Path: utzoo!mnetor!uunet!husc6!ut-sally!utah-cs!defun.utah.edu!shebs
From: shebs%defun.utah.edu.uucp@utah-cs.UUCP (Stanley T. Shebs)
Newsgroups: comp.lang.misc
Subject: Re: Poor Algorithms
Message-ID: <5330@utah-cs.UUCP>
Date: 8 Mar 88 20:18:56 GMT
References: <3821@ihlpf.ATT.COM> <2791@enea.se> <404@tub.UUCP> <14476@oddjob.UChicago.EDU>
Sender: news@utah-cs.UUCP
Reply-To: shebs%defun.utah.edu.UUCP@utah-cs.UUCP (Stanley T. Shebs)
Organization: PASS Research Group
Lines: 22

In article <14476@oddjob.UChicago.EDU> matt@oddjob.UChicago.EDU (My Name Here) writes:
>Oliver Laumann quotes Butler W. Lampson:

>[...] A friend of mine had worked for a certain electronic
>instrument company doing numerical analysis for some engineers.  One
>engineer gave him a 200x200 matrix to invert.  (This was in the early
>70s when that was a big matrix.)  My friend did the job and noticed
>that the inverse looked a little familiar.  He asked the engineer,
>"Could this matrix represent a rotation of some sort?"  Answer: "Why,
>yes, you could consider it as one."  (The inverse was the transpose.)

I wonder if this is an example of a "software urban legend".  Forman Acton
mentions a very similar situation in his 1970 book on numerical analysis,
although there it is purported to be a first-person experience, and it
was claimed that 15 minutes of analysis were necessary to determine that
the matrix was orthogonal (thus representing a rotation).

Incidentally, this same book has some amusing flaming passages about
people who use too many cycles, and about the evils of recursion...

							stan shebs
							shebs@cs.utah.edu