browsingContext.navigate[interactive] should not stop the progress listener if navigation has not started
Categories
(Remote Protocol :: WebDriver BiDi, defect, P1)
Tracking
(firefox109 fixed)
Tracking | Status | |
---|---|---|
firefox109 | --- | fixed |
People
(Reporter: jdescottes, Assigned: jdescottes)
References
(Blocks 1 open bug)
Details
(Whiteboard: [webdriver:m5])
Attachments
(1 file, 1 obsolete file)
See detailed analysis at https://bugzilla.mozilla.org/show_bug.cgi?id=1785544#c18
The call to stop
at https://searchfox.org/mozilla-central/rev/92d3050ecbbdd077ffa4cb78e0dffc4bd5021109/remote/webdriver-bidi/modules/root/browsingContext.sys.mjs#399-401 will stop the progress listener regardless of which document becomes interactive.
Normally the progress listener for navigate should only be stopped for an explicit new navigation (expectNavigation=true, waitForExplicitStart=true), but external callers can still call stop() and bypass all the state logic from the progress listener.
We should expose a safer stop
, that will still check that the progress listener state is correct, and maybe make the current stop
private. Or maybe stop
should also check the progress listener state, which feels a bit redundant since right now it's done case by case before calling stop
within the ProgressListener
Assignee | ||
Comment 1•2 months ago
|
||
Updated•2 months ago
|
Assignee | ||
Updated•2 months ago
|
Updated•2 months ago
|
Pushed by smolnar@mozilla.com: https://hg.mozilla.org/mozilla-central/rev/4c736eef7473 [bidi] browsingContext.navigate[interactive] should not stop the progress listener if navigation has not started r=webdriver-reviewers,Sasha,whimboo
Comment 3•2 months ago
|
||
bugherder |
Assignee | ||
Comment 4•2 months ago
|
||
Depends on D163993
Comment 5•2 months ago
|
||
Comment on attachment 9306961 [details]
Bug 1801461 - [wdspec] Add wdspec tests for browsingContext.navigationStarted
Revision D163994 was moved to bug 1756595. Setting attachment 9306961 [details] to obsolete.
Description
•