[117773] in Cypherpunks
Re: Build a better OTP?
daemon@ATHENA.MIT.EDU (Bill Stewart)
Fri Sep 10 04:26:51 1999
Message-Id: <3.0.5.32.19990910011009.009a8870@idiom.com>
Date: Fri, 10 Sep 1999 01:10:09 -0700
To: Anonymous <nobody@replay.com>, cypherpunks@cyberpass.net
From: Bill Stewart <bill.stewart@pobox.com>
In-Reply-To: <199909091824.UAA17065@mail.replay.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Reply-To: Bill Stewart <bill.stewart@pobox.com>
At 08:24 PM 9/9/99 +0200, Anonymous wrote:
>> >> [...RNG whitening code...]
>> > Well, putting it on the chip sure ain't the right place.
>The RNG currently consists of two thermal resistors, two oscillators,
>and the simple state machine to detect 1-0 and 0-1 transitions. The
[SHA big]
>In the face of these engineering realities, what is the argument for
>putting the whitening on the chip? It is fundamentally ideological.
There are two separate decisions -
1) whether to do whitening on the chip
2) whether to allow access to the Raw Bits as well
The reason for 1) is straightforward - if you don't put the
whitening on the chip, clueless users will fail to do it themselves,
and people will blame Intel that their gambling programs and
crypto systems and fuzzy-logic stock market predictors work badly,
and they'll blame Intel for collaborating with the NSA by
providing bad randomness.
On the other hand, 2) is bad - some people really want access
to the raw bits, and it shouldn't take much resources to provide it.
Thanks!
Bill
Bill Stewart, bill.stewart@pobox.com
PGP Fingerprint D454 E202 CBC8 40BF 3C85 B884 0ABE 4639
