26 January 2001: See John Gilmore's version of this file and reader's comments:
http://www.toad.com/gnu/whatswrong.html
19 January 2001
To: cryptography@c2.net
Cc: Ron Rivest <rivest@theory.lcs.mit.edu>, gnu@toad.com
Subject: What's Wrong With Content Protection
Date: Thu, 18 Jan 2001 17:06:07 -0800
From: John Gilmore <gnu@toad.com>
Ron Rivest asked me:
> I think it would be illuminating to hear your views on the
> differences between the Intel/IBM content-protection proposals
> and existing practices for content protection in the TV
> scrambling domain. The devil's advocate position against your
> position would be: if the customer is willing to buy extra, or
> special, hardware to allow him to view protected content, what is
> wrong with that?
There is nothing wrong with allowing people to optionally choose to buy copy-protection products that they like.
What is wrong is when people who would like products that simply record bits, or audio, or video, without any copy protection, can't find any, because they have been driven off the market. By restrictive laws like the Audio Home Recording Act, which killed the DAT market. By "anti-circumvention" laws like the Digital Millennium Copyright Act, which EFF is now litigating. By Federal agency actions, like the FCC deciding a month ago that it will be illegal to offer citizens the capability to record HDTV programs, even if the citizens have the legal right to. By private agreements among major companies, such as SDMI and CPRM (that later end up being "submitted" as fait accompli to accredited standards committees, requiring an effort by the affected public to derail them). By private agreements behind the laws and standards, such as the unwritten agreement that DAT and MiniDisc recorders will treat analog inputs as if they contained copyrighted materials which the user has no rights in. (My recording of my brother's wedding is uncopyable, because my MiniDisc decks act as if I and my brother don't own the copyright on it.)
Pioneer New Media Technologies, who builds the recently announced recordable DVD drive for Apple, says "The major consumer applications for recordable DVD will be home movie editing and storage and digital photo storage". They carefully don't say "time-shifting TV programs, or recording streaming Internet videos", because the manufacturers and the distribution companies are in cahoots to make sure that that capability NEVER REACHES THE MARKET. Even though it's 100% legal to do so, under the Supreme Court's _Betamax_ decision. Streambox built software that let people record RealVideo streams on their hard disks; they were sued by Real under the DMCA, and took it off the market. According to Nomura Securities, DVD Recorder sales will exceed VCR sales in 2004 or 2005, and also exceed DVD Player-only sales by 2005. (http://www.kipinet.com/tdb/1000/10tdb04.htm) So by 2010 or so, few consumers will have access to a recorder that will let them save a copy of a TV program, or time-shift one, or let the kids watch it in the back of the car. Is anyone commenting on that social paradigm shift? Do we think it's good or bad? Do we get any say about it at all?
Instead, consumers will have to pay movie/TV companies over and over for the privilege of time-shifting or space-shifting. Even if they have purchased the movie, and it's stored at home on their own eqiupment, and they have high bandwidth access to it from wherever they are. This concept is called "pay per use". It can't compete with "You have the right to record a copy of what you have the right to see". These companies can't eliminate that right legally, because it would violate too many of the fundamentals of our society, so they are restricting the technology so you can't EXERCISE that right. In the process they ARE violating the fundamentals on which a stable and just society is based. But as long as society survives until after they're dead, they don't seem to care about its long-term stability.
What is wrong is when companies who make copy-protecting products don't disclose the restrictions to the consumers. Like Apple's recent happy-happy web pages on their new DVD-writing drive, announced this month (http://www.apple.com/idvd/). It's full of glowing info about how you can write DVDs based on your own DV movie recordings, etc. What it quietly neglects to say is that you can't use it to copy or time-shift or record any audio or video copyrighted by major companies. Even if you have the legal right to do so, the technology will prevent you. They don't say that you can't use it to mix and match video tracks from various artists, the way your CD burner will. It doesn't say that you can't copy-protect your OWN disks that it burns; that's a right the big manufacturers have reserved to themselves. They're not selling you a DVD-Authoring drive, which is for "professional use only". They're selling you a DVD-General drive, which cannot record the key-blocks needed to copy-protect your OWN recordings, nor can a DVD-General disc be used as a master to press your own DVDs in quantity. These distinctions are not even glossed over; they are simply ignored, not mentioned, invisible until after you buy the product.
It isn't just Apple who is misleading the consumer; it's epidemic. Sony portable mini-disc recorders only come with digital INPUT jacks, never digital OUTPUTS. Sound checks in -- but only checks out in low-quality analog formats. Intel touts the wonders of their TCPA (Trusted Computing Platform Architecture). You have to read between the lines to discover that it exists solely to spy on how you use your PC, so that any random third party across the Internet can decide whether to "trust" you -- the owner. TCPA isn't about reporting to YOU whether you can trust your own PC (e.g. whether it has a virus), it doesn't include that function. It exists to report to record companies about whether you have installed any software that lets you make copies of MP3s, or any free software to circumvent whatever feeble copy-protection system the record company uses. Intel is pushing HDCP (High Definition Content Protection) which is high speed hardware encryption that runs only on the cable between the computer and its CRT or LCD monitor. The only signal being encrypted is the one that the user is sitting there watching, so why is it encrypted? So that the user can't record what they can view! If the cable is tampered with, the video chip degrades the signal to "analog VCR quality".
Intel is also pushing SDMI and CPRM (Content Protection for Recordable Media) which would turn your own storage media (disk drives, flash ram, zip disks, etc) into co-conspirators with movie and record companies, to deny you (the owner of the computer and the media) the ability to store things on those media and get them back later. Instead some of the stored items would only come back with restrictions wired into the extraction software -- restrictions that are not under the control of the equipment owner, or of the law, but are matters of contract between the movie/record companies and the equipment/software makers. Such as, "you can't record copyrighted music on unencrypted media". If you try to record a song off the FM radio onto a CPRM audio recorder, it will refuse to record or play it, because it's watermarked but not encrypted. Even when recording your own brand-new original audio, the default settings for analog recordings are that they can never be copied, nor ever copied in higher fidelity than CD's, and that only one copy can be made even if copying is ever authorized (if the other restrictions are somehow bypassed). Intel and IBM don't tell you these things; you have to get to Page 11 of Exhibit B-1, "CPPM Compliance Rules for DVD-Audio" on page 45 of the 70-page "Interim CPRM/CPPM Adopters Agreement", available only after you fill out intrusive personal questions after following the link from http://www.dvdcca.org/4centity/ . All Intel tells you that CPPM will "give consumers access to more music" (http://www.intel.com/pressroom/archive/releases/aw032300.htm). Lying to your customers to mislead them into buying your products is wrong.
What is wrong is when scientific researchers are unable to study the field or to publish their findings. Professor Ed Felten of Princeton studied the SDMI "watermarking" systems in some detail, as part of a public study deliberately permitted by the secretive SDMI committee, so they could determine whether the public could crack their chosen schemes. (SDMI would not allow EFF to join its deliberations, saying that we had no legitimate interest in the proceedings because we weren't a music company or a manufacturer. There are no consumer or civil rights representatives in the SDMI consortium.) Prof. Felten was in the New York Times last week, saying the SDMI people and Princeton's lawyers are now telling him that he can't release his promised details on what was wrong with these watermarking systems, because of the Digital Millennium Copyright Act. It's OK to tell the SDMI companies how easy it is to break their scheme, but it isn't OK to tell the public or other scientific researchers.
What is wrong is when competitors are unable to build competing devices or software, vying for the favor of the consumers in the free market. Instead those devices are banned or threatened, and that software is censored and driven underground. Such as the open-source DeCSS and LiViD DVD player programs. Such as DVD players worldwide that can play American "Region 1" DVDs. EFF spent more than a million dollars last year in defending the publisher of a security magazine, and a Norwegian teenager, from movie industry attempts to have them censored and jailed, respectively, for publishing and writing competing software that lets DVDs be played or copied but does not follow the restrictive contracts that the movie studios imposed on most players. The movie studios spent $4 million on prosecuting the New York case alone. Few or no manufacturers are willing to put ordinary digital audio recorders on the market -- you see lots of MP3 players but where are the stereo MP3 recorders? They've been chilled into nonexistence by the threat of lawsuits. The ones that claim to record, record only "voice quality monaural".
What is wrong is when the controls that are enacted to protect the rights reserved under copyright are used for other purposes. Not to protect the existing rights, but to create new rights at the whim of the copyright holder. Movie companies insisted on a "region coding" system for DVDs, because they would make less money if DVD movies were actually tradeable worldwide under existing free-trade laws. (They couldn't charge high theatre ticket prices if the same movie was simultaneously available on DVDs, and they couldn't combine the ad campaigns of the theatres and the DVDs if they waited a long time between releasing it to theatres and releasing it to DVDs.) This system results in the situation where a consumer can buy a DVD player legally, buy a DVD legally, and put the two together, and the movie won't play. The user has every legal right to view the movie, but it won't play, because if it did, movie companies might make less money. Similar controls exist in DVDs to prevent people from fast-forwarding past the ads or those nonsensical "FBI Warnings".
Microsoft built some deliberately incompatible protocols into Windows 2000 so that competing Unix machines could not be used as DNS servers in some circumstances. Microsoft released a specification but only under an encrypted file format that claimed to require that readers agree not to use the information to compete with them. When someone decrypted the trivial encryption WITHOUT agreeing to the terms, Microsoft threatened to use the DMCA to sue Slashdot, the popular free-software news web site, who published the results. (Luckily for us, Slashdot has a backbone and said "go ahead, we'll defend that suit" and Microsoft chickened out.) Copyright doesn't grant the right to prevent competition, or to restrict global trade -- but somehow the legislation that was enacted to protect copyrights is being used to do just those things.
What is wrong is when social policy is created in smoke-filled back rooms, between movie/record company executives and computer company executives, not by open public discussion, by legislatures, and by courts. The CPRM specification, for example, allows a distributor of a bag of bits (who has access to software with this capability) to decide that future recipients will not be permitted to make copies of that bag of bits. Or that two copies are permitted, but not three. This policy is not legally enforceable, it was not created by law. The law says something different. But the policy will be enforced by equipment built by all the major manufacturers, because they will be sued by the movie/record companies if they dare to build interoperating equipment that lets consumers make THREE copies, or copies limited only by their legal rights. Is it unexpected that such back-room policies end up favoring the parties who were in the room, at the expense of consumers and the public?
What is wrong is when the balance between the rights of creators and the rights of freedom of speech and the press is lost. Because any increase in the rights of creators is a DECREASE in the public's right of free speech and publication. Whenever copyrights are extended, the public domain shrinks. The right of criticism, the right to dispute someone else's rendition of the truth, is damaged. The First Amendment gives an almost absolute right to publish; the Copyright clause gives a limited right to prevent publication by others. Any expansion of the right to prevent publication diminishes the right to publish. For example, nothing that was created after 1910 has entered the public domain, because as the years went by, the term of copyright kept getting extended. But the copy-rights created by technological restrictions are not even designed to end. There is nothing in the SDMI or CPRM spec that says, "After 2100 you will be permitted to copy the movies from 1910".
What is wrong is that a tiny tail of "copyright protection" is wagging the big dog of communications among humans. As Andy Odlyzko pointed out, (http://www.research.att.com/~amo/doc/eworld.html, see "Content is not king" and "The history of communications and its implications for the Internet"), "The annual movie theater ticket sales in the U.S. are well under $10 billion. The telephone industry collects that much money every two weeks!" Distorting the law and the technology of human communication and computing, in order to protect the interests of copyright holders, makes the world poorer overall. Even if it didn't violate fundamental policies for the long-term stability of societies, it would be the wrong economic decision.
What is wrong is that we have invented the technology to eliminate scarcity, but we are deliberately throwing it away to benefit those who profit from scarcity. We now have the means to duplicate any kind of information that can be compactly represented in digital media. We can replicate it worldwide, to billions of people, for very low costs, affordable by individuals. We are working hard on technologies that will permit other sorts of resources to be duplicated this easily, including arbitrary physical objects ("nanotechnology"; see http://www.foresight.org). The progress of science, technology, and free markets have produced an end to many kinds of scarcity. A hundred years ago, more than 99% of Americans were still using outhouses, and one out of every ten children died in infancy. Now even the poorest Americans have cars, television, telephones, heat, clean water, sanitary sewers -- things that the richest millionaires of 1900 could not buy. These technologies promise an end to physical want in the near future.
We should be rejoicing in mutually creating a heaven on earth! Instead, those crabbed souls who make their living from perpetuating scarcity are sneaking around, convincing co-conspirators to chain our cheap duplication technology so that it WON'T make copies -- at least not of the kind of goods THEY want to sell us. This is the worst sort of economic protectionism -- beggaring your own society for the benefit of an inefficient local industry. The record and movie distribution companies are careful not to point this out to us, but that is what is happening.
If by 2030 we have invented a matter duplicator that's as cheap as copying a CD today, will we outlaw it and drive it underground? So that farmers can make a living keeping food expensive, so that furniture makers can make a living preventing people from having beds and chairs that would cost a dollar to duplicate, so that builders won't be reduced to poverty because a comfortable house can be duplicated for a few hundred dollars? Yes, such developments would cause economic dislocations for sure. But should we drive them underground and keep the world impoverished to save these peoples' jobs? And would they really stay underground, or would the natural advantages of the technology cause the "underground" to rapidly overtake the rest of society?
I think we should embrace the era of plenty and work out how to mutually live in it. I think we should work on understanding how people can make a living by creating new things and providing services, rather than by restricting the duplication of existing things. That's what I've personally spent ten years doing, founding a successful free software support company. That company, Cygnus Solutions, annually invests more than $10 million into writing software, giving it away freely, and letting anyone modify or duplicate it. It funds that by collecting more than $25 million from customers, who benefit from having that software exist and be reliable and widespread. The company is now part of Red Hat, Inc -- which also makes its living by empowering its customers without restricting the duplication of its work. It's no coincidence that the open source, free software, and Linux communities are among the first to become alarmed at copy protection. They are actively making their livings or hobbies out of eliminating scarcity and increasing freedom in the operating system and application software markets. They see the real improvement in the world that results -- and the ugly reactions of the monopolistic and oligopolistic forces that such efforts obsolete.
Converting the whole world to operate without scarcity is a huge task. Such a large economic shift would take decades to spread through the entire world economy, making billions of new winners and new losers. We will be extremely lucky if by 2030 we are prepared to end scarcity without massive social turmoil, including riots, civil unrest, and world war. If we are to find a peaceful path to an era of plenty, we should be starting HERE AND NOW, transforming the industries we have already eliminated scarcity in -- text, audio, and video. Companies that can't adjust should disappear and be replaced by those who can. As these whole industries learn how to exist and thrive without creating artificial scarcity, they will provide models and expertise for other industries, which will need to change when their own inefficient production is replaced by efficient duplication ten or fifteen years from now. Relying on copy-protection now would send us in exactly the wrong direction! Copy protection pretends that the law and some fancy footwork with industrial cartels can maintain our current economic structures, in the face of a hurricane of positive technological change that is picking them up and sending them whirling like so many autumn leaves.
This may be a longer discussion than you wanted, Ron, but as you can see, I think there are a lot of things wrong with how copy protection techologies are being foisted on an unsuspecting public. I'd like to hear from you a similar discussion. Being devil's advocate for a moment, why should self-interested companies be permitted to shift the balance of fundamental liberties, risking free expression, free markets, scientific progress, consumer rights, societal stability, and the end of physical and informational want? Because somebody might be able to steal a song? That seems a rather flimsy excuse. I await your response.
John Gilmore
Electronic Frontier Foundation
Date: Thu, 18 Jan 2001 21:35:39 -0500
From: Ron Rivest <rivest@theory.lcs.mit.edu>
To: gnu@toad.com
Cc: cryptography@c2.net
Subject: Re: What's Wrong With Content Protection
John --
Great essay... thanks for replying at such length!
I'm going to decline your (perhaps rhetorical) invitation to provide a devils-advocate counter-argument, because I'm not the right person to do so; I am far too liberal in my own views to be a fair representative of the "dark side". In any case, I was asking more for an education (which you have generously provided) than an argument.
Cheers,
Ron Rivest