- From: henbos via GitHub <noreply@w3.org>
- Date: Wed, 22 Oct 2025 08:29:41 +0000
- To: public-webrtc-logs@w3.org
Overall +1 on that explainer, just a few comments. > End users of game streaming services may experience increased latency, degraded quality, and battery drain when the browser switches from hardware to software decoding. [...] This gap makes it difficult to diagnose performance regressions and provide troubleshooting guidance. All of this is good, I'd just nit pick that we should go further and make it clear that this isn't just for debugging and troubleshooting - it is, of course, that too - but in addition to that, having this event is a must-have for the app to be able to recover from this issue in a timely manner, because only the app - not the user agent - has the power to ask the remote endpoint to send us a different codec. If the app can't do that, the user is best-case stuck in a suboptimal experience, and worse-case stuck without any SW decoder to fall back to (e.g. H.265), and the app would end up implementing getStats-based heuristics that could end up hurting web compat, e.g. looking at `decoderImplementation` strings being the prime example shipped in multiple popular apps. > This is not ideal because the UI prompt is invasive, it’s excessive since it grants access to the camera and mic hardware when apps don’t need it, and it has a high failure rate since users have little reason to grant the permission unless they want to use voice chat Not only is it not ideal, it's absolutely absurd to prompt the user for camera and microphone access in order to allow recovering from decoding issues. Apps doing this when they don't need to is an even bigger privacy implication than having this event and people not asking for it. > RESOLUTION: discuss proposal in more depth and prepare for Privacy review What is the process here @youennf @jan-ivar @guidou. Isn't the privacy implications identical to WebCodecs and can we piggyback on any of that or what is the next step? It seems to me that the finger printing surface isn't changing compared to that -- GitHub Notification of comment by henbos Please view or discuss this issue at https://github.com/w3c/webrtc-extensions/issues/146#issuecomment-3431073020 using your GitHub account -- Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config
Received on Wednesday, 22 October 2025 08:29:42 UTC