In article ,
Nova wrote:
Robert Bonomi wrote:

(Then you can just pick up the 'scrambled' deck, make a few passes through
the 'sorter', and have everything back in the right order.)

On the unit we had about 50% of the time it was a sorter. The other 50% it
was a shredder.

Well, there was the day I put a several-hundred card deck into the card reader
on the RJE station, hit the 'load' button, watched the cards go _into_ the
machine, and *NOTHING* come out.

Now, the path through the machine, from the input hopper, to the output
stacker was only about _four_ cards long.

The "impossible" had just happened.


I go report the matter to the computer operations staff, in the next room,
and the supervisor comes over (disbelivingly, I might add) to check out
the situation. (they knew me, *knew* I didn't 'make things up', but *this*
story _did_ stretch their credulity.)

He goes around to the back of the machine, opens it up, and breaks up,
laughing.The *entire* innards of the machine (*several* cubic feet) is
absolutely filled with crumpled up punch-cards. _MY_ job deck.

Apparently, the last 'pressure plate' covering the card path, had come up,
and as the cards 'shot' down the reader channel, they just flew up, past the
end of the channel, rather than being stopped at the end and pulled sideways
into the output hopper.

"Cards, Cards, *everywhere*, and not a byte to save."

I could (and *DID*) laugh about it at the time, because: (a) this happened
_after_ the cards went past the 'read' station in the machine, (b) the job
I was submitting was one that copied the data from the input cards to a
'permanent' disk file on the mainframe, and, most importantly, (c) that job
had run _successfully_.