Re: [Sks-devel] Keydump

[Daniel Kahn Gillmor]

On 07/14/2014 03:00 AM, Henning Kopp wrote:
> I am a Phd student and want to do some research on gpg-keys. Perhaps
> you know about factorable (https://factorable.net/) where some
> researchers scanned the internet for ssl/tls-certificates. They tried
> some trivial factorization methods, like computing gcd, etc. and had
> success. Also many of the keys were the same due to embedded devices,
> having not enough entropy.
> I want to do the same or something similar on gpg-keys. Do you know
> of any research which was already done?

Lenstra et al have done this work on the OpenPGP keyserver dataset, if i
understand the paper correctly:

  http://eprint.iacr.org/2012/064

says "almost 5.5 million PGP keys"

But you shouldn't let that stop you, you're sure to learn and better
understand the work from trying to replicate the results. :)

I'm glad to see you reading Heninger's et al's work
(https://factorable.net/).  Another interesting description of the
problem and the method is Seth Schoen's writeup:

  http://www.loyalty.org/~schoen/rsa/

hth,

        --dkg

signature.asc
Description: OpenPGP digital signature