Improving Random

[Bardur Arantsson]

(I certainly don't want to pile on, just want to emphasize a really
important point which I've learned over the years shipping lots of
software... which, while not perfect in any way, has certainly worked
"well enough")

On 03/06/2020 15.55, dominic at steinitz.org wrote:
> not wait another 5 years for perfection
> (which may never arrive).
> 

Aka. "perfect is the enemy of good".

Unless it can be shown that these changes to the 'random' library
*actively* prevent any later improvements[0], I'd say that this aphorism
applies rather... good.

[0] Specific suggestions would be necessary for credibility.

Regards,