After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 594738 - Windows often do not respond to events on dual-head
Windows often do not respond to events on dual-head
Status: RESOLVED FIXED
Product: gtk+
Classification: Platform
Component: Backend: Quartz
2.17.x
Other All
: Normal normal
: ---
Assigned To: gtk-quartz maintainers
gtk-bugs
Depends on:
Blocks:
 
 
Reported: 2009-09-10 11:57 UTC by Christian Hergert
Modified: 2009-09-16 06:54 UTC
See Also:
GNOME target: ---
GNOME version: ---


Attachments
Don't do a bounds check since windows can often be in negative space with dual-head (1.07 KB, application/octet-stream)
2009-09-10 11:57 UTC, Christian Hergert
Details

Description Christian Hergert 2009-09-10 11:57:46 UTC
Created attachment 142905 [details]
Don't do a bounds check since windows can often be in negative space with dual-head

After clicking on a window in a dual head setup, it often stops responding to
events such as button press and scroll events.

The attached patch forcefully fixes the problem.  Either its the right approach
because the bounds check is unnecessary or there is a better way to do it.

Comments appreciated.
Comment 1 Kristian Rietveld 2009-09-12 10:15:08 UTC
Looking into this now.  Tried on a dual-head setup, but I can also reproduce it in single head.
Comment 2 Kristian Rietveld 2009-09-13 19:46:58 UTC
Been looking in this over the weekend.  It seems that the issue I found in single head also exists in dual head, but is a separate problem from what you reported in the opening comment.  The issue I found is caused by changes in the grabbing code introduced due to the merge of client side windows.  Also when running with "native" GDK windows some things are broken.  There is a chance that the issue we see in dual head is also caused by this brokenness, if not the code in _gdk_quartz_window_find_child() should be made multi-head aware.

I will be working on getting to the bottom of both issues and fix them.  Because it is harder than I originally expected, it will take a few more days -- my spare time is limited, but this is on the top of my list.
Comment 3 Kristian Rietveld 2009-09-15 14:05:40 UTC
I have an idea for a patch that fixes, but I need to test that at home.  I don't have dual-head at this conference ;)  When it works I will commit and push the fix.
Comment 4 Kristian Rietveld 2009-09-16 06:54:37 UTC
Pushed.