RE: [Bug 20944] New: EME should do more to encourage/ensure CDM-level interop

> From: watsonm@netflix.com
> To: fredandw@live.com
> 
>> Yes, but it not clear to me how this could be possible when the CDM is
>> part of 'independently implemented' web browser, or when the CDM
>> runs in a context hosted by an independently implemented operating
>> system?
>
> I think you are assuming that 'independently implemented' implies
> 'does not contain non-user modifiable components'.

'Independently implementable' means that all components can be
independently implemented unless qualified.

If a system depends on components that can not be modified
then the system can not be fully independently re-implemented.

> As I have said before, we are faced with two groups of people.
> One group does not allow the software they author to be 'tivoized'
> (for want of a better term). We can also include those who choose
> not to use 'tivoized' software in this group. Another group
 does not
> allow the media content they author to be viewed on devices that do
> not include 'tivoized' components. And never the twain shall meet. 

This does not appear to be relevant, and can be ignored when we
discuss a competent person writing their own implementation.

A conflict over copyrights only occurs when the independently
implemented CDM is distributed, but the issues I raised occur
before this.



> This is fine. I fully respect and understand the choices these
> people have made. We should not take sides and indeed I cannot
> see how doing so would be compatible with the basic moral rights
> of authors with respect to their own works.

We do need to establish the scope of the ability to fully and
independently implement a working decoder as this affects
the market for which they can reach.  Decoders that can not
be independently implemented will exclude all software
implementations on user modifiable software stacks, or if
they are used on such stacks then they are trivially ineffective
and do not meet the use case and have 'dumb copyright'
impacts.

> So, lets assume that at least the stronger CDMs include non-user-modifiable
> components. This is not incompatible with independent implementation or
> with the source of those components being licensed under an open
> source license, though obviously not GPLv3.

So can we reach consensus that:

"Stronger CDMs are expected to depend on non-user-modifiable
components."

"Non-user-modifiable components are only expected to be
supportable in non-user-modifiable stacks."

"Thus stronger CDMs will only be supportable on non-user-modifiable
stacks."

Thus we can forget about users of most Linux based distributions
or Firefox OS etc being able to view content that requires stronger
CDMs, unless they include a proprietary hardware CDM module?

cheers
Fred