[dvd-discuss] my debate with Michael Shamos

[Dave_Touretzky(at)cs.cmu.edu]

Michael Shamos and I are both faculty members at Carnegie Mellon, and
we both testified at the 2600 trial -- for opposite sides.  Today we
held a debate on whether computer code should be protected speech.
People have been anticipating this event since the Chronicle of Higher
Education piece on our diagreement:
  http://chronicle.merit.edu/free/v47/i48/48a04501.htm

My slides from today's debate can be viewed here:
  http://www.cs.cmu.edu/~dst/DeCSS/Gallery/Debate

One of the points I made in my position statement was that the
argument that "computer code by itself doesn't DO anything", which has
recurred recently on dvd-discuss, is similar to the old saw, "Guns
don't kill people: people do."  Most folks do not find that a
compelling argument against firearms regulations, and hence we have
lots of laws controlling the sale, ownership, and carriage of
firearms.  The problem is that putting a loaded gun in someone's hand
leaves too thin a margin of protection for the general public.  So,
although firearms do have lawful uses, the law permits restrictions to
be laid on what guns people can own (no automatic weapons), and where
they can bring them (no concealed carry without a permit, no guns on
airplanes, etc.)

I argued that if congress and the courts are justified in treating
software like guns, there must be some "imminent peril" associated
with software that likens it to a loaded gun.  The court's initial
attempt at this was to count mouse clicks, but that's ridiculous.  How
many mouse clicks will the court find are required to establish a
"comfortable distance" from harm?  If I publish circumvention code on
a t-shirt instead of on the Internet, and so prospective circumventers
have to type in the code by hand, will that be enough of a safety
margin?  I doubt it.

The DoJ says it has no desire to prosecute Ed Felten for his Usenix
paper, but it refuses to promise not to prosecute him under the DMCA
for the *next* paper he writes, if it turns out they don't like it.
Their reasoning is correct.  No amount of mouse clicks will be enough
to deter people from circumventing access controls they find
irrational and offensive.  The only effective deterrent is to punish
the expression of IDEAS that could, with sufficient mouse clicks, be
turned into circumvention devices.  Any lesser measure will prove
ineffective, because computer folk are so darned inventive, as
evidenced by all the wonderful exhibits contributed to the Gallery of
CSS Descramblers.

And so we will be forced to recognize the DMCA as a restriction on all
forms of speech describing circumvention devices, not just speech that
is compilable computer code.

-- Dave Touretzky