Re: Random number access on the i5

Mark

No software=based so-called random number generator is truly random - they produce repeatable result sufficiently long before repeating, that they act like random numbers for all practical purposes. In fact, they are more correctly called pseudo-random number generators.

Hardware generators are better - take a look a this wiki article for more thoughts on this

http://en.wikipedia.org/wiki/Hardware_random_number_generator

Regards
Vern

At 11:00 PM 1/31/2008, you wrote:

I do not believe that a computer can generate a random number. There,
I said it. Hopefully you can prove me wrong.

I have done a bit of looking, especially with the database random
number functions. The DB route is really an eye opener.

Is there such a thing as a "certified random number"? I would like to
get this behind me and I have not found a methodology I can sign off
on yet. Ideally, any correct algorithm could be multiple platform. I
am under the impression that it should be CPU dependant and a CPU must
support it directly with an instruction.

Even if the number appears random, if I can reset the computer clock
over and over, re-submitting the command 10 or so times starting at
that exact moment, can I get the same random number? The odds are- no
in mathematics but yes on the computer. So, my theory is,
pseudo-random number generation has been more than adequate for
business solutions. i.e. picking the lucky winner of $millions.

I think it might be safe to go with a derivative of the Global Unique
Identifier GUID - but we know this GUID is not random either since a
computer can not generate a random number from pure logic, right?
-again, reset the clock, etc. etc. The only other seed that makes
sense to me is the exact moment in time, which of course, should never
occur again. This seed mixed with a couple of analog inputs might do
it. The CPU could be aware of it's temperature and something else that
is unlikely to be repeated at that moment in time. i.e. free
electrons.

I am not up on anything from the lab concerning random numbers but you
would think that this would be something that was solved long ago and
easily accessible in our code- i.e. for all practical purposes, it is
certified random.

I will enjoy reading and coding any responses,
Mark
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.