🚀 go-pugleaf

Another story of the PDP-11/20 at the University of Illinois.

During the semester I learned Pascal at the U of Ill in 1973, we used
Ian Stocks' barely debugged compiler on a PDP-11/23 with DEC-tape, disk,
card-reader and printer.  The disk was a head-per-track disk, and as
the semester wore on, more and more of the heads crashed, leaving us
with less and less space for temporary files and the like.  Eventually,
the system ground to a halt and DEC's service people were called.

I was there when they opened and disassembled the drive.  They took
the disk off its hub and the working surface was deeply grooved over
each head, with bare aluminum showing on each recording track.  The
service man looked at it, noted that the machine wasn't under service
contract, and offered two solutions:

  1) For lots of money, get a new disk and mount it on the drive.
      This might take a week or two.

  2) Turn over the platter.  The second surface wasn't certified, but
      it looked OK, and he could get it running the same day.

The man in charge chose the latter, so the maintenance man set the
disk aside, vacuumed the remains of the head crash out of the drive,
went over the heads with a felt pad to polish them up, and then, very
carefully, buffed and polished the never-used (and never intended to be
used) side of the disk platter before remounting it in the drive.

It worked like a charm!  Modern disks may store more megabytes, but
they arent as forgiving as the old kind.
						Doug Jones
						jo...@herky.cs.uiowa.edu