Remove "dom.mozBrowserFramesEnabled" pref
Categories
(Core :: DOM: Navigation, task)
Tracking
()
Tracking | Status | |
---|---|---|
firefox77 | --- | fixed |
People
(Reporter: kmag, Assigned: kmag)
References
Details
Attachments
(2 files)
Assignee | ||
Comment 1•4 years ago
|
||
Assignee | ||
Comment 2•4 years ago
|
||
Comment 3•4 years ago
|
||
RDM will continue to use mozbrowser until Bug 1585097 lands. I haven't confirmed yet that the changes in this patch break that, but I want to get the blocker noted here somewhere.
Pushed by maglione.k@gmail.com: https://hg.mozilla.org/integration/autoland/rev/ba0fb70d5b35 Part 1 - Get rid of most of the remaining mozbrowser API. r=nika,mtigley,bradwerth https://hg.mozilla.org/integration/autoland/rev/dcbfa257190e Part 2 - Get rid of the "dom.mozBrowserFramesEnabled" pref. r=nika
Comment 5•4 years ago
|
||
I'm concerned. I'm sorry that the blocker Bug 1585097 got assigned late; I thought we had communicated that requirement through e-mail channels but I obviously didn't follow up to assign the blocker bug to the right bit of code. If this rides the trains, it likely will break RDM (with unknown severity) in Dev Edition and Release. I haven't done a rebuild to check this patch with pref devtools.responsive.browserUI.enabled
set to false to confirm this. That pref is default true in Nightly only. Automated tests will not catch this failure because we made the decision to test only with the pref set as of landing Bug 1585121. That is looking to have been a very bad decision, now.
Assignee | ||
Comment 6•4 years ago
|
||
This doesn't remove anything that RDM uses, and in particular, RDM never used the dom.mozBrowserFramesEnabled pref.
Comment 7•4 years ago
|
||
(In reply to Kris Maglione [:kmag] from comment #6)
This doesn't remove anything that RDM uses, and in particular, RDM never used the dom.mozBrowserFramesEnabled pref.
Great, thank you. Would you please make the bug that is the "get rid of mozbrowser" work blocked by Bug 1585097? That will add some protection that we don't land code that breaks RDM.
Comment 8•4 years ago
|
||
(In reply to Kris Maglione [:kmag] from comment #6)
This doesn't remove anything that RDM uses, and in particular, RDM never used the dom.mozBrowserFramesEnabled pref.
I've confirmed this patch doesn't cause failures with RDM with devtools.responsive.browserUI.enabled
false. Sorry for the false alarm.
Comment 9•4 years ago
|
||
bugherder |
https://hg.mozilla.org/mozilla-central/rev/ba0fb70d5b35
https://hg.mozilla.org/mozilla-central/rev/dcbfa257190e
Comment 10•4 years ago
|
||
This patch is problematic for b2g because it also removes IsUserFocusIgnored() support, which we need. Can we get that part back in?
Comment 11•4 years ago
|
||
Note: Instead of using an <iframe mozbrowser> we now put the ignoreuserfocus attribute on a <xul:browser>.
Updated•4 years ago
|
Comment 12•4 years ago
|
||
(In reply to [:fabrice] Fabrice Desré from comment #10)
This patch is problematic for b2g because it also removes IsUserFocusIgnored() support, which we need. Can we get that part back in?
Apologies for the delay. After talking with the fission team it sounds like they'd be happy to r+ a patch restoring this functionality. Fabrice, would you mind putting together a patch that reverts the parts that broke ignoreuserfocus
?
Description
•