Re: PRNGs and testers.

[David Honig <honig@sprynet.com>]

At 11:30 AM 10/21/98 +0100, Mok-Kong Shen wrote:
>
>Not answerable.  (Reason: substitute car, HIFI, medical doctor, etc. 
>for 'software PRNG').

Nice metaphors.

>> 2. Are there any software implementations to test the randomness of a
>>    PRNG ? I've looked at Diehard - is there anything else?
>
>There are tests that one can implement with reasonable effort.
>A test relevant in cryptology is Maurer's test.

Recall that Ueli's Universal Statistical Test
is valid only for real sources of entropy.
PRNGs have zero entropy asymptotically ;-)