From: Christian Neukirchen <chneukirchen@...>
Date: 2005-12-07T00:41:39+09:00
Subject: Re: Shuffling an array, sort_by{rand}'s bias

Mauricio Fern�ndez <mfp@acm.org> writes:

> When called with no arguments, rand will use
>
> static double
> genrand_real(void)
> { 
>     unsigned long a=genrand_int32()>>5, b=genrand_int32()>>6; 
>     return(a*67108864.0+b)*(1.0/9007199254740992.0); 
> }
>
> so it would seem that two consecutive calls to rand() cannot return the
> same value, but I still have to think about the effect of qsort.

And that is a good thing?  How can the sequence be "random" if it
never contains the same two numbers in row?  (Remember Enigma...)

> Mauricio Fernandez
-- 
Christian Neukirchen  <chneukirchen@gmail.com>  http://chneukirchen.org