Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83 based; site homxb.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxr!mhuxt!houxm!homxb!afa
From: afa@homxb.UUCP (A.ACKERMAN)
Newsgroups: net.singles
Subject: Re: Mechanics and Musicians are Apt to Make Good Programmers
Message-ID: <1237@homxb.UUCP>
Date: Sun, 9-Feb-86 14:51:45 EST
Article-I.D.: homxb.1237
Posted: Sun Feb  9 14:51:45 1986
Date-Received: Tue, 11-Feb-86 06:23:54 EST
References: <1711@ittatc.ATC.ITT.UUCP> <461@hoptoad.uucp>
Organization: AT&T Bell Labs, Holmdel NJ
Lines: 70

********************************************************************

Although the name A. ACKERMAN appears in the header of this posting,
Bob McIlree is the author and is soley responsible for its content.

*********************************************************************

Well, at least the U.S. military found out something useful
with all the billions of $$ Ronnie dumps into its sewers..

I completely agree that the majority of programmers that are any
good at all DON'T HAVE A COMPUTER SCIENCE DEGREE.

I got my Bachelor's in Physics. I haven't done "physics" for a
living since I left school, but have worked in computers since
leaving campus. An old boss of mine once told me that the best
programmers hold degrees in a subject other than CS. This was
further borne out by the fact that my old employer hired math/physics
types for programming/systems prog., SW Engineer, etc. His primary
point: Programming languages can be taught to any reasonably
competent person in another subject, programming style as well.
The difficulty lies in the fact that progs. with exposure to
another field make better application programmers than persons
with straight CS degrees.

A typical CS degree program will teach its students much about
OS's, data base theory, the underlying mathmatical foundations,
etc. But they will also teach VERY LITTLE on exactly HOW this
information is supposed to be applied into "real-life" situations.
For example, my first job after graduating was to finish the
computerized control system that drove a very large nuclear
accelerator used for semiconductor implantation. This system
served as the operators console, and was a DEC LSI 11/23 under
RT-11 using Pascal.

To be fair, I had taken a couple of programming classes in
college, but that was it as far as CS was concerned. The primary
focus of that job was the fact that, with by backround in physics,
I was able to translate the needs of physicists and EE's , which
were expressed to me in THEIR terms, into working code onto the
computer. As I look back, a person with a CS-only degree would have
had a very difficult time with the position.

This is my third job since college, and in all three, a lot of fellow
employees are doing excellent work in all phases of computing
with educational credentails in another subject, not only
technical or physiccal science subjects but English, History,
and Political Science, to name a few.

As counterpoint, one of the programmers I supervise (he had the
demo blow up in front of us as I explained in my last posting)
has very high grades in his BSCS program from a respected school.
This guy couldn't program something useful if his life depended
upon it when he first got here, but with some time (and intense
tutorials, code reviews, etc.) he is making good progress. However,
he had to be taught the relevance of the application involved
in detail, and his CS degree was of very little help.

This is not to say that individuals with CS degrees have no
future in computing, quite the contrary. I would, however,
advise all soon-to-be CS graduates to have more than a
cursory knowledge of the type of work they want to do
after graduation. If that involves applications programming,
it's best to bone up on the subject thouroghly, then turn
yourself loose on the industry involved.


					Bob McIlree
					AT&T Bell Labs
					MIddletown, NJ