Re: A proposal for a unitary RIF phase 1

> On Tue, 2006-06-06 at 12:59 -0400, Michael Kifer wrote:
> > > On Sun, 2006-05-28 at 09:07 -0500, Dan Connolly wrote:
> > > > On May 27, 2006, at 6:34 PM, Michael Kifer wrote:
> > > > > Peter,
> > > > >
> > > > > It is no big deal to be unitary by restricting the language to Datalog.
> > > > > You don't even need to limit it to a function-free sublanguage. In our
> > > > > roadmap the language was unitary also up to this point.
> > > > >
> > > > > The issue is how to build such a system in an extensible way so that it
> > > > > could be extended to satisfy most of the RIF requirements.
> > > > 
> > > > Which requirements?
> > > 
> > > I'd really appreciate an answer, please.
> > > 
> > > Which (candidate) requirement(s) do you have in mind
> > > there, Michael?
> > 
> > Sorry, I thought it was a rhetorical question.
> 
> Oops; I'll try to be more clear in the future...
> 
> > The issues are integration with NAF, constraints, production and active
> > rules.  All these are requirements coming from the group members.
> 
> I'm having trouble finding them among the requirements materials on
> the ftf agenda
>   http://www.w3.org/2005/rules/wg/wiki/F2F3
> 
> which are, as far as I can tell...
>  http://www.w3.org/2005/rules/wg/wiki/Rulesystem_Arrangement_Framework
> and
>  http://www.w3.org/2005/rules/wg/wiki/UCR/Critical_Factors_Analysis

Both of them talk about all these issues.
For instance, in
http://www.w3.org/2005/rules/wg/wiki/UCR/Critical_Factors_Analysis there is
a link to open issues at the end. These are the issues that will need to be
incorporated into the main document. Most of these arise from multiple use
cases, which are on the wiki.
And, of course, requirements like the ones I mentioned, have been
extensively discussed on the mailing list.

> I see "FOL vs. other semantics" but that doesn't really state a testable
> requirement related to NAF.
> 
> Some of the others are easier to spot... "The RIF should
> support production rules." I don't know how to measure/test that one.
> 
> Does the "Extensible Design" proposal
>   http://lists.w3.org/Archives/Public/public-rif-wg/2006Apr/0068.html
> meet these requirements?

Yes, it attempts to tackle these points.

> 
> Hmm... I think I have a lot of study to do before I can understand
> that proposal very well.
> 
> Do you have a few minutes to give some examples of NAF, constraints,
> production, and active rules in that design?

I am leaving for the F2F in a few mins. If you are going to be there then I
will explain the problems in person.

> 
> 
> I feel that partial understanding
> aka
> http://www.w3.org/2005/rules/wg/wiki/Sound_reasoning_with_unknown_dialects
> is pretty important, and as far as I can tell, it conflicts with NAF.

Not with NAF -- it conflicts with everything we know about, because nobody
has any idea what this is supposed to be.

> As Sandro and are are the only evident supporters, I wonder if this
> partial understanding requirement will even get ftf time. I'll have to
> take another look at the use cases
> http://www.w3.org/2005/rules/wg/wiki/UCR/Use_Cases
> and see if any of them argues for partial understanding.

We can require anything we want, but at least we need to understand what it
is that we are requiring. In my mind, this is an interesting research
issue, but not more. With the other issues we, at least, have a fighting
chance, since they are well defined and the technologies are more or less
understood.

> 
> Darn; I don't see links from any of the use cases to requirements...

The cases were done before the requirements, and nobody updated them.
(Because the requirements are still in flux, I guess.)


	--michael  

> 
> -- 
> Dan Connolly, W3C http://www.w3.org/People/Connolly/
> D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E
> 
>

Received on Tuesday, 6 June 2006 18:20:42 UTC