What should be done with libbio-das-lite-perl

[Andreas Tille]

Hi,

On Tue, Jan 05, 2016 at 06:57:30PM +0100, gregor herrmann wrote:
> On Tue, 05 Jan 2016 14:35:40 +0100, Andreas Tille wrote:
> 
> > >  2) Make package properly maintained by debian-med, and find some other
> > >     uploader (the existing maintainer is no longer active).
> > I updated
> >    svn://anonscm.debian.org/debian-med/trunk/packages/ensembl/libbio-das-lite-perl/trunk/
> > in any case (regarding packaging and latest upstream version).
> 
> I see "source format 3.0 (quilt)" in svn log but the commit touches
> only d/changelog.

Just forgot to commit local change. (Fixed)

> > Questions to the pkg-perl team:
> > 
> >   1. Would you prefer to maintain this package?  This would be fine
> >      for me.
> 
> As you prefer.

I just uploaded for Debian Med now not because of an actual preference
but since it simply seems to cause less work to keep it as is.  Since I
do not expect frequent changes in future it does not really matter.
Feel free to take over if you see any good reason for it (as for any
Perl package inside Debian Med maintenance).

> >   2. If we should keep it in Debian Med I have some remaining questions
> >      about the packaging:
> > 
> >      a) Upstream tarball contains a copy of itself (Bio-Das-Lite-2.11.tar.gz)
> >         How to you deal with this usually (ignore or repack)?
> 
> I'd ignore it.

OK.
  
> Or even simpler: tell dh to use M::B:
> 
> -       dh $@
> +       dh $@ --buildsystem=perl_build

I've decided for this.
 
> t/00-distribution.t is for the upstream author to check if everything
> is fine before releasing, which obviously is not the case here
> (.tar.gz in .tar.gz, forgotten Makefile); I'm not sure what's wrong
> with MANIFEST in detail but it fits the whole picture (maybe it's
> build.tap). - I'd just skip the test, and the machinery for it is
> already in d/rules.

Yep.
 
> "Complete" svn diff attached.

Thanks for your help as always and uploaded to unstable

       Andreas.

-- 
http://fam-tille.de