Other Directory Sites: SeekWonder | Directory Owners Forum
The following archive was created by hippie-mail 7.98617-22 on Thu May 27 1999 - 23:44:21
mgraffam@idsi.net
Fri, 9 Apr 1999 13:33:01 -0400 (EDT)
On Fri, 9 Apr 1999, David Honig wrote:
> You would have to dump the *raw* bits coming in, measure their
> entropy[1], and look at how many bits in for each bit out. Then you
> would have a measure on physical-entropy-per-output bit.
Right. I'm actually looking into this.. now, the way its works is the
"entropy pool" gets stirred with a non-cryptographic hash every interrupt,
and then that pool gets passed through SHA when a user-space proggy hits
up one of the two character devices.
I'm looking into exporting another interface .. one that dumps the raw
contents of the pool, w/o SHA. Now.. the question becomes, do we want
the stirred pool (post-processed by the mixing function), or the virgin
data?
I'm leaning to both myself.. it isn't a big deal to export an interface,
so as soon as I am real comfortable with how it works, I'll add the code
in.. I just want to make sure that I don't end up dumping data that the
code might end up throwing away.. that wouldn't give accurate results.
> [1] Here MUST is more appropriate than Diehard.
Can someone email me an URL on MUST?
FWIW (completeness?) I've just generated a few meg of bits from
/dev/random and I'm going to run Diehard on 'em .. I'll post the results,
just so there is no doubt that the hash code is working :)
Michael J. Graffam (mgraffam@idsi.net)
I think that we should be men first, and subjects afterward. It is not
desirable to cultivate a respect for the law, so much as for the right.
Henry David Thoreau "Civil Disobedience"
The following archive was created by hippie-mail 7.98617-22 on Thu May 27 1999 - 23:44:21