On 2011-03-16, at 11:05 , Sander Steffann wrote: > The chance of this happening is *very* *very* small.
The chances that a MD5 hash of any data will ever be exactly zero (all octets are zero) is ridiculous tiny. It is probably so extremely unlikely that the world would rather be destroyed by a meteor tomorrow than this is ever going to happen. Still IPv6 implementations verify that the output of the MD5 hash function is not exactly zero when creating a temporary address as suggested by RFC 4941 (I verified this by source code). The scenario that I described is much more likely than a MD5 of zero, but here no code exists to prevent it. > I would worry more about people choosing such an IP address intentionally to > cause trouble... That would rather be security issue, since why would anyone do that, unless he's doing it by intention? And if someone does that by intention, he's probably planing to do something evil. Best regards, Markus -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
