Re: Good enough for crypto?

From: Scott Wilber (swilber_at_comscire.com)
Date: 12/02/03 

Date: 1 Dec 2003 22:14:47 -0800

Paul Crowley <paul@JUNKCATCHER.ciphergoth.org> wrote in message news:<877k1hohbv.fsf@saltationism.subnet.hedonism.cluefactory.org.uk>...
> swilber@comscire.com (Scott Wilber) writes:
> > This is certainly correct. Firstly, no analysis can determine the
> > entropy of a source by analyzing its output sequence, and secondly,
> > The Mersenne Twister followed by our infinitely recursive,
> > multiple-feedback shift register (IRMFSR) stirring function produces
> > sequences that have passed all our tests cumulative to 10's of
> > trillions of bits. (By the way, NO other generator has passed our
> > tests so far)
>
> From which we may conclude that either
>
> 1) you have a publishable attack
> on practically every cryptosystem in existence, or
>
> 2) your tests are
> so bogus that a PRNG which always passes them is in fact defective, or
>
> 3) there are a lot of very popular generators you haven't tried.
>
> Beware when reading replies to this that Wilber is not above modifying
> the text he quotes.

There are no questions or useful information in this post: B O R I N G

And of course, we have not tested every generator, although of the
"very popular" ones we have tested, all have failed - except, as
mentioned, Mersenne Twister. Our tester (RNGmeter) is online for free
download. Anyone can get the same results.

Bye.

Quantcast