Re: The emperor's new datatypes

At 07:57 AM 11/27/02 -0600, Dan Connolly wrote:
>This discussion is out of order since the question is
>decided, so I'm not copying the WG...

Good.  I don't intend to say more there.

>On Wed, 2002-11-27 at 07:23, Graham Klyne wrote:
>[...]
> > That is, the intent of the expression:
> >
> >     jenny age xsd:integer"10" .
> >
> > can be equivalently expressed as:
> >
> >     jenny age _:x .
> >     _:x xsd:integer "10" .
>
>While I believe that's true, that's actually not the point
>of the rdfs:format proposal. In the RDFS format proposal,
>it's just
>         jenny age "10".
>
>i.e. the age property takes a numeral, not a number.
>
>The RDFS format proposal is not incompatible with
>datatype properties, but it's orthogonal to it.

Agreed.

The point behind my analysis of CC/PP is that I think content negotiation 
works more cleanly given a consistent treatment for numbers associated with 
literals, and numbers that are genuine numeric values that one can test and 
express in other ways.  I suppose one could define, e.g., test:le that 
relates "1" to "2", "2" to "3" but not "3" to "2" in the numeral (text) 
domain, but my intuition is that that would ultimately lead to other 
confusions.

So I was explicitly looking at ways of expressing number values in the 
graph, as well as constraining literals.  This is something that the 
current datatyped literals give us.

#g


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

Received on Wednesday, 27 November 2002 15:09:56 UTC