Randomness is randomness. However, the important thing is that the randomness is not known, and the randomness cannot be retrieved.

If you're worried about passing the tests on a subset of the data, take the subset that you're thinking of and run the tests on it. If it's truly random, it will pass no matter what subset you take.

I don't know what the characteristics of the encrypted stream would be (to an eavesdropper) if the encrypted stream is completely random. Theoretically it shouldn't make a difference, but I haven't heard of any research on the topic. (Most research attempts to apply a cryptographically-generated sequence of apparently-random bytes to an information stream, thus making the stream appear random. It / should/ just appear like any other encrypted information stream.)

-Kyle H

On May 8, 2007, at 12:26 PM, Robert Welz wrote:

Hello!

I have a hardware rng for the parallel port of the pc. The device ( the rp1 from http://www.true-random.com ) produces a constant stream of random
bytes at 1 MB/s and it passes the diehard tests as well as some others
mentioned on the website.

My question now is: If I take the stream of data, slicing it into N chunks
of 500 kB over some kind of ring buffer and send the chunk over the
network to feed N computers /dev/random will that suffice and keep the
entropy so that those tests will still be passed? I am not the
mathematician, we had affine geometry instead of probabilistic theory at
school ;)

Second question: If I use ssh for the transmission, will that have
influence on the quality of the data; I mean encrypting random data will
that lower the entropy anyhow?

Thank you for your time,

Robert





______________________________________________________________________
OpenSSL Project                                 http://www.openssl.org
User Support Mailing List                    [email protected]
Automated List Manager                           [EMAIL PROTECTED]

______________________________________________________________________
OpenSSL Project                                 http://www.openssl.org
User Support Mailing List                    [email protected]
Automated List Manager                           [EMAIL PROTECTED]

Reply via email to