You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
#14531 adds the test /storage/estimate-usage-details-application-cache.https.tentative.html which was failing the Taskcluster stability check using ./wpt run --verify):
Subtest
Results
Messages
OK
estimate() shows usage increase after app is cached
FAIL: 5/10, PASS: 5/10
assert_greater_than: expected a number greater than 3364 but got 3364
The test isn't flaky in the sense of randomly failing, but it involves browser-global state, and ./wpt run --verify appears to run the test in such a way that the global state is shared across test invocations.
This could possibly be fixed in a manner similar to #15018.
The text was updated successfully, but these errors were encountered:
@jgraham have you seen cases like this with Firefox as well? Will tests that have this problem ever run into trouble in Mozilla's CI setup? What I'm wondering is whether it's too high a bar to require tests to involve no global state. I don't know the details of the ./wpt run --verify scheduling, but I guess the tests are run in sequence at least, never in parallel? (Parallel would make it very hard to write fullscreen tests.)
#14531 adds the test /storage/estimate-usage-details-application-cache.https.tentative.html which was failing the Taskcluster stability check using
./wpt run --verify
):estimate() shows usage increase after app is cached
assert_greater_than: expected a number greater than 3364 but got 3364
I confirmed this locally in #14531 (comment).
The test isn't flaky in the sense of randomly failing, but it involves browser-global state, and
./wpt run --verify
appears to run the test in such a way that the global state is shared across test invocations.This could possibly be fixed in a manner similar to #15018.
The text was updated successfully, but these errors were encountered: