Re: PRNG
- From: user923005 <dcorbit@xxxxxxxxx>
- Date: Tue, 3 Mar 2009 12:01:17 -0800 (PST)
On Mar 3, 7:02 am, Phoenix <ribeiroa...@xxxxxxxxx> wrote:
On 3 Mar, 14:16, Guy Macon <http://www.GuyMacon.com/> wrote:
Phoenix wrote:
I would like to now the opinion experts about this algorithm created by
me, that is a non periodic Pseudo Random Number Generator.
No such thing exists. You can make the time between repeats
so large that for all practical purposes it never repeats,
but you cannot write an non periodic Pseudo-Random Number
Generator algorithm on a computer that lacks infinite memory.
That's impossible.
--
Guy Macon
<http://www.GuyMacon.com/>
Non periodic, is matemathicly speeking.
We all now the computers limits.
Ex: X(n)=0.3526373.38474844...75858585..., or PI, or SQR(2)
is, or not, non periodic?
The digits of any irrational number are not periodic when expressed as
decimal.
http://en.wikipedia.org/wiki/Irrational_number
So (for instance) the function that prints out hex digits of pi can be
used to produce billions or trillions of digits with no repeat in the
pattern.
One might object, "Well, pi is *clearly* a pattern."
True, but so is the output of any other prng.
That's why they are prngs and not rngs.
A cryptographically secure prng is quite an achievement. For
instance, the Mersenne Twister, with a period of: 2^(19937) - 1 is not
recommended for cryptographic applications.
.
- Follow-Ups:
- Re: PRNG
- From: Mark Wooding
- Re: PRNG
- Prev by Date: Re: Encrypt and Authenticate (E&A)
- Next by Date: Re: Solving B from h(A || P) with h = MD5
- Previous by thread: Re: PRNG
- Next by thread: Re: PRNG
- Index(es):
Relevant Pages
|
