[uf-discuss] format for identifiers?

[C. Hudley]

Has any work been done on a format for identifiers?  By "identifier" I
mean "some kind of arbitrary string that identifies the object at hand
in some system".  Examples of the kinds of identifiers that might be
in scope would be ISBNs for books, PMIDs for pubmed, node ids for
weblog entries, DOIs for journal articles, flickr ids for flickr
images, etc.

I'd define the problem to be solved as:

  Thousands of different systems and resources publish records of some
normalized kind that each have their own identifier, and in many case
these identifiers can have utility outside of those systems.  In
particular, it may be useful to track an object's identifier for later
reference, citation, disambiguation, or service resolution. 
Unfortunately, although there are many standards for obtaining or
specifying identifiers themselves, there is no common practice for
indicating which values on a web page are content identifiers.


Perhaps something as simple as "class='identifier'" could go a long
way toward helping to address this problem, and it might be useful
embedded in other formats (like, say, the cite formats under
discussion).  I'd be happy to start documenting real world examples,
related specs, and potential benefits on the wiki if there's any
interest and it's appropriate.

I've looked in the wiki and through some of the recent list archives
for something like this, and not found it yet.  A brief discussion in
#microformats the other day indicated a msg to the list might be a
good next step.  I don't think an answer lives in the HTML specs
already.

My apologies if I've missed something obvious, please point me at
relevant backstory and I'll read up before posting again. :)

Many thanks.

--
C. Hudley
We Know The Truth, Inc.