What about software?
I guess the PDP-11 operating system TRANTOR would be my most unusual
piece.
"Trantor was created by Steve Orszag of the MIT Applied Math
department to access the CDC and Cray computers at NCAR for his fluid
dynamics research. NCAR expected people to access the systems with
an expensive piece of equipment that read punched cards, sent off the
data over a synchronous modem using a proprietary protocol, and then
sent output back to the printer. Not only was it expensive, who
wanted to use punch cards? So Orszag bought a PDP-11 and hired
undergrads to write software to use Emacs-like editing to create
programs on the local harddrive, submit them using the proprietary
protocol (which we sort of had to reverse engineer), and print and
graph the results. ECC, CBF, and I were the early developers (there
were a host of undergrad and grad student users as well, doing fluid
dynamics). Somehow this was cheaper than the archaic method.
Trantor was then a communications OS with built-in applications. We
did the development on MIT-MC using the PDP-11 assembler and
emulator."
I recently found a copy and sent it to the original authors.