- From: Danny Ayers <danny.ayers@gmail.com>
- Date: Fri, 15 Jan 2010 10:36:33 +0100
- To: Jiří Procházka <ojirio@gmail.com>
- Cc: Pat Hayes <phayes@ihmc.us>, Geoff Chappell <geoff@sover.net>, Dan Brickley <danbri@danbri.org>, Steve Harris <steve.harris@garlik.com>, Semantic Web <semantic-web@w3.org>
2010/1/15 Jiří Procházka <ojirio@gmail.com>: > PS: Actually there is one thing which would aid bootstrapped nature of RDF: > > ex:contains rdf:type rdf:Property ; > rdfs:comment "property for stating that a reified triple is part of > graph" ; > rdfs:range rdf:Statement . That's rather neat. Pat, I must admit I'm finding myself more aligned with Jiří's attitude here, I don't really see a problem with RDF being a LISP-like language, though I take the point of it being declarative rather than procedural. We do have things like XSLT to guide us (which isn't too dissimilar to SPARQL CONSTRUCTs). While there does seem to be utility to OWL/OWL2 and full-blown DLs in general, I would posit that the thing that is more immediate is tying this stuff to the Web, instead of using programming cycles to discover inferences, doing a HTTP GET and getting more information that way. Clearly it's possible to expose RDBMSs as triples, but ordered Lists & things are still a pain - might it not be possible to sprinkle a little more sugar over the syntax to make things like n-ary relations/n-tuples more straightforward, without any major hacking on the model? (caveat - IANAL either). Cheers, Danny. -- http://danny.ayers.name
Received on Friday, 15 January 2010 09:37:07 UTC