- From: Jeremy Carroll <jjc@hplb.hpl.hp.com>
- Date: Thu, 15 Nov 2001 14:50:24 -0000
- To: "Brian McBride" <bwm@hplb.hpl.hp.com>, "rdf core" <w3c-rdfcore-wg@w3.org>
Sorry Brian, I know you're trying to simplify life but ...
>
> o The WG believes that extending the RDF/XML syntax so that it
> can respresent
> all RDF graphs is beyond the scope of its current charter
> and resolves
> to postpone consideration of this issue.
I have to date agreed with this, but .... I'm wavering.
I'm wavering for three reasons:
+ with a developer hat on, a developer who controls both an RDF/XML reader
and an RDF/XML writer I note that I have a *responsiblity* to my users to
fix this problem, at least between my writer and my reader. With WG hat on,
maybe I can duck that responsibility, but I, at least, need to have a
solution; even if it doesn't have WG blessing.
+ this is really a bug, a significant bug, not just an academic bug, a bug
with the spec. I think it is in charter to fix bugs.
+ there are a number of adequate fixes with adequate backwardly compatible
behaviour. Viz -
a. mandate that all fragment URIs of a particular URI are in fact bNode's
and not real URIs e.g. http://www.w3.org/2001/11/rdf/bnode#b1
b. suggest use of processing instruction
e.g. <?rdf-fake-uri http://www.w3.org/2001/11/rdf/bnode?>
c. use of new attribute rdf:local
Anyone care to push me over the edge?
Jeremy
Received on Thursday, 15 November 2001 09:51:02 UTC