NETWORK@FRSAC11.BITNET (02/02/88)
If you are not into Atari ST, or the Icon programming language,
just go to the next message of your favorite mailer.
+++++++++++++++++++++++++++
ICON 6.5 (Implementation revision (e)) for the ATARI ST
has been sent to some people.
As the files are quite large, and I do not wish to abuse the local
mailer, I will ask any further request to be redirected to the nearest
person that already have it.
Here is the list where the binaries have been sent:
Vitas P. rochester!ritcv!vxp6840 (uucp)
Denise ? rochester!ritcv!dsh3059 (uucp)
Mark Storin ihnp4!uwmcsd1!lakesys!mark (uucp)
Marty Wiedmeyer ihnp4!uwmcsd1!lakesys!martin (uucp)
Rich ? ihnp4!uwmcsd1!lakesys!rich (uucp)
Plinio Barbeito acm@valhalla.cs.ucla.edu (edu)
Bertrand Decouty decouty@frcicg71 (earn)
Klaus Hahn hahn_k@dmrhrz11 (earn)
Greg Onufer ucbvax!ucdavis!uop.edu!root (uucp)
Andres F. Moreno moreno@umn-cs.cs.umn.edu (edu)
Bob Bright bright@dalac (bitnet)
Matthias Moritz u608017@hnykun11 (earn)
Ruud van der Plaats vdplaats@hnykun53 (earn)
Gordon Joly gcj%@uk.ac.qmc.maths (uk)
Rick Fonorow ihnp4!ihlpe!orf (uucp)
Wilfried Bloemberg fonetiek@hnykun52 (earn)
Jerry Nowlin ihnp4!ihuxy!nowlin (uucp)
The sources will not be distributed by me, They have been sent to
the Icon project at U. of Arizona, and to the 2 original implementors.
Some of you asked for some documentation on ICON or on this peculiar
version: I can send upon request the article on Icon version 6 from
U. of Arizona, and the various files that come with the former
implementation.
The reference book for the language beeing the book:
Griswold, Ralph E. and Madge T. Griswold. The Icon Programming
Language, Prentice-Hall, Inc., Englewood Cliffs, New Jersey.
1983.
Send all inquiries to icon-project@arizona.edu
++++++++++++++++++++++++
Notes on this (e) implementation:
-It is done using MCC Lattice 3.04 C compiler, with sources coming
from U. of Az (Rick Fonorow & Jerry Nowlin, made it for the 3.03 C).
I added the support for various missing stuff, the arithmetic overflow
and Co-expression assembly code was furnished to me by Ralph Griswold.
Ralph did answer to all questions that allowed me to bring this code up.
-All binaries are called .TTP in order to be called from the desktop,
without using a shell if you dont like it, or if you are short in memory.
I used/developped the thing on a 1040ST Monochrome, with a 400K
ram-disk. I think iconx.ttp will work on 520ST, with TOS in ROM.
Iconx.ttp takes 200K for the heap, and 10K for the stack, plus data
and text segment.
-If you want to use icont: it will attempt to find itran/ilink/iconx
by looking at the ICON environment variable, that should be set to
the disk/directory where these can be found, then it will try for the
PATH variable, then for the root directory of all currently mounted
disk drives, in alphabetical order.
If you are invoking icont.ttp from the desktop, you can tell where are
the other binaries by the "-p disk:\dir" option.
All options to icont can be in upper or lower case.
-If you want to use a very large Icon program, you may have to
increase some memory area, in this case, you will have to tell the
startup code of Iconx, with the % option, before all other arguments
on the command line. {iconx> is currently equivalent to <iconx %200000>,
you may have to try %300000 or more. The number after % is the total
heap size, in byte.)
-If you want to use the system() call, find a shell that can do what
you want... I have'nt yet...
Iconx will look into the SHELL environment variable, it should be set
to the complete filename of the shell you want to invoke.
I have tried PCOMMAND, ASH, MSH, COMMAND.TOS, and the lowly last one
gave me best results... (most shells are not able to understand
arguments passed to them, if this is an internal command, they
always try to "fork" it...)
(In fact it does not need to be a shell, just any program can be
"forked", args passed, and the exit code will be returned back.)
(I am still looking for the sources of a good shell that I can hack
and make useful on the ST, whatever be your keyboard religion.)
I would appreciate any hints of bugs, to be reported, altough
it may take time to fix... may be version 7...
Regards,
Jean-Pierre H. Dumas
network@frsac11 (bitnet)
network%frsac11.bitnet@mitvma.mit.edu (arpanet)
dumas@sumex-aim.stanford.edu (arpanet)