Closed Bug 517379 Opened 15 years ago Closed 14 years ago

restarting after quit from fullscreen presents fullscreen window with wrong controls

Categories

(Firefox :: General, defect, P2)

x86_64
Linux
defect

Tracking

()

RESOLVED DUPLICATE of bug 575195

People

(Reporter: karlt, Unassigned)

References

Details

(Keywords: regression)

STR:
1. With one tab in a single Firefox window, View -> Full Screen.
2. Quit by clicking on the close box in the urlbar.
3. Start Firefox again.

Expected results:
Probably a window with decorations, perhaps maximized.
Possibly a fullscreen window with fullscreen controls and "Full Screen" checked in the View menu.

Actual results:
A fullscreen window with normal toolbars, and "Full Screen" unchecked in the View menu.
Trying to check "Full Screen" in the View menu exits fullscreen mode.
Hardware: x86 → x86_64
probably my regression.  we did this on linux:

https://bug512575.bugzilla.mozilla.org/attachment.cgi?id=396836
This kind of stuff feels like it should be shared in nsBaseWidget.
Regression from http://hg.mozilla.org/mozilla-central/rev/ac92849658c3.
Bug 512529 comment 2 was also caused by that.
(http://hg.mozilla.org/mozilla-central/rev/8dc97b8f087f helped some of the other issues.)
Blocks: 510546
Flags: blocking1.9.2?
Assignee: nobody → doug.turner
Flags: blocking1.9.2? → blocking1.9.2+
Priority: -- → P2
Looks like we shouldn't return early here if GDK_WINDOW_STATE_FULLSCREEN has
changed:
http://hg.mozilla.org/mozilla-central/annotate/7d0c4369c0d1/widget/src/gtk2/nsWindow.cpp#l3252
Maybe the changes in Bug 512575 won't even be needed with that sorted.
Blocks: 527615
Depends on: 512529
(In reply to comment #4)
> Looks like we shouldn't return early here if GDK_WINDOW_STATE_FULLSCREEN has
> changed:
> http://hg.mozilla.org/mozilla-central/annotate/7d0c4369c0d1/widget/src/gtk2/nsWindow.cpp#l3252

I tried that (attachment 411575 [details] [diff] [review]) but it's not the (whole?) solution.

I don't know if this is really a Widget bug even.  (Either it's not or I don't know what Widget is meant to do.)
On MS Windows, the browser starts in normal size mode but the size covers the whole desktop.  That's not right either, but quite different.
nsGlobalWindow::SetFullScreen() dispatches a "fullscreen" event.
I'm guessing that delayedStartup in browser/base/content/browser.js has not yet been called to register the event listener, and var FullScreen doesn't seem to get initialized based on the value of window.fullScreen.

Product -> Firefox.
Assignee: mozbugz → nobody
Component: Widget: Gtk → General
Flags: blocking1.9.2+
Product: Core → Firefox
QA Contact: gtk → general
blocking1.9.2+ -> blocking-firefox3.6
Flags: blocking-firefox3.6?
No longer blocks: 527615
(In reply to comment #7)
> nsGlobalWindow::SetFullScreen() dispatches a "fullscreen" event.

that's dispatched here it looks like:

http://mxr.mozilla.org/mozilla-central/source/xpfe/appshell/src/nsXULWindow.cpp#1188

would moving registration for that even to the xul window's onload be early enough?
WANT. but not a blocker, would accept low-risk fix.
Flags: wanted-firefox3.6+
Flags: blocking-firefox3.6?
Flags: blocking-firefox3.6-
related: windows bug 575195. The browser.js patch there might fix this bug if linux's widget is sending a size mode event from Makefullscreen.
(In reply to comment #12)
> related: windows bug 575195. The browser.js patch there might fix this bug if
> linux's widget is sending a size mode event from Makefullscreen.

oh, and it would need the xpfe patch as well.
Depends on: 575195
Fixed by bug 575195.  Thank you!
Status: NEW → RESOLVED
Closed: 14 years ago
No longer depends on: 512529, 575195
Resolution: --- → DUPLICATE
You need to log in before you can comment on or make changes to this bug.