One very easy test is to compress the produced random number binary files with various compression algorithms or programs. If the data is truly random, it should not be possible to achieve any compression. At least not if you include the compression program data size in the calculation. Well, I'm not sure whether this is such a good practical test or not.(?) ++ J On Fri, 2 Oct 1998, Clifford Heath wrote:
We have been asked by a customer if we have any tests that demonstrate the randomness of the SSLeay random number generator (augmented by some sound-card random number seeding that we wrote).
I'd like to find some standard implementation for testing randomness, but Schneier offers no help (other than a reference to Knuth Vol 2), and I don't know where else to turn.
I realise that cryptographic randomness requires unpredictability, and this quality depends upon closed-world assumptions about unknown individuals' predictive powers, but we have to live with that.
-- Clifford Heath http://www.osa.com.au/~cjh Open Software Associates Limited mailto:cjh@osa.com.au 29 Ringwood Street / PO Box 4414 Phone +613 9871 1694 Ringwood VIC 3134 AUSTRALIA Fax +613 9871 1711 ------------------------------------------------------------ Deploy Applications across the net, see http://www.osa.com