- From: Jeremy Carroll <jjc@hplb.hpl.hp.com>
- Date: Thu, 03 Jul 2003 13:01:07 +0100
- To: Graham Klyne <gk@ninebynine.org>
- CC: pat hayes <phayes@ihmc.us>, w3c-rdfcore-wg@w3.org, Brian_McBride <bwm@hplb.hpl.hp.com>, Martin Duerst <duerst@w3.org>
Graham Klyne wrote: > > FWIW, I think Pat's proposal is marginally easier to implement than what > we have at the present, and considerably easier than some approaches > that have been suggested. > > (Just an opinion based on my implementation perspective.) For an implementor who is implementing datatypes then the current solution requires *no work* except in the parser, and the defn of an L2V mapping which is essentially the identity. Pat's solution requires work all over the place, and does not require less work in the parser. Implementing from scratch, without datatypes, and not implementing datatypes at all, I would concede that Pat's solution is possibly easier. Jeremy
Received on Thursday, 3 July 2003 08:02:01 UTC