Re: [dsfjdssdfsd] Any plans for drafts or discussions on here?

[Michael Hammer <[e-mail]>]

So, if you mix a non-random input with a random input, 
using a deterministic algorithm, the output will be more random?
That doesn't seem right to me.

We can't control the randomness of the input data, 
but the other inputs to an encryption or hashing algorithm are my concern.

Michael Hammer
Principal Engineer
[e-mail]
Mobile: [phone]
500 Yosemite Drive Suite 120
Milpitas, CA 95035 USA


-----Original Message-----
From: Krisztián Pintér [mailto:[e-mail]] 
Sent: Friday, January 24, 2014 9:02 AM
To: Michael Hammer
Cc: [e-mail]; [e-mail]
Subject: Re: [dsfjdssdfsd] Any plans for drafts or discussions on here?


Michael Hammer (at Friday, January 24, 2014, 12:19:03 AM):
> Hmmm...  that makes it sound rather subjective.
> If we don't have objective measures,
> then who is to say that one's randomness is better or worse than another?

as i said, we need to examine the physical processes. the best source of
entropy as of now is thermal noise. we understand thermal noise to a great
degree, we don't expect sudden breakthrough in modeling it, and it is
relatively abundant and easy to access. user input also contains noise, as
the user can control keystroke or mouse movement timing up to some 10's or
100's of milliseconds, below that, it is just noise from the equipment and
the "biological equipment".

> Was thinking in terms of how an app with access to alternate random 
> sources, some which might be from OS or from some software, might 
> choose one over another.

if you are adamant on doing homebrewed, why choose? you can combine them. if
your combinator is good, you can't lose.