Re: Issue 133, and permitting no body

> > Big stretch. 8-)  They also require additional agreement to be made over
> > the wire.  i.e., it's no longer sufficient that two parties agree on
> > what POST means, they have to agree on what the method name hidden in
> > the POST means.
> To some extent I thing this is a little over-played. In order to get
> something done we still have to know the 'significance' of a resource. To
> have a useful effect (or side-effect) we can't just POST any old thing... we
> have to know what the resource signifies and what vocab to use for its state
> representations... which in my book is an additional agreement over and
> above what HTTP provides.

Absolutely true.  The wonderful thing about the Web is that this is done
*with* the Web.  It can describe itself.  RPC cannot.

Take, for example, my suggestion for how to fix the convention about
additional encodings and how they communicate their conformance with our
default encoding;

Invoking a GET on, we might see;

<rdf:RDF xmlns:rdf=""
<rdf:Description about="">
<rdfs:subClassOf rdf:resource=""/>

Mark Baker, Chief Science Officer, Planetfred, Inc.
Ottawa, Ontario, CANADA.

Received on Monday, 4 February 2002 17:09:51 UTC