Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site cmcl2.UUCP
Path: utzoo!linus!decvax!harpo!floyd!cmcl2!ultra
From: ultra@cmcl2.UUCP
Newsgroups: net.columbia
Subject: Re: Computer question - (nf)
Message-ID: <809@cmcl2.UUCP>
Date: Tue, 13-Dec-83 00:31:55 EST
Article-I.D.: cmcl2.809
Posted: Tue Dec 13 00:31:55 1983
Date-Received: Wed, 14-Dec-83 02:28:06 EST
Sender: ultra@cmcl2.UUCP
Organization: New York University
Lines: 24

#R:isrnix:4300001:cmcl2:13800001:000:1009
cmcl2!ultra    Dec 13 00:31:00 1983

The shuttle processors are standard avionics processors, 16 bit, whose
design was fixed around 1975.  I.e., off-the-shelf (if your shelf contains
various parts for F-15s).  It is programmed in HAL/S, an ALGOL-like language.
There are no special features in the processor design itself (other than
MIL-spec chips): reliability is in the 4-way voting bus, and the 5th
backup processor running operating system and code from another manufacturer
(maybe Rockwell).  Like a PDP-11, but probably more like an AN/YUK-7.

There is a "Space Shuttle Operator's Handbook" you can get in bookstores
which has a reasonable level of detail, I think more than you can get
by paging through back issues of Aviation Week.  It is in general a good
book to get (I gave mine away to a project at MITRE which is doing some
knowledge-based systems demos for shuttle tasks such as crew activity 
planning and fuel loading).

	Lars Ericson 

	(direct personal queries to
		ARPANET:  Lars.Ericson@CMUA
		UseNet: ..cmcl2!csd1!ericson)