2009/3/31 Gregor Hagedorn g.m.hagedorn@gmail.com:
2009/3/23 Paul Kirk p.kirk@cabi.org:
In fungi we have MycoBank and citing MB510023 (for Quambalaria coyrecup T. Paap 2008, for humans to read) and having a true GUID (not yet implemented) or LSID (implemented: http://lsid.tdwg.org/urn:lsid:indexfungorum.org:names:510023) 'underneath' is closer to the GenBank model which everyone accepts.
Paul
Paul,
Who will educate all biologists about the difference between the lsids and their use? And I specifically mean the non-taxonomists that use species names. In "normal" resources like PDF published articles, html species pages, Drupal or Wiki-based information, urn:lsid:indexfungorum.org:names:510023 is useful only for human display. But it will be used.
http://lsid.tdwg.org/urn:lsid:indexfungorum.org:names:510023 will work as a clickable link to a web resource and can be read by semantic web machine reasoners. But to create this, external knowledge about resolution mechanism is expected from every biologist in the world. Some help material to be found somewhere may instruct them: if in your publication you want to create a link that is usable by humans or semantic web machine reasoners, you have to convert the lsid by prefixing it with http://lsid.tdwg.org/, http://lsid.gbif.org/, or perhaps lsid.ncbi.nlm.nih.gov/.
The mistrust in http current URL management practices is certainly valid. But this is a social issue, not a technical. The mistrust into http as a protocol I consider invalid: who is willing to bet that http will disappear before urn:lsid? And we would simply introduce resolver THEN (p2k5://http.tdwg.org/http://persistent-identifier.indexfungorum.org/names/510023).
I agree. The current Bio2RDF URI's, which look like mini-lsid's by design, could be resolved using a similar process, ie, p2k5://http://bio2rdf.org/names:510023 would be just as good a way to resolve http://bio2rdf.org/names:510023 as http://bio2rdf.org/urn:lsid:bio2rdf.org:names:510023 would be with the lsid urn:lsid:bio2rdf.org:names:510023
I have been doing some work on the Bio2RDF LSID resolution process today and it should be operating properly soon for the http://bio2rdf.org/urn:lsid:bio2rdf.org:* LSID URI resolutions. They were a bit neglected, as noone has been using them as far as I know and hence there were no bug reports to trigger my attention, but since there is still some interest we should have an alternative interface for use with LSID systems
Cheers,
Peter Ansell PhD Student Queensland University of Technology