[pkg-bioc] BioConductor

[Andreas Tille]

On Mon, 16 Mar 2009, Dirk Eddelbuettel wrote:

> ...

Thanks for the clarification.

> Given that we are talking about 1600 source package and a growth rate of over
> 40% per year (see my presentations page and the recent MFA talk for a chart),
> I (personally) have no intention of pushing these into Debian.  You just
> can't babysit that many packages on all the autobuilders, library
> transitions, ... or even walking them through the NEW queue.

Well, I think I see your point but I have heard rumors that the Perl
team is auto building a lot of CPAN packages.  So I wonder in how far
there is a difference in principle between CPAN and CRAN.

> | BioConductor and I have seen that there is a really interesting collection
> | of R packages at
> |
> |      http://www.bioconductor.org/packages/release/bioc/
>
> Yes, BioC is very impressive, and very widely used in Bioinformatics. Lots of
> nice things there eg for visualization.

I just realised thet the link on our tasks page is just missleading.  My
previous assumption was that BioConductor is a single package and not the
name for an effort that includes 100+x potential packages.  I have to think
about how to reflect this.

> Check out hexbin, for example. I
> think that is finally on CRAN too.

Thanks for the hint.

> There is none. See above.

Do you think it makes sense to clean up SVN and Wiki a bit?  I see no
point in keeping old unmaintained code there and outdated docs in the
Wiki which explain things in detail which finally do not work.
The only think which seems to be maintained now is r-cran-qtl and
this might fit into the Debian Med SVN as well (seems a bit hidden
under the BioC project - at least I never made out the connection
inbetween.

> The new cran2deb is on R-Forge.r-project.org, and one fellow already managed
> to set up an instance local to his lab. Now Charles and I think it is too
> early for that but heck.

I admit I'm keen on seeing a bioc2deb and I would even consider
baby sitting the resulting packages through new etc - perhaps Charles
Plessy might share some experiences from Debian Perl group how they handled
the grunt work.

Many thanks for the clarification

        Andreass.

-- 
http://fam-tille.de