- From: Jan-Ivar Bruaroey via GitHub <sysbot+gh@w3.org>
- Date: Tue, 27 Jun 2023 18:48:59 +0000
- To: public-webrtc-logs@w3.org
> Calling the method on the ICE controlling peer causes the ICE agent to begin selecting the candidate pair for transport using the procedure described in [section 7.2.5.3.4](https://www.rfc-editor.org/rfc/rfc8445.html#section-7.2.5.3.4) of RFC 8445. Does RFC 8445 allows this? Reading through that section there seems to be rules that govern when the nomination flag is supposed to be set, that are not entirely up to the user agent. With https://github.com/w3c/webrtc-extensions/issues/166 we only enabled apps to intervene in user agent decisions, whereas here it's less clear to me what the rules must be followed in order to not violate RFC 8445. Can you clarify? It seems to me that one way around it is to only allow app choices where the a user agent has discretion in the process. This might resolve any conflict between the user agent and the IETF spec. But there's perhaps a more basic question between the app and the user agent here is: who's in charge now? Here I confess a level of ignorance wrt ICE, but let's say hypothetically that a user agent is allowed to change the selected pair in response to changing network conditions (is it?) — would/should it act any differently based on whether this method as been called (once) or not? I.e. would any of these methods switch the agent from automatic to manual? If not, what is the scope of these application-initiated actions? -- GitHub Notification of comment by jan-ivar Please view or discuss this issue at https://github.com/w3c/webrtc-extensions/issues/171#issuecomment-1610045566 using your GitHub account -- Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config
Received on Tuesday, 27 June 2023 18:49:01 UTC