[Fwd: [Fwd: Re: Extensibility Design Choices]]

Dear all,

I would like to ask for feedback on a page I started on the wiki to
revive and substantiate the abstract model discussion in connection with 
extensibility:

http://www.w3.org/2005/rules/wg/wiki/AbstractModel

I still think an abstract model is good for dialect extensions/ 
restrictions. But I think the idea to link the abstract model
tightly with the BLD XML syntax so far was maybe not the right way to
go. That is why I was starting with an attempt to save the abstract
model, using RDFS (some bits of OWL, but not really necessary)
to describe it and subsequently adding a few missing pieces.

My intention is to define dialect specific restriction in terms of
constraints on RDF instances there. I currently tried 3 options to 
define such constraints, using:
  1)SPARQL ASK QUERIES (boolean queries formulating constraints on 
rulest instances of the abstract model)
  2) RIF itself (with naf added): Rules to formulate those constraints, 
this is very much along my earlier propposal in [1]
  3) OWL (axioms read as constraints along the lines of [2])

as possible such constraint languages.

The reason for it is that I still believe that
a) an abstract model is a better basis for extensability than an XML
syntax only
b) in XML Schema itself, not all constraints on the model can be formalized
c) you get an RDF exchange syntax for RIF for free by this exercise,
still keeping the XML syntax the central and normative one (for those 
who fancy RDF exchange or embedding of rules in RDF, etc.)

I would kindly ask you to have a look over the wiki page and give me 
some initial feedback, whether you share the view that this is a 
worthwhile path to follow and hope to be able to discuss it alogn with 
other extensability issues in the next telecon or at the F2F.

best,
axel

1. http://lists.w3.org/Archives/Public/public-rif-wg/2007Feb/0083.html
2. http://www.cs.man.ac.uk/~bmotik/publications/papers/mhs07bridging.pdf


-- 
Dr. Axel Polleres
email: axel@polleres.net  url: http://www.polleres.net/

Received on Thursday, 25 October 2007 21:00:31 UTC