about:support mistakenly shows presence of user.js in Firefox for Android
Categories
(Toolkit :: General, defect)
Tracking
()
People
(Reporter: philipp, Assigned: robwu)
References
(Regression)
Details
(Keywords: regression)
Attachments
(1 file)
Updated•11 years ago
|
Updated•11 years ago
|
Updated•3 years ago
|
| Assignee | ||
Comment 2•5 months ago
|
||
Some elements were shown when they should not be on Android,
because these elements rely on the hidden class name introduced in bug
948883, but that declaration is missing for the mobile-specific
aboutSupport.css style sheet.
This patch fixes the visibility by relying on the hidden attribute
instead, and for consistency replaces all uses of display:none with
the hidden boolean. I verified that there are no display rules in
either aboutSupport.css file that would supersede the display:none
declaration of the [hidden] attribute.
Moreover, to avoid errors about "content-analysis-active" element that
is non-existent on Android due to ifndef in the XHTML file, an early
return is added to its usage.
I also audited all uses of ifndef in the aboutSupport.xhtml file and
added/changed the condition in the aboutSupport.js file for consistency.
| Assignee | ||
Comment 3•5 months ago
|
||
Comment 2 references two bugs:
- This reported regression was introduced in bug 948883.
- "content-analysis-active" error / logspam was introduced in bug 1882606.
Backed out for causing bc failures @ browser_aboutSupport_places
| Assignee | ||
Comment 7•5 months ago
|
||
The test explicitly checked inline style.display states, but because the patch refactored to use the hidden attribute, that does not work.
I've switched to ContentTaskUtils.isHidden and isVisible, and the test now passes again.
As an extra sanity check I looked for all tests that reference "about:support" and ran them locally, they all pass.
So I'll reland the patch with the fix.
Comment 9•5 months ago
|
||
| bugherder | ||
Updated•5 months ago
|
Updated•4 months ago
|
Description
•