Re: Modelling HTTP GET with RDF

Graham et al.

> At 08:00 AM 2/26/02 -0500, Jonathan Borden wrote:
> >What is returned after an HTTP GET is a literal, or perhaps represented
as a
> >"data" URI, this HTTP transaction might be represented _in a very
> >fashion_ as an RDF statement (and without URI escaping)
> >
> ><> http:GET
> ><data:text/html,<html><title>example</title><body><p>This is a
> >document</p></body></html> > .
> What an interesting idea!  Instead of trying to formalize RDF in terms of
> web operations, formalize web operations in terms of RDF.
> [Thinks... contexts... modalities... not easy, but I wonder if it could

It occured to me that the RDF representation of a MIME message (XMTP) gets us much of the way there in terms of a
full fledged RDF representation of an HTTP transaction:

I've started to describe this:

So roughly an explanation of the RFC 2396 terms:

"resource" is a node labelled with a URI of scheme "http"
"entity" is either a node labelled with a URI of scheme "mid" _or_ an
anonymous node (bNode in the new terminology?) with rdf:type = http:Response

Pat, perhaps the presented example is easier to parse, and less ambiguously
stated, than the English text descriptions the RFCs :-))

Brian, this goes well along the way toward a proper defintion of these web
architecture terms as you have requested from the TAG ... now only to get
everyone to agree on these statements...


Received on Friday, 1 March 2002 19:37:35 UTC