On Fri 26/Apr/2013 02:53:58 +0200 Mark Nottingham wrote:
Personally, I don't have a firm position on these issues, but I couldn't let
this pass by.
On 25/04/2013, at 7:38 PM, Alessandro Vesely <vesely(_at_)tana(_dot_)it>
wrote:
DRMs are obviously designed to be non-interoperable, and EME is a
standard for managing such non-standard stuff. That is going to
break interoperability, as any given browser will inevitably miss
some decryption modules.
The same could be said about the Content-Type header in HTTP;
allowing for new, even non-standard formats in browsers is by
design, not counter to the Web (or Internet) architecture.
Correct, and Content-Type is in mail too. In that case,
standardization is encouraged for each type and subtype.
"All implementations moving in lockstep" is not the same as
"interoperability," and we have plenty of examples of such
extension points in our protocols here.
Yes.
Injecting DRM through EME is a disservice to web standardization,
since the latter is supposed to foster the Internet revolution.
What does that *mean*? I'm wary of waving around banners like "the
Internet revolution", since they can so easily be misused.
The Internet revolution is the big step after the industrial
revolution. I didn't mean to misuse that banner: It is a typical
effect of industrial revolution to bring many workers to some large
factories. Free-software development, for a counter example, doesn't
fit into that model. I'd agree that the phenomenon is still young and
that the economics of the new model definitely need to be improved.
However, that cannot be done while sticking to the old production
model. Which direction is the media industry heading to?
Watch-on-demand could have been implemented on private networks and
proprietary equipment. However, cables had not been laid until the
Internet took root, and that was pushed by the web. The direction of
history seems to be clear enough to allow taking a firm position.
--
If you haven't done so already, please sign the FSF petition:
http://www.defectivebydesign.org/no-drm-in-html5