Joe Teagarden wrote:
...
> I'm retired from IBM and go so far back into the world of computing
> that carbon dating might be more in order than a calendar.
>
> I'm probably the guilty party for those crazy printer music programs.
>
> It started in 1953 with the IBM 701.  Several of us worked the mid-
> night shift with little to do and noticed that some of the diagnostic
> routines produced a frequency buzz.  Probing around with a small
> speaker attached to a o-scope output we discovered that several
> instructions could be used to reliably produce, and control to some
> extent, a "tone".  (We used the MQ bit in the multiplier instruction).
...

I was hit by the same disease, Joe, at a later age, again, in high
school circa 1974.

We had a Digital Equipment Corp. (DEC) PDP-8/E minicomputer.  From
somewhere we acquired a music program that would play music.  Similar
to Joe's program, for an A note, it would execute the CAF (Clear All
Flags) instruction 440 times a second, evenly spaced over time.  Since
the CAF instruction sent a pulse down a unique circuit line (the
initialize line), we could hear music simply by placing an AM radio
next to the front panel.

The program had a degree of sophistication to it.  It was capable of
playing four-part harmony!  Furthermore, it offered rudimentary volume
control for the individual voices.  Somehow, it figured out how the
multiple sine waves of four simultaneous notes should be played by
executing CAF instructions with various spaces in-between.  (I always
wondered about the algorithm.)

Like line-printer music, it was more like scratchy noise (I still have
a cassette tape somewhere), but I spent hours keying in everything from
Bach to Scott Joplin.  I probably drove the computer office workers
nuts!

C. Jim Cook