Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!rutgers!ames!ucbcad!ucbvax!decvax!ima!johnl
From: johnl@ima.UUCP (John R. Levine)
Newsgroups: comp.arch
Subject: Re: Definition:  Kludge
Message-ID: <545@ima.UUCP>
Date: Thu, 16-Apr-87 10:28:17 EST
Article-I.D.: ima.545
Posted: Thu Apr 16 10:28:17 1987
Date-Received: Sun, 19-Apr-87 06:25:11 EST
References: <1736@im4u.UUCP>
Reply-To: johnl@ima.UUCP (John R. Levine)
Followup-To: misc.misc
Distribution: na
Organization: Javelin Software Corporation
Lines: 21

In article <1736@im4u.UUCP> suhler@im4u.UUCP (Paul A. Suhler) writes:
>Someone asked for a definition of "kludge," but I never saw the
>real one.  ...

The term goes way back, Datamation ran a funny series about the Kludge
Komputer Korp. in the 1950s. In any event, a kludge is like pornography in
that, as the justice said, it's hard to define but you know it when you see
it. (There are those who would say that a true kludge is by definition
pornographic, but that's beside the point.) It seems to me that for
something to be a kludge it 1) has to work, at least sort of, and 2) has to
work for the wrong reason. Anything which is elegant or generalizable is
immediately disqualified. Hardware kludges frequently involve things like
sending signals over spare ground wires. Software kludges involve the
equivalent, e.g. passing a pointer from one routine to another as a
floating point number because there was already a routine to do that and
you found 9 bytes of patch space close enough to it to hack in a 9-byte
sequence using short addresses.
-- 
John R. Levine, Javelin Software Corp., Cambridge MA +1 617 494 1400
{ ihnp4 | decvax | cbosgd | harvard | yale }!ima!johnl, Levine@YALE.something
Where is Richard Nixon now that we need him?