Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!linus!decvax!decwrl!glacier!hplabs!ames!eugene
From: eugene@ames.UUCP (Eugene Miya)
Newsgroups: net.cse
Subject: Re: Value of Computer Science degree
Message-ID: <1404@ames.UUCP>
Date: Mon, 17-Feb-86 22:57:18 EST
Article-I.D.: ames.1404
Posted: Mon Feb 17 22:57:18 1986
Date-Received: Wed, 19-Feb-86 07:27:08 EST
References: <4514@kestrel.ARPA> <3407@nsc.UUCP> <4588@kestrel.ARPA>
Organization: NASA-Ames Research Center, Mtn. View, CA
Lines: 61

<3422@nsc.UUCP> <1405@gitpyr.UUCP>

A comment from a person without a completed CS degree (I've math degrees)
formerly working toward a phd who's gone to bat for CS degrees.

The CS degree is a valuable degree.  It should not be a math degree or
a EE degree or anyother degree.  I've have taken on more than two assignments
where the people hiring me didn't want a programmer/hacker/CS person,
what THEY really wanted was a radar engineer, or XYZ jockey who happened
to know how to program.  Letters have been posted to the CACM saying they
wanted their CS people to have experience with chemistry, physics, and English.
Fine.  But some of these people really want physicists who know programming,
chemists who know programming, and so forth.

There is no one single problem to pin point.  The problems lie in
the political structure of universities (departments), the nature of
the teaching materials, what's taught and so forth.

First, CS degrees are important, but they are not math degrees.  I know
several mathematicians who would make terrible programmers.  Knuth wrote
an AMM paper last March I think on the differences.  Basically, the 
"fault" here can be found in early programming books where people believed
they knew how to program after reading a FORTRAN book (or Pascal, etc.).
This is not true, and it is where the CS major should be able to step in.
I learned to "program" years ago with learning about data structuring.
I learned about DS and their tradeoffs.  It was really neat to ask a
friend (physicist) who asked about sorting techniques [after giving
him a couple of simple choices]: have you considered trees?  No,
what are those?  Well, let me illustrate...  The CS major should
be able to help educate us old dogs.

Second, too many CS departments are wedded to the EE or math dept.
This is a historical trend which dates to the first CS departments
in the late 60s and early 70s but is still found in many smaller
schools such as the Cal State system (I've found).  Why create
a new department?  What biases are created.  One of the biggest
problem areas for the next 20 years for those math centered
departments will be the emergence of stronger experimental programs.
CS is not very empirical by it's structure and composition.
I've had some discussons with Denning (PJ) here about it.
He wrote two editorials.  Great.  Does someone teach an experimental
methods class in a CS department?  Let me know.

Also, many of the tools used to teach algorithms, programming, design
of compilers, OSes, etc. are too low-level.  We teach "top-down
programming" in a bottom-up fashion!

There are other biases and factors, and spoke of two as well as an important
reason for you guys to exist: education.  If you only could see the number
of highly paid programmers I have met who don't know what a stack is,
what semaphores (or rendevzous) are, why studying compilers are useful
for doing numerical methods, that languages and operating systems and
networks are merging.....   What's worse is when people such as
these get into positions of decision making. This is why CS majors should
exist.

From the Rock of Ages Home for Retired Hackers:
--eugene miya
  NASA Ames Research Center
  {hplabs,ihnp4,dual,hao,decwrl,allegra}!ames!aurora!eugene
  eugene@ames-nas.ARPA