Just back from TDWG2010 where I gave a talk on the use of standards in PESI and the prospect for stable identifiers for taxa. The take home message was that if we bind the identifiers to nomenclature then the identifiers will not be stable. You can see the presentation as a PDF here: PESI Standards & GUIDs, Roger Hyam (Natural History Museum London) & Yde de Jong (University of Amsterdam)
In side conversations with the guys from Encyclopedia of Life and Barcode of Life I wondered whether we will eventually gravitate to using EoL page URIs as identifiers for taxa globally – combined with barcodes for normative identifiers for those taxa. It would be simple. It would remove the need for complex nomenclatural databases.