From: Xcott Craver <caj@math.niu.edu>
Date: Mon, 4 May 1998 13:04:56 -0700 (PDT)
To: cypherpunks@toad.com
Subject: Re: Bad, bad, bill
In-Reply-To: <19980501153418.25294.qmail@hotmail.com>
Message-ID: <Pine.SUN.3.91.980504142845.22358A-100000@baker>
MIME-Version: 1.0
Content-Type: text/plain


On Fri, 1 May 1998, Corvus Corvax wrote:

> "WASHINGTON - The Senate Judiciary Committee on Thursday adopted the 
> Digital Millennium Copyright Act, which takes a new step toward 
> protecting creative content in cyberspace by outlawing the equipment 
> used to commit a copyright infringement, not just the act or intent of 
> illegally copying material. 
> ...

	Yah.  This same was mentioned on a digital watermarking forum I
read.  A great deal of my own research centers around attacking & finding
weaknesses in such schemes, and as it's science I am obliged to write code
to prove my hunches.  This bill may make me a criminal!

	This struck a nerve at the time, because the fella who mentioned
the bill was (a) an employee of a watermarking company and (b) mad as Hell
at the rest of us for developing watermark attack software.  If that law
was on the books (and the guy was the CEO or such) at the time, I have no
doubt a few of us would be in big legal trouble. 

	Now, I should disclaim that that particular company is much more
understanding than that one guy.  I met a bunch of their employees (as did
a couple other attack-makers) at a conference, and they were quite clueful
about the importance of watermark robustness research.  However, countless
other watermarking startups exist.  Many have the whole business bet on a
potentially breakable scheme.  Further, watermarking involves people from
non-crypto disciplines, such as image processing, where folks are more
likely to make security mistakes and less likely to appreciate the need
for an adversarial research mode.  This spell DANGER, boys and girls. 
Some company consisting of few people, lots of venture capital, little
experience and lots of tunnel vision is IMHO likely to resort to this law
when researchers find a big hole in their breadwinning technology. 

	I should add that watermark attack tools like unZign and StirMark
are now used as benchmarks by developers of new schemes.  It's only been,
what, half a year since StirMark was created?  A gaggle of watermarking
papers already cite it, and include it in a standard volley of tests.  
These are not evil warez from Hell here.

						-Xcott

==-  Xcott Craver -- Caj@niu.edu -- http://www.math.niu.edu/~caj/  -==
"Also note that elecronagnetic theory proves that if you microwave a
 bar of Ivory soap it turns into a REAL MARSHMALLOW THAT YOU CAN EAT."
                                                   -James "Kibo" Parry