Re: [webrtc-pc] Clarify addIceCandidate behavior when adding candidate after end of candidate

Reflecting back on this, I think it is not a good idea to restrict call to addIceCandidate after end of candidate is added. There may be multiple ICE generations as well as ICE restart, and quite frequently clients can get disconnected and change network. 

The ICE agent should be able to handle all different scenarios, so applications should be able to forward any ICE candidate event to the remote peer regardless of the end of candidate status. It is sufficient to say that adding past end of ICE candidate _may_ cause the ICE agent to fail and return `OperationError`, and let the ICE transport handle such state management.

Closing this issue then.

-- 
GitHub Notification of comment by soareschen
Please view or discuss this issue at https://github.com/w3c/webrtc-pc/issues/1216#issuecomment-324243336 using your GitHub account

Received on Wednesday, 23 August 2017 07:18:22 UTC