- From: Harald Alvestrand <harald@alvestrand.no>
- Date: Mon, 24 Sep 2012 16:01:06 +0200
- To: public-webrtc@w3.org
- Message-ID: <506067A2.3070302@alvestrand.no>
On 09/24/2012 03:49 AM, Wangyahui wrote: > > Hi, all > > I am new to WebRTC, and begin with initial study of WebRTC 1.0 > specificationhttp://dev.w3.org/2011/webrtc/editor/webrtc.html. > > There are some problems I encountered while reading this spec. I would > appreciate it if you could reply. > > 1£®¡°discard¡± means that codecs can abandon some channels? Or maybe it > should be ¡°decode¡± by mistake? > > */Location/*: ¡°4.1 Introduction¡±, the third paragraph and last > sentence. ¡°/All of the channels that a codec needs to encode jointly > MUST be in the same MediaStreamTrack and the codecs SHOULD be able to > encode, /*/or discard/*/, all the channels in the track./¡± > Sometimes it is permissible to lose information (can't think of a good example; possibly when you're sending audio + video to an audio-only participant?) - I read this as saying it is not permissible to crash because of non-handlable track types. in the stream. > > 2£®I saw (1) and (3) in this specification, but I didn¡¯t find any link > or reference. Besides, there is no (2), but many (3) > following¡±RTCPeerConnectionreadiness state > <http://dev.w3.org/2011/webrtc/editor/webrtc.html> is closed¡±. > This is a leftover. All states should now be enums, so "closed(3)" needs to be changed to just "closed". Thanks for pointing it out! > > */Location1/*: ¡°4.2.2 MediaStreamTrack¡±, the third paragraph. ¡°/A > track in a MediaStream , received with a RTCPeerConnection , MUST have > its readyState attribute [GETUSERMEDIA] set to muted //(1)//until > media data arrives./¡± > > */Location2/*: ¡°5. Peer-to-peer connections¡± under¡±NOTE¡± step 4. ¡°/If > the connection¡¯s RTCPeerConnection readiness state is closed //(3)//, > abort these steps/¡± > > 3. Step 12 of ¡°Simple Call Flow¡± isaddStream(data), but as I know the > parameterof addStream(MediaStream/stream/)is for audio/video not data. > I think this should be "createDataStream()" instead. This API hasn't been stable for all that long. > > By the way, is it possible to add some descriptions for steps of the > flow charts? > > */Location/*: ¡°9. Call Flow Browser to Browser¡±, ¡°10. Call Flow > Browser to MCU¡± > > In addition, some revisions for editorial errors are attached for your > information. Thank you for checking it. > > And thank you very much for your review!
Received on Monday, 24 September 2012 14:01:38 UTC