Hi Jonathan,
Jonathan said:
I think what is needed is some way to integrate the semantics associated
with a URI as provided by REST vs. that associated with RDF. Once that
has
been done -- it hasn't yet -- we will have a real foundation for the
"Semantic Web".
Didier replies:
Yes!!!
Jonathan said:
In the meantime it might be confusing to people that there might be
_different ways_ to associate semantics with a URI. This is true. Given
two
methods, they _might_ disagree. So it remains up to the "owner" of the
URI
not to be contradictory. In any case if the owner of a URI is
contradictory -- by whatever mechanism or mechanisms the contradictions
arise -- we have no way to make sense of what such a URI "means".
Contradictions are logically false, and unresolved contradictions in the
meaning of a URI, means that the URI has no meaning. That might be true
of a
namespace name, or a URI used to for any other purpose.
Didier replies:
Exactly. This is why we have to push hard and tirelessly to get this
spec fixed and made coherent with the vision of the semantic web.
Cheers
Didier PH Martin
