Summary of "What is missing for building real services" thread

Thanks to everyone who posted about what is missing in WebRTC. I attempted
to collate the results below, sorted into either "spec" or "implementation"
categories.

Basically, I think the key things that are causing trouble are being
actively worked on both in this WG and in implementations; we are on track
to resolve these problems, hopefully in the next few months.

Full list below:

Spec (in progress)

   -

   Bad error notifications
   -

   Lower image resolution without stopping the stream (RTCRtpSender or
   MST.applyConstraints)
   -

   API for capping bandwidth (RTCRtpSender)
   -

   Recording of streams (MediaStreamRecorder)


   -

   More debugging of candidate pair states (getStats)
   -

   Determine type of candidate (getStats)
   -

   List all the DCs on a PC (TBD if we need this or not)


Spec (v2)

   -

   Too attached for SDP, O/A
   -

   TURN auth failure does not cause an error
   -

   Better control of video mute behavior
   -

   Screen sharing without extensions (maybe)


Spec (future)

   -

   Access PeerConnection from Web Workers
   -

   Keep PeerConnection across reload/navigation


Implementations

   -

   Stable multi-stream support (working on this, some spec dependencies)


   -

   NAT/FW traversal, connection stability issues (Chrome working on this in
   Q1)
   -

   AEC performance issues (Chrome working on this in Q1)
   -

   BWE and handling of low-bandwidth situations (Chrome working on this in
   Q1)
   -

   Not all ICE states implemented/ICE never goes to failed (Chrome working
   on this in Q1)
   (Chrome: https://code.google.com/p/webrtc/issues/detail?id=1414)


Nontechnical

   - WebRTC support in other browsers (IE, Safari)