Folks,
This discussion about PURLs and LSIDs is very relevant, but, in my opinion, the most appropriate forum to discuss it is on the GUID mailing list and wiki.
To that effect, I have copied Gregor's original posts to the GUID wiki. Links to the two pages can be found at the end of the following page:
http://wiki.gbif.org/guidwiki/wikka.php?wakka=Technologies
I think cross-posting the two lists may be fine for the moment.
That brings me to the second point, which is that the GUID mailing list changed to:
tdwg-guid@mailman.nhm.ku.edu http://mailman.nhm.ku.edu/mailman/listinfo/tdwg-guid
THAT is off-topic for the TDWG TAG list. Sorry.
Cheers,
Ricardo
Roderic Page wrote:
Just when everything seemed settled... ;-)
For those wanting to revisit all this, there's also a nice series of presentations at http://www.dcc.ac.uk/events/pi-2005/ .
The ARK identifier is an example where appending symbols to the identifier determines what you get (e.g., '?' for metadata). I guess one could do something similar for PURLs.
Why not have, *ahem* "BioPURLs" (wince), that is, PURLs deployed by the biodiversity community with conventions for returning what we want?
I wonder whether, if we go down the PURL route, won't we eventually converge on Handles/DOIs, which have administration tools in place? Ultimately, centralisation requires good tools and good support.
In any event, is it possible to separate GUIDs from the whole metadata side of things? And given that every GUID system currently in operation uses (or can use) URLs, can't we postpone deciding on this until we have a few test systems built and we have a real idea of what's involved. In a sense, wrap everything in URLs, the GUID is either the URL or embedded in the URL, then see what happens.
And yes, I'm sure this pretty much contradicts everything in my earlier posts...
Regards
Rod