Re: upstream/metadata with references to software catalogs^Wregistries?
Hello,
On 22/06/16 15:06, Olivier Sallou wrote:
> On 06/22/2016 02:12 PM, Andreas Tille wrote:
>> [For readers of Blends list: We are discussing an additional field
>> in debian/upstream/metadata ]
>>
>> On Wed, Jun 22, 2016 at 01:12:46PM +0200, Steffen Möller wrote:
>>> Our plans are to offer reference to
>>> * the bio.tools registry (obviously)
>>> * the SeqWiki (since this community feeds the bio.tools and for now we
>>> would start upload to bio.tools what is not coming from the SeqWiki)
>>>
>>> The debian/upstream/metadata file would then for bowtie see the
>>> additional lines
>>>
>>> Registry:
>>> Name: bio.tools
>>> Entry: http://bio.tools/tool/DebianMed/bowtie/1.1.1
>>> Registry:
>>> Name: SEQwiki
>>> Entry: http://seqanswers.com/wiki/Bowtie
>> Just for the Syntax: This should rather be
>>
>> Registry:
>> - Name: bio.tools
>> Entry: http://bio.tools/tool/DebianMed/bowtie/1.1.1
>> - Name: SEQwiki
>> Entry: http://seqanswers.com/wiki/Bowtie
>> This would ensure that we have only one Registry field which might
>> contain a set of values.
Matúš and I am fine with that. We then just rename it to "Registrations"
and the "Name" becomes "Registry":
Registrations:
- Registry: bio.tools
Entry: http://bio.tools/tool/DebianMed/bowtie/1.1.1
- Registry: SEQwiki
Entry: http://seqanswers.com/wiki/Bowtie
While iterating Hervé's upload script on the first packages, we found
the same issue to appear for the publications. Those are "Reference"
today, which then most likely should be "References" and the content
should then also be itemised, right? For instance we had LTRsift changed
from
Homepage: http://www.zbh.uni-hamburg.de/LTRsift
Reference:
Author: Steinbiss, Sascha and Kastens, Sascha and Kurtz, Stefan
Title: >
LTRsift: a graphical user interface for semi-automatic classification
and postprocessing of de novo detected LTR retrotransposons.
Journal: Mobile DNA
Year: 2012
Volume: 3
Number: 1
Pages: 18
DOI: 10.1186/1759-8753-3-18
PMID: 23131050
URL: http://www.mobilednajournal.com/content/3/1/18/abstract
eprint: http://www.mobilednajournal.com/content/pdf/1759-8753-3-18.pdf
Repository: http://github.com/satta/ltrsift
to
Homepage: http://www.zbh.uni-hamburg.de/LTRsift
Reference:
- Author: Steinbiss, Sascha and Kastens, Sascha and Kurtz, Stefan
Title: >
LTRsift: a graphical user interface for semi-automatic
classification and postprocessing of de novo detected LTR retrotransposons.
Journal: Mobile DNA
Year: 2012
Volume: 3
Number: 1
Pages: 18
DOI: 10.1186/1759-8753-3-18
PMID: 23131050
URL: http://www.mobilednajournal.com/content/3/1/18/abstract
eprint: http://www.mobilednajournal.com/content/pdf/1759-8753-3-18.pdf
Repository: http://github.com/satta/ltrsift
to avoid trouble with the parser.
Though in subversion only artemis, muscle and trimmomatic are that
pedantic. Do you have a better interpretation of the parser results?
>>
>>> The bio.tools registry has the obvious problem that the version should
>>> not be passed along. Fixable, one tends to think.
>> ... fixable at bio.tools side you mean, right?
yes
>>> The motivation for a package maintainer to add those references could be
>>> * better visibility, i.e. hope for a reverse link, so the own work is
>>> found more quickly, at least for the bio.tools this is coming
>>> * the references may guide users to additional information on how to
>>> include the package in their workflow
>> Please note: Adding these values to upstream/metadata does not make
>> things user visible automatically. We need to patch tasks pages / other
>> things to let the data show up.
> would be nice indeed to get metadata info visible directly from the web
> or via a debian command like apt-meta package_source_name ;-)
Yes. For now I am mostly interested in getting the information we have
already collected
forwarded to the ELIXIR bio.tools database - so this gets some
visibility as a start.
And then certainly I'll check out how to get our own task pages adapted.
Best,
Steffen
Reply to: