RE: Do CURIEs represent a (namespace, localname) pair?

At 07:21 PM 4/26/2006 +0100, Misha Wolf wrote:
>In this mail you sometimes qualify "namespace" with "XML" and 
>sometimes not.

Yes, and quite deliberately so.

>I suggest that we all write "XML namespace" when we mean "XML 
>namespace" and just "namespace" when we mean "some namespace".

I endorse that suggestion.  RDF/XML uses the XML namespace
machinery -- for better or worse -- so I say "XML namespace" when
I specifically mean those things declared with an xmlns:... attribute.
When it is clear that I am talking about an RDF/XML document I
will also say "RDF namespace" which I take to be a synonym for
"XML namespace" in such contexts.

Tim has used the phrase "namespace" in his exposition of RDF/N3 [1]
but I don't believe he means "XML namespace".

   [1] http://www.w3.org/2000/10/swap/Primer.html

I do not believe that Mark intended his CURIE proposal to
be another way to declare (or use) XML namespaces.

Do you have an opinion on what you want a CURIE
syntax to represent?  Do you want CURIEs to have
any kind of namespace semantics?  Do you want all
CURIEs to have these semantics or only when the
containing document type is of a class that explicitly
adds such semantics to CURIEs?  I suppose that
the semantics "all URIs that have this string as their
first (left) part" is a kind of "namespace" semantics
but I'd prefer to call that something else, such as "URI prefix".

Received on Wednesday, 26 April 2006 19:28:29 UTC