[SCL] late as expected

[John F. Sowa]

Murray and Pat,

We must support XML for political and practical reasons.
But not even the XML fanatics claim that XML is readable.

It is absolutely essential to make the presentation of
the semantics readable.  If Pat finds it convenient to
write the spec in KIF notation, then I am all for it,
at least for the Dec. 12th version.

For political reasons, we can throw in a remark that there
is a simple one-to-one relationship between the KIF spec
and the XML spec and show that relationship in the document.

Re language:  I realize that logicians since Tarski have
used a convention that any change to the vocabulary implies
a different "language".	 Unfortunately, that convention is
totally confusing for everybody else in the world.

If we start talking about KIF or CGIF or any other notation
as a collection of infinitely many "languages", we are going
to lose readers and potential supporters instantly.

Recommendation:  let's avoid the word "language" altogether
and use the word "syntax" instead.  Whenever Chris uses the word 
"language", we make a global change to "syntax S plus lexicon L."
And instead of saying "infinitely many languages", we say
"an open-ended number of lexicons for any given syntax".
That also allows us to talk about KIF, CGIF, and CLML with
the same lexicon, if we choose to do so.

Re Z:  Core Z is nothing but traditional infix predicate calculus
with types.  Full Z has a very large set of operators in their
ontology, but we do not have to define or use any of them.

If we are going to present an infix syntax, I would suggest
that we use untyped Z notation for full SCL and the typed
version of core Z as an example of how we can support types.

John