Placeholders on the datetime widget ignore spoof English
Categories
(Toolkit :: UI Widgets, defect)
Tracking
()
| Tracking | Status | |
|---|---|---|
| firefox128 | --- | fixed |
People
(Reporter: pierov, Assigned: pierov)
References
(Blocks 1 open bug)
Details
Attachments
(1 file)
Steps to reproduce:
- set Firefox in a language different than en-US and enable RFP + spoof English
- open a page with a datetime widget (e.g., https://arkenfox.github.io/TZP/tests/applang.html)
- notice that the placeholder is using the en-US format, but the localized placeholders (e.g.,
mm / gg / aaaainstead ofmm / dd / yyyy)
| Assignee | ||
Comment 1•2 years ago
|
||
The datetime input is inconsistent with other inputs when using spoof
English: its placeholder is not translated, unlike the default values
and texts of all the other inputs.
Updated•2 years ago
|
Comment 3•2 years ago
|
||
Backed out changeset fab1526875d3 (Bug 1880108) for causing failures in browser_caching_attributes.js
Log: https://treeherder.mozilla.org/logviewer?job_id=448999171&repo=autoland&lineNumber=4504
browser-chrome: https://treeherder.mozilla.org/logviewer?job_id=448996825&repo=autoland&lineNumber=13133
Backout: https://hg.mozilla.org/integration/autoland/rev/81edde55cc8899af0acb636f545d91efcc6dc59c
| Assignee | ||
Comment 4•2 years ago
|
||
I can't reproduce these failures locally (but I'm on Linux, setting up a Firefox dev environment on a macOS machine would take me a while).
So, I tried to run another try build, but it doesn't contain the errors from the logs.
https://treeherder.mozilla.org/logviewer?job_id=448999171&repo=autoland&lineNumber=4504
This is a "Test timed out -", could it be due to something else?
https://treeherder.mozilla.org/logviewer?job_id=448996825&repo=autoland&lineNumber=13133
The root cause seems a missing datetimebox.ftl: https://treeherder.mozilla.org/logviewer?job_id=448999171&repo=autoland&lineNumber=4433
It says locale en, whereas I expected something like en-US instead.
I've tried to play a little bit with getWebExposedLocales(), but I always get en-US appended at the end (when using another language) or instead of a regional variant (e.g., in en-GB).
I don't think I know of to explain this failure.
Who could I redirect this NI to for some help?
| Assignee | ||
Comment 5•2 years ago
|
||
Is there anyway I could get this merged again?
I think the failed tests were false positives.
Comment 7•2 years ago
|
||
Backed out for causing multiple failures on OS X 10.15 WebRender. (Please check the: push with failures link)
Updated•2 years ago
|
| Assignee | ||
Comment 8•2 years ago
|
||
Okay, it seems the macOS test uses en as a locale (I don't know why).
In the new revision I'm forcing en-US to be there as a fallback, in case the webExposedLocales do not include any locale with the proper translation.
Also, I gated the patch on RFP.
| Assignee | ||
Comment 9•2 years ago
|
||
Hi, I sent some changes because the tests found some problems with my previous patch (basically, en-US wasn't automatically appended as fallback language).
Could you please review the patch again? Thanks in advance!
Updated•2 years ago
|
Comment 10•2 years ago
|
||
Went ahead and reviewed the patch, changes look good to me. Queued up the patch in Lando as well. Hopefully it'll stick this time!
Comment 11•2 years ago
|
||
Comment 12•2 years ago
|
||
FYI: test is at https://arkenfox.github.io/TZP/tests/elementforms.html, no visuals - you should get the same hash (or in this specific case, the details measurements) on your machine (because it can vary across machines) in any app language with spoof_english
I think we should add a test - RFP + en-US detail width == RFP + ga-IE detail width
for example: the first measurement is width
en-US for me = "details": [66.75, 21.566665649414062, -3.6666717529296875, -1.1833343505859375],
ga-IE for me = "details": [98.73333740234375, 21.566665649414062, -5.4166717529296875, -1.1833343505859375],
Comment 13•2 years ago
|
||
| bugherder | ||
Description
•