Re: Spec doesn't talk about two-valued relationships

Aaron,

I appreciate the problem you're trying to tackle here, but I wonder if 
having standard properties is the way to go.  What you suggest is getting 
close to rdf:type, and it's not clear to me that there is a clear distinction.

Also, introducing two-valued logic into the core of RDF would, I think, 
take us into FOL territory -- I've said elsewhere that I think this is 
something that I think should be built on top of an RDF foundation, not in 
the foundation.  E.g. how is the following pair to be interpreted:

   <http://www.aaronsw.com/> rdf:is    bob:ChocolateLover;
                             rdf:isNot bob:ChocolateLover.

Per core RDF, these are a valid pair of statements (i.e. describe a valid 
graph), even though they're _logically_ contradictory.  The current RDF 
core has no way to state contradictions, and I think it opens a can of 
worms to introduce it there.

#g

At 09:09 PM 3/8/01 -0600, Aaron Swartz wrote:
>The spec explains how to deal with relationships > 3 (rdf:value) but not
>those < 3 (i.e. two). I'd like to suggest we introduce two new properties:
>
>rdf:is
>rdf:isNot
>
>This saves us from having to do something klugey like:
>
><http://www.aaronsw.com/> bob:likesChocolate 0 .
>
>and also allows RDF processors to know that it's part of two-valued logic
>and treat it properly.
>
>--
>Aaron Swartz <me@aaronsw.com>|       The Info Network
>   <http://www.aaronsw.com>   |     <http://theinfo.org>
>AIM: JediOfPi | ICQ: 33158237| the way you want the web to be

------------
Graham Klyne
GK@NineByNine.org

Received on Friday, 9 March 2001 05:42:33 UTC