A couple of things come together ...
I was going to do a short post on the renaissance of interest in identifiers based on the approval by the IESG of the Info-URI, and on the growing awareness that we need to consistently identify the entities in our environment (institutions for example) if we are to more effectively tie together data and applications. Stu summarised some of the issues recently [ppt].
However, I came across David Weinberger's note on identifiers earlier, so I will point to that instead and come back to some specifics later.
"Web 2.0" is one of those terms with lots of precise meanings, none of them entirely consistent with the others. To me, it refers to the way in which data and applications can be integrated across the Web, building new apps out of snippets of old. (I'm not nearly as fond of the implication that only with Web 2.0 did users come to have a voice on the Web. User voice has driven the Web since it began.) Web 2.0 takes what were monolithic apps and breaks them apart so they can be stitched together in new ways. Tags break apart the world of hyperlinked pages so that we can pull them together around meanings that we, the readers, supply.
But none of this restitching is possible without thread. That's where unique IDs come in. [JOHO - March 21, 2000]He goes on to discuss the case of books and Hamlet, and reports on a discussion with my colleague Thom Hickey about Hamlet and FRBR. And, hey, yet another mention of xISBN as a nice example of a service which usefully rolls up identifiers for the different manifestations of works.