Re: [rtcweb] WGLC of draft-ietf-rtcweb-use-cases-and-requirements-11

["Karl Stahl" <[e-mail]>]

Hi Xin,
> permission to access the device is orthogonal to consent to transmit the
media to whom....

Thanks for the draft pointer, but what about the common usage that I pass
you a link: "Click here so we can discuss further", You click and run _my_
application och you are prompted by your browser to permit Mic&Camera, and
get no further option at all to restrict to only voice?

Doesn't _the browser_ really have to show an "Only Mic"-permission also,
i.e. three buttons "Deny", "Video Call", "Voice Call"?

/Karl

-----Ursprungligt meddelande-----
Från: Chenxin (Xin) [mailto:[e-mail]] 
Skickat: den 26 september 2013 14:59
Till: Karl Stahl; [e-mail];
[e-mail]
Ämne: RE: [rtcweb] WGLC of draft-ietf-rtcweb-use-cases-and-requirements-11

Hi karl,

>-----Original Message-----
>From: Karl Stahl [mailto:[e-mail]]
>Sent: Wednesday, September 25, 2013 4:09 AM
>To: Chenxin (Xin); [e-mail];
>[e-mail]
>Subject: SV: [rtcweb] WGLC of 
>draft-ietf-rtcweb-use-cases-and-requirements-11
>
>
>
>-----Ursprungligt meddelande-----
>Från: Chenxin (Xin) [mailto:[e-mail]]
>Skickat: den 24 september 2013 13:56
>Till: Karl Stahl; [e-mail];
>[e-mail]
>Ämne: RE: [rtcweb] WGLC of 
>draft-ietf-rtcweb-use-cases-and-requirements-11
>
>Hi Karl,
>>>While reading the draft-ietf-rtcweb-use-cases-and-requirements-11,
>>>here are a few "telephony related" WebRTC things I think should be 
>>>clarified in the use cases.
>>>
>>>
>>>3.2.1.  Simple Video Communication Service 3.2.1.1.  Description ...
>>>The invited user might accept or reject the session.
>>>[Suggest adding] The invited user might accept only audio, rejecting 
>>>video (even if a camera is enabled). A user may also select to 
>>>initiate an audio session, without video.
>>>
>>>And in API requirements:
>>>   ----------------------------------------------------------------
>>>   A1      The Web API must provide means for the application to ask the
>>>browser for permission to use cameras and microphones, individually 
>>>as input devices. (One must be able to answer with voice only - 
>>>declining
>>video.)
>>>   ----------------------------------------------------------------
>>
>>>Same under
>>>6.2.  Browser Considerations
>>>...
>>>The browser is expected to provide mechanisms for users to revise and 
>>>even completely revoke consent to use device resources such as camera 
>>>and microphone. [Suggest adding] Specifically, a user must be given 
>>>the opportunity to only accept audio in a video call invitation.
>>>
>>[Xin] it is a common use case to accept only audio call and reject the 
>>video and quite useful. But I am doubt that this function should be 
>>mixed with video or audio device access permission . Do I 
>>misunderstand
>your proposal?
>>I think we could just disable the video stream when signaling. So we 
>>could make video call with one and reject it with other in the same 
>>web-service. I think the audio and video device access permission is 
>>not for each call(peer connection).
>>
>>Thanks,
>>   Xin
>>[Karl] Try using a WebRTC application with Chrome and you will see: 
>>The Permission/Allowance to use Camera and Microphone comes up at a 
>>bar at the top of the browser window and is the actual answering of a
call.
>
>[Xin] yes, it come up because invoking the getUserMedia API. When we 
>write the webrtc app, we could call getUserMedia API just once and use 
>the same mediaStream later. The webrtc app could decide to send the 
>video stream to the other side or not by configure the signaling(SDP or 
>other).  It is trivial to click the Permission bar at the top every 
>time when I get a call, even terrible when join a p2p conference.
>That is the reason I think the use case you mentioned should not mix 
>with permission, which should be a signaling configuration problem. Now 
>in Chrome,we could control it by using creatOffer or createAnswer and 
>setting the OfferToReceiveVideo constraint.
>
>Thanks,
>   Xin
>[Karl] If you permit the Browser to use the camera, then later "answer 
>with only audio in the application" (as I understand you suggest)
[Xin] this is the function of PeerConnection API in W3C. 

- Can we trust
>that the application isn't sending our video anyway - not even informing
us?
>
>I don't think we can trust any video WebRTC application, but will learn 
>to only trust well known browsers not to view us, when we don't want to 
>be seen.
>
>Isn't this a necessary security thing?
>
>/Karl
>
[Xin] I agree with you that the app is possible to steal user's video or
audio stream. 
But it is still hard to controI if the stream from getUserMedia could be
used by other call(peerconnection) as designed in API now.

What I want to say is that the permission to access the device is orthogonal
to consent to transmit the media to whom. The first one is for a site. The
second one is for a user, it could be guaranteed by other ways, such as
identity, media-encryption and so on, which need other consent method. 

 There are some details in
http://tools.ietf.org/html/draft-ietf-rtcweb-security-05#section-4.1.2.1