Path: utzoo!attcan!uunet!shakti!shri
From: shri@ncst.ernet.in (H.Shrikumar)
Newsgroups: comp.lang.forth
Subject: One Closure for myFORTH was Re: CATCH and THROW
Summary: One closure syntax is enuff ...
Message-ID: <955@shakti.ncst.ernet.in>
Date: 26 Sep 90 10:21:52 GMT
References: <9009191309.AA25742@ucbvax.Berkeley.EDU>
Reply-To: shri@shakti (H.Shrikumar )
Organization: National Centre for Software Technology, Bombay, INDIA
Lines: 53

Mikael Patel asks:

mip>> But why use exception in interpret state after all. Do you have an
mip>> example of this, Mitch?

Mitch Bradley replies:

wmb>The classic example is debugging.  When debugging, it is a pain to
wmb>have to remember that there are certain things that you can't just type
wmb>in at the keyboard and have them work the same way that the work inside
wmb>a colon definition.
wmb>
wmb>It is even more of a pain to *explain* to novices why certain things
wmb>don't work.  It is nicer if everything just works. 

   Seconded.

   Both from the stand-point to easy education, as well as being
more consistent. A different "Interpret" and "Compile" mode is not
too different from vi's "command" and "insert" modes.

   To add to wmb's observations about ' and ['], I also dislike "No
DO ... LOOP from Interpret Mode" and things like that. I enjoy UNIX
shell just due to these constructs, and I want them on my system too.
(Write a command, add a pipe or a while... (so-on) .. and when happy
bottle it in a shell-script.)

   The other gripe is the fact that IF ..THEN, DO ... LOOP etc.  are
all sort-a pre-fix-ish. Could that syntax be given a post-fix sugar ?

   I do like PostScript's { } construct, which makes the syntax much
cleaner.

   For a home-brew I am building, I had initially intended to do a 
quick-and-dirty FORTH, but now I am playing with a syntax cleaned up
according to the above.

   I have a feeling that precisely *one* context-free struture, (a
parenthesis, which is the only closure really needed) which *also* is
given lazy-evaluation semantics, (everything else being strict and
immediate) would give me what I want. (Interpret, Compile, block
structure, closures ...). I am experimenting with some example
pieces of code in myFORTH, a prototype of which I have built using
TILE (thanx Mikael Patel).

   Of course, this is non-standard ... but hopefully, I hope, FORTH
will evolve along these lines.

-- shrikumar ( shri@ncst.in )

PS: Now, if I can get myLOOP to be compiled as a tail-recursive
definition with myIF, with the tail-recursion being (trivially)
optimised out by the compiler ....