Some questions

David Megginson david at
Thu Dec 2 14:27:38 GMT 1999

Didier PH Martin writes:

 > Didier reply: 

 > But most of the RDF users set the description element "about"
 > attribute's value with a URL. In fact, this is OK because the spec
 > indicates that you are providing a description _about_ something
 > and the about value may be its location.

The advantage is that the URL is an absolute identifier (whether it
actually points to anything or not).  For example, imagine that uses id p0809764 to refer to the person David Bowie, while
Reuters uses the id p0809764 to refer to the person Bill Clinton.  If
I get some RDF

  <foo:Person rdf:ID="p0809764">

how do I know who I'm talking about?  On the other hand, if I have

  <foo:Person rdf:about="">

  <foo:Person rdf:about="">

then there's no room for confusion.  Certainly, local IDs have their
uses, but we're building a new environment where information has to be
useful across systems, and to accomplish that, we need to use some
kind of global identifiers, such as URLs or URNs (once the latter are
ready for Prime Time); local IDs are of little value outside of closed

 > I discovered that using this form, is, most of the time
 > bogus. Instead, I do what librarian discovered. Have the
 > classification card (i.e. description element) to be independent of
 > any properties.

Yes, but with the Web, a better analogy would be that you're in
Robarts Library in Toronto and have a card from the Bodleian in Oxford
that says to get the third book on the fifth shelf in the eighteenth
row.  It would have been better to have given you the ISBN so that you 
could find it in any library.

All the best,


David Megginson                 david at

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at
Archived as: and on CD-ROM/ISBN 981-02-3594-1
To unsubscribe, mailto:majordomo at the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo at the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at

More information about the Xml-dev mailing list