- From: henbos via GitHub <sysbot+gh@w3.org>
- Date: Thu, 08 Sep 2022 14:45:16 +0000
- To: public-webrtc-logs@w3.org
> Should we point out this may leak information about other concurrent user activities? E.g. If a certain hardware codec is suddenly no longer available, it could mean the user is watching video on another site. Sure, I added a fingerprint notice to [the PR](https://github.com/w3c/webrtc-stats/pull/670/files). > I find this argument compelling, and I think stronger language on what information UAs MUST and MUST NOT include in encoderImplementation/decoderImplementation would go along well with this change. Can we do that? If we take Chrome as an example, there are some strings like "libvpx" which we know are SW and other strings like "D3D11VideoEncoder" which we know are HW, but there are other strings like "libvpx, fallback from D3D11VideoEncoder" (or something similar) which is very useful for debugging and perhaps other strings where the HW/SW part is more ambiguous. One way it has been used is "if impl contains a known SW string then it is SW, otherwise it is HW" which would work most of the time. The string is very implementation-specific and I'm not sure what we should or should not say here, I'd file a separate issue if you want to discuss that further. -- GitHub Notification of comment by henbos Please view or discuss this issue at https://github.com/w3c/webrtc-stats/issues/666#issuecomment-1240818651 using your GitHub account -- Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config
Received on Thursday, 8 September 2022 14:45:18 UTC