Xref: utzoo rec.humor:18356 rec.humor.d:1513 comp.misc:4863
Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cornell!uw-beaver!tektronix!tekcrl!tekgvs!tekirl!jackg
From: jackg@tekirl.TEK.COM (Jack Gjovaag;6160;50-321;LP=A)
Newsgroups: rec.humor,rec.humor.d,comp.misc
Subject: Re: Looking for Computer Folklore
Message-ID: <4575@tekgvs.GVS.TEK.COM>
Date: 31 Jan 89 17:18:10 GMT
References: <7143@pyr.gatech.EDU> <532@geovision.UUCP>
Sender: nobody@tekgvs.GVS.TEK.COM
Reply-To: jackg@tekirl.TEK.COM (Jack Gjovaag)
Organization: Tektronix, Inc., Beaverton,  OR.
Lines: 38

In article <532@geovision.UUCP> geovision!pt writes:

>2) We had an IBM cluster controller controlling some 3270 terminals.  We
>paid $5000 for an upgrade that would allow more users to be connected to
>the controller.  The IBM service rep came in and REMOVED a board, that
>was put there to deliberately slow things down.

In a similar vein, the GE 415 and 425 CPUs were identical except that
the 415 had an extra wire that slowed the clock down a bit.  To upgrade
to the 425, after paying your money, the wire was removed.  Some
users knew about this and one of them made up a realistic looking
letter supposedly from GE saying something to the effect :
"CAUTION. Do not remove the wire from pin 4AB to 7FL in the CPU
enclosure.  This wire is located approximately 7 inches up
from the bottom of the backplane in bay 2 and should not be removed
by using a GE 112-3 wire unwrapping tool, first not removing the
wrapping from 4AB, then pulling the wire from under the other
wiring to its bound end at 7FL, followed by not unwrapping the
bound end from 7FL.  Not removing this wire will result in the
normal clockspeed which is 1.6 times slower than with the wire
removed and will not cause corresponding increases in system
throughput."  Naturally most of these wires got removed.

Another interesting but kludgy fix to a problem came from a user
of an IBM 7044.  The 7044 had a HALT instruction that stopped the
CPU clock.  The user was doing some realtime processing or something
of the sort and didn't like the idea of the CPU ever being able
to stop itself.  He asked IBM how much it would cost to disable
the instruction and they gave him some large quote which contained
the implicit message "We don't want to do it and this price is set
high enough to make you change your mind about the request."
The user didn't want to pay the money so he fixed up a photodiode
over the light on the console that was on when the CPU was running
and hooked it up to a solenoid that would push the RUN button whenever
the light went out.  The cost was a couple of dollars.

  Jack Gjovaag
  Tek Labs