On 29 Dec 2007, at 20:09, Aaron Stone wrote:
The risk of collision is very small, but is real -- every has acknowledged that. Then there's the evaluation of the risk, which we also all know to be extremely small, and the cost of mitigation, which involves a little decision making and some engineering time (to do byte-by-byte checking, or to double hash, or both, or some other approach to be thought up). It'llhappen, really.
It's usually the minimal risk events that materialises. I've had "ain't gonna happen" hash collisions more often than I care to remember in the past. Don't do it.
smime.p7s
Description: S/MIME cryptographic signature
_______________________________________________ DBmail mailing list [email protected] https://mailman.fastxs.nl/mailman/listinfo/dbmail
