- From: Graham Klyne <Graham.Klyne@MIMEsweeper.com>
- Date: Tue, 02 Apr 2002 12:29:09 +0100
- To: Aaron Swartz <me@aaronsw.com>
- Cc: Brian McBride <bwm@hplb.hpl.hp.com>, RDF Core <w3c-rdfcore-wg@w3.org>
At 05:05 PM 4/1/02 -0600, Aaron Swartz wrote:
>On 2002-04-01 3:57 PM, "Graham Klyne" <Graham.Klyne@MIMEsweeper.com> wrote:
>
> > Within an application/rdf+xml document, a fragment identifier identifies
> > the XML element having the given name for its rdf:ID attribute value.
>
>This would imply that rdf:type is an XML element, not the concept of type.
But in the context of the appropriate RDF schema, rdf:type *is* an XML
element, no?
Test case:
In Web-retrieval of the application/rdf+xml document:
http://www.w3.org/1999/02/22-rdf-syntax-ns
I think that http://www.w3.org/1999/02/22-rdf-syntax-ns#type refers to this
XML element:
<Property ID="type"
s:comment="Identifies the Class of a resource" />
(possibly with some namespace decorations...?)
>(dragons, dragons, dragons, dragons!)
Yes, I suppose, dragons.
For me, this is helping to clarify the distinction between fragment-ids
used in web retrieval, and fragment-ids used in RDF resource
identifiers. In the latter use, I think that:
http://www.w3.org/1999/02/22-rdf-syntax-ns#type
denotes a resource with a relational extension corresponding to the class
membership of RDF resources. Or, as you put it, the concept of
type. Which has only tangential relationship to the indicated part of the
indicated RDF schema document.
In writing this, I'm clinging to a clear terminological distinction:
referring, or reference, in the Web
denotation, a semantic concept
...
So, returning to my proposed wording, I'll suggest an revision:
[[[
Given a URI that dereferences to an application/rdf+xml document, that URI
with a fragment identifier refers to an XML element in the document having
the given fragment identifier for its rdf:ID attribute value.
]]]
#g
-------------------
Graham Klyne
<GK@NineByNine.org>
Received on Tuesday, 2 April 2002 06:26:41 UTC