[ih] Internet sounds

[Carsten Bormann]

On 2022-04-26, at 21:55, Scott Bradner via Internet-history <internet-history at elists.isoc.org> wrote:
> 
> I learned programming on the PDP-1 prototype at Information International Inc (Tripple-I) 
> it had a Macintosh amp tied to the low order bit of the accumulator register - easy to tell
> when you had a programming loop :-)

In this spirit — this wasn’t the “early Internet”, but still ~ 1984:

Our X.25/SLIP gateway (*) was running on a PC, and each time a packet entered, the code applied a voltage to the PC loudspeaker (~ 1 line of code), and each time processing was completed, the voltage was removed (1 more).  The resulting sound of a galloping herd (**) helped tremendously with debugging the code.

This hack really came to life when our Datex-P line (German X.25) developed a problem that occasionally made the line really slow.  
An intermittent problem, essentially unfixable.
But the sound was audibly different during these episodes, with breaks that indicated retransmission.
With that immediate feedback through the changing soundscape, we were able to pin down the problem:

The line went bad each time the snow on the streets was melting, and recovered when it became dry again.  No problems with rain, just with melting snow.
Deutsche Bundespost were incredulous about our problem report, but quickly found the defect and fixed it.

This may be not so impressive today with grafana and all, but at the time we were mightily impressed with how simple audio feedback was improving our problem solving skills.

Grüße, Carsten

(*) I know…  
but it was the quickest we could nail together with our 4.2BSD host.
(**) It helped that packets/s rates at the time still were in the audible range :-)