Path: utzoo!censor!geac!lethe!torsqnt!jarvis.csri.toronto.edu!mailrus!uunet!algor2.algorists.com!jeffrey
From: jeffrey@algor2.algorists.com (Jeffrey Kegler)
Newsgroups: comp.lang.c
Subject: Re: sorting algorithms
Message-ID: <1989Nov17.185514.15726@algor2.algorists.com>
Date: 17 Nov 89 18:55:14 GMT
References: <2835@phred.UUCP> <11564@smoke.BRL.MIL> <89Nov9.114339est.2758@neat.cs.toronto.edu> <6916@ficc.uu.net>
Reply-To: jeffrey@algor2.ALGORISTS.COM (Jeffrey Kegler)
Organization: Algorists, Inc.
Lines: 32

In article <6916@ficc.uu.net> peter@ficc.uu.net (Peter da Silva) writes:
>I like Horowitz and Sahni, "Fundamentals of Computer Algorithms". They're
>a lot more readable than Knuth, and use a high-level pseudocode for
>everything. (whatever posessed Knuth to express algorithms in assembly?)

It is important to remember that Knuth's sorting book is over 15 years
old in a field that moves very quickly.  In particular, the decision
to use assembly, while by current standards, very ill-advised, looked
better then.  C use was not then widespead (the UNIX kernel had just
been rewritten in C from assembler), so what was Knuth to do?  Use
FORTRAN?  ALGOL?  Either probably would have been better, actually,
than MIX, but at least you can see that Knuth's choice was a rational
one at the time.

The "standard" answer to algorithms questions is "read Knuth" (a
variant of RTFM), but I feel this is more of a dismissal than an
answer.  Knuth is far more often cited or used as a status symbol then
read.  You need not have been around this business as long as I have
to have seen Knuth's volumes behind the desks of many a poor excuse
for a programmer.  Knuth is still an important source, but is no
longer up to date and was always hard to read.  I often wonder if his
problems do more to scare readers off than increase their
understanding of the field.

Knuth should not be anyone's first book on algorithms.  Learning
algorithms from Knuth is almost as bad as learning physics from Newton
in the Latin original.
-- 

Jeffrey Kegler, Independent UNIX Consultant, Algorists, Inc.
jeffrey@algor2.ALGORISTS.COM or uunet!algor2!jeffrey
1762 Wainwright DR, Reston VA 22090