[OT] Combinatronics WAS Re: /dev/random and linux security issues (kinda long)

[Dan Jenkins]

aluminumsulfate at earthlink.net wrote:

>  For a "truly random" source, it might be best to use a good white
>  noise source, such as a properly chosen and biased point contact
>  silicon diode or the like. For something in between and completely
>  accessible to a software approach, the computer clock could be
>  factored into a prospective algorithm.
>
>  A point contact what?? Info on this would be welcome. Any links on
>  how to build one/attach it to Linux? I was thinking of hooking up an
>  AM radio to my sound card, tuning in some static, hitting record,
>  and post-processing the results. Other ideas I had included running
>  edge detection algorithms on pictures of cooked rice, hooking some
>  kind of scintillation detector to a radiation source (like a smoke
>  detector), and measuring the interval between barks of my neighbor's
>  (annoying) dog.

Barks of a neighbor's dog aren't random. They are continuous 
(continuously annoying, that is). :-)

A bit of googling yielded:

An explanation/source of noisy diodes:
    http://www.avtechpulse.com/faq.html/IV.8/

How to use an audio card for random number generation:
    http://www.av8n.com/turbid/paper/turbid.htm

And a compendium on random number generation with some hardware designs 
and products:
    http://www.cs.berkeley.edu/~daw/rnd/

-- 
Dan Jenkins (dan at rastech.com)
Rastech Inc., Bedford, NH, USA --- 1-603-206-9951
*** Technical Support for over a Quarter Century

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.gnhlug.org/mailman/private/gnhlug-discuss/attachments/20050515/bce70129/attachment.html