Archives for 2005

The Center for Democracy and Technology just released a new copyright policy paper. Derek Slater notes, astutely, that it tries at all costs to take the middle ground. It’s interesting to see what CDT sees as the middle ground.

Ernest Miller gives the paper a harsh review. I think Ernie is too harsh in some areas.

Rather than reviewing the whole paper, I’ll look here at the section on DRM. Here CDT’s strategy is essentially to wish that we lived on a planet where DRM could be consumer-friendly while preventing infringement. They’re smart enough not to claim that we live on such a planet now, only that people hope that we will soon:

While DRM systems can be very restrictive, much work is underway to create content protections that allow expansive consumer uses, while still protecting against widespread distribution.

(They footnote this by referring to FairPlay, TivoToGo, and AACS-LA, which all fall well short of their goal.) CDT asserts that if DRM systems that made everyone happy did exist, it would be good to use them. Fair enough. But what should we do in the actual world, where DRM that everyone loves is about as likely as teleportation or perpetual motion?

This means producers must be free to experiment with various models of digital distribution, using different content protection technologies and offering different sets of permissions and limitations. [Government DRM mandates are bad.]

Consumers, meanwhile, must have real options for purchasing different bundles of rights at different price points.

Producers should be free to experiment. Consumers should be free to buy. Gee, thanks.

Actually, this would be fine if CDT really meant that producers were free to experiment with DRM systems. Nowadays, everybody is a producer. If you take photographs, you’re a producer of copyrighted images. If you take home movies, you’re a producer of copyrighted video. If you write, you’re a producer of copyrighted text. We’re all producers. A world where we could all experiment would be good.

What they really mean, of course, is that some producers are more equal than others. Those who are expected to sell a few works to many people – or, given the way policy really gets made, those who have done so in the recent past – are called “producers”, while those who produce the vast majority of new copyrighted works are somehow called “consumers”. (And don’t say that big media produces the only works of value. Quick: Which still images do you value most in the world? I’ll bet they’re photos, and that they weren’t taken by a big media company.)

Here’s the bottom line: In the real world, DRM policy involves tradeoffs, and requires choices. Wishing for a magical DRM technology that will please everyone is not a strategy.

The big tech news today is that Apple will start using Intel microprocessors (the same brand used in PCs) in its Macintosh computers, starting next year. Some have speculated that this might be motivated by DRM. The theory is that Apple wants the anticopying features that will be built into the hardware of future Intel processors.

The theory is wrong.

Though they’re not talking much about it, savvy people in the computer industry have already figured out that hardware DRM support is a non-starter on general-purpose computers. At most, hardware DRM can plug one hole in a system with many holes, by preventing attacks that rely on running an operating system on top of an emulator rather than on top of a real hardware processor. Plenty of other attacks still work, by defeating insecure operating systems or applications, or by exploiting the analog hole, or by capturing content during production or distribution. Hardware DRM blocks one of the less likely attacks, which makes little if any difference.

If DRM is any part of Apple’s motivation – which I very much doubt – the reason can only be as a symbolic gesture of submission to Hollywood. One of the lessons of DVD copy protection is that Hollywood still seems to need the security blanket of DRM to justify accepting a new distribution medium. DVD copy protection didn’t actually keep any content from appearing on the darknet, but it did give Hollywood a false sense of security that seemed to be necessary to get them to release DVDs. It’s awfully hard to believe that Hollywood is so insistent on symbolic DRM that it could induce Apple to pay the price of switching chip makers.

Most likely, Apple is switching to Intel chips for the most basic reason: the Intel chips meet Apple’s basic needs better than IBM chips do. Some stories report that Intel had an advantage in producing fast chips that run cool and preserve battery power, for laptops. Perhaps Apple just believes that Intel, which makes many more chips than IBM, is a better bet for the future. Apple has its reasons, but DRM isn’t one of them.