4 ms·
FWIW, Theo did email me a brief note which indicated he had given serious thought to my analysis. He sounded really busy: http://bit.ly/eVNck1 http://bit.ly/eVN
by marshray 16y ago
FWIW, Theo did email me a brief note which indicated he had given serious thought to my analysis. He sounded really busy: http://bit.ly/eVNck1 http://bit.ly/eVNck1
The IV chaining is obviously broken, but it doesn't seem like it would make a make a good backdoor.
RFC 2405 (1998-11) says "The IV MUST be a random value."
RFC 3602 (2003-08) says "The IV MUST be chosen at random, and MUST be unpredictable." So the protocol knew about it at that time.
So it was a learning experience for the protocol developers at that time, not just OpenBSD.
If there is a smoking-gun type backdoor found, I suspect it will have to do with unexpected or out-of-order IP headers or ipcomp.
- marshray 16y agoOn the other hand, IV chaining would make a fantastic backdoor if you were able to brute-force 56-bit keys. Let's see, EFF did that for $250,000 in 1998? That leaves one budget cycle between it being public knowledge and the alleged contract work. http://en.wikipedia.org/wiki/EFF_DES_cracker http://en.wikipedia.org/wiki/EFF_DES_cracker
- djmdjm 16y agoI don't think OpenBSD ever recommended the use of single-DES. All the effort of having development outside the USA was to ensure that OpenBSD supported strong crypto without export encumbrances.
- marshray 16y agoGood point, 56 bit crypto is so broken anyway it hardly needs a backdoor. But looking at the dates on the RFCs, 3DES had only been standardized for a couple of years. It's plausible that the intended users didn't know, or they could be convinced, that they were required to use 56-bit DES.