- From: Bijan Parsia <bparsia@cs.man.ac.uk>
- Date: Tue, 30 Nov 2010 15:29:42 +0000
- To: Enrico Franconi <franconi@inf.unibz.it>
- Cc: Birte Glimm <birte.glimm@comlab.ox.ac.uk>, SPARQL Working Group <public-rdf-dawg@w3.org>
On 30 Nov 2010, at 14:51, Enrico Franconi wrote: > On 30 Nov 2010, at 15:45, Birte Glimm wrote: > >> Yes, and going from one regime to a more expressive one should just >> give you more answers I think. > > Any logical, philosophical, formal argument for this? You know, I'd find it helpful if you moderated your tone just a bit. The general intuition I find people have is that going "more expressive" means "more answers". This intuition is quite robust across many sorts of users 9in my experience). It even works for various groups of languages in the semantic web family. Typically where it breaks down (e.g., RDFS to OWL DL) it does so in reasonably obscure parts which can be usefully neglected. However, people tend to use bNodes as "local" names quite reliably and they use them a fair bit. People do use SPARQL to interrogate graphs with bNodes. When moving to a more expressive entailment regime I think it's quite baffling to find out either that your query is no longer syntactically valid, or to lose answers. Thus, from a overall usability perspective, I don't think violating this principle, esp. on a common operation, is a good idea. There should be a large user gain for that pain. Thus far, I have only one bit of hearsay evidence (from you) that there are any users of any OWL profile who make routine good use of non-distinguished variables. I would be interested to know more about such users. Cheers, Bijan.
Received on Tuesday, 30 November 2010 15:29:59 UTC