Closed Bug 1215476 Opened 10 years ago Closed 10 years ago

[Test] task_manager_icon_test Marionette test failing locally

Categories

(Firefox OS Graveyard :: Gaia::System, defect)

ARM
Gonk (Firefox OS)
defect
Not set
normal

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: apastor, Assigned: mcav)

References

Details

(Whiteboard: [systemsfe])

Attachments

(2 files)

Despite gaia-try doesn't seem to show it, running locally: make test-integration TEST_FILES=apps/system/test/marionette/task_manager_icons_test.js results in a test failing. 1) Task Manager - Icons Firefox Apps use icon from app manifest: AssertionError: false == true at Context.<anonymous> (/Users/alberto/Projects/gaia/apps/system/test/marionette/task_manager_icons_test.js:102:7)
Assignee: nobody → apastor
Comment on attachment 8674822 [details] [review] [gaia] albertopq:1215476-task-manager-test > mozilla-b2g:master Hey Marcus, if I'm not wrong, the assertion in this test is incorrect. Given https://github.com/mozilla-b2g/gaia/blob/master/apps/system/js/card.js#L59 the icon size taken will be 64 and not 32 as the test expects. Could you please take a look? Thanks!
Attachment #8674822 - Flags: review?(m)
Blocks: 1179723
Whiteboard: [systemsfe]
Attachment #8674822 - Flags: review?(m) → review+
Status: NEW → RESOLVED
Closed: 10 years ago
Resolution: --- → FIXED
Does this test pass locally for you, Michael? I believe this doesn't fail in gaia-try? Any idea why this would fail in b2g-inbound?
Flags: needinfo?(mhenretty)
(In reply to Alberto Pastor [:albertopq] from comment #5) > Does this test pass locally for you, Michael? I believe this doesn't fail in > gaia-try? Any idea why this would fail in b2g-inbound? I tried running task_manager_icons_test.js locally before I reverting, and it did indeed fail locally. When I reverted it passed locally. So I can at least reproduce It is intermittent (even on b2g-inbound), so my thinking is we should fix it locally, submit a PR, re-trigger a bunch of times in the PR, and then land and watch b2g-inbound.
Flags: needinfo?(mhenretty)
Seems to be working now..
Status: REOPENED → RESOLVED
Closed: 10 years ago10 years ago
Resolution: --- → INVALID
I think this is a devicePixelRatio-related issue; it passes on my Linux VM, fails on my Mac, with the same source tree. Taking.
Status: RESOLVED → REOPENED
Resolution: INVALID → ---
I think this is a devicePixelRatio-related issue; it passes on my Linux VM, fails on my Mac, with the same source tree. Taking.
Assignee: apastor → m
Attachment #8676517 - Flags: review?(apastor)
No longer blocks: 1179723
Comment on attachment 8676517 [details] [review] [gaia] mcav:icon-size-test > mozilla-b2g:master I think that code will only work if devicePixelRatio == 2. I think we need to find the closest size to 32 * devicePixelRatio in the expectedSizes list, instead.
Attachment #8676517 - Flags: review?(apastor)
Blocks: 1215518
Comment on attachment 8676517 [details] [review] [gaia] mcav:icon-size-test > mozilla-b2g:master You were right that the previous-patch devicePixelRatio test would only work for integer values. This updated patch just mocks devicePixelRatio to always equal 1, with the following rationale: The devicePixelRatio-dependent icon-selection algorithm is deep within IconsHelper, which has a bunch of unit tests already determining which icon will be selected based on dPR. I had attempted to duplicate the logic within task_manager_icons_test, but it seemed much more brittle compared to mocking dPR and letting IconsHelper be responsible for its own dPR-specific wishes.
Attachment #8676517 - Flags: review?(apastor)
Comment on attachment 8676517 [details] [review] [gaia] mcav:icon-size-test > mozilla-b2g:master Thanks!
Attachment #8676517 - Flags: review?(apastor) → review+
Status: REOPENED → RESOLVED
Closed: 10 years ago10 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: