OwnXORed

A year ago, I wrote "It's still rockin' XOR to me.", which I thought was a clever title for a post until it became popular. That post was about a program which used XOR to create files with ambiguous copyright properties. The gist of my post was mainly that "this is nothing new." Then I linked to very similar ideas that had floated in years before. Besides my "nothing new" assertion, I characterized an earlier debate which concluded, basically, "this won't work."

That post has been seen many times, largely because Copyfight linked to it.

Recently, proponents of yet another technical system for ambiguating copyrights came across it and posted a comment. I replied that they should read Matthew Skala's much better articles on the subject: "What Colour are your bits?" and "Colour, social beings, and undecidability." I think he really puts his finger on the misunderstandings at work here, and he completely convinced me that making a point about copyright with encoding hocus pocus is a dead end.

Still, I don't want to be closed minded, so I read their paper anyway. I find that the system they're talking about is pretty close to a system I linked to from five years ago, but with more automation.

Instead of focusing on the contents of files, the OFF System folks focus on ownership. They've basically shown that there's no sane way to assert ownership of blocks in their system, so there must be something wrong with the idea of ownership.

Here's an example of the thinking:

The separation of possession from violation is counter intuitive to many people. It is often attacked as follows: If block Z is copyrighted by Brittney, and you without permission possess a block B and a block C such that they XOR together to reproduce block Z, then you possess an encoding of block Z and have violated Brittney's copyright.

This logic is easily shown false by the following scenario:

[details omitted]

In the above case Morgan can legitimately hold blocks B, C, D, and E in order to reproduce blocks X and Y. Holding these blocks in no way implies that Morgan has ever reproduced Z, intends to reproduce Z, or knows he can reproduce Z.

I want to focus on that last statement.

I'm not a lawyer, but it seems to me that intent and knowledge are important. Courts have ways of determining what was intended and what was known, and passing judgments based on those things. That's not to say they can do this perfectly, but in some sense, that's their job. I think a court would agree that Morgan hasn't done anything wrong (but I'm still not a lawyer).

If you have a computer throwing randomly generated data around willy-nilly with no criminal intent, that's fine. If your system, however, is some sleight-of-hand, obviously designed to confuse a judge, I doubt the law will look favorably on that.

In short, the law doesn't care about encodings. If your system is used to produce something perceived as identical to a copyrighted work (without permission), that's illegal. Of course, the folks who made the system have heard this before.

That is why, the team took the time to actually put together content and burn it to CD. According to traditional understanding, there must exist things that are on the "Shock CD" and things that are not on the Shock CD.

This is still the wrong focus. Copyright isn't concerned so much with possession as it is with reproduction. If you look at the rights copyright grants, "possession" isn't among them. They're all to do with production, distribution, display, and performance. What the designers of the OFF System have done is create a system that reproduces something without "having" it anywhere, but the fact that there's no possession of copyrighted works is irrelevant as long as they're produced.

Deniability is not the same as legality, and it's especially not the same as innocence. Reproducing copyrighted works without permission is illegal, and the "no copyrighted works up my sleeve" bit doesn't change that.