- From: Michael Kifer <kifer@cs.sunysb.edu>
- Date: Wed, 03 May 2006 11:46:52 -0400
- To: Francois Bry <bry@ifi.lmu.de>
- Cc: "Peter F. Patel-Schneider" <pfps@inf.unibz.it>, paula.patranjan@ifi.lmu.de, public-rif-wg@w3.org, francois.bry@pms.ifi.lmu.de
>
> Michael Kifer wrote:
> >> We have also just seen a proposal to add counted quantifiers.
> >>
> >
> > This was thrown in but to make it fly one would have to incorporate these
> > quantifiers into the answer-set and well-founded semantics. (My understanding
> > is that Francois main interest is in CWA-based dialects.)
> > I think a better and a more natural way to express this kind of features is to
> > use aggregate functions. (For CWA-based dialects.)
> >
> C pounting quantifiers" can be expressed by aggregate functions. As
> shorthand, or syntactic sugaring, they aree very uimportant *in practice*.
>
> François
If you just mean a shortcut for the count aggregate function then it is
fine. The only question is whether this new syntax for counting quantifiers
is really needed. IMO, expressions like 3 > count{?X|condition(?X)} or 7 <
count{?X|condition2(?X)} are much clearer and don't require any new syntax,
since aggregates are going to be there anyway.
--michael
Received on Wednesday, 3 May 2006 15:47:07 UTC