Seth Schoen raises two interesting issues in his response to Richard Stallman’s essay on “trusted computing.” (To see Seth’s posting, click here and scroll down to the “Trusted computing” heading.)

Stallman says

[Trusted computing] is designed to stop your computer from functioning as a general-purpose computer.

Schoen responds:

Neither of these concerns is applicable at all to Palladium (as Microsoft has described it to us) or to TCPA (as the TCPA has specified it and as it has been implemented). While Microsoft could be misleading us about Palladium, the TCPA specification is public and implementations of it have already been made.

It’s possible that some other trusted computing system could have such a misfeature, but the design of TCPA and Palladium doesn’t require these properties at all, as far as I can tell, and they seem to be more or less independent.

Schoen is right here – Palladium and TCPA do not do what Stallman says it does. Stallman seems too eager to blame Microsoft for the sins of others.

The conversation then moves on to the connection between Palladium and the Hollings CBDTPA. The Hollings bill mandates that some kind of “trusted computing” restrictions be made mandatory in essentially all digital devices. But what kind of restrictions would be mandated?

Stallman implies strongly that the CBDTPA would mandate the use of Palladium. Schoen disagrees, saying that he is “not convinced that something like Palladium is the infrastructure contemplated by the CBDTPA.”

Here I don’t know who is right. The CBDTPA is cleverly constructed so that it doesn’t say what it is mandating – it leaves that to be decided later, either by the FCC or by a vaguely-specified industry consortium. This gives CBDTPA advocates a way to dodge hard questions about the bill’s effects, by invoking a hoped-for perfect technical solution that is just around the corner. Given the track record of copy restriction and its advocates, I think we should insist on taking a test drive before we buy this used car.

Alex Halderman, a senior here at Princeton, has written a very interesting paper entitled “Evaluating New Copy-Prevention Techniques for Audio CDs.” Here is the paper’s abstract:

Several major record labels are adopting a new family of copy-prevention techniques intended to limit “casual” copying by compact disc owners using their personal computers. These employ deliberate data errors introduced into discs during manufacturing to cause incompatibility with PCs without affecting ordinary CD players. We examine three such recordings: A Tribute to Jim Reeves by Charley Pride, A New Day Has Come by Celine Dion, and More Music from The Fast and the Furious by various artists. In tests with different CD-ROM drives, operating systems, and playback software, we find these discs are unreadable in most widely-used applications today. We analyze the specific technical differences between the modified recordings and standard audio CDs, and we consider repairs to hardware and software that would restore compatibility. We conclude that these schemes are harmful to legitimate CD owners and will not reduce illegal copying in the long term, so the music industry should reconsider their deployment.

The paper will appear in the proceedings of the ACM’s DRM Workshop. It’s currently available, but only in PostScript format, on the Workshop’s site. (It’s available in PDF format here.)

It’s rare to see a workshop like this accept a single-author paper written by an undergraduate; but this paper is really good. (Grad schools: you want this guy!)

Seth Finkelstein and Eric Albert criticize my claim that the fallacy of the almost-general-purpose computer can best be illustrated by analogy to an almost-general-purpose spoken language. They make some good points, but I think my original conclusion is still sound.

Seth argues that speech (or a program) can be regulated by making it extremely difficult to express, even if it isn’t strictly impossible to say. I’m skeptical of this claim for human languages, since it seems to me that no usable language can hope to prevent people from creating new words and then teaching others what they mean. I think my skepticism is even more valid for computer languages. If a computer language makes something difficult but not impossible, then some programmer will create a library that provides the difficult functionality in more convenient form. This is the computer equivalent of creating a new word and defining it for others.

Eric argues that advancing technology might make it possible to restrict what people can say online. I’m skeptical, but he may be right that restrictions on, say, porn may become more accurately enforceable over time. Still, my point was not that mega-censorship is impossible, but that mega-censorship necessarily causes huge collateral damage.

There’s another obvious reason to like the 1984 analogy: using it puts the anti-computer forces into the shoes of the 1984 government. (I don’t think they’ll spend a lot of time comparing and contrasting themselves with the 1984 government.)

You may say that this is cheap rhetorical trick, but I disagree. I believe that code is speech, and I believe that its status as speech is not just a legal technicality but a deep truth about the social value of code. What the code-regulators want is not so different from what the speech-regulators of 1984 wanted.