ContainerTiming: reduce WPT waits in buffering cases
In the buffering test cases we were using step_timeout to wait first to set up
the paint before setting the observer, then to actual check if a paint was
observed. Instead of this, use the elementtiming wait as in other tests, and
only wait 1 second in the second buffering disabled case.
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
I think this okay, but has a merge conflict. Waiting for rebase in case the previous patches changed.
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Merge conflict seems to be gone now.
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Code-Review | +1 |
Added one concern but it doesn't need to block the patch. I think this is an improvement either way.
img.src = '/container-timing/resources/square100.png';
Nit: is this `src` path required to be the same as the one used in `addPaintingElementTimingAfterDoubleRAF` or is it just coinidence?
It seems to me that you are racing the network fetch of the container timing and element timing elements, and the race is (mostly?) resolved because they use the same resource.
Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |