On Tue, 2014-01-14 at 23:57 -0500, grarpamp wrote:
> I believe there is sufficient demand to merit deployment of a
> good mix network. As well as perhaps web/other intake frontends
> due to the now prevalent a) dwindling free email b) demand by
> mail providers for phone authentication. As for ope
D. J. Bernstein (at Monday, January 20, 2014, 8:01:20 PM):
> Brainpool, in particular, generated curves using a method with _some_
> details that are not explained anywhere in the Brainpool documents.
let me add that nothing prevents the authors from publishing their
design rationale, especiall
Peter Gutmann writes (on one of the harder-to-use mailing lists):
> Some of their objections seem pretty subjective though, I mean they
> don't like the Brainpool curves
Actually, the Brainpool curves _meet_ the rigidity requirement that
you're alluding to. The SafeCurves site displays this in the
Ondrej Mikle wrote:
> The HW RNG is unsuitable to be used as constant entropy source, because number
> of times it can be used is limited. The limit exists because the ATSHA204 uses
> some sort of EEPROM update while generating random numbers (which suggests
> that
> something like PRNG is used in
Hi,
I have an embedded Linux (openwrt) device here, that might suffer from the known
entropy hole that caused predictable keys to be generated in past. Fortunately,
the device has HW RNG (ATSHA204 to be precise).
The plan is to seed /dev/urandom from the HW RNG upon boot, by feeding 512 bytes
int
