- From: Graham Klyne <Graham.Klyne@MIMEsweeper.com>
- Date: Thu, 14 Mar 2002 10:20:08 +0000
- To: "Tim Berners-Lee" <timbl@w3.org>
- Cc: <w3c-rdfcore-wg@w3.org>
At 04:31 PM 3/13/02 -0500, Tim Berners-Lee wrote: >I accept the form of these literals only in the context of an attempt to >formalize the >RDF spec as it is, but not as an endorsement of the way forward. For the >future, >I hope that RDF will move toward unicode strings as primitives, and >langauges as properties. { "chat"en = [lang:en "chat"].} I, for one, am beginning to feel as if I'm losing sight of the larger picture here (WG-fatigue, or something?). I, too, very much prefer this "interpretation property" approach and do not exactly recall what considerations have lead us down the other route... ... my concern, and the reason I am responding, is that I think the current approach to datatyping semantics will make it more difficult to make the future transition suggested here. I.e. that literals always denote themselves in the semantic domain. It would be easier to see how the transition could be effected under datatyping schemes that have literals denote values, as then one could imagine defining an equivalence between the following: ex:someSubj ex:value "chat"en . and ex:someSubj ex:value [ lang:en "chat" ] . I raise this with some trepidation, as to fix this might involve pulling up our stake-in-the-ground regarding datatyping. So, my question to the WG is: is this something we should be concerned about? #g ------------------- Graham Klyne <GK@NineByNine.org>
Received on Thursday, 14 March 2002 06:34:58 UTC