Lava Lites: Easy to Break, Hard to Crack

Joe Kiniry (kiniry@cs.caltech.edu)
Fri, 28 Mar 1997 12:42:16 -0800


tbyars@earthlink.net writes:
> Lava Lites: Easy to Break, Hard to Crack
> by Mark Frauenfelder

... text deleted ...

> Shortly afterward, the three scientists applied for a patent, which
> claims ownership for any process that uses one of more chaotic sources
> plus zero or more nonchaotic sources, digitizes them, crypto-hashes the
> digital file, then uses the file as a seed for a random-number generator.

this is exactly the kind of broad-based patent that drives me crazy as
a professional in the industry. can anyone explain to me how one/they
can possibly justify this patent? _it is plainly obvious_ that the
use of chaotic real-world sources is _a_ way to go for the generation
of seeds and people have been doing it in various forms for years (be
it random key-presses, voltage fluctuations, or scanning your bunnies
doing the wild thing). anyway, i'll quit ranting and move on to my
next fork.

> Lava Lites are just one specific application of the more general patent.
> "Using Lava Lites is unique and novel," says Noll, "but also sound,
> resulting in strong cryptographic data. To reverse the process, you'd
> have to come up with the exact same picture of the Lava lamps. If even one
> pixel has a different shade you end up with a completely different hash."

joe