Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!clyde!rutgers!mit-eddie!uw-beaver!ubc-vision!alberta!jim
From: jim@alberta.UUCP
Newsgroups: comp.sys.mac
Subject: Re: kludge
Message-ID: <779@cadomin.UUCP>
Date: Tue, 28-Apr-87 17:49:26 EDT
Article-I.D.: cadomin.779
Posted: Tue Apr 28 17:49:26 1987
Date-Received: Fri, 1-May-87 01:02:00 EDT
Organization: U. of Alberta, Edmonton, AB
Lines: 48

> Help!   Help!   Help!
>
> Will someone explain it to me what exactly a "kludge" is?
>
> Thanks,
>
> Paul

Sometime around 1960 IBM produced the 1620.  There were many things
wrong with that machine but nothing wrong enough to prevent it from
being used.

It had core memory, which was just about it's only saving grace.  Its
arithmetic was decimal to arbitrary precision from arithmetic tables which
were accessible to the program.  Mag tape was 7 track even parity, memory
was odd parity, paper tape was odd parity which meant that tape feed (ie.
blank tape) was "all holes punched"  (this weakened it so much that it
usually didn't manage to survive more that two or three readings).  The
paper tape reader was a classic Rube Goldberg device (It read from the
inside of the roll as I recall).

Shortly after Datamation published an article entitled "How to Design
a Kludge".  It was an hilarious article, which I fervently wish I had
saved (I didn't).  In it they defined a "kludge" - central to that
definition was the property that it was distressing to use but would
do the job (thereby ensuring that it would indeed be used).

Other properties of Kludges were that they had to be expensive to
produce and maintain.  The ideal Kludge had inexpensive modules (eg.
mag. tape, cpu etc) but that the box required to make the whole mess
work together cost ten times the rest put together (just part of its
being distressing)  One of the design rules was "rigid departmentalization"
eg. never let the person designing the mag tape talk to the guy who's
designing the main memory"

The term stuck and more articles were published about the progress of
the "Kludge Komputer Korporation".  eg. shortly after TSS was announced
we were treated to an article about the "Super Kludge Time Sharing System"
which could support one interactive user.  Data rate was measured in BPF
(Bauds per fortnight).  Then there was the Kludge OS which reached the
epitome of all operating systems - viz. it spent 100% of its time on system
activities.

I haven't seen any recent article on the "Kludge Komputer Korporation",
but then I haven't been looking either.  Certainly there's lots of material
to work with.

	Jim Easton (..!alberta!jim)