From: Mike Dean ([email protected])
Date: 07/29/03
I started to update [1] to include the candidate objectives listed in [2], imports, and modules, but fear that it might obscure what I consider to be the main issue currently facing us. Over the last few months, it's become apparent that there have been 2 different objectives within the membership of the Joint Committee: extending Semantic Web ontology languages with rules developing a rule language that can work with the Semantic Web An ideal solution would meet both sets of objectives with a minimum of mechanism. We're not quickly converging on such a solution. The Joint Committee operates independently, but many of us have a responsibility to the DAML program. To help achieve the programmatic and technical objectives of the DAML program, I think the major responsibility for those folks is to keep progressing up the Semantic Web stack in a way as compatible with the underlying layers (XML, RDF, and OWL) as possible. I think this is also the approach most likely to gain acceptance in the larger Semantic Web community. I think many of the issues currently being addressed by RuleML are ahead of what the larger Semantic Web community is ready to accept. RuleML 0.7 provided most of the facilities we're likely to need in an initial phase (I also think we'll need simple built-ins such as arithmetic functions). Features such as support for n-ary relations (motivating a roled syntax), support for ordered and unordered atoms, and complex mappings to URIs detract from the goal of a simple rule extension for ontologies. I realize that a lot of work has been invested in these features. A simple syntax supporting only binary relations would also be more compatible with DQL. I'd like to discuss this during today's telecon. A few specific questions: 1) Do others share this view? 2) Are the folks that have been working the most on rules willing to defer even more of their work past phase 1? Phase 1 might also identify a long list of "postponed" issues with possible solutions. Mike [1] http://www.isi.edu/~stefan/rules/20030325/ [2] http://www.daml.org/listarchive/joint-committee/1390.html
This archive was generated by hypermail 2.1.4 : 07/29/03 EST