Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84 (Fortune 01.1b1); site graffiti.UUCP
Path: utzoo!linus!philabs!cmcl2!seismo!ut-sally!ut-ngp!shell!graffiti!peter
From: peter@graffiti.UUCP (Peter da Silva)
Newsgroups: net.arch
Subject: Re: Stack architectures - why not?
Message-ID: <226@graffiti.UUCP>
Date: Sun, 22-Sep-85 09:44:20 EDT
Article-I.D.: graffiti.226
Posted: Sun Sep 22 09:44:20 1985
Date-Received: Wed, 25-Sep-85 10:35:28 EDT
References: <172@myriasa.UUCP> <12200021@orstcs.UUCP>
Organization: The Power Elite, Houston, TX
Lines: 16

> 1) A very heavy reliance on RPN.  Although routines can be written to get 
> around this, and some people actually like RPN, this is still a dubious 
> feature at best.  Algebraic and other forms of higher math don't work well
> in RPN, and (although I don't know about you) RPN makes logic notation far
> more confusing than it needs to be.  Since I use logic functions in a lot of
> situations, this is a hassle.  Even the If...Then structure is made more
> complex than necessary because of forcing it into RPN.

You seem to be confusing the language you used and machine architecture.
That's a job for the compiler writer. Have you ever written one? I have done
a couple and I'd like to say that for sheer ease of code generation from
algebraic expressions nothing beats a stack machine. You can compile to
FORTH in nothing flat. A great way of getting the language design working
before dealing with grubby registers for the real thing. Unfortunately the
project was killed before we could do the real thing, but that was in spite of
being able to show managers the thing working in FORTH.