Re: RSA moduli sizes

pubkeybreaker <pubkeybreaker@xxxxxxx> writes:

Instead, do what I told you to do several posts back.
Generate p,q uniformly at random from an interval such that
their product is GUARANTEED to have 2n bits. This is trivial.
[...]
May I suggest that you people stop presenting your own 'pet'
ideas and go out and STUDY this subject???? Read some books.
Read IEEE 1363 and FIPS-140 and X9-80 and X9-31, etc. etc.
Stop prattling! Leave design of algorithms (such as the
erroneous one given above) to people who have studied how to
do it.

In practice, primes are chosen by picking a random starting point and
finding the next smallest prime which satisfies some criteria. Such
primes are certainly not uniformly distributed -- indeed, they're biased
towards primes preceded by large contiguous regions of composite
numbers.

X9.31 demands that you choose the primes like this -- and NIST's
validation system verifies that you do it like X9.31 says, hopelessly
quaint notions of strong primes and everything. (They give you starting
points and tell you to generate RSA keys. If you generate the wrong
ones, you don't get a certificate.)

-- [mdw]
.

Relevant Pages

  • Re: The METHOD of creating RSA key
    ... frieds with you and Pubkeybreaker. ... I will assume that the primes are in a pre-computed table, ... You could also treat sas just representing the ODD numbers in the ... I gave me myself an example according to your instructions. ...
    (sci.crypt)
  • Re: The DNA of prime numbers
    ... On Feb 17, 2006 6:46 AM CT, Pubkeybreaker wrote: ... about it is a mystery. ... ...because the primes have DNA!!! ... *wipes drool off face and takes off tinfoil hat* ...
    (sci.math)
  • Re: The METHOD of creating RSA key
    ... frieds with you and Pubkeybreaker. ... I will assume that the primes are in a pre-computed table, ...
    (sci.crypt)
  • Re: Primes of the form 2^(2^n)+1
    ... Pubkeybreaker wrote: ... If a is allowed to vary, the truth of this proposition would follow ... e.g. Schinzel's Conjecture, or more generally, the Bateman-Horn ... no primes)? ...
    (sci.math)
  • Re: SF: Back to theory
    ... the product of two primes. ... M, M/q = p ints in 0..M-1 in the same residue class as f mod q, and 1 int ... That's actually a bit less than the probability of picking a winning k ... finding f-g winners than picking f and g at random. ...
    (sci.math)