W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > October 2002

Re: Typed literals: current status

From: Graham Klyne <Graham.Klyne@MIMEsweeper.com>
Date: Mon, 21 Oct 2002 20:45:39 +0100
Message-Id: <5.1.0.14.2.20021021203937.03256030@127.0.0.1>
To: Brian McBride <bwm@hplb.hpl.hp.com>
Cc: Jeremy Carroll <jjc@hplb.hpl.hp.com>, w3c-rdfcore-wg@w3.org

At 03:34 PM 10/21/02 +0100, Brian McBride wrote:

>At 15:16 21/10/2002 +0200, Jeremy Carroll wrote:
>
>[...]
>
>>Motivations are:
>>- uniform framework
>>- addresses TBL's desire that XML is not built-in at the lowest level to RDF
>>- provides argument why lang tags are part of literal
>>- gives an example of a non-XSD type system that Brian is prepared to
>>defend.
>
>I'm sorry, I'm maybe being contradictory on this. This proposal means that 
>either:
>
>  1) A datatyped literal denotes a value, in which case RDF datatypes map 
> a pair (lex, lang) to a value which is contrary to the xsd datatyping model
>
>  2) A datatype literal denotes a pair (val, lang) and then we have 
> (speaking loosely) French integers being different from English integers, i.e.
>
>   <jenny>  <age> "10"-"fr"-<http://...#decimal> .
>   <johnny> <age> "10"-"en"-<http://...#decimal> .
>
>does not entail
>
>   <jenny>  <age> _:l .
>   <johnny> <age> _:l .
>
>I really don't want to go anywhere near 2.

Yes, (2) looks like a minor disaster to me.

Suppose I have a URI that denotes the integer 42, nothing more.  Let's call 
it urn:number:42.

If I say:

   <jenny>  <age> "42"-"fr"-<http://...#decimal> .
   <johnny> <age> urn:number:42

does that entail the following, given knowledge that urn:number:42 denotes 
the same number that is represented in the identified ...#decimal datatype 
mapping of "42":

   <jenny>  <age> _:l .
   <johnny> <age> _:l .

I think the answer should be yes.  Then what about:

   <jenny>  <age> "42"-"en"-<http://...#decimal> .
   <johnny> <age> urn:number:42

I think the answer should still be yes.  Then, I think your original 
entailment (above) must also be allowed.  Taking language ids on numbers 
into the domain of interpretation seems all wrong to me.

#g
--

>No one wants to declare the existing Nokia data illegal, but I currently 
>see a choice between:
>
>  o following the xsd datatyping model (except we play a little fast and 
> loose on the legacy)
>  o or blessing the current Nokia data
>
>I suggest that if we choose the latter, we are in for heavy last call 
>comments.  I doubt that the schema datatypes decision that lang was not a 
>factor in the mapping was taken lightly.
>
>Brian

-------------------
Graham Klyne
<GK@NineByNine.org>
Received on Monday, 21 October 2002 15:30:03 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:52:28 EDT