The problem in talking about identifiers is encapsulated in the Spittoon Joke. If you're not familiar with this joke, I'm sorry, but it is too tasteless to relate away from the flicker of a campfire. The essence is that there's no easy place to stop once you start, and the starting place isn't always obvious either. This is the dilemma I've been struggling with, having agreed to comment on a blog post by Bruce D'Arcus on identifiers: URIs as Names.
Bruce approaches the question from the perspective of tools for scholars, with the normal sets of problems that scholars have, including persistent citation. Among the things that we want from citations is a convenient handle for any arbitrary resource, a handle that we can use to hang the resource on our scholarly pegboard, pick up the resource, pass it to others, let them hang it on their scholarly pegboard, and so forth. Since it is easier to arrange our pegboards in standard ways, it is best if the handles are the same size and configuration, or barring that, that the number of differentstyles of handles is small.
We'd all be fine if we could agree on a single style of handle for all the resources that we want to manage, right? And Lo! We now all live on Planet Web, as Norm Walsh puts it, and as Web denizens, we know that http:URIs are the obvious and most useful form of identifiers, and hence our problem is solved, now and always, and for all manner of resources ever to be conjured for scholarly or other purposes.
We are nearly 15 years downstream from the New York Times article that served to awaken many of us to Tim Berners-Lee's marvelous creation. The URL, the now-discredited moniker that has been displaced in discussions such as these by the term http:URIs, has indeed become the most widely-used identifier in the world.
For all that, we still don't enjoy the identifier heaven that the Web promises. I'd like to explore some of the reasons i think this is the case, and perhaps even argue a particular perspective or two. Bruce's post on the subject points to Norman Walsh's blog post in which he argues that http:URI schemes are entirely sufficient to the need, and to deviate from this is harmful. Norman's post points also to a W3C TAG finding under development by Henry Thompson and David Orchard, presumably giving voice to the official W3C position on Web naming issues. I will begin my exploration of these issues by reprising the arguments in these documents, in case you want to read ahead. Reaching, now, for the spittoon....
Image: Seattle evening skyline, featuring the Smith Tower (the short, pointy white building), for many years the tallest building west of the Mississipi River. Taken from the Alaskan Way Viaduct, August 27