[Debwebid-discuss] rdf coordination

[chrysn]

On Wed, Aug 12, 2015 at 06:55:47PM +0200, Olivier Berger wrote:
> Any pointer on that suggestion, for context and possible discussions on
> other aspects of the question ?

One aspect would be how we can offer the data we're accumulating. Iain
has mentioned issues with the dataset growing out of hand, so how can we
serve it? Should we at all collect it in a single place, or have (in the
spirit of the diverse services that are not UDD) the information spread
breadcrumb style over the services? Can we, then, still offer it as a
dump download and/or a SPARQL endpoint?

Once we collect the information, what can we offer (at least in the
style of a demo) where the linked data nature of it can be leveraged?
(Everything that only queries Debian internal data can be queried from
UDD as well, and SPARQL might or might not be a more accessible approach
than SQL on UDD). Which systems can we mash it up with? PyPI?

> I've just added a pointer to our paper that discusses some of the
> rationale, ideas, potential for such efforts.
> 
> > * set up tags for semantic tools with enrico in debtags, and go through
> >   everything in the archive that looks like rdf and tag it.
> >
> 
> Hmmm... whereas "semantic data" probably makes sense, it seems to me
> that nowadays, the "Semantic Web" discourse if more generally coined as
> "Linked Data"... it's mainly a marketing issue, since the concepts are
> quite the same, AFAIU.

I thought about both, but at the level of abstraction the works-with,
IMO semantic data fits better. It might also include semantic data tools
that are not URI/RDF based; those tools would be recognized by not being
tagged works-with-format::rdf:* but some other data format.

> I'll be in Heidelberg between 18th and 21st, and would love to discuss /
> hack on these topics.

Great :-)

> Btw, I've seen that Matthias Klumpp intends to present efforts on
> AppStream in Debian at the Debconf (see
> http://blog.tenstral.net/2015/08/appstreamdep-11-for-everyone-beta.html
> ) and I feel like there's some overlap between metadata in AppStream and
> Linked Data / Semantic Web which is mainly a means to more
> interoperability for me.

Yeah, there might be an overlap.

I was not very convinced of appstream when I had a first look at it (I
even roughly remember having had a discussion on it about what's present
in the metadata's copyright statement, which I didn't follow any further
after it was suggested that it could be stored in an XML comment).

Given that it's still a thing, I should probably have another look.


Best regards
chrysn

-- 
To use raw power is to make yourself infinitely vulnerable to greater powers.
  -- Bene Gesserit axiom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: Digital signature
URL: <http://lists.alioth.debian.org/pipermail/debwebid-discuss/attachments/20150812/795a64a4/attachment-0001.sig>