On Thu, 23 Dec 2010 19:31:23 +0100 mouss <mo...@ml.netoyen.net> wrote:
> if you're worried about performace, don't hash at all. would you use a > cesar/base64/... ? either you need security and you use an algorithm > that's not considered broken, or you don't. The breaks in md5 would allow an attacker to generate a second email address that collides with a given address. I don't see how that compromises anything since presumably the intent is to avoid an attacker inferring an address from a hash. From the security point of view the scheme itself is far more broken than md5 is. A secure hash function can only protect addresses that are both secret and contain a cryptographically secure amount of entropy. I'm curious as to the point of this. Phishing/fraud contact addresses might be better left to AV software that already have the infrastructure to push this kind of information without any side-channel leakage. Abusive marketers use fixed from addresses but their status is often subjective. If the intent is to catch lazy spammers, I think it'll be a very short-term win.
