Add support for handling BiDi connections
Categories
(Remote Protocol :: WebDriver BiDi, task, P2)
Tracking
(firefox91 fixed)
Tracking | Status | |
---|---|---|
firefox91 | --- | fixed |
People
(Reporter: whimboo, Assigned: whimboo)
References
Details
(Whiteboard: [bidi-m1-mvp])
Attachments
(3 files)
With the work on bug 1693802 we should have a connection module that should allow us to implement accepting an incoming connection, and closing it.
Note that it won't have to handle the session yet. Another bug will be filed for that.
Assignee | ||
Updated•4 years ago
|
Assignee | ||
Updated•4 years ago
|
Assignee | ||
Updated•3 years ago
|
Assignee | ||
Comment 1•3 years ago
|
||
While the constructor should only do the initialization
of the class, anything else related to the HTTP server
should be delayed until the protocol gets actually
started.
Further the protocol constructor needs the reference
to the Remote Agent itself instead of only the
HTTP server.
Assignee | ||
Comment 2•3 years ago
|
||
Depends on D119264
Assignee | ||
Comment 3•3 years ago
|
||
Depends on D119265
Comment 5•3 years ago
•
|
||
Backed out for causing multiple failures.
Backout link: https://hg.mozilla.org/integration/autoland/rev/744f95915c784e236b10c18d1ae71a90ded9a76a
Assignee | ||
Comment 6•3 years ago
|
||
This failure came in due to the rebase that I did before pushing to autoland. With that we set the recommended preferences in the constructor of the CDP module, which gets immediately loaded now. As such this needs to be delayed until CDP actually gets started.
The underlying problem needs to be fixed, and I filed bug 1719667 for that work.
Comment 8•3 years ago
|
||
Backed out 3 changesets (Bug 1693805) for causing bc failures in browser_all_files_referenced.js
Backout link: https://hg.mozilla.org/integration/autoland/rev/6fa18226d738435952f76e2eeb7ae69ebdca1e2f
Push with failures, failure log.
Assignee | ||
Comment 9•3 years ago
|
||
That's sad, but WebDriverBiDiConnection.jsm
isn't used at this point. It means this browser chrome test always fail, and I have to wait with the push until bug 1691446 has a reviewed patch.
Comment 10•3 years ago
|
||
Assignee | ||
Comment 11•3 years ago
|
||
Julian had the idea to just remove the inclusion of the module from the jar.mn
file for now. That actually fixes the test. I'll try to land again, and hopefully it works this time.
Comment 12•3 years ago
|
||
Comment 13•3 years ago
|
||
bugherder |
https://hg.mozilla.org/mozilla-central/rev/5d23409efae3
https://hg.mozilla.org/mozilla-central/rev/12bc4a9548c2
https://hg.mozilla.org/mozilla-central/rev/7102bff7059a
Description
•