3 ms·

It does not necessarily stop the web being an open standard in that sense. An open standard is really just a way of saying "if you want to do X in a way that is

by jiggy2011 14y ago

It does not necessarily stop the web being an open standard in that sense. An open standard is really just a way of saying "if you want to do X in a way that is compatible with Y then your implementation must follow these rules and these rules are published in such a way that any person will be equally able to produce an implementation". You can propose an open standard for DRM that everyone is free to implement or not. If you want to have a DRM free browsing experience you can simply use a browser that implements every part of the standard apart from the DRM part. The only thing that changes is that stuff that is built to require the DRM component will not work. It would be just like disabling <canvas> support in your browser or something like that.

michaelt 14y ago
You can propose an open standard for DRM that everyone is free to implement or not. At the core of DRM is this: The video is encrypted, and you only give the decryption key to trusted software that promises not to let the user save anything. Unfortunately open source software cannot make that promise - because anyone can modify it, anyone can modify it to let the user save things. So you either let open source implementations have the decryption key, in which case the DRM is ineffective; or you refuse to let open source implementations have the decryption key, in which case there's no video on Firefox or Chromium. Perhaps you ask "Why can't the open source software check itself for modifications any only decrypt the video if no modifications were present?" - unfortunately, the check could be modified just as easily as the rest of the software. Perhaps you ask "Surely we could have a little bit of closed source software as a plugin, like flash player?" - but the whole point of HTML5 video is to get away from shitty closed-source plugins like flash. If you want flash, stick with flash. Perhaps you ask "What's the big deal with open source anyway? Firefox only has 20% market share" - but Chrome, Safari and (recently) Opera are all based on the open source Webkit rendering engine. That leaves IE and nothing else. Before open source browsers became popular, Internet Explorer 6 had 90%+ market share and went 5 years without a substantive upgrade due to lack of competition. A return to the bad old days would not be a net win, even for the likes of Netflix, Google and the BBC. The fact of the matter, as I understand it, is DRM barely works when it's closed source and if it was open source it wouldn't work at all. If you think there's some way to implement open source DRM I'd love to hear how.
- jiggy2011 14y ago
  Not sure of the specifics of implementation proposed but I imagine that the actual DRM decryption would have to happen somewhere in the guts of the kernel and the HTML part would mainly be metadata. DRM only implemented at browser level would be laughably bad as screen capture software would defeat it. Fundamentally though this isn't really any different from current Silverlight based DRM. Things like netflix don't work on Linux (without hacks) as it stands. Other non-drm video (which constitutes the overwhelming majority of web video) would continue to work as it does today. Besides even closed source DRM enforcement mechanisms can be defeated by someone with strong asm skills and enough time.
- acdha 14y ago
  > The fact of the matter, as I understand it, is DRM barely works when it's closed source and if it was open source it wouldn't work at all. If you think there's some way to implement open source DRM I'd love to hear how. Fundamentally, the problem has nothing to do with open or closed software: it's all about access to the decryption key. Open source doesn't really change this dynamic much as hackers have routinely retrieved keys from anything which they can access. You could implement something like HDCP where the content is encrypted all the way to the display device, at which point it's easy to have a completely OSS stack since it's just copying bytes to a specific location. I'd like to think that at some point people will realize that making digital data uncopyable is, as Bruce Schneier famously quipped, like trying to make water unwet. In the meantime, I'd prefer that the software millions of people depend on be standardized and developed by conscientious security-minded professionals – i.e. not Flash.

michaelt 14y ago

You can propose an open standard for DRM that everyone is free to implement or not. At the core of DRM is this: The video is encrypted, and you only give the decryption key to trusted software that promises not to let the user save anything. Unfortunately open source software cannot make that promise - because anyone can modify it, anyone can modify it to let the user save things. So you either let open source implementations have the decryption key, in which case the DRM is ineffective; or you refuse to let open source implementations have the decryption key, in which case there's no video on Firefox or Chromium. Perhaps you ask "Why can't the open source software check itself for modifications any only decrypt the video if no modifications were present?" - unfortunately, the check could be modified just as easily as the rest of the software. Perhaps you ask "Surely we could have a little bit of closed source software as a plugin, like flash player?" - but the whole point of HTML5 video is to get away from shitty closed-source plugins like flash. If you want flash, stick with flash. Perhaps you ask "What's the big deal with open source anyway? Firefox only has 20% market share" - but Chrome, Safari and (recently) Opera are all based on the open source Webkit rendering engine. That leaves IE and nothing else. Before open source browsers became popular, Internet Explorer 6 had 90%+ market share and went 5 years without a substantive upgrade due to lack of competition. A return to the bad old days would not be a net win, even for the likes of Netflix, Google and the BBC. The fact of the matter, as I understand it, is DRM barely works when it's closed source and if it was open source it wouldn't work at all. If you think there's some way to implement open source DRM I'd love to hear how.

jiggy2011 14y ago
Not sure of the specifics of implementation proposed but I imagine that the actual DRM decryption would have to happen somewhere in the guts of the kernel and the HTML part would mainly be metadata. DRM only implemented at browser level would be laughably bad as screen capture software would defeat it. Fundamentally though this isn't really any different from current Silverlight based DRM. Things like netflix don't work on Linux (without hacks) as it stands. Other non-drm video (which constitutes the overwhelming majority of web video) would continue to work as it does today. Besides even closed source DRM enforcement mechanisms can be defeated by someone with strong asm skills and enough time.
acdha 14y ago
> The fact of the matter, as I understand it, is DRM barely works when it's closed source and if it was open source it wouldn't work at all. If you think there's some way to implement open source DRM I'd love to hear how. Fundamentally, the problem has nothing to do with open or closed software: it's all about access to the decryption key. Open source doesn't really change this dynamic much as hackers have routinely retrieved keys from anything which they can access. You could implement something like HDCP where the content is encrypted all the way to the display device, at which point it's easy to have a completely OSS stack since it's just copying bytes to a specific location. I'd like to think that at some point people will realize that making digital data uncopyable is, as Bruce Schneier famously quipped, like trying to make water unwet. In the meantime, I'd prefer that the software millions of people depend on be standardized and developed by conscientious security-minded professionals – i.e. not Flash.