Re: datatypes and MT

At 02:56 AM 11/4/01 +0100, jos.deroo.jd@belgium.agfa.com wrote:
> > I'm sorry, I'm being thick again.  What's the advantage exactly?  To whom?
>Does
> > this advantage differentiate DanC's proposal from the others?
> >
> > What I see above is the A & B => A.  Which is true, but I miss the
>significance.
>
>   [ rdfs:str "10" ]
>is what we call the 'least common unifier' of
>   [ rdfs:str "10" ] and [ rdfs:str "10"; rdf:type dt:decimal ]
>and we found that a useful thing for inferencing
>that's basically all we wanted to say
>(so 'advantage' was rather subjective)

Let's see if I read this correctly:  sometimes, the only thing one knows 
about some property, and the only thing one needs to know, is that it has a 
value with a given lexical representation.

So, returning to my DTLS, DTVS, DTLV musings [1]:

- Sometimes, we know/express a value in DTLS (the literal space) -- the 
case noted above.

- Sometimes, we know that the value of some node is in DTVS (the value 
space) -- this corresponds to the view of data type as describing a value 
space (The view I think Brian is expressing).  Of itself, this doesn't help 
us express a particular value.

- Sometimes, we know a particular value in DTVS;  but to *express* this 
value, we need a corresponding value in DTLS, and knowledge of the mapping 
DTLV.  (But also note there may be different, possibly overlapping, lexical 
spaces with different mappings to the value space.)

#g

[1] http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Nov/0041.html



------------------------------------------------------------
Graham Klyne                    MIMEsweeper Group
Strategic Research              <http://www.mimesweeper.com>
<Graham.Klyne@MIMEsweeper.com>
------------------------------------------------------------

Received on Sunday, 4 November 2001 07:00:08 UTC