Eric Rescorla wrote: > Jim Jagielski <[EMAIL PROTECTED]> writes: > >> Eric Rescorla wrote: >> > >> > Glenn Olander <[EMAIL PROTECTED]> writes: >> > > 5) The strength of the PRNG is largely irrelevant >> > > >> > > As a user, I wouldn't trust any solution which lacks a check for >> > > duplicate session id's, regardless of the strength of the PRNG. >> > This doesn't seem to me to be a plausible position in view >> > of the fact that all of our security mechanisms absolutely >> > depend on statistical uniqueness of randomly generated large >> > numbers. >> > >> >> These are 2 different points I think. If you randomly generate numbers >> between 1 and 1,000,000 you will, after a point in time, have >> duplicate numbers. > Yes, but if you randomly generate numbers between 1 and 2^128, you'll > have to generate roughly 2^64 random numbers to have a good chance of > getting a duplicate. Sure, over time you'll get a duplicate, > but in this context over time needs to be measured over a > time scale far in excess of the time scale that is interesting.
Adding the startup time doesn't hurt. Or enough bits of the startup time to guarantee a duplicate won't happen in our life. We'll still have the probability that 2 random number generated at the same time ( same millis ) will be identical. :-) Costin -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
