RE: details of rdf:datatype?

> For any datatype one could define a canonical form if one so chose.
> So requiring RDF datatypes is a possibility. (Technically I have required
> certain transfinite type systems to invoke the axiom of choice!

very interesting, any pointer?

> but I
> suspect all realistic type systems are safely countable).
>
> The text in the minority version is intended to utilize a canonical form
if
> one is provided, and not if not.

agreed
which could make an engine incomplete, but that's fine
(and maybe we should express that incompleteness fact in RDF???)

-- ,
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/

Received on Thursday, 17 October 2002 04:12:24 UTC