[1534] in cryptography@c2.net mail archive
Re: Access to Plaintext: An Obvious Consequence
daemon@ATHENA.MIT.EDU (Dave Koontz)
Thu Sep 18 16:30:40 1997
From: koontz@netapp.com (Dave Koontz)
Date: Thu, 18 Sep 1997 13:07:49 -0700
To: karn@qualcomm.com, spector@zeitgeist.com
Cc: cryptography@c2.net, rah@shipwright.com
>But then they would also have to license/track common ICs in Radio
>shack... after all, one could build some pretty serious encoders out
>of a UART, a EEPROM and a DSP, no...? In a pinch your car's
>flight-control computer would make a good real-time system -- its got
>a fast MC88000-type (or the equivilent) CPU & lots of i/o ports....
Some of us are capable of building crypto out of non-CPU devices (as
in MSI).
I'm also toying around with the idea of building an electro-mechanical
implementation of DES as a project for my new milling machine.
Having worked on a KW-26/TSEC in the mid '70s, It would be a lot harder
to keep someone from building crypto that you think. These were based
on saturated core transformers driven by tube pulse drivers. The only
semiconductors in them were signal diodes. Of course, the MILSTD-188B
interface line drivers had some transistors as well. The tubes could
have FET transistors substituted. Working in 3D, I think you could implement
IDEA in around a cubic foot...
