[Gllug] Questions arising from the random discussions...

[Nix]

On 2 Aug 2010, Martin A. Brooks outgrape:

> On Mon, August 2, 2010 22:52, Dylan wrote:
>> Hi All,
>>
>> So, where does the stream for /dev/random and /dev/urandom come from and
>> what's the difference?
[...]
> The former blocks on exhaustion, the latter give a continuous stream of
> "psuedo-random" data when you run out of entropy.

Well, /dev/urandom just keeps extracting entropy from the pool even when
there isn't any left: it uses exactly the same method as /dev/random
does: it just never stops. Since the entropy figure is only an estimate
anyway, there's no guarantee that /dev/random will never return you data
which is theoretically predictable given knowledge of enough values it
has previously handed out. It's just going to block before it gives you
very many of them.
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug