RE: win32 seamonkey march 17 build platform: win_nt 4.0 Env: HP X86 family 6, 233 mhz, 32 MB Problem: A blinking vertical bar appears on top of the left-hand corner of the screen. Steps: 1. Execute apprunner.exe 2. From the Task menu, click on Messenger to bring up the 3 pane message window 3. Click on New Msg to bring up the compose window 4. Compose and send a message 5. Click X to close the compose window, a vertical bar immediately appears on the top left-hand corner of the screen. 6. Repeat step 3 to 5, each time you close a compose window, a new vertical bar will come up. 7. A maxiumn of 3 blinking vertical bars will come up. And if you drag any seamonkey window over the blinking bars, the window will collect these vertical bars. 8. However, if you exit the seamonkey application, the vertical bars go away.

-> kostello

Simon -- all yours

This is almost certainly because of bug 4127. The entire web shell and pres shell get leaked, so the caret never knows to go away and just keeps blinking. You're lucky it does not crash ;-)

*** Bug 4560 has been marked as a duplicate of this bug. ***

Changing platform to all. According to the just-marked dupe, this bug now includes the fact that closing the compose window on Linux causes an immediate crash. Changing self to qa-contact; jeffu is outta here.

Correcting summary to reflect the crash on Linux.

*** Bug 5211 has been marked as a duplicate of this bug. ***

This is really a dup of 4127, but I want to keep it open in case we need a workaround.

M5, but blocked by 4127.

Adding to QA Blocker radar.

*** Bug 5717 has been marked as a duplicate of this bug. ***

This problem has been fixed by the fix for 4127.

This one is a M5 blocker? Seems I can't test it till 5793 is fixed, and that one isn't a M5 blocker. Jan/Lisa - might want to consider making that one a blocker, or unmaking this one a blocker....

Having problems with closing the compose window showed itself during M4. As sfraser states, this was manifested by bug 4127. Bug 4127 was fixed, but further problems surfaced described in bug 5793. Bug 5793 was made into a M5 stopper, but was changed to M6 due to resource priorities (ducarroz was working on another stopper bug). So, this bug has the right target milestone; it's just that bug 5793 was changed to a later milestone as described above. Just mention in the "status whiteboard" that you can't verify until bug 5793 is fixed. Thanks.

Looks fixed to me, 05-04-08 Win32. Marking as verified.