- From: Steve Harris <S.W.Harris@ecs.soton.ac.uk>
- Date: Sat, 4 Jun 2005 10:04:44 +0100
- To: public-rdf-dawg@w3.org
On Fri, Jun 03, 2005 at 06:21:08PM +0100, Andy Seaborne wrote: > >Here I'm not only using the "representation" terminology, but > >also following Andy's later suggestion that FROM and FROM NAMED > >completely specify the dataset (which seems somewhat > >responsive to Yoshio's proposal, though it's not quite the same). > > > > > > > >>>Does any one have an example where it woudl nto be GETted in some way? > >> > >>Yes, in the case that the service is itsself a cache, and the service has > >>chosen to identify graphs by thier resolved URIs, or where there is > >>application level knowledge that the service should not resolve the URI > >>for security or performance reasons. > > > > > >Those cases seem consistent with the "representation" terminology, to > >me. > > Steve - is there any text you can suggest to b eclear about this. I've > noted Dan's suggested text in the document so it gets incorporated. Does > it address the point you raised? > > In particular, there shouldn't be an implication that the graph is obtained > at query execution time. Caching can always happen and one web cache is > the local agent. Yes, that seems fine. - Steve
Received on Saturday, 4 June 2005 09:04:59 UTC