Tom Duff ran across a FAQ that describes the evils of Intel's TCPA and Microsoft's Palladium. If you still think Digital Rights Management is in your best interest, take the time to read this FAQ.
I'm still waiting for more details. The corporate wet dream that you can give users full details to view content at full quality and then take away that knowledge again on hardware which is "known" based on what it sends over 10BaseT... Dunno. I'll believe it when I see it. They're welcome to claim that nobody will figure out how to fake a "Fritz Chip" encryption algorithm, just as I'm welcome to claim that I'm the king of all pink flamingos. I suppose they could try to make sure that a sufficiently wide variety of licenses were in use, so not all Linux-based not-a-TCPA boxes could use the same signature, but then the question becomes not breaking a single Fritz chip once, but breaking the way that keys are assigned. The keys don't have to be real or unique, they just have to credible enough not to be immediately discredited as fake over the course of a couple of transactions. In fact, one obvious modification to a Linux-based Fritz-spoofing box would be to have periodic key randomization so that you could periodically say "oh, and change me to some other random identity." Rather like what Zero Knowledge Systems' products did. Granted, all of this is questionable under the DMCA. Conceivably, they may have the legal right to say "and nobody is allowed to build a competing product ever again." I suspect that that, too, will cause a few ruffled legal feathers.
I'm still waiting for more details. The corporate wet dream that you can give users full details to view content at full quality and then take away that knowledge again on hardware which is "known" based on what it sends over 10BaseT... Dunno. I'll believe it when I see it. They're welcome to claim that nobody will figure out how to fake a "Fritz Chip" encryption algorithm, just as I'm welcome to claim that I'm the king of all pink flamingos.
ReplyDeleteI suppose they could try to make sure that a sufficiently wide variety of licenses were in use, so not all Linux-based not-a-TCPA boxes could use the same signature, but then the question becomes not breaking a single Fritz chip once, but breaking the way that keys are assigned. The keys don't have to be real or unique, they just have to credible enough not to be immediately discredited as fake over the course of a couple of transactions.
In fact, one obvious modification to a Linux-based Fritz-spoofing box would be to have periodic key randomization so that you could periodically say "oh, and change me to some other random identity." Rather like what Zero Knowledge Systems' products did.
Granted, all of this is questionable under the DMCA. Conceivably, they may have the legal right to say "and nobody is allowed to build a competing product ever again." I suspect that that, too, will cause a few ruffled legal feathers.