[rdfweb-dev] Describing FOAF using OWL: enumerating allowed

[Dan Brickley]

* Danny Ayers <danny666 at virgilio.it> [2003-06-22 20:11+0200]
> 
> > We use the mechanism I outlined before, to control the literals.
> > Morten is
> > showing us how to use OWL to make generalisations about all those things
> > that have some specified myersBriggs property value. Which is pretty cool.
> 
> Right, thanks - it hadn't clicked. It's a shame about the godawful list
> syntax, but being able to (effectively I suppose) select based on the value
> of a literal is very cool indeedy, (words like 'rules' and 'queries' spring
> to mind!).

Exactly! I really don't think this aspect of OWL has been fully appreciated 
by enough people. It really is pretty powerful. This (dons dayjob hat) 
is one of several other reasons why I am cautious about launching new 
formal W3C work in this area, while OWL is still 'settling in'. Not only OWL,
but XML Query too. RDF folks tend to have a knee-jerk reaction to XML Query,
along lines of:

  "XML Query is all about elements and attributes; RDF is all about the 
   stuff we happen to exchange in those elements and attributes, and 
   querying in terms of surface syntax simply doesn't make sense for RDF."

Unfortunately I don't think things will be that simple. XML Query allows a 
lot of the angle-bracket stuff to be hidden away behind function definitions.
That's why lots of industry folks are happy using it to encapsulate non-XML
data (SQL etc.). 

So we have RDF, OWL, RDF query systems, RDF rule systems, and XML Query to 
think about. Urk! All the more reason to get some RDF data out there in the 
public Web, imho. 

I'm curious as to how this will turn out, whether OWL, or some
conventions expressed in XML query, or Squish-like languages will prove 
useful for querying FOAF repositories...

Dan