Digg.com Insanity

[Cogito]

Right, so when we move into the domain of the standards of evidence, that's where I think the smoking gun resides. I haven't seen a random or representative sample of evidentiary issues, but I've looked at a handful of cases where inferences about intent were essential and I gotta say that the mismatch between evidence and conclusion can be appalling. A lot of it hinges on inferences based on statements, where the prosecution is simply wrong about what follows from a given statement.

Can you give some examples?

[DragonMaci]

If there is at least one legitimate use for such knowledge...

Such as? I do not think there is any but if you think you know one I would love to see it.

[DavidOdden]

Can you give some examples?

The book Language Crimes (Roger Shuy, 1996, Blackwells) has a high density of examples, and Forensic Linguistics (John Gibbons, 2003, Blackwells) also has a good supply of bad-inference cases. The case of Steve Suyat in Hawaii (Shuy book) is a good example of a bad inference about intent, centering around a surmise about his intent in interpreting "organizing" as a technical legal term (The prosecutor basically pulled the old razzle-dazzle on Suyat, and the judge disallowed probative expert evidence that the testimony was not perjurous, as judges are wont to do when they feel professionally threatened). Gibbons has a number of analogous example involving amateurish legal interences based on a person saying "Yes" (simple word, how could you screw up, and yet...). I especially recommend Shuy's book which gives a fair overview of the problem.

[Korthor]

Am I right in thinking that the codes in question don't decode the DVD format, but rather unlock the copy-protection mechanism? If so, then the codes are just a password to unlock a lock imposed by the DVD manufacturers. Why are passwords patentable? Is it a violation of property rights to unlock a lock without using the key provided by the manufacturer?

What if a conventional locksmith designed and patented a really clever lock. You put this lock on your house, but lose your key. You then call a locksmith. In order to unlock it, he has to insert tools that copy the metal patern of the original key. Is that a violation of IPR?

Maybe my claim that you should be able to patent locks but not keys seems strange to some, but it makes sense to me. After all, a the key's only function is to bypass an artificially imposed condition--the lock. Does that meat the "usefulness" standard required for patents?

Edited May 7, 2007 by Korthor

[DavidOdden]

After all, a the key's only function is to bypass an artificially imposed condition--the lock.

Similarly, a lock's only function is to enforce an artificially imposed condition, a closed door. A computer password's only function is to bypass an artificially imposed condition -- limited access to an account. Etc. Since man left his natural state on the African savannah, he has been living with artificial conditions, ones that do not occur in pure nature and only exist because of man's interference with Original Nature.

[Qwertz]

Locks and keys:

The lock manufacturer has a patent not only on the lock, but on the whole locking system, which includes the lock, they key blank, and the method for cutting the key blank. If your house has a Kwikset lock, the locksmith must use a Kwikset keyblank and a Kwikset cutting adapter for his universal key cutter. The Kwikset key blank has a unique shape to its grooved profile to allow it to fit into the Kwikset lock. Kwikset licenses the key design to third-party keyblank manufacturers and the cutting adapter to third-party universal key cutter manufacturers. The locksmith then determine the depths to cut into the keyblank and then uses licensed technology to cut the key. The depth information (the lock's 'password') is completely useless without the keyblank and key cutting machine. He could machine his own Kwikset keyblanks and cut them to depth by hand, but in doing so, he'd be infringing the keyblank patent.

That keys and cutting technologies are also patented is readily observable when you look at certain car keys, or tubular keys, the Zeiss-type keys used in The Club, or any other specialty key that cannot be cut on a universal cutter. Manufacturers of these types of keys generally are more restrictive about licensing out the cutting technologies, and they don't sell blanks. A talented locksmith could still determine the depths that would need to be cut, but he couldn't produce the key himself. Again, the 'password' is useless unless he infringes on the patent by machining a key from scratch.

AACS passwords:

AACSLA holds a patent on the encryption and decryption algorithms. If your HD-DVD is locked by AACS, it requires a volume ID (the password) and the decryption algorithm. The volume ID is stored on the disc using specialized hardware, so it can only be read using special hardware, much like a locksmith has to use special hardware to determine key depths. Once you have the volume ID (by looking it up on the internet, by cracking your player's firmware, or whatever), you still don't have enough to decode the content. You need an unlicensed player that will accept a volume ID input from a source other than the disc. You also need an unlicensed player because licensed players don't give access to the decrypted data stream. If you don't want access to the dercypted data stream, then why wouldn't you just play the disc in a licensed player? So you have to build or buy an unlicensed player, containing an unlicensed implementation of the decryption software and hardware. No licensed player accepts input of the ID from a source other than from the disc itself.

If you're trying to play a bit-for-bit copy of an HD-DVD, it won't play in licensed players because the volume ID is missing from the disc, and cannot be written to it without patented hardware that AACSLA is very restrictive about licensing. In order to play it, you need some way to put the password into the machine. Licensed players don't accept IDs from sources other than the disc itself, so you either will need to infringe the ID-writing patent by making a writing machine, or the decryption algorithm patent by making a player that accepts ID input from a third-party source.

It is true that a string of numbers cannot be patented. But the technology needed to turn that string of numbers into something useful can.

-Q

Edit: wrong verb.

Edited May 7, 2007 by Qwertz

[DragonMaci]

Qwertz has a very good point, one that you should bare in mind Korthor.