[Debian-med-packaging] Debian Med! (Was: Debian Med?)

[Tim Booth]

Hi,

> > In terms of Debian vs. Ubuntu there are various differences between the
> > two systems that mean you can't necessarily package for Ubuntu and
> > assume everything is right for Debian - but, as Andreas says, you are
> > not the first person to come across this issue.
> 
> As far as I know about those differences it is because we might be a bit
> picky about lintian issues and such things.  My guess is that the
> differences which are needed in Precise would be quite small (I'd be
> keen to know if I'm wrong here.)

There are few outright incompatibilities but there are a few things that
matter for Ubuntu but not for Debian and vice versa - for example I've
been tweaking various .desktop files to work nicely in the Unity
environment (eg for ARB).  I've not committed anything back to SVN yet
as I know Unity isn't a specific issue for Debian but I will, so as to
keep everything neat.
On the other hand, I think Debian still wants me to provide entries
under /etc/menu which I'd not normally bother with for an Ubuntu
package.

> BTW, I could imagine some autobuild process for BioLinux.  I know that
> NeuroDebian team does autobuild packages for a set of Debian and Ubuntu
> releases.  You might like to talk to them to create BioLinux releases
> more effective.  These people are lurking on Debian Med mailing list.

I'm spending more time right now on "problem solving" than on repetitive
tasks, so auto-building is a nice-to-have but not something that would
speed up the arrival of BL7.  I am, however, trying to do things in a
way that can be automated - eg. using build scripts and packages rather
than making ad-hoc changes.

> > I've just reached "M" on my TODO list of packages so Qiime is right in
> > my sights for today.  I have a backported version for Ubuntu Lucid that
> > incorporates various fixes but is totally forked from the version in
> > Debian Med (as Lucid doens't support the python2 build helper).  Now I'm
> > working on Precise I'm planning to grab the latest version from Alioth
> > SVN and then see if I need to apply any fixes to it to get the same
> > functionality as I had before.  If I see the fixes are also relevant to
> > Debian I'll work with Charles to merge them back into Debian.
> 
> Hint: The new package recommends python-biom-format which is for sure in
> Debian unstable - you might need to take this over as well. 

Cheers - Soon also mentioned that to me as he's been using the new Qiime
(which won't run on BL6 due to needing updated Python) recently.

> Arb is as well in Debian non-free.  It might make sense to settle with
> one single packaging.  I admit I did some things years ago which I would
> not do these days.  Please have a look into this and change whatever you
> feel worth changing.  (BTW, arb is my most hated Debian Med package -
> I never was able to simply build it but always had to deal with certain
> issues and patches. :-()

I'll have some minor fixes to push but I'm not changing your overall
setup - it was clearly a heroic effort given how wacky the whole
codebase is.
 
> It would be worth checking.  We should try to reach a zero diff between
> Debian Med and BioLinux packaging code if possible.  Perhaps nagging
> some upstream about freeing their stuff would be helpful as well.

I'll see what I can do.  One obvious change would be to change the
default clusterer from uClust to a free alternative.  Qiime does support
this, but I need to consult the developers on how sensible it is to set
this by default.

Cheers,

TIM

-- 
Tim Booth <tbooth at ceh.ac.uk>
NERC Environmental Bioinformatics Centre 

Centre for Ecology and Hydrology
Maclean Bldg, Benson Lane
Crowmarsh Gifford
Wallingford, England
OX10 8BB 

http://nebc.nerc.ac.uk
+44 1491 69 2705
-- 
This message (and any attachments) is for the recipient only. NERC
is subject to the Freedom of Information Act 2000 and the contents
of this email and any reply you make may be disclosed by NERC unless
it is exempt from release under the Act. Any material supplied to
NERC may be stored in an electronic records management system.