RE: random number generator oddity

Quoting "McKown, John" <John.Mckown@xxxxxxxxxxxxxxxxx>:

-----Original Message-----
From: midrange-l-bounces@xxxxxxxxxxxx
[mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of
ChadB@xxxxxxxxxxxxxxxxxxxx
Sent: Thursday, December 04, 2008 3:51 PM
To: Midrange Systems Technical Discussion
Subject: Re: random number generator oddity


I'll be interested to hear on this one... I don't know any
specifics on
your question, but my reading has always led me to believe
that it's very
difficult to have a program that gives truly random results...

Very true. Most are really called "pseudo-random" generators. They are
called pseudo random because for a given seed value, they will always
return the same series of numbers. They are declared "good enough" when
the series of numbers has a high cardinality before repeating and when a
number of statistical functions declare them "random" (I don't remember
the functions).

I vaguely remember that Intel was going to have an instruction which
would give a nearly truly random number by measuring something
(temperature?) that simply could not be predetermined.

--
John McKown
Systems Engineer IV
IT

Administrative Services Group

HealthMarkets(r)
Administrative Services Group
9151 Boulevard 26 * N. Richland Hills * TX 76010
(817) 255-3225 phone * (817)-961-6183 cell
john.mckown@xxxxxxxxxxxxxxxxx * www.HealthMarkets.com

Kind of ancient history here. But the 8 bit CPU, the Z-80 from Zilog, had an
internal register that could be read that was primarily used to generate a
memory refresh signal. Just counted up and restarted from zero. Still pseudo
random, but hard to get two computers to have the same value at the same
moment.

John McKee