DeCSS, crypto, law, and economics

[John S. Denker]

John Gilmore wrote in part:

 > firmware refuses to write in the key area of the disk, and the blank
 > disks are shipped with the key area obliterated.  (And, DVD readers
 > will only let you read out the keys from a disk after you've reverse
 > engineered some simple bits that the industry wouldn't reveal.)  So
 > you can't do any bit-for-bit copying of DVDs unless you have a very
 > expensive (and restrictively licensed) DVD mastering press.

That's an interesting bit of technology, but it
cannot be considered a cryptographically-strong
anti-piracy scheme.  Any (unPalladiated) software
player could be patched to remap the "key area" to
another part of the disk (with an insignificant
loss of total disk capacity).

As a general principle, it's hard to talk about
security unless you are clear about what the security
boundary is.  As applied to DVDs, this means they
didn't have a chance of being secure unless all the
security-related processing steps were performed
inside tamper-resistant hardware.

Also note that if your goal is to make pirate disks
that play on living-room-grade players, it is not
even sufficient to have DeCSS||remapping, because of
low-level hardware incompatibilities.  You cannot
play a home-made unencrypted disk in many living-room
grade players.

To summarize, the necessary condition for piracy is
something like
    (DeCES||remapping) && low_level_compatibility

The compatibility matrix for unencrypted disks is
something like:
                              living-room   computer
                                player      RW drive
                              -----------------------
   commercially-pressed disks    yes         yes

   home-made DVD+R or DVD+RW   often no      yes


 > bit-copying ability ....  No other major computer
 > storage technology has lacked it

It depends on what you mean by "major" and "computer".
The video-game industry is huge.  Since its earliest
days it has distributed many games on cartridges for
which most consumers lack bit-for-bit copy capability.

Not-easily-copyable media do not actually cause the
earth to fall out of its orbit.  Some people may
object to them on political or philosophical grounds,
but that's another matter.

I wrote in part:

 >> We should try to avoid overwrought arguments about the
 >> "morality" of market segmentation and/or arbitrage.

> Unfortunately you set the wrong tone by starting as apologist for it.

What's wrong with my "tone"?  I was trying to
set a tone of moderation.  That will seem "wrong"
to extremists on both sides, but complaining
about the tone won't change the facts.

I'm not even sure what "it" refers to (segmentation
and/or arbitrage) but since I'm not an apologist
for either I can reject this accusation without
needing to fully parse it.

Sometimes I'm a buyer.  Sometimes I'm a seller.  I
see no reason to be an apologist for either extreme.
IMHO extremists do everyone a disservice.

>> In fact it is easy to demonstrate that _some_ market
>> segmentation is good for society as a whole.

How does this make me an "apologist"?  It looks
to me like an inalterable objective fact.

> The kind of segmentation your graphs rely on can easily be created
> by *time* segmentation.  

Yes.  And by other means as well, such as (gasp!)
genuine product differentiation.

And I agree with all who have noted that it is
highly ironic that while one part of the US
government is busily negotiating free-trade zones,
to remove geography-based barriers to trade, other
parts of the government are enacting and enforcing
new laws (!criminal! laws) that the studios exploit
to erect novel geography-based barriers.

======================

>> The "created wealth" is given by the length of the
>> magenta line, the difference between the customer-value
>> curve and the producer-cost curve.

Wayne Allen Simpson wrote:
> This starts the thread, and the labels (and concepts) are wrong.  

The concepts are correct.
Various labels are used by various authors; there is
no standardization.

The purpose of terminology is to facilitate communication.
Given no terminology of any kind (standard or otherwise)
that was a_priori well-known to the readership, I have
tried to be clear about what I mean by the terms I use.
I believe anybody who wants to understand will be able
to understand.  (Those who want to not understand will
always find a way to not understand;  there's nothing
I can do about that.)

 >>  http://www.monmouth.com/~jsd/econ/val-cost.gif

>> Note that in the absence of market segmentation,
>> the society as a whole is worse off.  ...  If the
>> buyers insist on buying every unit at the lowest
>> possible price, society loses.
> 
> Not so!  As long as the price is greater than the cost (far far left of 
> your tick mark), there is no reason that the goods cannot be produced 
> FOREVER.  Society benefits FOREVER.  

Whether the goods can be produced forever is irrelevant,
because the _demand_ is saturated at the point in question
(left of the point of max created wealth, in the absence
of segmentation).  There is no benefit to anyone in producing
goods for which there is no demand.

Please do not misread my curves.  My ordinates represent
cumulative value and cumulative cost.  Most "econ 101"
textbooks present something else, namely per-unit price
curves, which are the derivative of my cumulative curves.
I greatly prefer the cumulative curves, because they can
readily represent such things as
  -- sunk costs (e.g. the Y-intercept of the blue line), and
  -- the created wealth (the length of the magenta line).

.. whereas the per-unit curves cannot reasonably represent
such things.

> What you have indicated is where 
> the *seller* maximizes gains, not 
 > where society maximizes gains.  That is
 > the optimum profits, not the optimum wealth.

I cannot imagine how my presentation could be so
grossly misinterpreted.

The market I described (the green line in the figures)
can be clearly seen splitting the created wealth (CW)
more-or-less evenly between the producers and the
consumers.  Maximum producer profit would be described
by a much higher green line, hugging the red line.

> Rather, the optimum wealth will be created by the lowest possible price, 
> as more consumers will be able to afford the commodity and more of the 
> commodity will be produced!

There is no rational basis for such an assertion.

First of all, when market segmentation is being
discussed, there is no such thing as "the" price.
There is an ensemble of prices, one per segment.

Rational economic principles require the "incremental"
unit -- the very last unit -- to go for the lowest
possible price.  This is a statement about the slope
of the green line at the endpoint of the green line.
  -- This condition on the slope does not seriously
     constrain the height of the green line, i.e.
     the way in which the CW is divvied up between
     producers and consumers.
  -- This condition on the incremental unit does
     not require other units to sell for the same
     price.

================

I remind everyone that I said

>> it is easy to demonstrate that _some_ market
>> segmentation is good for society as a whole. 

When I said _some_ segementation I meant _some_
segmentation [emphasis in the original].

I did not say we should put the fox in charge of
the henhouse.

  -- When I'm selling, I might hope for a seller's
   market (green line near the top of the feasible
   region).
  -- When I'm buying, I might hope for a buyer's
   market (green line near the bottom of the feasible
   region).

But I'm not so hypocritical as to think I'm entitled
to whatever most benefits me.  In the real world,
commerce revolves around _mutually beneficial_ bargains.
The seller benefits and the buyer benefits.

I'm not interested in bogus economic "principles"
that buyers invent in order to "prove" that all
markets must be buyers' markets ... or sellers
invent in order to "prove" that all markets must
be sellers' markets.

==========================

Eric Rescorla wrote:

> When there is a conflict between liberty and Pareto 
 > dominance, economists get a headache.

Really?  Maybe some of them do, but I suspect most of
them wouldn't formulate it as a conflict at all;  they
would just ask "how much do you want to pay for your
liberty?"

Example:  Suppose you have the choice of either carpooling
to work or taking your own car, solo.  The latter gives
you more liberty as to when you drive home.  But it comes
at a cost.

>> There are two not-quite-separate sensible questions:
>>  -- Find a way to maximize the created wealth.
>>  -- Decide how to divvy up the created wealth among
>> the various stakeholders:  inventors, authors, performers,
>> publishers, manufacturers, wholesalers, retailers,
>> consumers, et cetera.

Pete Chown wrote:

> Would you rather maximise wealth or maximise competition? 

I stand by what I wrote before.  We should maximize
the created wealth, but that is not the whole story;
we also need to arrange that the created wealth is
divvied up in a reasonable way.

Competition is not an end unto itself.  It is
_sometimes_ a means toward an acceptable divvying.
  -- Examples abound where it works amazingly well.
  -- Examples abound where it doesn't work and
   other means must be employed.

====

To be explicit about what I'm claiming and what I'm
not claiming:  A low-level supply versus demand
analysis (what-does-it-cost versus what-is-it-worth)
produces the curves shown in the figure and predicts
how much created wealth there will be.

But that's not the whole story.  We need to figure
out how the created wealth will be divvied up, and
the low-level analysis does not address that.

If you want to predict the market price (the slope
of the green line) the low-level analysis makes a
definite prediction about the incremental unit, but
for every other unit it predicts a _range_ of feasible
prices, i.e. a _solution set_ rather than a unique
solution to the price question.

There are ways to take the analysis to the next level,
in hopes of more fully describing/predicting the green
line, but then things become much more dependent on
details.  We can go there if people are interested, but
for now I don't need to go there;  my original statement
was crafted in terms of "society as a whole" precisely
so that it could be fully appreciated using only a
low-level analysis:

 >> In fact it is easy to demonstrate that _some_ market
 >> segmentation is good for society as a whole.


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