Re: How to signal loss of consent freshness from Stefan Hï¿½kansson LK on 2016-02-19 (public-webrtc@w3.org from February 2016)

From: Stefan Hï¿½kansson LK <stefan.lk.hakansson@ericsson.com>
Date: Fri, 19 Feb 2016 06:51:24 +0000
To: "public-webrtc@w3.org" <public-webrtc@w3.org>
Message-ID: <1447FA0C20ED5147A1AA0EF02890A64B37499DA2@ESESSMB209.ericsson.se>

On 18/02/16 21:24, Nils Ohlmeier wrote:
> Hi,
>
> when implementing STUN consent freshness per RFC 7675 in Firefox I
> failed to locate any spec or draft which explains how to signal a
> loss of consent from the ICE stack up through the JS APIï¿½s.
>
> As the ICE Connection State is pretty much the only way to
> communicate back up Iï¿½m currently planing to switch the state to
> ï¿½disconnectedï¿½ when a single component fails the 30s timer from 7675
> (or the remote side revokes consent). The other option I see is to
> switch the ICE Connection State to ï¿½disconnectedï¿½ immediately when
> any of the consent freshness STUN transactions times out (and switch
> it back to ï¿½connectedï¿½ after receiving a late response or response to
> a subsequent try) and then switch to ï¿½failedï¿½ once the 30s timer
> pops.
>
> Should I open an issue against the webrtc-pc spec on that?

Please do.

Br,
Stefan

>
> Best regards Nils Ohlmeier
>

Received on Friday, 19 February 2016 06:51:57 UTC