Re: Open Source implementations Re: Encrypted Media proposal (was RE: ISSUE-179: av_param - Chairs Solicit Alternate Proposals or Counter-Proposals)

On 2/28/2012 11:27 AM, Glenn Adams wrote:
>
> On Tue, Feb 28, 2012 at 12:10 PM, Charles Pritchard <chuck@jumis.com 
> <mailto:chuck@jumis.com>> wrote:
>
>     The issue has been with JS implementations, not hardware.
>
>
> I beg to differ. In the case of TVs (as oposed to STBs or HPs), the 
> amount of RAM and CPU capacities available for JS implementations are 
> significantly less (sometimes orders of magnitude). The traditional 
> focus on absolute minimization of BoM (bill of materials) cost 
> continues to produce a performance barrier for most TVs. Sure, more 
> capacity could be added until it becomes merely a JS/VM implementation 
> issue, but that has not happened in general in this device class.

Give me some numbers, and I'll do what I can to benchmark.

If you're strictly talking about hardware based implementations, then 
yes, it ought to be something like a JS crypto API, as Tab brought up.

If it's a "smart TV" with a JS engine on it, then it ought to have 
enough power to grab an ArrayBuffer via XHR and push it into the media 
element, otherwise, it just doesn't have enough juice to run what the 
web considers HTML5.

If this is about making the Nintendo Wii+opera the baseline of 
performance, we can do that. I need something tangible to work with.
I realize that my mobile phone is more powerful than the devices of 
yesteryear, and perhaps, some of the devices coming out next year.


-Charles