3 ms·

You totally hit the nail on the head: DRM 'works' if it makes it difficult to get access to content, not impossible. And of course, I say that having spent the

by daeken 15y ago

You totally hit the nail on the head: DRM 'works' if it makes it difficult to get access to content, not impossible. And of course, I say that having spent the majority of my teenage years reversing DRM. Nobody believes that DRM will hold up forever, but there's a huuuuge imbalance here that favors the DRM creators, which is that it's really, really easy to put together modifications on a DRM system that make a reverser's life hell. For a while, iTunes was rolling out a new version of their store page crypto with every release (and deprecating the old one) so I'd have to go and reverse it to keep my iTMS client working; I may have kept up for a while, but eventually I stopped because it simply took too much damn time, and they won in the end with a fairly small amount of work from their side. That imbalance means that DRM will always work, even if it doesn't 'work'.

mikeash 15y ago
This actually seems to paint a nice picture where everybody wins. Content producers win because DRM only needs to work for a short time. Everybody else wins because it can still be cracked in relatively short order.
- shabble 15y ago
  If you merge these 2 aspects: DRM first, free(er) one later, you could get most of the money, and the PR bonus of being DRM free, which, as more people get annoyed by cumbersome DRM, could even give you a 2nd sales bump. The problem would be to avoid the situation where it's released, and nobody buys it for the first week/month/whatever, and then it's immediately distributable. The answer might lie in the 'ransom model', which is a bit like kickstarter, except "Once we've made $n million or sold 5M copies, we'll open it up" You then get that money almost guaranteed (avoiding the simple delayed purchase option), and if done carefully, you might even hit your targets before it's cracked, which if you can do that with any regularity, crackers may just stop bothering (Or they'll take it as even more of a challenge - hard to say)
tptacek 15y ago
While I'm pretty sure you already know this, I doubt many people on this thread do: The BD+ scheme that protects Blu-Ray disks was designed by very, very smart people (outside the content industry) to make it maximally easy to update the protection scheme, potentially on a title-by-title basis, without ever having to ship new players (and ideally never having to update player firmware).
- mikeash 15y ago
  I read the Wikipedia entry on BD+ to see what made it special. Sounds like it's essentially a way to implement DRM on the disc rather than in the player by having the player provide a virtual machine which executes arbitrary DRM code on the disc. However, it seems like this would be defeated permanently by simply implementing the virtual machine in the hypothetical DRM unlocker. Has it simply been too difficult to fully reverse engineer, or is there some other hurdle in the way beyond that?
  - tptacek 15y ago
    I can't comment in any detail here, but the problem of implementing the particular VM that BD+ programs are written to is not a trivial one, more akin to implementing a simulator for an entire X86 execution stack including the chipset, microarchitecture, and all the MSRs than it is to writing a JVM. Get anything wrong and you fail to derive keys.

by daeken 15y ago

mikeash 15y ago

This actually seems to paint a nice picture where everybody wins. Content producers win because DRM only needs to work for a short time. Everybody else wins because it can still be cracked in relatively short order.

shabble 15y ago
If you merge these 2 aspects: DRM first, free(er) one later, you could get most of the money, and the PR bonus of being DRM free, which, as more people get annoyed by cumbersome DRM, could even give you a 2nd sales bump. The problem would be to avoid the situation where it's released, and nobody buys it for the first week/month/whatever, and then it's immediately distributable. The answer might lie in the 'ransom model', which is a bit like kickstarter, except "Once we've made $n million or sold 5M copies, we'll open it up" You then get that money almost guaranteed (avoiding the simple delayed purchase option), and if done carefully, you might even hit your targets before it's cracked, which if you can do that with any regularity, crackers may just stop bothering (Or they'll take it as even more of a challenge - hard to say)

tptacek 15y ago

While I'm pretty sure you already know this, I doubt many people on this thread do: The BD+ scheme that protects Blu-Ray disks was designed by very, very smart people (outside the content industry) to make it maximally easy to update the protection scheme, potentially on a title-by-title basis, without ever having to ship new players (and ideally never having to update player firmware).

mikeash 15y ago
I read the Wikipedia entry on BD+ to see what made it special. Sounds like it's essentially a way to implement DRM on the disc rather than in the player by having the player provide a virtual machine which executes arbitrary DRM code on the disc. However, it seems like this would be defeated permanently by simply implementing the virtual machine in the hypothetical DRM unlocker. Has it simply been too difficult to fully reverse engineer, or is there some other hurdle in the way beyond that?
- tptacek 15y ago
  I can't comment in any detail here, but the problem of implementing the particular VM that BD+ programs are written to is not a trivial one, more akin to implementing a simulator for an entire X86 execution stack including the chipset, microarchitecture, and all the MSRs than it is to writing a JVM. Get anything wrong and you fail to derive keys.