done
via meta packages or the tasks pages in the web etc.

Kind regards from Argentina

         Andreas.

-- 
http://fam-tille.de