- From: Miguel París Díaz <mparisdiaz@gmail.com>
- Date: Thu, 20 Oct 2016 11:57:39 +0200
- To: Stefan Håkansson LK <stefan.lk.hakansson@ericsson.com>
- Cc: Bernard Aboba <Bernard.Aboba@microsoft.com>, "public-webrtc@w3.org" <public-webrtc@w3.org>
- Message-ID: <CAEn+E3jzTwX=qjUwL6HcbqVO-95zv4_PTm2pWW+KirocuKwXkg@mail.gmail.com>
Hello, if I am not wrong, focusing on the NOT negotiation needed case, we are discussing about stop/restart sending media from the RtpSender, typically for the switching on/off feature. So, from my point of view: - From the media perspective, the video should freeze as if the RtpReceiver does not receive media due to another situation (network problems, etc.) - Is application dependent knowing that the RtpSender has been stopped, signal this properly and manage the UI to set a black frame, hide the video, put a photo of the user or whatever it wants. Best!! 2016-09-29 14:46 GMT+02:00 Stefan Håkansson LK < stefan.lk.hakansson@ericsson.com>: > On 28/09/16 18:32, Bernard Aboba wrote: > > Stefan said: > > > > "Is that the way we should go? And in case of simulticast, should > > the displayed video freeze if encoding[0] is inactive, if any of the > > encodings are inactive, or only if all of them are (assuming the > > receiving browser is able to receive simulcast)? > > > > Or should we add a RtpSender global 'active' attribute? Or something > > else?" > > > > [BA] I would expect a "video freeze" to be displayed if a receiver > > stops receiving all video packets, however that happens. > > > > If the SFU is only sending a single video stream to the receiver, > > then freeze would occur if that stream ceases to be received (or > > there is very high loss that RTX/FEC cannot repair). > > > > If the sender stops sending an encoding to a receiver, that would not > > cause a freeze if there were other encodings being sent to that > > receiver , or the SFU decided to stop forwarding all encodings to > > that receiver. > > > > If the receiver supported simulcast reception, it should not freeze > > video if it was receiving at least one encoding. > > To me the above makes sense, the question is what we need to spec, and > where. > > > > > > -- Miguel París Díaz ------------------------------------------------------------------------ Computer/Software engineer. Researcher and architect in http://www.kurento.org http://twitter.com/mparisdiaz ------------------------------------------------------------------------
Received on Thursday, 20 October 2016 09:58:17 UTC