[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: debian med packages in the ELIXIR registry

Hi Steffen,

On Thu, Feb 05, 2015 at 10:48:50AM +0100, "Steffen Möller" wrote:
> > I actually was considering to split up the (IMHO currently way to
> > large) med-bio into say:
> > 
> >     bio-alignment
> >     bio-analysis
> >     bio-annotation
> >     bio-classification
> >     ...
> > to match the "Operation" category of Edam[1] or also possibly
> > 
> >     bio-data-resources
> >     bio-nucleic
> >     bio-protein
> >     bio-sequencing
> >     bio-structure
> >     bio-phylogenetics
> >     ...
> > 
> > the Edam topics[2] (since tasks and topics might be the best match).
> Very reasonable. I cannot tell which would be more appropiate.

I hope to find some time for coding Blends issue at DebCamp / DebConf.
Once I will be able to implement a plan (I think from now ten years ago)
to "resolve metapackages dependencies" on the tasks pages (which came a
lot closer at hand since last years sprint) I might come up with some
sensible implementation for a pure Debian Bio Blend containing those
fine grained Bio tasks while Debian Med simply stays with med-bio.
> > You might remember that I was seeking for a sensible split of the now
> > more than > 150 dependencies in med-bio for a long time and it might be
> > that Edam gives some good reason here.
> Liking it.
> The assignment for the blend metapackages is a static one, but would
> you fancy task pages that allow for two different views on the same
> packages - one for topics and (upon press of some magic button) the
> switch to an operation-based view?

This goes beyond the current technique and I have no idea how to
implement this for the moment.
> @Kristoffer,Jon,Hervé,Matus et al. - our website adopting your ontology
> based approach please consider to interpret as an early success for
> your dissemination work.

BTW, if we really establish the ontology specification for *all* our
packages - which is *work* I can not do since I'm simply lacking the
necessary knowledge - we could base our tasks design also on this data.
While this is looking far into the future I'd consider this a nice
application of the edam data for Debian Med.

Kind regards



Reply to: