Joint Committee Minutes 18 September 2001

This page summarizes the telecon for the Joint US/EU Committee on Agent Markup Languages held from 1300 to 1400 PDT on 18 September 2001. These minutes were prepared by the chairman, and were approved during the 25 September telecon.

Participants

Announcements

No announcements this week.

Last Telecon's Minutes

There was no telecon last week due to the terrorist events in New York City and Washington.

The minutes from September 4 were approved by those present.

RDQL

Frank forwarded an announcement of RDF Data Query Language (RDQL), an implementation of Squish Query Language for Jena by HP Labs Bristol, along with a quick comparison with RQL done by one of his Ph.D. students.

We'll plan to discuss this next week.

DAML+OIL Submission

Peter has had to do quite a bit of HTML reformatting to get the submission to pass the automated Note acceptance criteria. For any future work, he recommends using the Link Checker interface to validator.w3.org.

The Microsoft Word indentation of axioms in the axiomatic semantics has been lost, making it very difficult to read. Stefan volunteered to work on the indentation problem. Once everything is fixed, we should maintain the HTML version. This will require updates to the axiom extractor, which is currently based on Microsoft Word.

There was some discussion regarding the packaging as 4 separate documents vs. a single document with 4 title pages. Deborah expressed some concern about pointing to outside documents. To keep the submission self-contained, we'll add the KIF-only document as an appendix.

The submitter can presumably provide new versions of the documents, if necessary.

The Director generally takes 1-4 weeks to decide whether to publish a new Note.

UnambiguousProperty

Several independent groups have recently inquired why daml:UnambiguousProperty is a subclass of daml:ObjectProperty rather than just rdf:Property. Could we provide a parallel mechanism (e.g. UnambiguousDatatypeProperty) that worked for DatatypeProperties?

Peter noted that we can't support UnambiguousProperty on datatypes because UnambiguousProperty is a shorthand for a restriction on inverse, and DatatypeProperties don't support inverses. The pre-March 2001 semantics would have produced an inconsistency or incoherency (because inverse was only defined on objects).

Next Week

We'll continue our discussion of query languages:

Links

IRC
last week's minutes
Joint Committee home page

$Id: 2001-09-18.html,v 1.5 2001/10/13 03:53:13 mdean Exp $