[148903] in cryptography@c2.net mail archive
Re: [Cryptography] Dual_EC_DRBG backdoor: a proof of concept
daemon@ATHENA.MIT.EDU (=?iso-8859-15?Q?Kriszti=E1n_Pint=E)
Fri Jan 3 12:05:18 2014
X-Original-To: cryptography@metzdowd.com
Date: Fri, 3 Jan 2014 18:02:04 +0100
From: =?iso-8859-15?Q?Kriszti=E1n_Pint=E9r?= <pinterkr@gmail.com>
To: Theodore Ts'o <tytso@mit.edu>
In-Reply-To: <20140103155740.GC31411@thunk.org>
Cc: Thierry Moreau <thierry.moreau@connotech.com>,
Cryptography Mailing List <cryptography@metzdowd.com>,
Jon Callas <jon@callas.org>, ianG <iang@iang.org>
Errors-To: cryptography-bounces+crypto.discuss=bloom-picayune.mit.edu@metzdowd.com
i think you put too much burden on a prng. all prngs need a secret
seed. it is not an argument against them. the question is, what comes
after? rephrasing: supposed that we have some secret (for example true
random), how can we expand that into a random stream in a way that we
don't introduce *new* vulnerabilities. it is not the task of the prng
to solve the seeding problem, that should be handled separately.
in that sense, BBS has the benefit of having a proof in the standard
model. as opposed to AES based generators, that have formal proof
against some attacks only, while have a general proof in the random
oracle model. again, *in addition* to the problem of the seeding,
which they also have.
i'm not claiming that this is a practical advantage, or i would pay a
dime to get that. but it certainly represents *some* value.
that said, as i heard, dual-ec does not have a security proof. correct
me if i'm wrong.
Theodore Ts'o (at Friday, January 3, 2014, 4:57:40 PM):
> Um, where are you going to get the true entropy from? If you are
> willing to assume that you can seed the device with true entropy, then
> you can just use an AES-based CRNG.
_______________________________________________
The cryptography mailing list
cryptography@metzdowd.com
http://www.metzdowd.com/mailman/listinfo/cryptography
