Re: debian-science and science-* packages

[Frederic Lehobey <Frederic@Lehobey.net>]

Hi,

Andreas Tille <tillea@rki.de> (2007-08-28 23:29:44) :
> On Tue, 28 Aug 2007, Christian Holm Christensen wrote:
> 
> >The problem could be, that `-physics' is just a tad too broad - maybe
> >(and only maybe) it needs to be split up into parts (-nanoscience, -hep,
> >-theory, -bio, -solid, ...) with some `-physics-common' recommended
> 
> We all know that every science has several flavours.  If you want to
> reflact this in meta packages we will end up with many meta packages
> with zero to two dependant packages which just makes no sense at all.

I agree with this. We should not want to split into too many
categories and recreate with metapackages what is properly done with
debtags and facets.

Currently, for the names of the proposed science-* metapackages I have
deliberately chosen similar names as existing debtags (see
http://wiki.debian.org/DebianScience/Classification).

So to create a new metapackage for a scientific field I would suggest
to be very conservative and first convince debtags people that a facet
with this name is relevant with respect to the number of packages for
this field that are already in Debian.

For example, I am unsure yet it is even really needed to create
science-*-dev packages.

An other possible source of inspiration for the names of the
*relevant* scientific fields is the new menu categories:
http://lists.debian.org/debian-devel-announce/2007/07/msg00000.html

> >On the other hand we could also use the DebTAGS mechanism more
> >aggressively.
> 
> Sure - did you start debtagging your favourite package?

Definitely something very useful for the distribution as a whole. I am
even usertagging/debtagging the ITPs...  :-)
See http://wiki.debian.org/DebianScience/Usertags

Best regards,
Frédéric Lehobey