4 ms·
We offered to open a discussion about a covenant, modelled on the W3C's own patent covenant, that would mean that W3C members would agree to only prosecute thos
by dannyobrien 9y ago
We offered to open a discussion about a covenant, modelled on the W3C's own patent covenant, that would mean that W3C members would agree to only prosecute those who had broken the DRM within a W3C standard, if they had another cause of action. That's to say, those companies could prosecute you if you broke the DRM and used it infringe copyright, but not if you only broke the DRM for lawful purposes -- such as security research, for accessibility, to protect privacy, and so on.
There was a sizeable component within the W3C (I am still unsure as to what degree I can reveal how much, given that the W3C holds members to a confidentiality agreement.) who felt this was an important step to take before making EME a standard. In the end, the W3C executive team overrode those objections, and decided to go ahead without any such covenant.
You can now for the first time be prosecuted for revealing a flaw in a W3C standard, overriding a browser "feature" on behalf of its user, or making a HTML video accessible without first asking permission of the DRM vendor who obfuscated it.
- zerocrates 9y agoThe covenant would have been a nice step forward and it's unfortunate to see that it didn't happen, but I think you overstate the consequences here. > You can now for the first time be prosecuted for revealing a flaw in a W3C standard, overriding a browser "feature" on behalf of its user, or making a HTML video accessible without first asking permission of the DRM vendor who obfuscated it. Presumably you're referring to the DMCA anti-circumvention provisions? Your second two examples are just as illegal now as they were yesterday: EME was already a browser "feature," already being used for HTML video. The only thing that's changed is the "W3C standard" part. The vendors presumably would have been fine to take EME, already a de facto standard and just leave it as-is, or "standardize" it in another body. Forcing that would have been nice for W3C's purity and image in some circles, and maybe even a good move, but it doesn't seem that it would have made any practical difference.