From: Ian Sparkes <isparkes@q9f47.dmst02.telekom.de>
Date: Thu, 8 Jan 1998 17:07:52 +0800
To: Adam Back <billp@nmol.com
Subject: Re: brute forcing combination locks (Re: Bruce Schneier, Sandia, FBI and the REAL WORLD)
In-Reply-To: <34B3DFB4.6396@nmol.com>
Message-ID: <3.0.2.32.19980108100339.00722fc8@q9f47.dmst02.telekom.de>
MIME-Version: 1.0
Content-Type: text/plain



>So I hacked up some code to compute the minimal universal door entry
>sequence number...
>
>The sequence looked something like this:
>
>01234567890124568902346780...

We have the same system at my [x], and I have played with the same idea
coming up with similar results (my value was somewhat lower - but this was
fag packet maths and therefore subject to error). In my fantasies I have
also connected this sequence to an array of solenoids which punch in the
number while you wait. The locks seem to be able to handle a throughput of
around 2 digits per second, meaning the average search time of 10 minutes.
Bear in mind that there are 8 of these locks - with the appropriate
hardware you could guarantee to be in in less than three mins.

The reason that this has remained in my fantasies is that any key changes
are predictable - the company is the (until last Wednesday) state owned
telephone company =) and the keyspace consists of the area codes for the
surrounding regions.

And this lock is the only thing that stands between the outside world and
25 Servers and 300 workstations.

Makes you think...