4 ms·

"DRM is a "smart contract", a protocol for enforcing a contract without laws or violence." No it isn't, because DRM fundamentally cannot enforce the contract i

by voyou 12y ago

"DRM is a "smart contract", a protocol for enforcing a contract without laws or violence." No it isn't, because DRM fundamentally cannot enforce the contract it is supposed to; you can't allow people to view something while preventing them from copying it. DRM only "works" if circumventing DRM is made illegal; DRM is an attempt by copyright holders to introduce legal restrictions on fair-use copying and on copying devices with substantial non-infringing use. DRM is entirely legal and political, not technical.

dchest 12y ago
That's a good argument, thank you! I'd argue that for DRM to work reasonably well (without coercion), it doesn't need to be perfect, it can work in the same way password stretching works: you don't get the absolute protection, but influence the cost (value). For example, when applied to software, DRM can be stripped, but then you'll be receiving the executable from a possibly untrusted source. However you are spot on that currently DRM is mostly legal and political, but I don't think that's the inherent property of it. * * * Unfortunately, seeing my comments downvoted makes me uncomfortable to continue discussion here on HN, as (when applied to reasonable comments) it provides instant feedback that you're going against the popular opinion, and I'd rather avoid such feedback, so I'll go think about it more and then write a blog post or something.
- ultramancool 12y ago
  DRM doesn't work nearly as well as password stretching. I can sit down for a few hours with IDA and crack nearly any DRM scheme. With password stretching, the result of how long attacks will take is predictable and they can actually be made difficult or require more hardware (and at larger expense) rather than simply requiring a more skilled attacker.
- ultramancool 12y ago
  DRM doesn't work nearly as well as password stretching. I can sit down for a few hours with IDA and crack nearly any DRM scheme. With password stretching, the result of how long attacks will take is predictable and they can actually be made difficult or require more hardware (and at larger expense) rather than simply requiring a more skilled attacker. For example, the Adobe DRM scheme has been cracked for ages, go Google around, there are tools to dump the keys and decrypt books using it. All the decisions are made by a few instructions executed on the user's CPU. With cryptography, this is not possible. I can publish as open source any good cryptographic algorithm and it will stand up to public scrutiny. I cannot do this with a DRM scheme as publishing its workings would inherently mean allowing it to be broken. Very, very different. DRM is evil. It is an attempt to restrict your machine to not listen you, the owner. Do not tolerate that.
- betacat 12y ago
  It's not a good argument. DRM is an executable contract. Contracts cannot enforce themselves. That is why we have police, courts, etc. The fact that it is an executable contract makes it more difficult to circumvent than a paper one, but circumvention is still circumvention.

by voyou 12y ago

dchest 12y ago

That's a good argument, thank you! I'd argue that for DRM to work reasonably well (without coercion), it doesn't need to be perfect, it can work in the same way password stretching works: you don't get the absolute protection, but influence the cost (value). For example, when applied to software, DRM can be stripped, but then you'll be receiving the executable from a possibly untrusted source. However you are spot on that currently DRM is mostly legal and political, but I don't think that's the inherent property of it. * * * Unfortunately, seeing my comments downvoted makes me uncomfortable to continue discussion here on HN, as (when applied to reasonable comments) it provides instant feedback that you're going against the popular opinion, and I'd rather avoid such feedback, so I'll go think about it more and then write a blog post or something.

ultramancool 12y ago
DRM doesn't work nearly as well as password stretching. I can sit down for a few hours with IDA and crack nearly any DRM scheme. With password stretching, the result of how long attacks will take is predictable and they can actually be made difficult or require more hardware (and at larger expense) rather than simply requiring a more skilled attacker.
ultramancool 12y ago
DRM doesn't work nearly as well as password stretching. I can sit down for a few hours with IDA and crack nearly any DRM scheme. With password stretching, the result of how long attacks will take is predictable and they can actually be made difficult or require more hardware (and at larger expense) rather than simply requiring a more skilled attacker. For example, the Adobe DRM scheme has been cracked for ages, go Google around, there are tools to dump the keys and decrypt books using it. All the decisions are made by a few instructions executed on the user's CPU. With cryptography, this is not possible. I can publish as open source any good cryptographic algorithm and it will stand up to public scrutiny. I cannot do this with a DRM scheme as publishing its workings would inherently mean allowing it to be broken. Very, very different. DRM is evil. It is an attempt to restrict your machine to not listen you, the owner. Do not tolerate that.
betacat 12y ago
It's not a good argument. DRM is an executable contract. Contracts cannot enforce themselves. That is why we have police, courts, etc. The fact that it is an executable contract makes it more difficult to circumvent than a paper one, but circumvention is still circumvention.