[RC5] Statistics of Key Distribution
Joe Zbiciak
jzbiciak1 at ti.com
Wed Apr 15 13:29:11 EDT 1998
'gindrup at okway.okstate.edu' said previously:


 Well... Actually... The Normal Distribution is what you get in the
 limit of infinite numbers of dice. It's convenient however that the
 sequence of distributions converges somewhat rapidly to its limit.

 This result is called the "Central Limit Theorem".
[...]
 What this has to do with Distributed Computing, Cracking, RC5, DES,
 programming, UI issues, or DCTI is entirely beyond me.
Random factoid which may make this (only very slightly) relevant to
cryptography:
It is this theorem which highlights why checksums based just on adding
together a series of numbers (esp. modulo a large accumulator size) don't
work nearly as well as a stronger hashing function such as a CRC or a
cryptographically strong hash such as RMD160 or MD5.
Regards,
Joe

+ Joseph Zbiciak +
     jzbiciak1 at ti.com      Don't answer, don't ask,
 http://www.primenet.com/~im14u2c/  Don't try to make sense.
   Texas Instruments, Dallas     "Numb", U2
+#include "std_disclaimer.h"+

To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net
rc5digest subscribers replace rc5 with rc5digest
More information about the rc5
mailing list