Re: ruleml and RDF

> >> SWRL assigns a different meaning to documents written in RDF/XML than that
> >> provided by the RDF model theory.  
> > 
> > Which of course means they are not written in RDF/XML, right?
> 
> No, they are written in RDF/XML (syntax), but they are not
> RDF documents. 

I don't really see the distinction you are making.  Labeling something
X when it does not conform to the published specification of X is a
bad practice.  There may be exceptions, but it looks to me like it's
either done by mistake or with malice, but never with well-informed,
well-considerd, unselfish motivation.  In the SWRL case (having been
in many of the discussions), I'm pretty sure it was just not thought
through carefully enough.

> This is like using the language of classical
> predicate logic, but assigning a non-classical (e.g. intui-
> tionistic) semantics 

Somewhat, but anyone doing that and not labeling their use
appropriately for their audience would surely be castigated.  In an
academic paper, imagining and discussing a language with the syntax of
RDF/XML but different semantics would be fine; doing so in a document
which offers an industry standard is not.

    -- sandro