Closed Bug 406342 Opened 17 years ago Closed 17 years ago

Clicking Done in toolbar customization panel hides parent window

Categories

(Core :: Widget: Cocoa, defect, P1)

PowerPC
macOS
defect

Tracking

()

VERIFIED FIXED

People

(Reporter: philor, Assigned: enndeakin)

References

Details

Attachments

(1 file, 1 obsolete file)

After the landing of bug 395334, clicking the "Done" button in the toolbar customization panel hides the parent window.

STR:
1. Right click toolbar, choose Customize...
2. In the customize panel, click Done
3. Panic, because the whole app window disappeared
4. Cmd+tab to another app, then Cmd+tab back to your apparently unharmed Minefield

Not quite sure how to tell whether it's really just from the widget: cocoa parts of bug 395334, or a cross-platform bug that we only exercise on OS X.
Flags: blocking1.9?
Isn't this and 403942 the same? So the Add Toolbar button doesn't actually factor in?
This seems to do the trick. Do I need to call SendSetZLevelEvent or the parent as well?
Attachment #291216 - Flags: review?(joshmoz)
Yeah, I'd call SendSetZLevelEvent on the parent. Otherwise this looks good to me.
Flags: blocking1.9? → blocking1.9+
Priority: -- → P2
Attachment #291216 - Flags: superreview?(roc)
Can you explain what's going on here? I'm lost.
(In reply to comment #1)
> Isn't this and 403942 the same?

Not to someone who doesn't know widget code anyway - the bug 403942 problem was that you could interact with the panel while it had opened a sheet, and at that time when you closed them in the order panel, then sheet, you were left with a visible but non-functional window, while closing them sheet, then panel worked perfectly. Now, no matter what order you close them, or whether or not you open the sheet, closing the panel hides the browser window.
Flags: in-litmus?
(In reply to comment #4)
> Can you explain what's going on here? I'm lost.
> 

The part of the code does: when Show(PR_FALSE) is called for a panel, unhook the panel from its parent window. The patch adds a makeKeyAndOrderFront call which ensures the parent window is ordered on the front and a key window (key window means focused window). I don't know why the window isn't already the front one, but this patch fixes the issue.
 
This is better I think. No need for the makeKeyAndOrderFront call and no need for the z-level event.
Attachment #291216 - Attachment is obsolete: true
Attachment #291690 - Flags: review?(enndeakin)
Attachment #291216 - Flags: superreview?(roc)
Attachment #291216 - Flags: review?(joshmoz)
Per mac meeting discussion, this will block beta 2.  We need to get this landed ASAP once Neil reviews.
Priority: P2 → P1
Attachment #291690 - Flags: review?(enndeakin) → review+
Attachment #291690 - Flags: superreview?(vladimir)
Attachment #291690 - Flags: superreview?(vladimir) → superreview+
landed on trunk
Status: NEW → RESOLVED
Closed: 17 years ago
Resolution: --- → FIXED
verified fixed using  Mozilla/5.0 (Macintosh; U; PPC Mac OS X 10.5; en-US; rv:1.9b2pre) Gecko/2007120604 Minefield/3.0b2pre. I verified using the STR in Comment 0.
Status: RESOLVED → VERIFIED
https://litmus.mozilla.org/show_test.cgi?id=5048 is the test case added to Litmus.
Flags: in-litmus? → in-litmus+
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: