[Coco] Is the RND command really generating a random number?

[James Jones]

This gets us into algorithmic information theory, where the amount of
information in some data is measured by the length of the program it takes
to generate it (reading the data itself from a file and writing it out
doesn't count); that can also be taken as a measurement, or really a
definition, of randomness. So, viewed that way, linear congruential random
number generators, being very short pieces of code, by definition can't
actually generate random numbers.

(Linear congruential random number generators were the kind popular back
when I was in college, where

s(n+1) = (s(n) * a + b) mod c

for some carefully chosen a, b, and c.  I'm sure Color BASIC uses one,
precisely because of the size issue. A notoriously bad example of one is
IBM's "RANDU", from the OS/360 SSL (Scientific Subroutine Library); see
https://en.wikipedia.org/wiki/RANDU.)

The usual "randomness tests" are things that random sequences pass... but
other sequences may pass, too, depending on the test.

On Fri, May 19, 2017 at 3:34 PM, Dave Philipsen <dave at davebiz.com> wrote:

> Now *that* seriously is a profound statement!  We could veer off into a
> discussion about intelligent design from that but probably not in this
> forum...
>
> Dave
>
>
> On 5/19/2017 2:02 PM, Bill Gunshannon wrote:
>
>> But the truth is there is no such thing as a
>> random number as it is truly doubtable that anything in nature is truly
>> random.
>>
>> bill
>>
>>
>>
> --
> Coco mailing list
> Coco at maltedmedia.com
> https://pairlist5.pair.net/mailman/listinfo/coco
>