Generalization and opacity (was "duri" as a URI-based URN scheme)

I think there's an infinite number of interesting and abstract kinds of
things that we'd like to identify that *could* be broken out into their
own URI scheme.

We could have stock quotes;

  stockquote:sunw

Real time stock quotes;

  realtimestockquote:sunw

15-minute delayed stock quotes;

  15minutedelayedstockquote:sunw

Yesterday's closing stock quote;

  yesterdaysclosingstockquote:sunw

And even Unicorns;

  unicorn:bob

And then we can build (and deploy) application protocols or WSDL
documents that define an interface to these things.  We can do this,
if we want to.

Alternately, we can have a semantic-free URI scheme (opaque identifiers)
that can be used to identify absolutely anything, and we can bind
generic application semantics to that scheme that can be used to access
and manipulate each of those things.  Any "type" information, such as
whether one of these things is a stock quote, a unicorn, or some other
resource at some point in time, could be communicated via the application
semantics associated with that scheme.

So rather than this;

"urn:tdb:2001:http://www.ietf.org"

The IETF could choose to identify itself at the beginning of 2001 as;

http://www.ietf.org/foobar/

or if they don't or can't, then Larry could with;

http://larry.masinter.net/bar-foo-whiz/

But in both cases, a GET on the URI could return some HTML and/or RDF
describing the relationship between the resources identified by those
two URI.

  "There can be only one."
    -- not a line from the Matrix 8-)

P.S. for the Web services equivalent (roughly) of this email, see;

  http://lists.w3.org/Archives/Public/www-ws-arch/2002Jun/0085

MB
-- 
Mark Baker, CTO, Idokorro Mobile (formerly Planetfred)
Ottawa, Ontario, CANADA.               distobj@acm.org
http://www.markbaker.ca        http://www.idokorro.com

Received on Monday, 7 October 2002 13:27:42 UTC