- From: Jos De_Roo <jos.deroo@agfa.com>
- Date: Thu, 29 Jul 2004 00:30:51 +0200
- To: "Dan Connolly <connolly" <connolly@w3.org>
- Cc: RDF Data Access Working Group <public-rdf-dawg@w3.org>, Steve Harris <S.W.Harris@ecs.soton.ac.uk>
DanC wrote: > On Wed, 2004-07-28 at 10:10, Steve Harris wrote: >> On Wed, Jul 28, 2004 at 09:49:41AM -0500, Dan Connolly wrote: >>> On Wed, 2004-07-28 at 09:40, Steve Harris wrote: >>>> I agree with these points wholeheartedly, but I have concern over the >>>> difficulty and scalability issues of implementing anything to handle RDF >>>> collections. Are there any large-ish RDF stores that handle these >>>> intelligently or as a special case? >>> >>> cwm has a list:in built-in; i.e. it supports queries over >>> a virtual graph that has list:in triples inferred. >>> It's no speed daemon, though. >> >> This is the kind of solution I was worried about, a list of lenght N needs >> N*(N+1) / 2 inferred triples, > > Only if you're forward-chaining. Euler would only infer the ones it > needs, I think. Yes, but for instance asking for all list:in solutions and their proofs we get memory trouble with lists above 80 items, at least while using recursive rules such as {?L rdf:first ?X} => {?X list:in ?L}. {?L rdf:rest ?B. ?X list:in ?B} => {?X list:in ?L}. > Anyway... I'm not arguing strongly in favor of adding special support > for collections. I just think it merits a place in our issues list. > Folks who wonder "why doesn't the new-fangled DAWG-BARKING-QL from W3C > support collections explicitly?" should get an explicit answer if > they go poring thru the WG's records. agreed >> which quickly becomes unmanageable, eg. >> MusicBrainz has containers with several thousand members. Admittedly thier >> not collections, so the problem is not as bad, but its a reasonable >> thing to want to store in a collection. >> >> - Steve > -- > Dan Connolly, W3C http://www.w3.org/People/Connolly/ -- Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/
Received on Wednesday, 28 July 2004 18:31:47 UTC