[tdwg-tag] SourceForge LSID project websites broken - role for TDWG?
leebel at netspace.net.au
Sat Apr 11 02:00:39 CEST 2009
Thanks Rod. Very constructive.
From: Roderic Page [mailto:r.page at bio.gla.ac.uk]
Sent: Friday, 10 April 2009 1:30 AM
To: Lee Belbin
Cc: 'Technical Architecture Group mailing list'
Subject: Re: [tdwg-tag] SourceForge LSID project websites broken - role for
In a rare attempt at being constructive, here are a few thoughts.
LSIDs and linked data
If adoption of LSIDs proceeds, then we should make efforts to see that
they play nicely with Linked Data efforts. For example, a HTTP
resolver would need to support 303 redirects and content negotiation.
This help avoid us creating our own ghetto, but still exploit whatever
advantages LSIDs have.
Roger set up something along these lines to handle Biological
Collections Index (BioCol) LSIDs. There is a nice tool at
to check whether a URI behaves as Linked Data tools expect. Sadly
the proxied BioCol LSIDs (e.g.,
) don't validate, but this should be easy to fix. The TDWG resolver
I've implemented a simple resolver at bioGUID that returns either raw
RDF or a clumsily formatted HTML version of the XML, but which passes
the http://validator.linkeddata.org tests. An example URI is
, which validates http://tinyurl.com/cgje5n
So, my first recommendation is to ensure that a TDWG HTTP proxy passes
. This means we can play with the Semantic Web crowd with LSIDs.
Note that getting HTTP URIs to play with Linked Data is not trivial,
so whatever technology we adopt we'll need clear guidelines as to how
to use it. As an aside bioGUID can make DOIs play nice as well (they
don't by default), and Kinglsey Idehen http://www.openlinksw.com/blog/
~kidehen/ of OpenLink Software is supporting LSIDs in the Linked Data
tools he's developing.
As part of my experiment to wikify taxonomic names, literature, etc.,
I've been playing with the TDWG vocabularies. I've a few grizzles, but
in general they've been really useful, and I think these will be key
(as Donald and Lee have emphasised).
Ironically one of the examples Lee listed when defending the TDWG's
resolver (urn:lsid:gdb.org:GenomicSegment:GDB132938) seems to have
disappeared (I think TDWG has a cached copy). This raises the ongoing
problem of service availability. TDWG's resolver could help here, in
that could be used to generate reports on service quality and notify
providers when something's wrong. Whatever GUID technology adopted
this will be an issue, and the challenge is to build tools and
mechanisms to manage this.
I've nothing useful to say here, other than to suggest that clearly
the integration of biodiversity data sales pitch hasn't (yet?)
succeeded. I think us techies get it, but we've not made that vision
real or compelling. If we had, I think we'd have institutions falling
over themselves to ensure the infrastructure exists and persists.
Naive, I know, but we could ask why we haven't managed to convince
those with the purse strings that this stuff matters.
One quick and dirty way that might help is if the TDWG LSID resolver
stored all the metadata in the LSIDs it resolves in a triple store and
exposes a SPARQL query interface to that metadata. We could then start
to look for interesting links between data.
Professor of Taxonomy
Graham Kerr Building
University of Glasgow
Glasgow G12 8QQ, UK
Email: r.page at bio.gla.ac.uk
Tel: +44 141 330 4778
Fax: +44 141 330 2792
AIM: rodpage1962 at aim.com
Home page: http://taxonomy.zoology.gla.ac.uk/rod/rod.html
More information about the tdwg-tag