Re: ΄πΈ΄: new scenario suggestion: aquire media capture without user interaction

On 7/6/2012 3:25 AM, Sunyang (Eric) wrote:

>>> We should let getUserMedia has a way to tell the UA we do not need user
>>> permission to capture video, and the UA need to have some configuration
>>> switch about it.
>>
>> This would break the security model - it is essential that the app can
>> not start capturing without the user giving consent.
> 
> Even the app, browser, device and camera are all belong to company, and this scenario must recognize the user's face at any time?

That's not "generic WebRTC in a browser", that's a custom
application/browser.

Chrome doesn't have this as an option because it's a browser.  You could
add such a command-line switch (for example) to a source clone of a
browser.

We're also investigating how to give "trusted by the user" apps ability
to avoid the security requesters; this likely would leverage the
installed-app model.  (This would let you develop VoIP-like apps where
you have a dial and answer buttons in the app.)  If your app is a
trusted, installed app, then this model would likely work for you.

Of course, an embedded device like you describe could be far dumber.
And the smarts don't have to be in that device, they can be in a faster
PC that "calls" the cameras.


-- 
Randell Jesup
randell-ietf@jesup.org

Received on Friday, 6 July 2012 07:35:20 UTC