Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/5/84; site baylor.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!allegra!mit-eddie!think!harvard!seismo!rochester!rocksanne!sunybcs!kitty!baylor!peter
From: peter@baylor.UUCP (Peter da Silva)
Newsgroups: net.lang.forth
Subject: Re: FIGIL DIGEST
Message-ID: <603@baylor.UUCP>
Date: Wed, 30-Oct-85 14:38:28 EST
Article-I.D.: baylor.603
Posted: Wed Oct 30 14:38:28 1985
Date-Received: Sun, 3-Nov-85 06:04:29 EST
References: <8510110938.AA09626@UCB-VAX.ARPA> <8510211813.AA09575@UCB-VAX>
Organization: The Power Elite, Houston, TX
Lines: 27

> I would suggest polyForth from Forth, Inc as a professional, consistent,
> and well-documented implementation of the language for the IBM-PC. I have
> several years of experience using various Forths in different projects

So have I, and PolyForth fails badly in one important respect:

	THE DEVELOPMENT SYSTEM DOES NOT RUN UNDER MS-DOS.

This means that:

	- You cannot run it under DoubleDos, or with Sidekick, or use any
	  of the other tools that try to make the PC a usable software
	  development environment.
	- You cannot easily transfer files between MS-DOS and Forth screens.
	- You cannot use it on anything but an IBM or a good clone.

Fig-Forth, while it requires you to write more software, helps you by letting
you work under MS-DOS... or whatever other O/S you are using. I wouldn't dream
of using another Forth that requires you to discard the native O/S after my
experiences with PolyForth.

Otherwise, you're quite right. The multitasker is nice, and the fact that the
same system is available for a wide variety of processors is extremely useful.
But if you want respect from your system, it's a total loss.
-- 
-- Peter da Silva
-- UUCP: ...!shell!{baylor,graffiti}!peter; MCI: PDASILVA; CIS: 70216,1076