more on HLLs

Jecel Mattos de Assumpcao Jr. jecel@lsi.usp.br
Wed, 8 Mar 1995 00:20:56 -0300


Strange... I still haven't received back the last few messages
I sent. Anyway, I used Mosaic to take a peek at the latest HLL
pages. I am really impressed.

Here is a reference that you might find interesting ( I haven't
read it yet, so I can't be sure :-)  :

  Irons, E. T.
  "Experience with an extensible language"
  Communications of the ACM, vol. 13 no. 1
  pp. 31-40, January 1970

And this is about a very *cool* language that I haven't heard
of since ( I read it, but never fully understood it ) :

  Barber, C. Bradford
  "Echonet - Part 1: A Flexible Programming System"
  Byte Magazine, vol. 8 no. 9
  pp. 356-373, September 1983

  Barber, C. Bradford
  "Echonet - Part 2: The Compiler"
  Byte Magazine, vol. 8 no. 10
  pp. 384-395, October 1983

Check it out - you will be amazed! Of course, I would not recomend
something like that for Tunes, but it can give you neat ideas. I
think I will ask in comp.lang.misc if anyone knows what became of
it.

The first reference inspired Alan Kay in the design of Smalltalk-72.
This language allowed you to define your own syntax! Unfortunately,
this created some problems of context dependent code and slow
interepreters. In Smalltalk-76 this idea was abandoned with the most
popular "home grown" syntaxes becoming built in.

I think it is possible to use the flexible syntax idea of Smalltalk-72
without its problems using a two step process. In the first step, the
text is parsed into LLL via a "symbolic interpretation" using special
methods in the objects themselves ( not quite a meta-level ). So objects
would be able to extend the syntax ( once past a mind twisting
bootstrap, of course ) used in their, and other object's, methods. I
did something like this in my language Troy but mixed this up with
LLL execution. That had the Smalltalk-72 problem of making source
impossible to compile.

This idea would be the perfect way to win obfuscated code contests.
At least it would make macros obsolete :-)

-- Jecel Mattos de Assumpcao Jr
   Laboratorio de Sistemas Integraveis - University of Sao Paulo - Brazil
   jecel@lsi.usp.br    http://www.lsi.usp.br/~jecel/jecel.html