web-platform-tests
Regression in htmlallcollection.html in Edge despite no test changes
Compare these two: https://wpt.fyi/html/infrastructure/common-dom-interfaces/collections/htmlallcollection.html?sha=13eaad17a4 https://wpt.fyi/html/infrastructure/common-dom-interfaces/collections/htmlallcollection.html?sha=03d67ae5d9
In the second run, Edge is timeout. Reporting this because it seems like quite often some random thing is timeout in Edge, looking at the stability of results over time might reveal how much of a widespread problem this is.
This and https://github.com/w3c/wptdashboard/issues/389 together made the dashboard not as useful as it could have been for understanding https://github.com/whatwg/html/pull/3392, and I had to stitch together different results to get the full picture.
https://wpt.fyi/html/infrastructure/common-dom-interfaces/collections/htmlallcollection.html
It was pretty easy to find a timeout in the latest run as well, one that wasn't a timeout in the run before: https://wpt.fyi/generic-sensor/idlharness.https.html?sha=03d67ae5d9 https://wpt.fyi/generic-sensor/idlharness.https.html?sha=7a4c0d6617
So I think the issue is "random timeouts on Edge" rather than a specific test.
https://wpt.fyi/webrtc/historical.html is another case that I wanted to look at just now and noticed it's timeout. Also a regression from the previous run: https://wpt.fyi/webrtc/historical.html?sha=03d67ae5d9 https://wpt.fyi/webrtc/historical.html?sha=7a4c0d6617