- From: Jan-Ivar Bruaroey via GitHub <sysbot+gh@w3.org>
- Date: Tue, 25 Feb 2025 22:23:57 +0000
- To: public-webrtc-logs@w3.org
Lack of tests is a red herring. Thanks for contributing them, but tests have nothing to do with the risk of breakage we're discussing. WPT only tests browser conformance to behaviors we _specify_. A decision here would inform tests, not the other way around. What I said was that track order dependency has been _"effectively tested in the real world"_ by Safari and Firefox behaving differently for years (in the case of Safari having no reliable order from one track to the next). Therefore, we know _interoperable_ websites—by definition—do not rely on track order. Kudos to Safari for proving this point and taking one for the team here! > Your "works in Firefox therefore no risk" theory was very recently proven when libWebRTC changed how it assigned rtx payload types (not intentionally). It turned out that folks relied on "rtx pt == pt + 1" even though that is non-spec and Firefox does not follow that pattern. That was far from "no cost". In case it's unclear to readers, by _"folks"_ we're down to maintainers of non-interoperable chromium-only websites. I would venture these websites represent a much larger "cost" to us than to you. -- GitHub Notification of comment by jan-ivar Please view or discuss this issue at https://github.com/w3c/mediacapture-main/issues/1028#issuecomment-2683430568 using your GitHub account -- Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config
Received on Tuesday, 25 February 2025 22:23:58 UTC