David G. Wiseman

Death Watch on Wall Street

This really happened, long ago in the early Seventies, at a stock
brokerage house that has long since been merged or bought out or
eaten by the larger sharks.

We had a pair of IBM 370/155's (1 mip, 1Mb mainframes) to run the
business applications of the brokerage house.  When plug-compatible
memory and plug-compatible disk drives came along, we were one of the
first shops east of the Rockies to install both.  All at once, no
trials, no testing, no phased cutover.  Over a weekend, all of the
memory and all of the disk drives were removed and the OEM parts cabled
in. This was the vendor's first installation outside California, and
there were a few bugs that needed to be found that were being
found at the customer site.

As the bugs were being found, system reliability declined, and the
discetionary parts of the workload were shed -- no programmer jobs
were run until the daily batch processing was complete. If bug-shooting
took a while, the programmers sat around .... or we went out to drink
lunch (as we knew that we wouldn't have to perform any more useful
work that day). The reliability was a concern to the programmers,
as there was a rule that had been made in the great Wall Street
backoffice crunch of the late Sixties, that any brokerage that did not
process its daily work for so many days would be out of business.
If the processing did not get completed, we would lose our jobs.
Hence, a certain amount of gallows humor . . .

One morning, on my way into work, I stopped at the local head shop (a
common business in the early seventies) and picked up a black candle.
If we weren't going to have any system availability, we might at least
have an altar .... With the knack for ceremony that only programmers
with untested code can muster, we stuck the candle in someone's ashtray
and placed it unlit on top of the 1403 printer (q.v.) that had been placed
with a card reader in a utility room next to the "programmer's area"
as a way to keep the unwashed hordes out of the machine room and the
production areas. When we had figured out that the daily cycle from
the day before was still processing, we headed out early to drink lunch.

As we came boisterously back in the middle of the afternoon, one of the
secretaries pulled me aside and asked if I was the one who had brought
in the candle.  When I said yes, she told me to never admit that,
in her best, "don't ever say this if you want to keep your job here" tone.

I found out later . . . .

After we had gone to lunch, the office hippie had waltzed in, and found
the candle on the printer and the bulk of the progamming staff gone. He
closed the door to the utility room, lit the candle, turned off the
lights, and lay down on the carpet to enjoy the silence and the shadows
that the candle flame cast on the acoustic tile. After he had done this,
the daily cycle finished, and the operators started the programmer's card
reader and printer.

After the operators had started the programmer's card reader and printer,
the VP came downstairs to check out programmer morale. Instead of finding
the programmers, he found our office hippie on his back on the floor
with the lights out and a candle burning on top of the printer. The card
reader was reading cards (it had an input tray that held 2500 cards,
and all of our tests had been stacked up), and the printer was printing.
The VP knew that the cover went up automatically on the printer when it
ran out of paper or jammed. So, for the sake of safety, he went to blow
out the candle. He cupped his hand around the candle flame to blow it out.
It was a **very** cheap candle, not a dinner taper.  When he blew it out,
he got wax on his hand, wax on the sleeve of his silk shirt, and wax on
the sleeve of his wool suit. And the machine chose that moment to go
down, and was down for two days solid.

The moral . . . .
						- Elliott Frank

Ha, ha, ha. Take me back to [ the alphabetic list ] [ the date-ordered list ].