[Cryptography] Generating random values in a particular range

[Sidney Markowitz]

John-Mark Gurney wrote on 17/08/16 7:00 AM:
> This looks very close to the way RFC 2631 specifies how to generate
> DH parameters.

RFC 2631 is simply quoting FIPS PUB 186-1 and so in section 2.2.1.1 steps 5-7
(the second 5 and 6, the linked document has typos in its numbering) it has
the vulnerability that was fixed in the FIPS Pub 186-2 Change Notice.

The patent claims the steps in that Change Notice where instead of taking the
mod in step 5, if q is too large they iterate on advancing the seed and
hashing again.

The point of the patent near as I can tell seems to be to claim that fix for
the vulnerability, then submit the fix to FIPS who accepted it into their
standard.