Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!rutgers!im4u!ut-sally!bcm!svedberg!rick
From: rick@svedberg.bcm.tmc.edu (Richard H. Miller)
Newsgroups: comp.edu
Subject: Re: Computer Science: where it belongs
Message-ID: <421@uni2.bcm.tmc.edu>
Date: Fri, 13-Nov-87 17:44:13 EST
Article-I.D.: uni2.421
Posted: Fri Nov 13 17:44:13 1987
Date-Received: Sun, 15-Nov-87 16:01:31 EST
References: <469@ndsuvax.UUCP> <16118@clyde.ATT.COM> <180@spock.UUCP> <1759@pdn.UUCP>
Sender: usenet@bcm.tmc.edu
Lines: 29
Summary: bridges & architects & large systems

In article <1759@pdn.UUCP>, reggie@pdn.UUCP (George W. Leach) writes:

> >Here is thought for what it's worth.  I have always felt that one of the
> >failings of the software field is that people don't spend much time studying
> >large existing programs.  There are two points here.

Quit true. One of the worst failings I have found with recent CS graduates 
it their total lack of exposure to maintenance of large systems. They have
never really seen a large system. (Such is usually found on most mainframe
systems.) They really don't seem to grasp the concepts required to maintain
this type of system as well as understanding the requirements for modifying 
software projects. They also have never really been exposed to a variety of
different software systems. This seems to be especially true of CS departments
which teach ONLY Unix. CS should try to expose graduates to a variety of
systems and styles. (Much of the time, we have to spend some months to 
teach them that UNIX does not always provide the best way of solving a system
software problem.)

>        One of the worst aspects of comparing software to anything else, like
> building bridges or cars is that each and every bridge that has ever been
> build will serve the same function.

One minor point here, bridge building (or design actually) has usually 
been a very individualistic approach. Although the function is the same,
(enable traffic to cross an obstacle), the design is VERY individual (what
type of bridge, environment of the bridge, etc.) It is similar to our problem
in that we have to provide a functional system but the exact environment
will vary. Thus bridge design is actually a fairly good analogy to software
engineering, but has several decades of development beyond Sy d