RE : [Debian Science Workshop] Preliminary results of fdmnes packaging
Hello Andreas
> Hello,
> I would like to inform you about one result of the "Debian packaging
> hands-on tutorial" which was integrated into the Debian Science
> sprint[1] kindly hosted by ESFR[2] (thanks to the organisers). In the
> tutorial session I was busy with Rainer Wilcke and a colleague of him.
thanks for your help during this tutorial session and also to dominique Dumont.
It should not have be possible without you.
> As previosely written in private mail I uploaded preliminary packages to
> my webspace at people.d.o[3] and I now also have injected the package
> into Debian Science Git. You can build the package as follows:
> 1. gbp-clone git+ssh://<yourAliothId>@git.debian.org/git/debian-science/packages/fdmnes.git
> (if you do not have an Alioth ID yet, you could also clone via
> gbp-clone git://git.debian.org/git/debian-science/packages/fdmnes.git
> 2. cd fdmnes
> 3. git-buildpackage
> The resulting package is usially created in the directory
> ../build-area
great work :)
> TODO for ESRF:
> Easy: Those people who will do the actual packaging work on ESFR
> should be added to the Uploaders field and copyright for
> debian/* files
> Harder:
> Ask upstream author for a license of the code and the
> documentation as well as the "source" of the documentation
> (be it Word or LaTeX)
> See below for an example mail to ask for this.
> Possibly hardest:
> It seems the data files spacegroup.txt and xsect.dat
> are not created by the FDMNES author himself but rather
> by Ralf W. Grosse-Kunstleve, Laboratory of Crystallography,
> ETH Zurich, Switzerland (ralf@kristall.erdw.ethz.ch)
> For uploading the package to official Debian also a
> license for these files is needed.
> In short: The technical work to create a Debian package could usually be
> done when sitting behind your desk. The very boring legal part is
> something which is no fun and depends from the good will of others.
we also manage to build an xds[2] package which is sort of non-free :( but hey this is used on the beamlines.
XDS is free of charge for non-commercial applications and available here
for downloading. Note, that the
executables of the package will expire on December 31, 2012.
For industrial usage of XDS a license is required
(e-mail enquiry : Wolfgang.Kabsch@mpimf-heidelberg.mpg.de).
Andreas do you think that this should be part of non-free ?
> Frédéric, it would be a very good idea to add a
> Depends: fdmnes
> Pkg-URL: http://people.debian.org/~tille/packages/fdmnes/
I have a concern about the PAN Blends. andreas, is it possible to work on this Blend on a sort of staging area, which could be available on the web
a sort of work in progress. that way I should see the generated web pages and I could ask also to the ESRF guyes if this is ok for them.
there is a PANdata initiative [1], and I think that this blend should be a prototype for the distribution of PAN software in the Debian world.
the difficult point is that fields of science covered by PAN softs have also their pendant in the debian-science. The meta package could be more techno centric but I have not yet a great opinion about this.
exafs type of experiment
saxs experiments
coherent diffraction
etc...
then their is plenty of software use for the data analysis, image analysis, nano materials software etc...
TODO ESRF:
could you provide a list of all software that need to be package from your point of view on your beamlines and also the machine
this way I should use thoses information to populate the PAN blends and I should propose a meta package organisation.
> to the relevant tasks file - this should fetch all relevant information
> and puts the data accordingly onto the web sentinel page. From my
> perspective the work I planed to do on this package is done. I'll stay
> idle until I hear more from you.
thanks for this preliminary work, now the packaging effort should continue at the ESRF :)
do you think that we should coordinate with the ESRF guyes (also girls ;) about this effort on the debian-science mailing list. (it is also good for statistics ;)
but this implie that peoples of ESRF should subscribe to this mailing list. They should also obtain some packaging tips from us.
or is it better to create our own mailing list.
I prefer to work on the debian-science list as we do not know if other peoples of the debian-science mailing list should not be interested in a few packages used at the ESRF.
> Kind regards and many thanks for the invitation to ESRF (specifically
> also the interesting visit) - it was a big pleasure to me
same for me
Frederic
[1] http://pan-data.eu/
[2] http://xds.mpimf-heidelberg.mpg.de/
Reply to: