Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site ucbcad.UUCP
Path: utzoo!linus!decvax!tektronix!ucbcad!notes
From: notes@ucbcad.UUCP
Newsgroups: net.politics
Subject: Re: Re: Do: Re: Mi: BTL takes the lead - (nf)
Message-ID: <891@ucbcad.UUCP>
Date: Sun, 4-Dec-83 02:00:58 EST
Article-I.D.: ucbcad.891
Posted: Sun Dec  4 02:00:58 1983
Date-Received: Thu, 1-Dec-83 04:36:02 EST
Sender: notes@ucbcad.UUCP
Organization: UC Berkeley CAD Group
Lines: 33

#R:pucc-h:-36800:ucbesvax:7500055:000:1511
ucbesvax!turner    Nov 20 17:03:00 1983

I had a different reaction to Hoare's Turing Award essay.  He was saying
that Ada (tm) was so huge and complex that it would be impossible to
verify a compiler to the level of tolerance required in applications
like nuclear weapons systems.

Well, I say that Ada is a damn convenient target for him, and critics like
him--it means that he doesn't have to criticize *programmers*.  He need
not say to weapons-systems developers "look: how do you know you're smart
enough?"  He can just ask "how do you know your compiler will carry out
your specifications," and thus dodge the issue of the quality of those
specifications--and the minds that produced them.

That sort of indirect criticism is, I think, self-invalidating.  In a way,
he *was* criticizing programmers--for thinking that they could properly
define and implement Ada compilers.  But the examples he gave from his
own experience were pretty unconvincing.  (I urge everyone who has not
read his Turing Award lecture to read it, by the way.)

Certainly, he can't argue that systems developed in Ada will be more
reliable than the systems currently developed and deployed in assembly
language!

I think he should have been a little more forthright, even at the risk of
offending a larger segment of the CS community.  If he really doesn't
think that programmers should work on nuclear weapons systems, he should
*say* so--not criticize them when they propose order-of-magnitude improve-
ments in methodology.
---
Michael Turner (ucbvax!ucbesvax.turner)