[SCL] Re: Report on Common Logic

[John F. Sowa]

Murray,

I agree with Pat that we have to get a document finished
by Dec. 17th and the version of the semantics posted on the
CL web site is an adequate basis for that.

>Are things actually stable enough for that yet? Don't questions
>about seqvars have to be answered definitively first? Are there
>any other outstanding issues that you guys need to agree upon?

We always fight about issues.  But we understand one another
well enough that we can sit down and work with the other guy's
semantics if we decide to do so.

I'm the one who has been resisting seqvars for a long time,
and I admit that I would prefer to see sequences defined as
first-class entities.  But I can define a CG-like syntax for
either version, and the same abstract syntax could be used as
a basis for an XML version.  After we finish the Dec 17th
document, we can take a look at other options, which would
replace one well-defined version of the semantics by another
equally well-defined version.  Both versions would map to
concrete syntaxes in a way that most casual users would not
even notice a difference.

>After that's done, an XML syntax still isn't completely
>straightforward. Like any language, you need to decide exactly
>what you're trying to do with it. Just *one* of the questions
>I brought up early on was "web-enabled" or not? Use of URIs or
>not? Any extensibility, so it can be the basis of other languages?

Those answers were decided long ago:  web enabled, yes;
URLs, yes; extensibility, yes; and basis for most common
declarative languages, yes.

>Does it have to be RDF-based, or can it be a new language?

The RDF triple-style is too restrictive to support full CL.
But the RDF triples can be adopted as a subset of CLML, which
would need a more general syntax to support full FOL.

>If there's to be an XML version of CL, a list of requirements
>should be the first thing created, which, if you remember, was
>the first part of the XCL proposal I created way back when, in
>the section on "Design Goals":

Two important requirements:  support full CL, and support
the RDF subset.  Whether the OWL subset can be supported
in the OWL notation is not clear, since the full FOL notation
is much more general.  One option is to inlcude OWL as a
redundant option, which could also be supported as a subset
of the full notation for FOL.

If you have a list of requirements, please send them
to SCL list.

> ... Until then, there could be three or four or
>more XML grammars that would "do" CL. First question is, what do you
>really want from an XML version? Hopefully not just to be able to
>say "there is an XML version." What are the goals?

For political purposes, I regard the ability to say
"There is an XML version" as extremely important.

But as I have said in a note to CG list, I expect the N3
notation to overtake the RDF notation as the most popular
way of marking up web pages.  See the N3 primer:

   http://www.w3.org/2000/10/swap/Primer

As long as N3 is defined by a concrete syntax that conforms
to the CL abstract syntax, I really don't care whether the
programmers choose N3 or RDF (and I would bet that they are
more likely to choose N3).

John