- From: Pat Hayes <phayes@ihmc.us>
- Date: Tue, 26 Feb 2013 19:30:10 -0600
- To: Peter Patel-Schneider <pfpschneider@gmail.com>
- Cc: RDF WG <public-rdf-wg@w3.org>, Richard Cyganiak <richard@cyganiak.de>
On Feb 26, 2013, at 2:11 PM, Peter Patel-Schneider wrote: > The new semantics document doesn't really use datatype maps (but still has vestiges). > > The new treatment is new, different from Concepts, and, I think, more confusing than the old treatment. Why is it more confusing? It simply singles out a set of IRIs which are assumed to denote datatypes. This is how datatypes are identified already, in fact. We all talk about (for example) xsd:string, and we know what is being referred to. We don't see any need to say, "the datatype d which such that <'xsd:string', d> is in the datatype map". The old treatment allowed for D-interpretations where (for example) the xsd datatype IRIs were interpreted to mean non-xsd datatypes in D. I don't think we need this much generality, and can rely on external specifications to define what datatype is denoted by a datatype IRI. I have been trying to simplify the semantics document to eliminate needless formality and 'mathematical' style, and this is part of that simplification. The new treatment is much simpler and IMO superior. The datatype map is simply part of the interpretation mapping, which is what it always was in any case. > I suggest going back to the old treatment (with modifications). I suggest modifying Concepts to conform to the new treatment. It will be simpler and easier to understand. (If it would be helpful, I could draft the relevant section.) Pat > > peter > ------------------------------------------------------------ IHMC (850)434 8903 or (650)494 3973 40 South Alcaniz St. (850)202 4416 office Pensacola (850)202 4440 fax FL 32502 (850)291 0667 mobile phayesAT-SIGNihmc.us http://www.ihmc.us/users/phayes
Received on Wednesday, 27 February 2013 01:30:39 UTC