Re: Summary: Section 2: What does a URI identify?

Paul Grosso wrote:

> At 14:55 2002 03 18 -0500, Jonathan Borden wrote:
> [many things to which I do not take issue, and then...]
> >The next point:
> >[[
> >A URI that consists of only a fragment identifier (i.e, one that begins
> >a " # ") always points into the document that contains the URI,
> >of the effective base URI.
> >]]
> >
> >I disagree. Again, particularly with RDF applications that do not
> >necessarily maintain the concept of a "current document", a bare fragment
> >identifier might be converted into an absolute URI + frag id, via the
> >URI. Indeed RFC 2396 explicitly licences such use when appropriate.
> No!  People keep getting this wrong, and unfortunately major browsers
> do too which is really making a mess of the use of base URLs to the
> point where one cannot use <base> and <a href="#xxx"> sorts of references
> together, because this misinterpretation of #xxx refs with respect to
> base URL destroys the ability to maintain intradocument xrefs.

As I said, the usage depends on the application.

For example:


<daml:Class rdf:ID="foo">
    <daml:subClassOf rdf:resource="#bar"/>

and then...

<daml:Class rdf:about="#foo">
    <daml:subClassOf rdf:resource="#baz"/>

We all (should) agree that both daml:Class descriptions are describing the
same Class.

The question is: what are the RDF statements that are emitted as a result of
parsing the document?

subject predicate object .

<> rdf:type daml:Class .
<> daml:subClassOf
<> .

and later

<> daml:subClassOf
<> .

> >
> >Read 'however' ...
> However, you are talking about empty URI references which isn't
> the current topic.
> Instead, read the appropriate part about "5.2. Resolving Relative
> References to Absolute Form".  Specifically, the second step of
> the algorithm:
>   2) If the path component is empty and the scheme, authority, and
>       query components are undefined, then it is a reference to the
>       current document and we are done.
> I don't find that ambiguous at all.

Well I guess RDF has a problem then, eh? I don't see why an application
cannot _choose_ to turn a bare fragment identifier into an absolute URI ref,
particularly when perhaps many documents are being parsed and stored into an
RDF triplestore, linkbase etc.


Received on Monday, 18 March 2002 18:35:26 UTC