Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!linus!decvax!harpo!utah-cs!shebs From: shebs@utah-cs.UUCP (Stanley Shebs) Newsgroups: net.politics Subject: Re: refuse the cruise - no thanks Message-ID: <2044@utah-cs.UUCP> Date: Sat, 22-Oct-83 21:50:50 EDT Article-I.D.: utah-cs.2044 Posted: Sat Oct 22 21:50:50 1983 Date-Received: Tue, 25-Oct-83 01:26:30 EDT References: uccsrg.221 Lines: 87 Rely on cruise missiles for a deterrent? Perhaps I can relate a little of my impressions gained from some time spend working (at Boeing) on the software for the Air-Launched Cruise Missile (ALCM). 1. They are not invincible. The original cruise missile theory (from 1972 or thereabouts) was to have many small slow missiles flying low to the ground, to escape radar detection. Unfortunately, we came up with the idea of AWACS, among other things, and that meant that flying low was no longer such a good idea (hence the stealth cruise missiles). Also, a large collection of portable radar-equipped anti-aircraft batteries can shoot down the missiles, and the missiles are pre-programmed, so anyone with access to SAC's library of missions can find out where they are going to be flying (the missions are top secret, of course, but human pilots can deviate from a course if attacked - the missiles can't). Too, the cruise missiles depend on a set of navigation maps (stored onboard) for accuracy. If a Russian farmer digs a fishpond where a map is supposed to be, any missiles overflying it have a substantial chance of getting lost - and once it's disoriented, there's almost no chance of regaining the correct course (few bomb worries though - the warhead won't go off anyplace but the target, after a fairly complicated set of conditions are met (at least that's the theory (ask Ted Jardine (ssc-vax!tjj) for details))). 2. They are not reliable. The test flights had some rather entertaining failures - missiles not starting and falling into the sea, running into hillsides, etc, and some of these were due to the software, which is all written in assembly language. My particular niche was in some of the ground planning software development, which tries to determine if a given mission can actually be flown successfully. For various reasons, this is all probabilistic, and so one expects that some percentage of the missiles will crash on the way. The lower you try to fly at, the worse (the ALCM has less performance than the average Cessna, and can't really climb over hills without crashing). The software we worked on was awful. The original authors snuck away without documenting their code, and it was strictly guesswork as to how things actually worked. There's nothing quite like writing large documents about code which one does not understand, and to know that nobody really cares whether the document is correct or not. Our released software gave rise to over 5000 bug reports, which works out to one bug for every several lines of Fortran code. The flight software is mostly debugged now, but it's been worked on for quite a few years.... but there's no sort of proof that it's correct, and most of the original authors are long gone. On the other hand, as far as I know the ground software was only responsible for one test flight failure - the missile used a marginal navigation map and got totally lost... In fact, we told each other that, in a nuclear war, the safest place in the world was at the North Pole, and the worst was 0 0, somewhere in Equatorial Africa. Why? Well, none of the missiles have been tested over the North Pole (for obvious reasons), but for navigation, it's a singularity. We were sure that all the fancy inertial navgation systems and their associated computers would divide by 0 or overflow or something, reset themselves, and fly to 0 0. There were a lot of reliability rumors, but they were mostly about the hardware, and I'm not competent to speak of them. The original article raised some other issues which are more general in nature, relating to how cruise missiles are to be used. For instance, If they are to be launched from airplanes (B-52, B-1B, whatever - 747's are impractical - don't meet the requirements for maneuverability, range, etc), either the airplanes must be kept continually airborne, or else sit on runways. If the Russians launch a strike against airbases from subs, then there's 5-10 minutes (?) to get the planes airborne and out of range of the blast (that's where they get the term "scrambling" :-) ). Continually airborne launchers were part of the "Big Bird" proposal for MX, but it would require many planes to be aloft at one time, and that they each have a great range. This kind of thing is just the tip of the iceberg. Cruise missile accuracies mean that they can be used against silos, but that's a ground burst, which means fallout all over the world, but they're so slow that they could only be used in a retributive strike, which means air bursts over cities ("city-busting", as one old engineer put it), but then the Russians have reloadable silos, so ground strikes are necessary anyway, and so on ad nauseam. Much time was spent at work arguing strategies, and methods, and defenses, and it made me aware of the enormous complexity of the whole issue - not to mention the moral angles. There were a surprising number of "fellow travelers" among the software group, which is perhaps another reason to doubt the reliability of ALCM... stan the l.h. utah-cs!shebs