eWeek: Cryptography Guru Paul Kocher Speaks Out

[Thor Lancelot Simon]

On Thu, May 01, 2003 at 12:24:39PM -0400, Ian Grigg wrote:
> 
> > ...  This would seem to suggest that raising the
> > level of investment and technical sophistication required beyond mere
> > triviality is, in fact, the correct target to shoot for, from the
> > studios' point of view; and that Kocher is probably really doing a
> > reasonable job at that with his proposal.
> 
> No, the point with print is that the cost of each
> additional print is very high.  In fact, the cost

No, it's not; there are three ways the work gets cheap:

1) You automate it; there's tons of last-generation automated equipment
   available for duplicating 35mm long-roll film and always has been,
   because it's a fairly cutthroat business and equipment is upgraded
   frequently to keep costs down.

2) You do it semi-manually and pay your workers a pittance.  The operations
   involved (loading film, loading chemistry, simple time and temperature
   control) are simple enough that they can essentially be done by
   minimum-wage workers.

3) You ship your first internegative to India and let them deal with how
   to make your cheap, stolen prints.

The cost is higher than that of burning a DVD-R, I'll grant you, but it
is not so high as to dwarf the setup costs, or even come close.  I'm
persuaded by your argument that the "keep duplicating forever" property
of the attacks we've seen described on Kocher's system is more serious
than I originally suggested, but generally speaking, I think the analogy
to unauthorized duplication of actual film prints is a useful one, because
it demonstrates that what the industry really has to fear is unauthorized
duplication of one or two copies each by a billion anklebiters; large-
scale duplication by organized criminals is a nuisance, but pretty much
only that, when compared to billions of copies circulated without payment
to billions' of folks one-friend-each.

Thor

---------------------------------------------------------------------
The Cryptography Mailing List
Unsubscribe by sending "unsubscribe cryptography" to majordomo at metzdowd.com