92210 - Window steals focus if you switch to another app while "Transferring data from..."

Status: VERIFIED FIXED

(Core :: DOM: UI Events & Focus Handling, defect)

Description

[Brian Ryner (not reading)]

On a Linux branch build (200107120, which has saari's patch), I can get the
window to steal focus back from another app using the following steps:

1. Type in a URL2. During the time that it says "Transferring data from ..." in
the status bar (the old document is still showing at this point), alt-tab to
another app.
3. When painting is unsuppressed, the browser window steals focus.

Comment 1

[Brian Ryner (not reading)]

Ok, so it looks like we aren't getting a deactivate event in the ESM for some
reason...

Comment 2

[Brian Ryner (not reading)]

Ok, I have a patch and better steps to reproduce this:

1. Create a personal toolbar bookmark to query Bonsai for "all checkins in the
last day".
2. Load www.mozilla.org
3. Click the bonsai toolbar button
4. Wait for the status bar to say "Transferring data from bonsai.mozilla.org". 
You may have to move the mouse around a bit while it's connecting to get this
message to show up.
5. Alt-tab to another app
6. Alt-tab back to mozilla
7. Alt-tab back to the other app

Result:

When painting unsuppresses on the new page, the mozilla window steals focus.

Comment 3

[Brian Ryner (not reading)]

Attachment: patch

Comment 4

[saari (gone)]

r=saari

Comment 5

[David Hyatt]

sr=hyatt

Comment 6

[Brian Ryner (not reading)]

checked in.

Comment 7

[Simon Fraser [no longer active]]

This checkin caused a regression: bug 95243

Comment 8

[Jacek Piskozub]

This checkin may be the reason for another hang, bug 95483.

Comment 9

[kinmoz]

Reopening this bug. I had to back out the patch for this bug because it caused
blocker bugs 95243 (Crash when typing in body of mail compose window)
and 95297 (session history doesn't restore scroll position and form controls).

Comment 10

[Jacek Piskozub]

After applying the patch to bug 95243, I see that it is really fixed but blocker
bug 95297 and bug 95483 are not. It means the last two bugs were not caused by
this checkin.

Comment 11

[Brian Ryner (not reading)]

Ok, I think I have a patch for this bug that does not cause the regressions.

Comment 12

[Brian Ryner (not reading)]

Attachment: patch #2

Comment 13

[David Hyatt]

sr=hyatt

Comment 14

[saari (gone)]

okay, code looks good, and I'll assume hyatt knows what he's talking about 
(bwahahah)  r=saari

Comment 15

[Brian Ryner (not reading)]

checked in.

Comment 16

[Madhur Bhatia]

verified fixed - buildID : 2001-08-22-08 trunk. - Linux  7.1