system entropy pool (was Re: [PLUG] Do it yourself moderation)

Chris Jantzen chris-plug at maybe.net
Sun Nov 30 13:04:01 UTC 2003


On Sun, Nov 30, 2003 at 12:52:40PM -0800, Kyle Hayes wrote:
> Still, on a more serious note, has anyone actually looked at using something 
> like Usenet to generate random data?  It has to be a huge pool of entropy.  
> Perhaps the alt.binaries.* lists should be avoided, but mailing lists and 
> non-binary news groups should have significant entropy.  I guess you'd also 
> want to remove all quoted material (repetition can't be that good for the 
> entropy level).
> 
> You wouldn't get very many bits of entropy per message, but you'd have so
> much raw data!

I suspect linguistic character and word 'distance' would pretty
predictable from language to language. If anything you *would* want to
include binaries. Now, as a matter of *active* (i.e., beginning right
now) random data analysis, you could measure message injection rates
and latency (here or on USENET) on a per-group/list basis.

-- 
chris kb7rnl =->
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.pdxlinux.org/pipermail/plug/attachments/20031130/95614594/attachment.asc>


More information about the PLUG mailing list