Last year, I wrote about an Ars Technica article covering a new IEEE study group looking at the idea of “Digital Personal Property”. The idea is to use DRM to make electronic property act more like physical property—you can lend it to a friend, but you lose access to it while your friend has it and your friend might choose not to give it back.
Yesterday, I received an e-mail from Paul Sweazy, the chair of the IEEE working group that is taking over for the study group in question, pointing me to the group’s press release (PDF), call for participation (PDF), and website. Where the study group studied the issue, the working group is actually going to try to write a standard.
The press release states:
The Working Group is developing a standard for the creation, distribution, and perpetual consumer-ownership of downloadable items of copyrighted works such as movies, music, books, and games. We will specify the required behavior of online and device-embedded services and of content player devices and applications in order to achieve the following goals: robust barriers against item counterfeiting, content ripping, and stranger-sharing; de-tethering of product items from suppliers; full consumer freedom of private content usage including backup and restore, annotation, and editing; reformatting and re-encoding for consumer-preferred player devices; simultaneous and concurrent sharing; ownership transfer including lending and borrowing, giving and taking, donating and reselling, bequeathing and inheriting; protection of consumer privacy and autonomy; leveraging of the natural social mechanisms that limit the sharing of personal property to private, trusted parties.
They’re certainly nothing if not ambitious.
But I thought back when I wrote the original piece and I still think now: they’re trying to solve the wrong problem. Why force digital media into the Procrustean bed of imitating physical media? It’s antithetical to the whole idea of progress. Rather, the publishing industry should be looking at new business models that are robust enough to survive people making the uses of digital media that they want to make, and drop the DRM boondoggle altogether.
And whether they want to prevent “content ripping” or not, if someone decides they want to hack the system to allow it they will. DPP is still DRM, whether they want to call it that or not—and if the entire DVD consortium couldn’t prevent people ripping their DRM, I have a hard time imagining how an IEEE working group will do better.
But on the other hand, perhaps considering all possible alternatives is a part of progress. If we don’t try it, then we can’t say for sure it wouldn’t work. Maybe it’s worth a shot after all—though I’ll admit I won’t be holding my breath.
From the IEEE P1817 website: Send one of two (moveable but uncopiable) playkeys from the vendor to an online playkey bank of your choosing.
Bwaahaahahah! Movable, but not “copiable,” digital files.
Ow, my brain hurts. Someone failed at computers 101.
(Oh, I know. They mean “the copy isn’t usable until/unless it’s been validated by a playkey bank, and only one works at a time.” But it’s hideously sloppy language, and that’s a bad sign from a programming group, especially one that’s trying to jump into something as linguistically complex as IP law.)
What a travesty! You’re 100% right to criticise these people for setting out to cripple modern technology in order to prop up a failed business model. Most workers try to “add value” through their work, some, sadly, add no value at all. These people’s mission is to REMOVE value from the economy. Well good luck to them because quite frankly they haven’t got a hope in hell of success, and thank goodness for that.
There is value in negative information. The *assumption* that matter-based property attributes can’t/shouldn’t be translated onto the mass-less digital plain is just that: an assumption. Untested, unverified, and little more than a widely held opinion.
I say, let them try.
It’s their money, their time, their efforts.
If/when they converge on a scheme and the scheme is implemented and tested in real-world competition with the hypothetical business models that are supposed to emerge, *then* we will know if their ideas are any good. So far most everybody is converging on either selling limited usage right via DRM’ed files or betting the farm on google ads, with only a few betting on unsecured content. Maybe these folks *can* come up with something.
Let’s not dismiss a new approach simply because it runs counter to the sacred cows of “free content”; that is: “DRM is *always* evil” and “ad money is infinite and can support all businesses”.
Until somebody comes up with truly new paradigms there are worse things than trying to adapt a familiar one to a new domain.
The horse might sing.
Johnny takes his coins and dollar bills and deposits them a bank account. His wealth is now represented by bits in a digital system. Then Johnny hacks the system and changes his balance to a much larger number. I’m sure you would agree that it is not “antithetical to the whole idea of progress” to build obstacles to Johnny’s hack, or for that matter, to allow banks to use digital technologies?
There is room for both new and old business models, and it has been irresponsible of us to not provide for the model that consumers know best. It would be Procrustean to insist that emulation of the singular nature of physical objects must be amputated just because it doesn’t fit in the “all bits are free” bed.
1. DRM systems emulate unownable rental services.
2. Unprotected files emulate unownable public goods.
3. P1817 emulates ownable private goods.
Providing this new alternative seems to us to fit with the IEEE’s tagline… Advancing Technology for Humanity. And we would like to assure everyone that we have no intention of emulating the disadvantages of physical products by limiting the advantages of digital ones.
I’m glad that you consider the possibility that it will work. It will work for the same reason that we provide and use bank accounts in spite of fraud, counterfeiting, and hacking – because it serves the balance between our individual and collective self-interests.
There’s a big difference between DRM systems and online banking, thank goodness. No-one will give away access to their online bank account, but people will happily buy access to digital information and then illegally give copies away. So the undoubted success of online banking is not itself evidence that this new DRM mechanism will work in practice.
There are still a lot of people too used to getting free content online. I’m not sure if they represent the mainstream consumers, or have fully considered the economic implications of not having DRM. Can you imagine a writer or a singer seeing millions of copies of his work in circulation but only received money for a few copies? What’s he going to live on?
I agree that finding a balance is essential, but am also concerned that it should not be ahead of its time, or it will likely fail. Paul do you think the world is ready?
Felix: “The *assumption* that matter-based property attributes can’t/shouldn’t be translated onto the mass-less digital plain is just that: an assumption. Untested, unverified, and little more than a widely held opinion.”
On the contrary – DRM systems have been tried again and again and again (CSS for DVDs, SDMI in music, Adobe’s ADEPT for ebooks, all of the many, many software-copy-protection schemes in the 80s, etc). Whenever they have achieved widespread use, it has become economically valuable for others to crack them. In general they can avoid being cracked only by not ever achieving much economic importance.
In principle, such schemes have to “square the circle”: they have to provide digital information to the end user but somehow prevent them from using that information in various ways. They generally do this by encrypting the information, and embedding a secret decryption key into a system which presents that information. Hence the information can only be accessed through the system with the embedded key. The fundamental flaw of this is that the system with the embedded key is IN THE HANDS OF THE END USER. So rather than being strictly secret (like the keys used in online banking), the key ends up being merely difficult to access.
So, not only is there direct empirical evidence that the project is flawed, it’s also clearly flawed in principle, in information-theoretic terms. Yet people still try, again and again and again … why? Because of ignorance of previous attempts; because of ignorance of the theoretical problem, and because of a conservative desire to cling to what they know; the features of hardware.
It’s interesting to note that “copy-protection” schemes in the shrink-wrapped software market in the 80s and 90s declined partly BECAUSE they were too much hassle for end users. Unencumbered software which was more easily piratable was more acceptable, and hence more valuable in the market, to end users. By lumbering their products with DRM, vendors actually REDUCED the market value of those products to the point that they could not compete.
It’s certainly not too early. There is a case to be made that it is too late, although I remain hopeful. Every day that goes by without digital private goods sees more people sufficiently offended by DRM that can rationalize the complete abandonment of copyright for all things digital. Now is the time. Ten years ago would have been better.
Your quite right about the the “flaw” as you defined it. Of course, every real-world security system and every monetary system is similarly flawed – currency gets counterfeited, online systems get hacked, DRM-protected content get ripped. The difference between the first two and the last one is a matter of motivation. The first two usually get defeated to to defeat a fair system. The last usually gets defeated to remedy an unfair situation. (It creates another unfair situation, but it appears to only hurt the suppliers that were already acting distrustful or selfish or greedy.) Since every system can be gamed, the keys to public acceptance are that (1) the system be judged by most consumers to be fair and (2) attacks on the system by individuals be seen as damaging to consumers in general. If we fail, it will be because we ignored those principles of fairness.
P1817 is no more a DRM system than the protections that keep people from changing the balance on their bank accounts. You can use that term if you wish, but I can assure you that DRM proponents don’t think it deserves the title.
P1817 is a DRM system; unlike bank account security, it’s not being offered to *consumers* to keep their content safe; it’s being offered to producers, to restrict the uses of digital content. It may limit (it can’t prevent) unauthorized copies and piracy; it will also prevent some fair use aspects.
Bank accounts are voluntary. You can keep your money as cash if you don’t trust a bank. Or you can use a different bank if you don’t like the policies of one.
Tech-savvy consumers won’t trust P1817 for the same reasons they haven’t trusted other DRM methods–the claim of “it’s very simple; you just give us all your identifying digital information and then we authorize you to have access to your purchase” falls a bit flat, and those who’ve been burned by the last several rounds of “oops, we don’t support that DRM method anymore” aren’t going to trust that this one will stick around.
That’s aside from the practical issues of server maintenance and continued support in new platforms/hardware, and release of content that’s no longer covered by copyright. If a person buys an ebook in the US, then moves to Canada where that text is in the public domain, can he get it unlocked? In another decade or so when the US public domain starts growing again, how will content that was formerly locked be opened for use?
Paul: I disagree: actually not every security system is similarly flawed.
Many security systems can remain secure by simply not disclosing their private cryptographic keys to end users. In the banking system these keys are protected by the physical security which banks have always used: locks, walls, armed guards, etc. Of course these keys can be stolen or leaked, or they can be exposed through carelessness, but the operation of the system doesn’t require those keys to be disclosed.
Whereas, on the contrary, DRM systems have the inbuilt and inescapable flaw that they must actually disclose the information they purport to protect. You must be able to read your eBook or watch your DVD, and if you can do that then you must be able to copy it.
As I see it, P1817 is not different in that regard to other forms of DRM: it’s an attempt to use some “Technological Protection Measure” to implement a particular kind of property system in digital information (“digital personal property”), but in principle it must be as problematic as any other DRM system.
It is common (less so at Teleread) for folks to assume that a system is busted, then to flame on about how dumb it is because it’s busted. Another option is to ask how it works.
If a consumer wishes to avoid online bank accounts entirely, that is his choice. Playkeys come in pairs (one online and one in a consumer device), and one can choose to only keep and use the one that resides in one of his electronic devices. However, the benefits of reaching a point of trust with at least one online entity (perhaps the EFF or ACLU may host playkey bank accounts someday) is that the online playkey makes content play possible using any player with online access. Limiting one’s self to just the twin device playkey requires an electronic connection to that particular device. Very private but not so convenient. The online playkey can be used to restore a device playkey that is lost when its host device fails or is destroyed or lost, so backup and restore works both for the data folder and for the playkeys. These sound like benefits to me.
We have almost all aspects of fair use covered already, in particular the “private fair use” issues like backup copies and resale rights. The hard one is “public fair use” in which content is extracted in unprotected form and republished. No purely technical solution is possible because this aspect of fair use is always a human judgement call. But we’ve got a crew of attorneys who are volunteering their time to help us technologists craft a solution to this too. At this time, and of the problems we have thought of so far, public fair use is the only one that we don’t have a solid solution for yet.
The purpose of the playkey banking system is to provide an isolation layer to assure that, while any consumer with a playkey can locate and contact the its vendor and its publisher, the vendor and publisher can’t locate the playkey or monitor its use or its users’ activities.
Every owner of digital personal properties will have the right to reformat them for current or future devices, and to re-encode them for emerging standards.
Regarding the unlocking of copies once they are in the public domain, I think that this will become a near non-issue. First, if it’s public domain, there are likely to be other sources of the unprotected content. We’ll need to lobby to make sure that draconian anti-circumvention laws can’t be illogically applied to public-domain content. and lastly, there is precious little reason to need to jailbreak a title you already own, whose use and sharing are unrestricted, and which may retain resale or sentimental value. Like printed books, they may have greater value then; they may retain annotations from a renowned previous owner or hold a digital personal dedication signed by the author or performer.
IEEE P1817 isn’t perfect. It can’t be, since we haven’t even held our first Working Group meeting yet. But the preceding Study Group has determined feasibility, and we are excited about our potential for success. The more smart people we attract, the better we will be at avoiding errors and covering corner cases. Luckily, it is far simpler to just reject usage and sharing restrictions than to try to invent user-friendly ones – they are so hard to find.
Limiting one’s self to just the twin device playkey requires an electronic connection to that particular device.
And this is what makes it just another form of DRM. Without the intervention of a third party, use of one’s purchase is restricted to a single device. It’s like saying “you can only read this book under one lamp. Anytime you like, you can call us up, and we’ll change the registered lamp for your book.”
It does sound much more practical than most current forms of DRM, but it comes with the same set of problems–public fair use restrictions, need for server reliability (who is going to pay for hosting all that bandwidth? for the entire length of copyright?), forward compatibility, and unlocking/releasing the content when it enters the public domain. (Perhaps once a title enters the public domain its playkey could have infinite copies instead of just one, since there’s no longer a reason to restrict it.)
And it’s likely to make publishers twitchy, because they’re currently very invested in a “one purchase = 1 user” model; P1817 would make it possible for online groups to share books around without buying them (just like physical book clubs–but globally, and without losing quality with each exchange). They have no motive for supporting a DRM system that gives them less control than they currently have.
While we desperately need a way to simply & legally transfer ownership of digital content, handing management of those transfers to a third party, even one as noble as the ACLU or EFF, isn’t going to satisfy that need. If the purpose of this is to prevent unauthorized copies, the playkey database properly should reside in a government department. (Which would serve the dual purpose of “stable location not subject to bankruptcy or weird lawsuit problems” and “remind the public that copyright enforcement *costs money*, and that’s part of why a large public domain is important.”)
It is feasible. As DRM plans go, it’s a very good one. It’s just subject to the same problems as every other DRM plan, all of which have been either cracked or abandoned. Being crackable doesn’t make it useless; Amazon makes plenty of money off their crackable Kindlebooks.