- From: Dan Brickley <danbri@w3.org>
- Date: Mon, 28 Jul 2003 10:32:56 -0400
- To: Tim Bray <tbray@textuality.com>
- Cc: Jonathan Borden <jonathan@openhealth.org>, Norman Walsh <Norman.Walsh@Sun.COM>, www-tag@w3.org, pat hayes <phayes@ihmc.us>
* Tim Bray <tbray@textuality.com> [2003-07-27 20:55-0700] > > Jonathan Borden wrote: > > >So where does that leave us with respect to "document about weather in > >Oaxala" vs "concept weather in Oaxala" -- it seems to depend on how we > >define it. What prevents someone from saying: > > > ><http://weather.example.com/oaxaca> rdf:type web:document . > >or *instead* saying: > ><http://weather.example.com/oaxaca> rdf:type ex:weatherLocation . > > > >The SW treats URIs as opaque. The current Web doesn't care about what the > >range of HTTP URIs is. What is the actual physical purpose of making this > >distinction? > > > >Pat suggests that SW agents need to know. If so we can use assertions to > >tell them. > > Yes. This is *exactly* the question I've been trying to ask, but > haven't put it as clearly as Jonathan did. I just think that > inferencing the range of a URI based on its scheme feels fragile and > inflexible. Especially since the taxonomic division between what > Jonathan calls a web:document and other kind of things feels like just > one of a hundred interesting assertions one might want to make about > whatever a URI names, and for damn sure we can't infer all of them from > the syntax. -Tim Well said. Dan (hoping for a minimally constraining decision that allows RDF classes and properties within http:'s range)
Received on Monday, 28 July 2003 10:32:57 UTC