Re: Encrypted Media proposal: Summary of the discussion so far

On Tue, Mar 13, 2012 at 3:07 PM, Glenn Adams <glenn@skynav.com> wrote:
> On Tue, Mar 13, 2012 at 3:35 PM, Tab Atkins Jr. <jackalmage@gmail.com>
> wrote:
>> On Tue, Mar 13, 2012 at 2:25 PM, Glenn Adams <glenn@skynav.com> wrote:
>> > Wrong. Other than Clearkey, individual CDMs are not part of what is
>> > specified in the proposal.
>>
>> And, *once again*, the fact that the CDMs that will *actually be used*
>> aren't specified in the spec means the spec is incomplete.  If this
>> spec moves forward, I will formally object to this lack as well,
>> because it is not possible to implement the spec in practice without
>> the CDMs that will actually be used.
>
> You can FO to whatever you like, but you are wrong in claiming that "the
> spec is incomplete" merely because some CDM instance that is expected to be
> used is not specified by the proposal or by W3C.
>
> That is equivalent to FOing to canvas.getContext("x-my-context") because
> x-my-context is not defined by the W3C and is expected to be used. A rather
> absurd proposition, but that is exactly what you are doing.

If we were newly speccing canvas and expected an "x-my-content"
context to be used, the spec *would* be incomplete without it.  I'm
not sure why you think this is absurd.

Such a context wouldn't necessarily have to be described in the *same
spec*, of course.  It could be in some other spec, even some other
standards body, as long as it had good qualities (namely, being
royalty-free and implementable in open-source without licenses).  This
is precisely the situation with the "webgl" context.

So, I don't know how you think that this analogy with <canvas> is
helping you at all.  The CDM situation is clearly materially different
in important ways.

~TJ