On 2014-05-22 02:44, Robert Kaiser wrote:
Jim schrieb:
You have still not proven your claim that the CDM will be robust

I think that can only be proven once the code exists, and it still to
be written. Once it's there, I'm sure everyone will be happy if you
inspect it for that robustness.

Enough details have been supplied to make it clear that the CDM has no chance of being confident in the state that the sandbox supplies to it when the sandbox passes control the the CDM.

The only protection offered to the CDM is that it can look over the memory of the process it runs in. However the sandbox can change the process memory after it is initialized and before the CDM is loaded and run, so the CDM has no guarantee of the provenance of the state passed to it from the sandbox when it starts. The idea that the CDM can check that the sandbox code matches it's expectation to guarantee provenance of the device identifier is flawed. If Mozilla can not even get this right then I have no confidence that the CDM would be robust.

What is Mozilla going to do when Adobe are unable to convince content owners and distributors that this is robust? Will Mozilla use the same propaganda to justify supporting platform CDMs? Nothing Mozilla has said suggests otherwise. Mozilla do not know where to draw the line.

What will Mozilla do if in future the content owners and distributors withdraw support for the CDM, after the EME API has been well established and after DRM APIs have been convincingly added to the open web? Mozilla will have caused great damage and will be in an even weaker position to fight.

All Mozilla will have done is nurture and promote the EME API, and in doing so will have set back alternatives and damaged the open web.

Given that the CDM has no control there will be Firefox derivatives offering control over device identifiers and able to save the content. Are Mozilla going to support content owners and distributors if they prosecute open web developers and distributors adding innovative features to the open web, such as EME content saving functions? The notion that Mozilla management would act as expert witnesses for the prosecution, claiming that the open web supports DRM restrictions, is just absurd given the mission. If this is what Mozilla management claim then they do not represent the open web, they do not get to claim they are part of the open web, and they do not get to claim they champion user security and privacy.

What are Mozilla going to do when some CDM innovations allows HTTP requests to be passed to the CDM and received and presented in the web browser? This will effectively add DRM to any web content. The EME solution requires JS to complete the player and this along with the DOM gives all the flexibility to implement this. Add a JS engine to the CDM and there goes even more of the web. Add HTML rendering support to the CDM and it's game over, and Mozilla are making it happen.

Jim

_______________________________________________
governance mailing list
governance@lists.mozilla.org
https://lists.mozilla.org/listinfo/governance

Reply via email to