Re: preventing sequence number guessing

[proff () suburbia apana org au (Julian Assange)]

> > ~/scratch/md5 $ time ./md5drivr -t
> > MD5 time trial. Digesting 1000000 8-byte blocks ... done
> > Digest = 2278bf63bfa354c582138cde1233fd15
> > Time = 7 seconds
> > Speed = 1142857 bytes/second
> > 6.776s real  6.680s user  0.090s system  99% ./md5drivr -t
>
> > So it takes about 7 u-seconds to MD5 hash a small block on a fairly
> > wimpy Sun.  [It's faster on faster boxes, of course.]
>
> Perhaps not.  MD5 has relatively high overhead; that is, in the
>
>       time = overhead + (perbyte * bytes)

Don't forget alsothe that above example also most likely ran entirely within the
cache on an unloaded machine. In a real-life circimstance it would be inter
spliced with both much other kernel level networking code and the user-level 
processes that are the initial/final senders/recpients of the TCP packets. The
reality is that cryptographically secure sequence numbers are not needed. There
are ample other sources for truely random variables of small enough granularity
within most kernel data spaces.

Proff