Path: utzoo!attcan!uunet!nih-csl!lhc!mimsy!midway!linac!pacific.mps.ohio-state.edu!zaphod.mps.ohio-state.edu!usc!chaph.usc.edu!aludra.usc.edu!wilber
From: wilber@aludra.usc.edu (John Wilber)
Newsgroups: comp.edu
Subject: Re: Textbook for Intro. Comp. Theory course
Summary: Theoretical Computer Science != Computer Science
Message-ID: <12234@chaph.usc.edu>
Date: 28 Sep 90 06:54:08 GMT
References: <392sis-a@massey.ac.nz> <12007@chaph.usc.edu> <29541:Sep1820:32:0790@kramden.acf.nyu.edu>
Sender: news@chaph.usc.edu
Organization: University of Southern California, Los Angeles, CA
Lines: 58
Nntp-Posting-Host: aludra.usc.edu

In article <29541:Sep1820:32:0790@kramden.acf.nyu.edu> brnstnd@kramden.acf.nyu.edu (Dan Bernstein) writes:
>In article <12007@chaph.usc.edu> wilber@aludra.usc.edu (John Wilber) writes:
>> In article <392sis-a@massey.ac.nz> C.Eagle@massey.ac.nz writes:
>> >I  have been asked  to  survey netters  on textbooks  for  a course on
>> >introductory computer  theory.
>> As far as I have been able to tell, the concept of a "good" textbook on 
>> "computer theory" is an oxymoron.  Since such books are invariably
>> written by computer science theoreticians and computer science theoreticians 
>> invariably know almost nothing about computers (as one would expect 
>> from mathematicians) the textbooks I have seen have all been dismally
>> uninformed about the realities of computer technology (and even good
>> theory for that matter).

>Be serious. The classic example of a computer text written by a
>mathematician is Knuth's Art of Computer Programming, which is arguably
>the best set of books in either mathematics or computer science. Do you
>seriously believe that ACP is ``uninformed about the realities of
>computer technology''?

Well, though ACP does have some useful stuff in it, it certainly DOES 
display a disregard for the way things really work.  Why are his examples
all in languages nobody really uses?  It's kind of like a sex manual
written by a nun.  Sure, there's some useful information in there, but
it leaves some things to be desired that a non-mathematician would
have put in there.

>I don't want to start a flame war on the merits of mathematics versus
>computer science, the irrelevance of computer science to computer
>programming, or the common belief that the half of math degrees that
>migrated to CS represent the bottom half. I just want to point out that
>your snubs will be resented by mathematicians around the world. 

Do you think that I didn't expect that? ;-)

>You may
>be right that computer science theoreticians invariably know almost
>nothing about computers, 

In fact I *AM* right about it.

>but you shouldn't imply that theoretical CS has
>anything more to do with mathematics than linguistics does.

Why not?  I don't see that there is a stronger link.  Could you explain
why you think this is true?

I would like to clarify my position here again to prevent unnecessary flames.
I don't think that theoretical mathematics is always completely 
useless to the advancement of computer science.  I believe it has a place.
My gripe is that many mathematicians place theoretical mathematics at the
center of the academic study of computer science, a position not warranted
by the practical impact or importance of mathematical theory.  The other
gripe is that these mathematicians have a very limited understanding and
appreciation for the rest of the field (both the electronics end, and the
software engineering end (and any other end you care to mention)).  I believe
this results from the fact that people who can build circuits and write
big programs go out into the world and do those things, while the
theoreticians remain in academia to "define" computer science as they see fit.