| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
Let's prioritize either always including transition_scoped.css or moving all of these rules to html.css, considering that we shouldn't be able to disable view transition (non-scoped) anymore
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Exportable changes to web-platform-tests were detected in this CL and a pull request in the upstream repo has been made: https://github.com/web-platform-tests/wpt/pull/58299.
When this CL lands, the bot will automatically merge the PR on GitHub if the required GitHub checks pass; otherwise, ecosystem-infra@ team will triage the failures and may contact you.
WPT Export docs:
https://chromium.googlesource.com/chromium/src/+/main/docs/testing/web_platform_tests.md#Automatic-export-process
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
<meta name="fuzzy" content="maxDifference=0-2; totalPixels=0-48600">that's a lot of pixels... is this an animation that started running already and began a crossfade? if so, this is going to be flaky.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
<meta name="fuzzy" content="maxDifference=0-2; totalPixels=0-48600">that's a lot of pixels... is this an animation that started running already and began a crossfade? if so, this is going to be flaky.
Cross-fade has finished, however, pixels in final image could be from compositor or main thread. On other tests, we've noticed rounding errors. a maxDifference of 0-1 is likely fine, but wasn't sufficient on some wide-gamut tests. Though the number of pixels is high, the colors have to be very close. The pixel count and difference would be sell over the threshold if not properly clipped. Erring on the side of caution, against test flakes.
<meta name="fuzzy" content="maxDifference=0-2; totalPixels=0-48600">that's a lot of pixels... is this an animation that started running already and began a crossfade? if so, this is going to be flaky.
It's finished the crossfade. Don't require fuzzing when the test is run locally, but as we have seen rounding errors in other tests, decided to experiment here to determine if it resolved the problem.
It turned out that the actual difference was significantly higher due to sub-pixel
antialiasing, and rounding of the clip rect position. As the animation is paused at the finish line, I'm hopeful we have precise pixel values in the interior of each box, and that adding a margin to push painted pixels away from the border resolves the anti-aliasing issue.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Still wrestling with anti-aliasing artifacts.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |