- From: Bijan Parsia <bparsia@cs.man.ac.uk>
- Date: Fri, 27 Jul 2007 22:17:55 +0100
- To: Harry Halpin <hhalpin@ibiblio.org>
- Cc: Bruce D'Arcus <bdarcus@gmail.com>, bnowack@appmosphere.com, Garret Wilson <garret@globalmentor.com>, Semantic Web <semantic-web@w3.org>
On Jul 26, 2007, at 2:39 PM, Harry Halpin wrote: > Bruce D'Arcus wrote: [sni[] >> I'm now totally lost. >> >> Everyone I have talked to has discouraged use of rdf:Seq. AFAIK, you >> cannot really query it reliably with SPARQL either, though am not >> sure >> of that. > Is that true? [snip] Which? That everyone bruce has talked to discourages use of rdf:Seq? I don't know. I certainly do so discourage. But I discourage use of rdf:List too :) I don't know about "reliably". Someone later in the thread gave a pretty good query for getting at the values of containers. Ordering generally is done by term sorting, so _1 _2, etc. should work fine. There are implementation specific extensions (both actual and possible) e.g. http://thefigtrees.net/lee/sw/sparql-faq#transitiv8 Back in the day, some engines had trouble with large numbers of properties, so the coined properties in a seq could cause trouble if one had no special ordering of them (e.g., if you naively mapped properties to tables, that could suck). But that strikes me as implementor lameness, which is also to be discouraged. :) Cheers, Bijan.
Received on Friday, 27 July 2007 21:18:03 UTC