- From: Sandro Hawke <sandro@w3.org>
- Date: Fri, 13 Jul 2007 09:19:40 -0400
- To: Axel Polleres <axel.polleres@deri.org>
- Cc: "Boley, Harold" <Harold.Boley@nrc-cnrc.gc.ca>, Sandro Hawke <sandro@w3.org>, public-rif-wg@w3.org
> >> class NAF
> >>
> >> subclass StratifiedNaf
> >>
> >> subclass WellFoundedNaf
> >>
> >> subclass StableModelNaf
>
> Hmmm, you mean to say that I need to specify which naf I use ALWAYS?
> In this case we rather call the first class
> PerfectModelNaf
> since this is what it is on the pure semantic basis. The perfect model
> semantics is simply only defined for stratified programs. Systems which
> implement it -- and it seems that there are no systems which only allow
> strat naf and do check is -- norlmally deploy perfect model semantics
> and just fault in case the program is unstratified.
Ah, that makes sense. One might also call it "StratifiedOnlyNaf",
maybe...
> To call it stratnaf sounds like stratnaf would be something *different*
> from well-founded or stable negation, which it is not in case the
> program is stratified. Why I wanted to have the subclassing in first
> place was because my understanding was that I wanted to express that in
> the stratified case both well-founded and stable coincide.
I don't think the class hierarchy gives you that in any formal way. Can
you say what you were thinking/hoping the class hierarchy would provide
here? Did you want RIF documents to sometimes be written not saying
which kind of NAF was intended?
-- Sandro
Received on Friday, 13 July 2007 13:20:43 UTC