RE: Correction Re: The case against URNs

> -----Original Message-----
> From: Paul Prescod [mailto:paul@prescod.net]
> Sent: Sunday, October 06, 2002 7:56 PM
> To: www-tag@w3.org
> Cc: Champion, Mike
> Subject: Correction Re: The case against URNs
> 


> If "now:" is really dereferencable through a string-rewriting 
> then it is 
> really more like "http:.../#" than like "urn:". I expect that Roy 
> Fielding will have the same complaints about the former that he does 
> about the latter: that it is unnecessary because Web URIs can already 
> represent abstract things.

I really should shut up because this is all a couple levels of abstraction
above my head.  Still, it would seem to me that the "abstract things" a HTTP
URI represents in the real Web are something akin to "documents."  There is
some abstraction of a document-thing at http://www.whitehouse.gov that may
be represented with text, HTML, SVG, VoiceXML, Flash, etc., and an HTTP
request gives the server enough information to decide which to actually
send. I'm deeply impressed with the thinking that made this a reality, and
with Roy Fielding's formalization of that thinking. But it seems to me that
the "thing" is some sort of language-based description of whatever the owner
wants you to see/hear when you visit the website.  That's pretty abstract,
and the distinction between the abstract "resource" thingie and the concrete
TXT/HTML/etc. representations is extremely powerful.  Still, this is
MASSIVELY less abstract than the notion that http://www.whitehouse.gov
identifies the "the White House" (or the US Presidency, or whatever) and the
HTML I get when I point my browser there is a "representation" of that
abstraction.

Sheesh, we've been here before .... sorry.  But that's my point ... we keep
coming back to the same place.  Somebody has to break out of the conceptual
rut we're in, and I thought Micah's question/proposal did just that.

Received on Sunday, 6 October 2002 20:29:39 UTC