On Sep 6, 2013, at 10:39 AM, Stephane Bortzmeyer <[email protected]> wrote:
> On Thu, Sep 05, 2013 at 03:03:28PM -0700, > Lucy Lynch <[email protected]> wrote > a message of 23 lines which said: > >> may be in order: >> >> https://www.propublica.org/article/the-nsas-secret-campaign-to-crack-undermine-internet-encryption > > Well, it seems this story (a better presentation of the > standardization problem is > <http://www.wired.com/politics/security/commentary/securitymatters/2007/11/securitymatters_1115>) > is a strong argument in favor of the IETF practices: openness of > standardization (even security standardization), "show me the code", > "explain what your proposal does or I won't consider it", do not > blindly believe the authority, etc. This works OK for protocols. But does this really work for parameters? Take a look at section 3 in RFC 3526 ( http://tools.ietf.org/html/rfc3526#section-3 ) or similar sections in RFC 5114. I'll quote it in full here: 3. 2048-bit MODP Group This group is assigned id 14. This prime is: 2^2048 - 2^1984 - 1 + 2^64 * { [2^1918 pi] + 124476 } Its hexadecimal value is: FFFFFFFF FFFFFFFF C90FDAA2 2168C234 C4C6628B 80DC1CD1 29024E08 8A67CC74 020BBEA6 3B139B22 514A0879 8E3404DD EF9519B3 CD3A431B 302B0A6D F25F1437 4FE1356D 6D51C245 E485B576 625E7EC6 F44C42E9 A637ED6B 0BFF5CB6 F406B7ED EE386BFB 5A899FA5 AE9F2411 7C4B1FE6 49286651 ECE45B3D C2007CB8 A163BF05 98DA4836 1C55D39A 69163FA8 FD24CF5F 83655D23 DCA3AD96 1C62F356 208552BB 9ED52907 7096966D 670C354E 4ABC9804 F1746C08 CA18217C 32905E46 2E36CE3B E39E772C 180E8603 9B2783A2 EC07A28F B5C55DF0 6F4C52C9 DE2BCBF6 95581718 3995497C EA956AE5 15D22618 98FA0510 15728E5A 8AACAA68 FFFFFFFF FFFFFFFF The generator is: 2. How do we know that these are good parameters? We have Tero's word that he checked them, and those who know enough can repeat the test, while the rest of us can only wonder where those numbers came from, what a real number (pi) is doing there, and whether it's OK that the top and bottom 64 bits are all set. And this is for MODP which is well understood. For EC, we mostly use the NIST curves, which we know are OK because NIST said so. Luckily we have NIST to protect us from the NSA. Oh, wait… So I have a lot of faith in the IETF process for analyzing protocols, but I think the actual pool of people who understand cryptography well enough to make comments about the security of algorithms and the safety of parameters is rather small. Yoav _______________________________________________ perpass mailing list [email protected] https://www.ietf.org/mailman/listinfo/perpass
