W3C home > Mailing lists > Public > www-rdf-interest@w3.org > December 2001

Re: Namespaces wihtout "#" Was: Few CWM Bugs

From: Mark Baker <distobj@acm.org>
Date: Wed, 26 Dec 2001 23:12:37 -0500 (EST)
Message-Id: <200112270412.XAA08625@markbaker.ca>
To: danbri@w3.org (Dan Brickley)
Cc: timbl@w3.org (Tim Berners-Lee), jborden@mediaone.net (Jonathan Borden), sean@mysterylights.com (Sean B. Palmer), connolly@w3.org (Dan Connolly), www-rdf-interest@w3.org (www-rdf-interest)
Dan Brickley wrote;
> Am I right in thinking one can use HTTP 1.1 (the protocol) to ask an HTTP
> service about non-HTTP URIs? (Don't web-caches basically work this way?)

For sure.  GET is suitable for retrieving a representation of
anything with identity (ditto for other HTTP methods, because
they were designed to be this general).

> I'm still of the opinion that many HTTP-named resources aren't network
> serializable, most notably web services (online thermometers, coffee pots
> etc, alongside the usual databases, web indexes...).

The coffee pot itself isn't serializable, but representations of it
may be; images, HTML, CoffeePotML, etc..  I'm not sure what you mean
by 'web service' in this context, but all of those examples have
representations that are serializable.

> HTTP's notion of
> document seems pretty weak, if it allows services and the like to count;
> weaker still if it includes all the stuff that one can ask an HTTP server
> about regardless of the species of name used.

As Aaron stated, HTTP has no notion of document.  It knows of resources
and representations of those resources, and it has a pretty strong
notion of both.  I would really like to see this terminology replace
the (IMO) confusing document-centric one that's also in use.

Mark Baker, Chief Science Officer, Planetfred, Inc.
Ottawa, Ontario, CANADA.      mbaker@planetfred.com
http://www.markbaker.ca   http://www.planetfred.com
Received on Wednesday, 26 December 2001 23:12:38 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:07:38 UTC