3 ms·

No, that would be unenforceable. If your implementation doesn't honor the DRM, they won't provide you with the encryption keys in the first place.

MichaelGG 11y ago
Then you end up with a key distro problem. And unless the idea is to do it with hardware-protected keys, it's pretty much DOA for such low-bandwidth, simple-format as a still picture. Otherwise, as an example, Firefox wouldn't be able to decode these images without some plugin. And that plugin would be just as subject to key-dumping.
- ekimekim 11y ago
  Yes, all DRM systems are inherently impossible to make fully secure with only local data. However, several other options exist: * Needing to contact an external server to view an image (of course, you could always take the opportunity to save the now-viewable image then share it) * Charging a large fee / legal contracts for the information on how to decode (security through obscurity) * Making circumvention illegal and enforcing this with fines or jail time. This is probably the only option that "actually works" as you've turned the technical problem into a social problem. People would still break it, but you could simply arrest them if they try to help others do the same.
  - MichaelGG 11y ago
    I'm not sure these approaches work for images. Certainly web comics could require special clients even now, yet they don't. I'm guessing they want the ability for their pictures to still work normally, but "without saving".
    - ekimekim 11y ago
      Yes, that none of these systems will ever actually work is the point ;)
- pdkl95 11y ago
  > subject to key-dumping Not if you are running a skylake CPU with the new SGX instructions that encrypt RAM. https://software.intel.com/en-us/blogs/2013/09/26/protecting-application-secrets-with-intel-sgx https://software.intel.com/en-us/blogs/2013/09/26/protecting... 6, Enable software vendors to deliver trusted applications and updates 8. Enable applications to define secure regions of code and data that maintain confidentiality even when an attacker has physical control of the platform and can conduct direct attacks on memory. By "trusted", Intel almost certainly means "trusted computing"/"trusted execution environment". It took a decade, but Palladium (NGSCB) is here. Firefox already caved and ships the EME plugin, so obviously they would cave to an image-DRM plugin as well. /* this is why it's so incre4dibly important to take a hard-line stance against DRM; it's a lot harder to reverse course once you cave and start accepting some DRM */
  - MichaelGG 11y ago
    SGX doesn't provide a way to secure output at the moment. It's awesome for remote attestation. Like, you could run a probably secure Bitcoin tumbler. (Well I'm not sure how the key verification part works.) But it does not provide a way to stop me from pressing PrintScreen. I agree that they may expand it (MS's waste of engineers during the Vista period in their rush to provide Protected Media Path comes to mind.) But I'm also saying that web comics could do this today. But I'm guessing that they don't want to sacrifice the UI of just loading images.

MichaelGG 11y ago

Then you end up with a key distro problem. And unless the idea is to do it with hardware-protected keys, it's pretty much DOA for such low-bandwidth, simple-format as a still picture. Otherwise, as an example, Firefox wouldn't be able to decode these images without some plugin. And that plugin would be just as subject to key-dumping.

ekimekim 11y ago
Yes, all DRM systems are inherently impossible to make fully secure with only local data. However, several other options exist: * Needing to contact an external server to view an image (of course, you could always take the opportunity to save the now-viewable image then share it) * Charging a large fee / legal contracts for the information on how to decode (security through obscurity) * Making circumvention illegal and enforcing this with fines or jail time. This is probably the only option that "actually works" as you've turned the technical problem into a social problem. People would still break it, but you could simply arrest them if they try to help others do the same.
- MichaelGG 11y ago
  I'm not sure these approaches work for images. Certainly web comics could require special clients even now, yet they don't. I'm guessing they want the ability for their pictures to still work normally, but "without saving".
  - ekimekim 11y ago
    Yes, that none of these systems will ever actually work is the point ;)
pdkl95 11y ago
> subject to key-dumping Not if you are running a skylake CPU with the new SGX instructions that encrypt RAM. https://software.intel.com/en-us/blogs/2013/09/26/protecting-application-secrets-with-intel-sgx https://software.intel.com/en-us/blogs/2013/09/26/protecting... 6, Enable software vendors to deliver trusted applications and updates 8. Enable applications to define secure regions of code and data that maintain confidentiality even when an attacker has physical control of the platform and can conduct direct attacks on memory. By "trusted", Intel almost certainly means "trusted computing"/"trusted execution environment". It took a decade, but Palladium (NGSCB) is here. Firefox already caved and ships the EME plugin, so obviously they would cave to an image-DRM plugin as well. /* this is why it's so incre4dibly important to take a hard-line stance against DRM; it's a lot harder to reverse course once you cave and start accepting some DRM */
- MichaelGG 11y ago
  SGX doesn't provide a way to secure output at the moment. It's awesome for remote attestation. Like, you could run a probably secure Bitcoin tumbler. (Well I'm not sure how the key verification part works.) But it does not provide a way to stop me from pressing PrintScreen. I agree that they may expand it (MS's waste of engineers during the Vista period in their rush to provide Protected Media Path comes to mind.) But I'm also saying that web comics could do this today. But I'm guessing that they don't want to sacrifice the UI of just loading images.