[SCL] REQUEST FOR FEDBACK: core syntax

[Jay Halcomb]

Comments below.

Jay
----- Original Message ----- 
From: "John F. Sowa" <sowa at bestweb.net>
To: <scl at philebus.tamu.edu>
Sent: Sunday, December 21, 2003 18:18
Subject: Re: [SCL] REQUEST FOR FEDBACK: core syntax


> Pat, Jay, Tanel, and Murray,
>
> We should give a clear statement of who or what
> is expected to read or write the SCL core syntax.

I agree.

>
>   1. If it's intended for people to read and write,
>      then human readability is essential.

Yes.

>
>   2. If it is intended as an intermediate language
>      for programs to read and write then simplicity
>      in the grammar is more important than human
>      readability.

But humans will always have to debug such stuff at some times, so that
simplicity in the grammar is not the final desideratum. Furthermore, since
computers are today much faster than of old, there's much less point in
stripping down syntaxes merely to speed up parsing time. I suggest either an
infix or prefix KIF, to assist human readability.

>
> I have been assuming that SCL core syntax is primarily
> intended for computational purposes, and people would
> prefer to use many other notations, such as controlled
> English (or French, or Estonian, or whatever) or some
> other previously designed language, such as infix
> predicate calculus, CGs, TMs, UML, or whatever.
>