Re: CURIE proposal ...

> 
> Michael Kifer wrote:
> > What you are proposing is even worse than expanding the same
> > macro differently in different contexts. Now you are saying that foo:bar
> > really stands for "something-long-here", but in some contexts we are not
> > allowed to use "something-long-here" instead of foo:bar.
> 
> it just says standalone curies mean something different than
> curies after ^^ ... in the former case they denote an IRI, i.e. a 
> constant in the rif:iri symbol space and in the latter they denote
> the IRI for a symbol space.
> What is the problem with this?

you are using the term iri in two different senses: one is the RFC sense
and the other a rif:iri constant. So, rif:iri by itself is interpreted in
one way and when it occurs in "..."^^rif:iri in another.

Worse, the identifier for the rif:iri symbol space is said to be
"http://.../iri" and the definition says that the syntax for constants
is literal^^symspace-identifier, but you are telling me that I cannot
actually write that symspace-identifier as what it is (i.e., as
"http://.../iri"), but I must use a different representation for it.

If you do not think this is an ugly hack then I do not know how to explain
this better.

> The whole thing is not about CURIEs being a generic macro but about a 
> reasonable shortcut notation.

Exactly. This is not a reasonable notation in my view.


	cheers
	  --michael  

> 
> best,
> Axel
> 
> 
> 
> >>> As I said, your proposal had a couple of holes, which Jos was trying to
> >>> fix. Most of all, I do not like the fact that you are proposing that ":"
> >>> would macro-expand differently depending on where it appears (after the ^^
> >>> or elsewhere). My second proposal (<prefix:suffix>) was to fix that.
> >> The problem with using "<" and ">" like this:
> >>
> >>    In practice, URIs are delimited in a variety of ways, but usually
> >>    within double-quotes "http://example.com/", angle brackets
> >>    <http://example.com/>, or just by using whitespace:
> >>
> >>       http://example.com/
> >>
> >>    These wrappers do not form part of the URI.
> >>
> >> (from RFC 3986 [1]), and the various Semantic Web specs all use <...> in
> >> this way, to delimit URIs.   I think a good middle ground is something
> >> like:
> >>
> >>    1.  A "prefix" declaration syntax, as in Turtle:
> >>
> >>          @prefix ns:   <http://example.org/ns#> .
> >>
> >>        or SPARQL:
> >>
> >>          PREFIX ns:   <http://example.org/ns#>
> >>
> >>        or, maybe best, something more RIF-PS like:
> >>
> >>          PREFIX("ns", "http://example.org/ns#").
> >>
> >>    2.  The CURIE a:b syntax is the only syntax that can be used after
> >>        the "^^" operator.   Eg:
> >>
> >>          PREFIX("xs", "http://www.w3.org/2001/XMLSchema").
> >> 	 PREFIX("rif", "http://www.w3.org/2007/rif#").
> >>
> >>          ... "10"^^xsd:integer ...
> >>          ... "http://purl.org/dc/terms/creator"^^rif:iri
> >>
> >>    3.  The CURIE a:b syntax MAY also be used as a shortcut for rif:iri
> >>        terms, so givent his PREFIX declatation:
> >>
> >>        	 PREFIX("dc", "http://purl.org/dc/terms/").
> >>
> >>        the last term in #2 above could also be written as:
> >>
> >>           ... dc:creator
> >>
> >> I think that's all we really need to make IRI handling in the
> >> presentation syntax relatively comfortable and precise, no?
> >>
> >>            -- Sandro
> >>
> >> (Note that it seems BLD has the wrong namespace for Dublin Core.  It
> >> should be either "http://purl.org/dc/elements/1.1/" or
> >> "http://purl.org/dc/terms/", with the latter being in some sense
> >> preferred. [2].)
> >>
> >> [1] http://www.ietf.org/rfc/rfc3986.txt
> >> [2] http://dublincore.org/documents/dces/
> >>
> > 
> 
> 
> -- 
> Dr. Axel Polleres, Digital Enterprise Research Institute (DERI)
> email: axel.polleres@deri.org  url: http://www.polleres.net/
> 
> rdfs:Resource owl:differentFrom xsd:anyURI .
> 
>

Received on Tuesday, 22 April 2008 21:15:42 UTC