Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!ucsd!ucbvax!WORLD.STD.COM!bzs
From: bzs@WORLD.STD.COM (Barry Shein)
Newsgroups: comp.society.futures
Subject: (pssst...fortran?)
Message-ID: <9009131702.AA20599@world.std.com>
Date: 13 Sep 90 17:02:29 GMT
References: <1990Sep11.043926.26580@syd.dit.CSIRO.AU>
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 81


Chris Reynolds writes..
>All these languages have advantages and disadvantages - and one of
>the problems is that most of the critisms come from people who are
>wedded to a particular language - and this "determines what they
>can think about" in a manner which means they cannot see the
>other sides viewpoint. For instance, most academic critism of cobol
>is by people who have never worked (and have no wish to work)
>in the kind of environment for which cobol was designed.

Actually, that's not true. The MIS community is highly critical of
their current environment and have been for several years now, and a
lot of that criticism has been levelled at Cobol. Some developments
over the last few years have helped, but it seems people are still
quite desparate.

The killer problem in most MIS environments is called "backlog".

Traditionally, in these shops, if a manager wanted a printout of how
many widgets were being sold, broken down by region, and that wasn't a
standard report, then they'd have to get someone to write a cobol
program to produce the report. For simple requests like this, systems
such as CICS helped. Oftentimes your average middle manager did not
have access to things like CICS tho that has changed over the years.

So you had this pool of cobol programmers turning out report and other
requests. And the backlog (time from request to delivery of the
report) went from days, to weeks, to months in many shops. It became
something larger and more threatening in MIS terminology, from mere
backlog to "THE BACKLOG".

I have dealt with this sort of thing personally. A dean in need of
data to present in a few days (often with only that much warning) in
front of a court (e.g. an affirmative action suit against the
university) being told it would take at least 6 weeks to get the
necessary information. Sure, raise holy hell, unfortunately dozens of
such high-priority needs have already raised holy hell, 6 weeks *is*
high priority!

Database systems with report generators helped. Of course, in a lot of
shops middle-managers still weren't able to directly access these
systems so the same requests were sent to people who satisfied them
with database tools rather than cobol, producing about the same
reports but faster.

To hear the MIS crowd tell it, there are still enough requests that
database front-ends don't quite satisfy (e.g. on-line screen data
entry systems) that the database stuff only worked for a while and the
backlog simply moved to other areas (now that everyone had a screen on
their desk they needed a way to use those screens, data on demand was
becoming a reality.) These jobs, to this day, are still largely done
in Cobol with various packages (code generators for screen layouts
etc.)

I think if you polled the MIS management community on this current
mileu you'd get a resounding "it sucks".

So, just as much as it is a folly perhaps of CS types that one
programming language is somehow inherently better (or, more to the
point, that Cobol & Fortran are to be universally vilified), it is
also a fallacy to say "to each his own". It has a wonderful ring of
peacemaking, but ignores the facts of the matter.

I could draw similar pictures about the Fortran world. Besides
Fortran's weakness in being manageable for large projects (things like
COMMON and EQUIVALENCE just wreak havoc when things get large) there
is also, ultimately, a real problem with numerical analysis often
leading engineers and scientists not trained in such issues down the
garden (or, better, garbage) path.

There are real problems in these fields, the world has changed around
them and these tools just weren't designed for that changed world.

"To each his own" is not an answer either. The first thing one has to
shake is that any of this is a religious issue, only graduate students
think that.

        -Barry Shein

Software Tool & Die    | {xylogics,uunet}!world!bzs | bzs@world.std.com
Purveyors to the Trade | Voice: 617-739-0202        | Login: 617-739-WRLD