Write mochitests for Fenix hit testing regressions (bug 1637113, bug 1638655, bug 1638458, bug 1638441, bug 1637135, bug 1643212, bug 1638594)
Categories
(Core :: Panning and Zooming, task)
Tracking
()
People
(Reporter: botond, Assigned: botond)
References
Details
Attachments
(8 files)
47 bytes,
text/x-phabricator-request
|
Details | Review | |
47 bytes,
text/x-phabricator-request
|
Details | Review | |
47 bytes,
text/x-phabricator-request
|
Details | Review | |
47 bytes,
text/x-phabricator-request
|
Details | Review | |
47 bytes,
text/x-phabricator-request
|
Details | Review | |
47 bytes,
text/x-phabricator-request
|
Details | Review | |
47 bytes,
text/x-phabricator-request
|
Details | Review | |
47 bytes,
text/x-phabricator-request
|
Details | Review |
Timothy suggested in a review comment that it would be useful to add a test covering the scenario that bug 1637113 regressed to avoid regresseing it in the future.
Will add a test in this bug.
Comment 1•4 years ago
|
||
Testing the specific fix of bug 1637113 is good, having a general test in the tree that checks that "scroll down, pinch zoom, tap" actually hits the expected element under the tap would be great.
Updated•4 years ago
|
Assignee | ||
Updated•4 years ago
|
Updated•4 years ago
|
Updated•4 years ago
|
Updated•4 years ago
|
Assignee | ||
Comment 2•4 years ago
|
||
I have a mochitest for bug 1637113 that's working locally. It required making some improvements to coordinatesRelativeToScreen()
to handle iframes better, and these changes are causing some other mochitest failures which I need to investigate.
Assignee | ||
Comment 3•4 years ago
|
||
(In reply to Botond Ballo [:botond] from comment #2)
I have a mochitest for bug 1637113 that's working locally. It required making some improvements to
coordinatesRelativeToScreen()
to handle iframes better, and these changes are causing some other mochitest failures which I need to investigate.
Finally have a green Try push. The patches need a bit of cleanup as my understanding of the issue has evolved over the course of investigating the failures.
Assignee | ||
Comment 4•4 years ago
|
||
Assignee | ||
Comment 5•4 years ago
|
||
Depends on D83018
Assignee | ||
Updated•4 years ago
|
Pushed by bballo@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/461dea66f9f1 Fix coordinatesRelativeToScreen() to work better for targets inside an iframe. r=tnikkel https://hg.mozilla.org/integration/autoland/rev/92b59d89a40e Add a mochitest for bug 1637113. r=tnikkel
Assignee | ||
Comment 7•4 years ago
|
||
Pushed by rmaries@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/58000ee18336 Fix trailing whitespace in test r=rmaries
Comment 9•4 years ago
|
||
bugherder |
Assignee | ||
Comment 10•4 years ago
•
|
||
The test that was landed for bug 1637113 provides test coverage for bug 1638655 as well: this Try push shows the test failing on Android if the patch for bug 1638655 is backed out.
Assignee | ||
Comment 11•4 years ago
|
||
Assignee | ||
Comment 12•4 years ago
|
||
Depends on D84882
Assignee | ||
Comment 13•4 years ago
|
||
The test is also revised to fail immediately rather than via a timeout
if the expected event is fired but targets the wrong element.
Assignee | ||
Comment 14•4 years ago
|
||
Depends on D84891
Assignee | ||
Comment 15•4 years ago
|
||
Depends on D84892
Comment 16•4 years ago
|
||
Pushed by bballo@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/e2705663333f Fix non-unified bustage in nsDisplayList.cpp. r=tnikkel https://hg.mozilla.org/integration/autoland/rev/75650b9112fa Add a mochitest for bug 1638458. r=tnikkel
Comment 17•4 years ago
|
||
bugherder |
Comment 18•4 years ago
|
||
Pushed by bballo@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/729ccac1e832 Rewrite the test for bug 1637113 in the async/await style. r=tnikkel https://hg.mozilla.org/integration/autoland/rev/0b41d1d9ed6c Add a mochitest for bug 1638441. r=tnikkel https://hg.mozilla.org/integration/autoland/rev/2d3d30eb308b Add a mochitest for bug 1637135. r=tnikkel
Comment 19•4 years ago
|
||
bugherder |
Assignee | ||
Comment 20•4 years ago
|
||
We now have test coverage for 5 of the 7 bugs mentioned in the bug title.
The remaining two are:
- Bug 1643212, which is a dynamic toolbar issue. As discussed here, we don't really have the test infrastructure in place to test issues like this.
- Bug 1638594, which only affects content that's loaded in the parent-process, which again I'm not quite sure how to exercise in a mochitest environment.
I'm inclined to say that having test coverage for 5 of these 7 issues is good enough, and move on to other things.
Comment 21•4 years ago
|
||
(In reply to Botond Ballo [:botond] from comment #20)
- Bug 1638594, which only affects content that's loaded in the parent-process, which again I'm not quite sure how to exercise in a mochitest environment.
mochitest-chrome or browser-chrome test?
I'm inclined to say that having test coverage for 5 of these 7 issues is good enough, and move on to other things.
Sounds reasonable. Thanks for working on this!
Comment 22•4 years ago
|
||
This doesn't really block desktop zooming on nightly, so let's disconnect it.
Comment 23•3 years ago
|
||
The leave-open keyword is there and there is no activity for 6 months.
:botond, maybe it's time to close this bug?
Assignee | ||
Comment 24•3 years ago
|
||
I was going to write a browser chrome test for bug 1638594 as suggested in comment 21 but realistically I don't think I'm going to get to it.
Updated•3 years ago
|
Description
•