- From: Patrick Stickler <patrick.stickler@nokia.com>
- Date: Wed, 03 Jul 2002 20:42:01 +0300
- To: ext Paul Prescod <paul@prescod.net>, Joshua Allen <joshuaa@microsoft.com>
- CC: Jonathan Borden <jonathan@openhealth.org>, ext Tim Bray <tbray@textuality.com>, WWW TAG <www-tag@w3.org>
On 2002-07-03 20:05, "ext Paul Prescod" <paul@prescod.net> wrote: > Joshua Allen wrote: >> >> ... >> >> There are certain people (I assume yourself included?) who feel that >> URIs identify *nothing*, unless they are accompanied with ontology >> information; and say that therefore it is smart to use http: URLs to >> identify cars and butterflies, since "lots of people know how to >> dereference http: URLs". > > Let me rephrase that. URIs always identify something. And that something > can always be *represented* by some documentation, or schema or picture > or SOMETHING. It isn't that a URI without a dereference is useless. It's > that it is like software without documentation: inconvenient and > amateurish. Documentation about some resource is not an HTTP representation of that resource. While I agree that we should be able to describe anything that has URI denotation, and allow folks to obtain those descriptions easily and as transparently as possible, documentation/information about a resource is not the resource itself, and the real crux of this dilemma is that the present Web architecture makes no clear provision for information *about* things. So when we have URIs denoting things that cannot be accessed on the web -- i.e. no representation is possible because they are either abstract or non-digital -- then either you get a 404 error, or you get ambiguity, with the same URI denoting both the resource and something *else* that is put in lieu of the resource so folks can get *something* when they dereference the URI. The answer is not to make namespace names dereference to namespace documents (which even if they did, wouldn't do what they are supposed to do anyway) but rather to fix the web architecture to support information about resources and allow folks to both GET representations of resources as well as GET-META information about those resources. Saying namespace names should dereference is a short term hack, plain and simple, that just avoids or delays facing the real challenge. Regards, Patrick -- Patrick Stickler Phone: +358 50 483 9453 Senior Research Scientist Fax: +358 7180 35409 Nokia Research Center Email: patrick.stickler@nokia.com
Received on Wednesday, 3 July 2002 13:42:00 UTC