Re: [Hampshire] OpenSSL in Debian is broken

Peter Alefounder <p_alefounder@???> wrote:
> Why are pseudo-random numbers used anyway? Is it
> beyond the wit of man to devise a bit of
> hardware that would produce genuine random
> numbers?

What is a "genuine" random number?

I browse a cryptography mailing list. A while ago,
there was a long, heated, and highly mathematical
debate on what exactly constitutes randomness, and
how to test for it. It appears that there is no
definite answer to either question, even from
noted researchers in the field.

> I am not an electronic engineer, but I
> understand that a Zener diode can be used to
> generate white noise.

True - RF noise figure meters use wideband noise
diodes to generate a test signal. They're
exceedingly expensive, though, and need a highly
stable bias voltage to operate.

As Hugo noted, there are some ICs that incorporate
noise sources - I suspect they use some variant on
the biased diode technique. The difficulty in the
design, I expect, is keeping the bias and output
signal clean. It's no good making a wideband white
noise source if your output signal contains
interference from the decidedly non-random clocks,
and buses.

Nick.

--
Nick Chalk ................. once a Radio Designer
Confidence is failing to understand the problem.