May 12th, 2002

bswing

Gilmore on copy "protection"

Gilmore Thrashes Intel on "content" control
Date: Sun, 12 May 2002 02:03:11 -0400
From: "R. A. Hettinga" <rah@shipwright.com>
To: Digital Bearer Settlement List <dbs@philodox.com>

http://www.well.com/~doctorow/gilmoreintel.txt



I spoke with an audience of Intel executives and security folks in
November 2000 about its copy-prevention technologies, at the
invitation of Pat Gelsinger.  Brian Snow of NSA, a respected
colleague, spoke along with me (about building integrity into
products).  I have not been shy about telling Intel what I think of
their efforts to prevent users from making legitimate uses of their
products.

The heart of the difference between Intel and I on this point is on
what you called "the need to protect content".  I believe that there is
NO need to so-called protect so-called content.  In fact, both the
structure of our society (free speech and capitalism) and the structure
of our technologies (open standards that plug together in innumerable
ways to satisfy innumerable desires) require a lack of restrictions
on who can transmit what information to who.

Intel builds machines that process data.  "Content" is just data.
Every piece of data that an Intel processor or networking component
handles is copyrighted by somebody, under the Berne Convention.  It's
all "content".  You could talk about "protecting data" but people
would realize that preventing it from being copied does not "protect"
their data.  Frequently you NEED to copy your data -- e.g. onto a
backup tape -- to protect it.  So instead you use this made-up word
"content".  Since nobody knows a definition for "content", you can say
the most outrageous things about it and get away with it.

Intel's chips have no way to tell what permission that individual chip
owner has under the copyright law, for many reasons.  (The laws
change, copyrights expire, individuals or companies get more rights
than the general public does because they signed licenses, there are
things that everyone has the legal right to do with data whether or
not it is copyrighted, etc etc etc.)

If Intel really thought that there was a "need to protect content", it
would have built features into its chips to make sure that none of
Intel's OWN chip designs, hardware designs, software, documents, trade
secrets, and other intellectual property could ever be stolen or
copied using Intel equipment.  That's Intel's first and foremost
interest in intellectual property protection -- but it has expended
zero effort toward "fixing" its chips to provide technical barriers
against such unlawful copying.  Therefore I discount the claim that
Intel sees a "need to protect content".

What Intel is doing is a cynical scheme to buy off an oligopoly.

What really saddens me is that Intel has no need to buy off that
oligopoly.  In fact it's quite the opposite.  Intel's efforts to suck
up to that oligopoly have CREATED the perception among policy makers
that "something needs to be done", since Intel and the oligopoly
"agree on the problem".  Intel and other honest manufacturers should
stand fast and say, "We are not the world's policemen.  We sell
general purpose equipment and we make it as flexible as possible to
attract the broadest range of customers.  You can't hold the man who
makes pencils responsible because a bookie used a pencil to write down
a bet.  And you can't demand that he design a pencil that can't be
used to write down a bet."  If you answered the oligopoly demands in
those terms, there would be no political "problem".  And you would
have good Supreme Court law behind you -- the Betamax case.

Instead, you are working to undermine the Betamax case, the
competitiveness of your own industry, the interests of your
own customers, and the foundations of your own free society.

Intel has been running the wrong direction for years about this.
Les Vadasz claimed recently, both to Dave Farber and to Congress, that
Intel wants to protect consumer rights and not kowtow to the
oligopolies.  But he was, as I said, either lying or ineffective at
actually getting Intel to walk his talk.  Because Intel is at the heart
of mandating copy-prevention standards that would be forced into all
PC's, all TV's, and all software.

And the whole thing is coming to a head this week, in a farcial
"consensus" among a self-selected set of people with zero consumer
representation.  This "consensus" will be billed to Congress and the
FCC as a "consensus among all interested parties" that everybody
should be banned from competing to offer consumers better products
than what Intel is willing to offer.  Because Intel wants to make sure
that nobody can offer the consumer a better deal -- since Intel knows
in its heart that no consumer would buy the locked-down Intel product
if they had any choice in the matter.

John Gilmore


[The addendum, about a week later:]

You may not realize the pervasiveness of the requirement being
demanded by Hollywood.  A friend of mine is planning to build digital
speakers (he's an EE who builds his own speakers).  Just as carrying
an analog signal to a video monitor over a cable degrades it, analog
speaker cables also degrade the signal.  Sending the audio to the
speakers digitally would carry the signal more faithfully, and it's
cheap these days.  It has other advantages: you could plug all of your
speakers into your house Ethernet, with little selectors on them to
pick which packets each speaker would listen for.  Then you wouldn't
have to run separate pairs of wires from your stereo to each room;
you'd just feed each room an Ethernet connection.

If the BPDG mandate goes through, digital speakers would have to
include copy-protection IN EACH SPEAKER.  If they didn't, they could
not be used to play the sound from HDTV or DVDs (or SDMI or whatever
comes next).  Each speaker would need two-way communication, so that
the copy-protection could negotiate keys.

If someone made digital headphones, you'd need a little cop, I mean a
little copy-prevention chip, in each HEADPHONE.

If someone made hearing aids that could receive wireless digital sound
(a godsend in concert halls or meeting rooms), you'd need a little
censor in each EAR.

To get these censorship chips, or the technical info needed to build a
speaker with them, my engineer friend would have to sign
nondisclosure, and sign a 50-page contract promising never to build
them into a system that could ever be used to copy music.  And the
technology in the chips would probably have a feature that would let
record companies later "self-destruct" the chip if anyone ever
discovered how to get the digital music out of it.  That is, as a
result of some teenager somewhere in the world cracking the chip's
scheme, a record or movie company would release hidden codes in their
products that when played would cause my friend's custom home-made
speakers to stop working -- permanently.

After reading a letter-to-the-editor in the Washington Post, sent by a
doctor who opposes the Hollings bills, I realized that many people
think that passing the "you must honor the 1-bit HDTV copy prevention"
measure would somehow make the world better than if the Hollings bill
was passed.  It won't.  Here's the letter:

> Hold On There a Minute, Jack!
> In response to Jack Valenti's May 1 editorial-page commentary "Stop Digital
> Piracy the DVD Way": I don't give a hoot whether or not digital video can be
> downloaded over the Internet, either now or in the future. What I do care
> about is that the computers I use in my medical laboratory practice continue
> to be designed by engineers and programmers rather than by government
> bureaucrats working as surrogates for copyright holders.
> The legislation introduced by Sen. Hollings that Mr. Valenti supports is
> arrogant and audacious, requiring the general public to both pay for and
> suffer performance degradation in order to reconfigure computing technology
> to serve the needs of the entertainment industry.
> Furthermore, it vastly exceeds the constitutional authority given Congress
> with regard to copyright regulation.
> Donn Fishbein, M.D., Ph.D.
> Coldwater, Ohio

Even if the Hollings bills die ugly deaths, but the BPDG mandate is
created by FCC and/or Congress, you'll see the exact same effect on
medical laboratory computers.  Computer monitors will only come with
copy prevention circuitry, because if they didn't, you couldn't plug
them in to computers that do copy prevention on their video outputs.
Computers will come with copy prevention on their video outputs,
because if they didn't, you wouldn't be able to plug in a PCI-bus
board to get HDTV (or play DVDs or etc).  Hard drives will come with
copy prevention, because if they didn't, software couldn't record local
TiVo-like copies of HDTV shows, for the consumer to view later, or save
because they want to see it again.

The whole system has to be seamless to implement even the SMALLEST
grain of protection ("honoring that copy prevention bit in HDTV
signals").  If the components allowed themselves to interoperate with
other components that have open interfaces with no prevention, then
the sacred bits could escape!

It's possible that mfrs would build two kinds of computers and
monitors and hard drives -- but only if people actively insisted on
buying NON COPY PREVENTED computers.  You would effectively have two
separate product lines.  The catch is that you could plug copy
prevented outboard subcomponents into an ordinary computer, and it
would work (e.g. a copy prevented monitor WILL play unencrypted
video).  But you could never plug any ordinary outboard component into
a copy prevented computer (e.g. an ordinary monitor WON'T display
video from a copy prevented video card -- the card will refuse to
talk to it).

When mfrs have to decide which product line to drop to cut costs, which
one will they drop -- the ordinary ("incompatible") one, or the copy
prevented ("compatible") one?

        John Gilmore

--
-----------------
R. A. Hettinga
[Error: Irreparable invalid markup ('<mailto:>') in entry. Owner must fix manually. Raw contents below.]

Gilmore Thrashes Intel on "content" control
Date: Sun, 12 May 2002 02:03:11 -0400
From: "R. A. Hettinga" <rah@shipwright.com>
To: Digital Bearer Settlement List <dbs@philodox.com>

http://www.well.com/~doctorow/gilmoreintel.txt



I spoke with an audience of Intel executives and security folks in
November 2000 about its copy-prevention technologies, at the
invitation of Pat Gelsinger.  Brian Snow of NSA, a respected
colleague, spoke along with me (about building integrity into
products).  I have not been shy about telling Intel what I think of
their efforts to prevent users from making legitimate uses of their
products.

The heart of the difference between Intel and I on this point is on
what you called "the need to protect content".  I believe that there is
NO need to so-called protect so-called content.  In fact, both the
structure of our society (free speech and capitalism) and the structure
of our technologies (open standards that plug together in innumerable
ways to satisfy innumerable desires) require a lack of restrictions
on who can transmit what information to who.

Intel builds machines that process data.  "Content" is just data.
Every piece of data that an Intel processor or networking component
handles is copyrighted by somebody, under the Berne Convention.  It's
all "content".  You could talk about "protecting data" but people
would realize that preventing it from being copied does not "protect"
their data.  Frequently you NEED to copy your data -- e.g. onto a
backup tape -- to protect it.  So instead you use this made-up word
"content".  Since nobody knows a definition for "content", you can say
the most outrageous things about it and get away with it.

Intel's chips have no way to tell what permission that individual chip
owner has under the copyright law, for many reasons.  (The laws
change, copyrights expire, individuals or companies get more rights
than the general public does because they signed licenses, there are
things that everyone has the legal right to do with data whether or
not it is copyrighted, etc etc etc.)

If Intel really thought that there was a "need to protect content", it
would have built features into its chips to make sure that none of
Intel's OWN chip designs, hardware designs, software, documents, trade
secrets, and other intellectual property could ever be stolen or
copied using Intel equipment.  That's Intel's first and foremost
interest in intellectual property protection -- but it has expended
zero effort toward "fixing" its chips to provide technical barriers
against such unlawful copying.  Therefore I discount the claim that
Intel sees a "need to protect content".

What Intel is doing is a cynical scheme to buy off an oligopoly.

What really saddens me is that Intel has no need to buy off that
oligopoly.  In fact it's quite the opposite.  Intel's efforts to suck
up to that oligopoly have CREATED the perception among policy makers
that "something needs to be done", since Intel and the oligopoly
"agree on the problem".  Intel and other honest manufacturers should
stand fast and say, "We are not the world's policemen.  We sell
general purpose equipment and we make it as flexible as possible to
attract the broadest range of customers.  You can't hold the man who
makes pencils responsible because a bookie used a pencil to write down
a bet.  And you can't demand that he design a pencil that can't be
used to write down a bet."  If you answered the oligopoly demands in
those terms, there would be no political "problem".  And you would
have good Supreme Court law behind you -- the Betamax case.

Instead, you are working to undermine the Betamax case, the
competitiveness of your own industry, the interests of your
own customers, and the foundations of your own free society.

Intel has been running the wrong direction for years about this.
Les Vadasz claimed recently, both to Dave Farber and to Congress, that
Intel wants to protect consumer rights and not kowtow to the
oligopolies.  But he was, as I said, either lying or ineffective at
actually getting Intel to walk his talk.  Because Intel is at the heart
of mandating copy-prevention standards that would be forced into all
PC's, all TV's, and all software.

And the whole thing is coming to a head this week, in a farcial
"consensus" among a self-selected set of people with zero consumer
representation.  This "consensus" will be billed to Congress and the
FCC as a "consensus among all interested parties" that everybody
should be banned from competing to offer consumers better products
than what Intel is willing to offer.  Because Intel wants to make sure
that nobody can offer the consumer a better deal -- since Intel knows
in its heart that no consumer would buy the locked-down Intel product
if they had any choice in the matter.

John Gilmore


[The addendum, about a week later:]

You may not realize the pervasiveness of the requirement being
demanded by Hollywood.  A friend of mine is planning to build digital
speakers (he's an EE who builds his own speakers).  Just as carrying
an analog signal to a video monitor over a cable degrades it, analog
speaker cables also degrade the signal.  Sending the audio to the
speakers digitally would carry the signal more faithfully, and it's
cheap these days.  It has other advantages: you could plug all of your
speakers into your house Ethernet, with little selectors on them to
pick which packets each speaker would listen for.  Then you wouldn't
have to run separate pairs of wires from your stereo to each room;
you'd just feed each room an Ethernet connection.

If the BPDG mandate goes through, digital speakers would have to
include copy-protection IN EACH SPEAKER.  If they didn't, they could
not be used to play the sound from HDTV or DVDs (or SDMI or whatever
comes next).  Each speaker would need two-way communication, so that
the copy-protection could negotiate keys.

If someone made digital headphones, you'd need a little cop, I mean a
little copy-prevention chip, in each HEADPHONE.

If someone made hearing aids that could receive wireless digital sound
(a godsend in concert halls or meeting rooms), you'd need a little
censor in each EAR.

To get these censorship chips, or the technical info needed to build a
speaker with them, my engineer friend would have to sign
nondisclosure, and sign a 50-page contract promising never to build
them into a system that could ever be used to copy music.  And the
technology in the chips would probably have a feature that would let
record companies later "self-destruct" the chip if anyone ever
discovered how to get the digital music out of it.  That is, as a
result of some teenager somewhere in the world cracking the chip's
scheme, a record or movie company would release hidden codes in their
products that when played would cause my friend's custom home-made
speakers to stop working -- permanently.

After reading a letter-to-the-editor in the Washington Post, sent by a
doctor who opposes the Hollings bills, I realized that many people
think that passing the "you must honor the 1-bit HDTV copy prevention"
measure would somehow make the world better than if the Hollings bill
was passed.  It won't.  Here's the letter:

> Hold On There a Minute, Jack!
> In response to Jack Valenti's May 1 editorial-page commentary "Stop Digital
> Piracy the DVD Way": I don't give a hoot whether or not digital video can be
> downloaded over the Internet, either now or in the future. What I do care
> about is that the computers I use in my medical laboratory practice continue
> to be designed by engineers and programmers rather than by government
> bureaucrats working as surrogates for copyright holders.
> The legislation introduced by Sen. Hollings that Mr. Valenti supports is
> arrogant and audacious, requiring the general public to both pay for and
> suffer performance degradation in order to reconfigure computing technology
> to serve the needs of the entertainment industry.
> Furthermore, it vastly exceeds the constitutional authority given Congress
> with regard to copyright regulation.
> Donn Fishbein, M.D., Ph.D.
> Coldwater, Ohio

Even if the Hollings bills die ugly deaths, but the BPDG mandate is
created by FCC and/or Congress, you'll see the exact same effect on
medical laboratory computers.  Computer monitors will only come with
copy prevention circuitry, because if they didn't, you couldn't plug
them in to computers that do copy prevention on their video outputs.
Computers will come with copy prevention on their video outputs,
because if they didn't, you wouldn't be able to plug in a PCI-bus
board to get HDTV (or play DVDs or etc).  Hard drives will come with
copy prevention, because if they didn't, software couldn't record local
TiVo-like copies of HDTV shows, for the consumer to view later, or save
because they want to see it again.

The whole system has to be seamless to implement even the SMALLEST
grain of protection ("honoring that copy prevention bit in HDTV
signals").  If the components allowed themselves to interoperate with
other components that have open interfaces with no prevention, then
the sacred bits could escape!

It's possible that mfrs would build two kinds of computers and
monitors and hard drives -- but only if people actively insisted on
buying NON COPY PREVENTED computers.  You would effectively have two
separate product lines.  The catch is that you could plug copy
prevented outboard subcomponents into an ordinary computer, and it
would work (e.g. a copy prevented monitor WILL play unencrypted
video).  But you could never plug any ordinary outboard component into
a copy prevented computer (e.g. an ordinary monitor WON'T display
video from a copy prevented video card -- the card will refuse to
talk to it).

When mfrs have to decide which product line to drop to cut costs, which
one will they drop -- the ordinary ("incompatible") one, or the copy
prevented ("compatible") one?

        John Gilmore

--
-----------------
R. A. Hettinga <mailto: rah@ibuc.com>
The Internet Bearer Underwriting Corporation <http://www.ibuc.com/>
44 Farquhar Street, Boston, MA 02131 USA
"... however it may deserve respect for its usefulness and antiquity,
[predicting the end of the world] has not been found agreeable to
experience." -- Edward Gibbon, 'Decline and Fall of the Roman Empire'