The interface between the CDM and the UA is not defined by EME. As a consequence, it is up to the UA vendor and the CDM supplier to define this interface as they see fit. Furthermore, you do not actually describe a use case, i.e., "compatible with the historical operation of the open web" is not a use case. Nor does this, if accepted as a use case, imply that "the CDM [must] return the decoded and decrypted streams to the user agent."

Please take this up with the Restricted Media CG if you wish, and, provided (1) the CG agrees, (2) creates a well defined use case, and (3) recommends it be brought to the attention of the HTML WG, then it may be brought back to the WG.

A well defined use case should take a form at least as complete as that defined in [1]. It should also suggest how it may be tested in a vendor neutral manner.

[1] http://en.wikipedia.org/wiki/Use_case#Martin_Fowler

(In reply to comment #0)
> Please add the following use case requirement for the EME specification: the
> EME API needs to be compatible with the historical operation of the open web
> and as a requirement must require the CDM to return the decoded and
> decrypted streams to the user agent.

As Glenn says, this is not a use-case. It is a proposal for new requirements.

Your proposal can be split into three parts:
(i) that the interface between CDM and UA should be defined, or at least constrained, within the EME specification
(ii) that this interface must support returning of decrypted and decoded media to the user agent
(iii) (maybe - I am not sure if you are proposing this) that all CDMs MUST return decrypted and decoded media to the user agent 

> 
> I would note that W3C has ruled that the EME is in scope and that work on
> the EME can proceed.  The W3C has not ruled on what the use cases and
> requirements are and the EME specification currently does not defined them
> in an object technical manner.  This bug adds use cases and requirements.

Yep. Your proposal (whichever it is) is in scope and should be considered by the WG.

> 
> Note that the W3C has also ruled that meta level discussion regarding the
> use cases and requirements is not within the scope of the working group and
> that such discussed should occur in the Restricted Media Community Group
> which is not charted to influence the technical process.  If there are any
> objections to meeting these uses case then please take it up in the
> Restricted Media Community Group.

In the WG, we have touched on all three of the above before and there has not been support for any of them. I doubt there will be support now, but let's let the bug run for a while and see.

In the event there is no support for these requirements, I believe you are claiming that this makes EME 'incompatible with the historical operation of the open web'. Debating that claim would clearly be a topic for the Restricted Media Community Group. Should that group conclude that the claim is indeed true there is then a question for the community of whether indeed we wish to break with that historical operation. I have comments on both of those topics, but I will leave those for discussion in the WG if indeed you wish to raise these things there.

Oops, I meant CG in that last comment.