- From: <Patrick.Stickler@nokia.com>
- Date: Thu, 15 Nov 2001 12:39:25 +0200
- To: phayes@ai.uwf.edu
- Cc: w3c-rdfcore-wg@w3.org
> -----Original Message----- > From: ext Pat Hayes [mailto:phayes@ai.uwf.edu] > Sent: 15 November, 2001 03:26 > To: Stickler Patrick (NRC/Tampere) > Cc: w3c-rdfcore-wg@w3.org > Subject: RE: DATATYPES: mental dump. > > > > > > > Yes, I do. I think that part of my overall feeling of > confusion comes > >> from the fact that we don't seem to have a crisp > definition of what > >> constitutes a datatype or a datatyping scheme > > > >How about the one defined by XML Schema, and also adopted by the X > >proposal and 'lit:' ontology, namely a paired value space and > >lexical space, the latter of which may have a canonical subset. > > Right, I am happy with that, as long as 'paired' includes the mapping > between the lexical and value spaces. That is pretty much what Peter > and I were using as the definition as well. > > Maybe we can all agree about something :-) Well.... I am still unsure just what you mean by "mapping". I am happy saying that a pairing of a lexical form (literal) with a data type (URI) together *denote* a value in the value space of the data type. But... ;-) I don't see how you could actually express the explicit *mapping* from lexical form to that actual value without employing a canonical *lexical* representation of that value, so what you are really defining is a mapping from non-canonical lexical form to canonical lexical form and thus just a relation between lexical space and canonical-lexical space for the data type. Furthermore, since a given data type may not define a canonical lexical space, RDF would have to assert one for it. I must be missing your meaning of 'mapping' here, because this all seems just way too impractical to me. Or, maybe we really are in agreement here and the vocabulary gremlin is up to its tricks... If by 'includes a mapping' you simply mean that we explicitly state that in the pairing of lexical form and data type the lexical form corresponds to one and only one member of the value space for the data type, with no attempt to offer any representation of that value, then I think we're in full agreement. Patrick
Received on Thursday, 15 November 2001 06:36:41 UTC