- From: Michael Kifer <kifer@cs.sunysb.edu>
- Date: Mon, 28 Apr 2008 13:04:55 -0400
- To: Christian de Sainte Marie <csma@ilog.fr>
- Cc: Sandro Hawke <sandro@w3.org>, "Boley, Harold" <Harold.Boley@nrc-cnrc.gc.ca>, public-rif-wg@w3.org
> Michael Kifer wrote: > >> > >>Just thinking about what is a reasonable scope for this WG, and where to > >>stop... > > > > Why are you talking with riddles? Are you proposing to junk FLD? > > Hey, wait! No, not for a second! > > I was just wondering how far we should go to make FLD cover... Well, > that is exactly the question: FLD could potentially cover many things, > and I wondered where we should just stop worrying about dialects that > some people may want to develop in the future and whether FLD would be > appropriate for their purpose (including wrt naming such as 'rule' vs > 'group', 'item' or what else)... > > More specifically: I wonder if we should not just stop right where we > are, add a Rule construct, maybe even a Fact one, and leave FLD 1.0 be > FLD 1.0 until we need a FLD 2.0 that corrects it and/or go further. But I did not propose to extend what we have. I am just saying that dialects must match (the existing) FLD. If rules in BLD must have a special wrapper, then they should also have it in FLD. But in FLD, there are other top-level formulas, which are not rules. By the same logic, they should also have a wrapper. It makes little sense to proliferate these wrappers, so we should converge on some neutral name (if at all to have a wrapper for rules). Sandro proposed "element". Could be "unit", "formula", "singleton", "solitaire" :-) This was a separate thread of discussion, which got mixed in with the thread of whether to have such a rule wrapper at all. --michael > Christian > >
Received on Monday, 28 April 2008 17:05:48 UTC