Path: utzoo!attcan!uunet!kddlab!trl!rdmei!iegva1!creamy!oucom2!digigw!gday
From: gday@digigw.lab.digital.co.jp (Gordon Day)
Newsgroups: comp.software-eng
Subject: Re: Assembler and teaching Software engineering
Summary: I have PDP-8 on MY resume and am damn proud of it!
Message-ID: <308@digigw.lab.digital.co.jp>
Date: 20 Mar 90 02:58:21 GMT
References: <891@gistdev.gist.com> <1003@dms.UUCP> <1940@opus.cs.mcgill.ca>
Organization: Digital Electronics Corp., Osaka, Japan
Lines: 50

In article <1940@opus.cs.mcgill.ca>, peterd@opus.cs.mcgill.ca (Peter Deutsch) writes:

[ an extremely long, but worthwhile set of observations removed]

>  So what do others think about using such
> a course as a first exposure to software engineering. And
> if it is such a bad idea, how _should_ it be done?
> 
> Enquiring minds want to know....
> 

Well, I'm not sure about the benefit to the understanding of software
engineering concepts of the type of course you describe (I'm one of the OOPS
crowd =:-)), unless we also assume that a clear understanding of the Von
Neumann architecture benefits the student as a base for other concepts in the
future (e.g. recursion, task switching, etc. (yes, I know, these are 
applications, not S.E. per se)).  I am a little concerned about your comment
about the 500 line assembly program.  In my experience, carefull design in
assembler code might get you partway there, but debugging (A REAL TIME WASTER
FOR STUDENTS!!) occupies most of the students time.  Now, you may have an 
effective debugging setup for them, in which case it's not so bad, but if
we are talking about S.E., isn't assembly teaching students the nasty things
as well?  global data space, zero security, "just stuff a jmp in there, son",
etc.  I don't necessarily disagree with you, but please define the area of S.E.
you are addressing.

Now on to the interesting bit.  I am a recent Univ. grad in Computer Science
and in my second year I had a course just as you describe.  We used PDP-8
assembler running on a simulator and our big project (in Pascal, NOT a.lang.)
was to write the assembler for a subset of the language (you can imagine how
many instructions were left...).  We worked up from handcoding in octal to
assembly language, culminating in designing an assembler.  This is one of 3
CS courses I took out of about 20 that still stays in my mind as one of the
most enjoyable and valuable in University.  I don't really list PDP-8 on my
resume, but then who really needs to list assembly on a resume now?  However,
given the background of the course I took I easily learned 68000 and 8088 in 
a very short time.  

My understanding of it all "really works" did _really_ help in my later O/S
courses and in my understanding the underlying stack operations in C (very
necessary to really master the language, IMHO).  One qualification however,
I learned my languages in the following order: BASIC, Pascal, PL/1, _then_
assembly.  I strongly feel that the first (or even second) course a student
hits should NOT be assembler.  It just ain't friendly, folks.

gday@digital.co.jp%uunet.uu.net (Japan's not so bad, mate) =:!