The point was that you can either make the binding cost no computational resources for the reader (by publishing and the binding in plaintext), or you can make it cost some mild amount of computational power (say, put it in a list of 20 fake numbers), or you could make it computationally infeasible by encrypting it. Is the second option illegal? At what computational difficulty does it become illegal?

Either way, the fact that the third is legal means that the law depends on compute power.

It's a judgement call. Which is why we have judges.

There are interesting paradoxes though when you mix "legalistic" and "geeky" worldviews. :P E.g. let's say somebody publishes the statement "one of the prime factors in Bill's SSN is 7"... and then someone else, independently... ;P You get the point! :)

Then you look at them, determine their intent (to publish Bill's SSN) and punish them or not accordingly. That's the point of the legal system and judges - as opposed to computer code they actually tend to try and determine intent as an important part of a ruling.

> Then you look at them, determine their intent (to publish Bill's SSN) and punish them or not accordingly.

That sounds like a very pragmatic view of the legal system. I'm more idealistic so I'd rather say that you have to "prove" (rather than "determine") their intent before you can punish them. ;P

Look, I'm not saying you're wrong, but there absolutely are times when the "geeks" win. For example consider export control for cryptographic algorithms which was thoroughly ridiculed by "geeks" [1] and later scrapped IIRC.

1. http://www.cypherspace.org/rsa/

The problem is, the intent of such action is, most probably, to just fuck with the law (requiring it to solve the task it can't work on, because its from the different realm[1]), not to publish Bill's SSN.

Legal system may try to handle the issue - in this particular case, it's possible to just ignore the primary intent and decide whenever an intent to publish did exist instead - but sometimes it just fails. I think that's, for example, how PGP source code was exported, working around munitions export laws.

[1] It must be dual. In a same manner formal logic can not be always applied to the realm of legal affairs, laws may malfunction when asked to handle problems from the domains they don't really map to.

Didn't Sony try to claim a patent on the specific number, rather than some kind of security argument?

Motorola has a patent—asserted in the Microsoft case—on a specific number used as an error correcting code.

No. To patent something you have to disclose it, so patenting a secret key would totally defeat the purpose.

Well, I remember that being part of the contention. Sony said, "we've patented it". Everyone else said, "how can you have patented it, when you won't tell anyone what it is!?". Or something along those lines.

If you can find the patent number that would make it easier to formulate a cogent response.

Some googling suggests it was not a patent they claimed, but that it was an "illegal number". Very possible I confused the two.

https://en.wikipedia.org/wiki/Illegal_number

I think at the time what I found ridiculous was that you were prohibited from "having" the number, but of course nobody could say what number it was you weren't allowed to have... seemed like something that belonged in Dr. Strangelove.