Re: RDF/XML Syntax Specification (Revised) W3C Working Draft published

From: Mike Moran <mmoran@netphysic.com>
Subject: Re: RDF/XML Syntax Specification (Revised) W3C Working Draft     published
Date: Thu, 20 Dec 2001 19:36:58 +0000

> Peter F. Patel-Schneider wrote:
> 
> [ ... ]
> >>I don't wish to troll, but it would seem like inheriting of  attributes 
> >>is a thing most people would expect to be able to do in RDF, somehow.
> >>
> > 
> > Why would one expect that? (And this is a serious question.)
> 
> [ ... ]
> 
> Simply for ease of representation. Why would one expect to be able to 
> inherit attributes in any type system? This is what RDF appears to be, 
> especially when there exists things such as rdf:subClassOf and the like.

I don't know why one would expect this.  I certainly don't expect to be
able to inherit attributes in *any* type system.  Some type systems allow
certain kinds of inheritance, C++ for example, while others don't allow any
kind of inheritance, ML for example.  Only a very few type systems allow
inheritance between values, which is what you seem to want.

> I'm coming from a programming background, so I may have a skew on this, 
> but that is what RDF `looks like' to me. When I see "subClassOf", or 
> "type", various connotations and expectations arise such as the 
> inheritance of attributes.

It is the case (or, actually, will be the case when the new RDF model
theory becomes official) that 

	Student rdfs:subClassOf Person .
	john rdf:type Student .

implies

	john rdf:type Person .

which is a kind of inheritance.

However, there is no way to say in RDFS that the friends of Persons are
Persons, so you can't inherit these sort of things between classes.


> Before I go off on the wrong track again, could a potted
> explanation be given of what type system RDF does have, and which 
> programming model is closest to it? Note that I'm fine if you say it 
> doesn't have one, just as long as I know.

I think that this is where your problem lies.  RDF is not a programming
language.  It does not have a programming model.  It, however, is, finally,
getting a badly-needed formal, unambiguous specification. 

[...]

> --
> Mike

peter

Received on Thursday, 20 December 2001 14:55:23 UTC