Path: utzoo!mnetor!uunet!husc6!cmcl2!brl-adm!umd5!eneevax!mimsy!aplcen!jhunix!ins_adgh
From: ins_adgh@jhunix.HCF.JHU.EDU (David George Heath)
Newsgroups: comp.lang.forth
Subject: Re: Novix chip with Computer Cowboys software by Calvin Moore
Message-ID: <6375@jhunix.HCF.JHU.EDU>
Date: 6 May 88 16:12:27 GMT
References: <17430@glacier.STANFORD.EDU> <460@ghidrah.tessi.UUCP>
Reply-To: ins_adgh@jhunix.UUCP (David Heath)
Organization: Johns Hopkins Univ. Computing Ctr.
Lines: 29

Recently someone posted an article on a silly use of the Novix forth chip.
My roommate built a system using the NC4000 and I thought you might be 
interested in it.

He has the chip running at ~7 mHz. Novix claims the chip will run at 8mHz,
but the clock timing starts to get very touchy. Even at 4mHz, the system
seems to run forth faster than a 6mHz AT runs assembly (I know, that's not
saying much).

He has added a floppy drive, a monochrome graphics board (using NEC7220),
and is adding a hard disk controller (using Hitachi HD63463). I have used
the system numerous times and enjoy it. The compiler is very fast -
loading takes almost no time. The problems with the multiply (and several
other bugs) aren't much of a problem, if you know they exist. Because
forth is so modular, fixes are easy. Also, the lack of a do...loop
construct is not a limitation - if you really want one, you can write it
yourself. The for...next loop is sufficient for most cases.

There is one serious limitation. Branches must always branch within the
same 4K block. Say, there is a branch at location 1000 hex. It can't 
branch back to 0fff hex! The compiler as supplied doesn't check for 
such problems. The compiler was meant to be small, however, so it's
not perfect.

All in all, the NC4000 is a nice chip, and is capable of being part of a
nice system.

"Wearing two gloves seemed so ordinary"
David Heath    (ins_adgh@jhunix.hcf.jhu.edu)