| Summary: | KWin behaves oddly with WM_HINTS.input == false (breaks cellwriter) | ||
|---|---|---|---|
| Product: | [Plasma] kwin | Reporter: | M G Berberich <kde> |
| Component: | general | Assignee: | KWin default assignee <kwin-bugs-null> |
| Status: | RESOLVED WORKSFORME | ||
| Severity: | normal | CC: | alex.merry, cedric.boutillier, cfeck, kde.org, kde |
| Priority: | NOR | ||
| Version: | unspecified | ||
| Target Milestone: | --- | ||
| Platform: | Debian testing | ||
| OS: | Unspecified | ||
| Latest Commit: | Version Fixed In: | ||
| Sentry Crash Report: | |||
|
Description
M G Berberich
2010-04-20 22:50:27 UTC
Problem confirmed on openSUSE 11.2 with KDE from trunk and Qt from 4.7 branch. It used to work with KDE 3 (that I remember), and I tried with KDE 4 again. It may have to do with extended input events, as I got it working when enabling them in cellwriter (see cellwriter home page), but it did not work reliably, i.e. at one point it stops working again. Yes it worked fine on KDE3. Switching on input-events make it work until the cursor leaves the cellwriter-window. Then it's not working at all until input-events are switched off again. it's a utility window and does not get focus ever. try settings "left button" to only "actiate" in the actions page of kwins configuration dialog. setting a special rule that enforces focussing (preferences, accept focus) will work around this. Thomas, cellwriter is a tool like xvkbd, so it should never get the focus. The focus should stay at the application window that the written input should receive. ok, makes quite some sense and indeed - it sets focus acception to false, Utility type handling is ok. interestingly, by shading the window i can _activate_ it, nevertheless the input fails. (unless forcing the focus) --- as the input seems timer related, i vaguely suspect the usage of CurrentTime instead xTime() in events.cpp - no idea why that was changed (for pointer inputs) or if it's the culprit at all, though :-\ *** Bug 244625 has been marked as a duplicate of this bug. *** Copying info from Bug 244625: When an application sets the input flag of WM_HINTS to false (preventing the window manager from passing focus to it), pressing a mouse button down on the window when running under KWin sends an XLeaveWindowEvent to the window. On other window managers this doesn't happen (the ButtonPress event gets sent, but not the LeaveWindow/Crossing event). See Bug 244625 for test cases. Here is a possible workaround for cellwriter that might(?) also help solve the issue: "I am running Kubuntu 11.04, and can confirm that launching with: GDK_NATIVE_WINDOWS=1 cellwriter fixed the single stroke problem for me also. Thank you so much. Is there a way to make this happen by default for cellwriter. I don't mean mine, I mean the program." https://code.google.com/p/cellwriter/issues/detail?id=34#c11 Cellwriter 1.3.5 does no longer show this issue. Not sure if it was a KWin bug, which has been worked around, or if this was a cellwriter bug. I don't see a 1.3.5 version, but archlinux/aur has this patch from launchpad invoked: http://launchpadlibrarian.net/18959127/cellwriter-1.3.4-cellwidget-dont-disable-xinput.diff So it likely works because they don't disable xinput anymore. The reason for kwin lies likely somewhere in the getWMHints() actions, no idea whether we or they behaved "incorrectly" (ie. for them on implicit expectations) I'll check whether there's some side unwanted effect (through urgency or whatever) nevertheless. I got the 1.3.5 version from http://code.google.com/p/cellwriter/ ahh, they moved. from the sources (cellwriter.c:2043) // Do not listen to leave_notify_event, after a certain GTK version // it fires for just about anything you do on the screen. I guess the fake event handling in regarding non native windows (GDK_NATIVE_WINDOWS=0) is a bit aggressive and will react to various (synthetic) configure events or even property changes (ie. kwin probably informs the client "you did not get the focus" and gtk maps that to "leave") - so we probably have to look for the rather "absurd" source of this ... is this still an issue? This issue report is quite old. Can you please confirm, that it still persists with KDE 5.23? Dear Bug Submitter, This bug has been in NEEDSINFO status with no change for at least 15 days. Please provide the requested information as soon as possible and set the bug status as REPORTED. Due to regular bug tracker maintenance, if the bug is still in NEEDSINFO status with no change in 30 days the bug will be closed as RESOLVED > WORKSFORME due to lack of needed information. For more information about our bug triaging procedures please read the wiki located here: https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging If you have already provided the requested information, please mark the bug as REPORTED so that the KDE team knows that the bug is ready to be confirmed. Thank you for helping us make KDE software even better for everyone! This bug has been in NEEDSINFO status with no change for at least 30 days. The bug is now closed as RESOLVED > WORKSFORME due to lack of needed information. For more information about our bug triaging procedures please read the wiki located here: https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging Thank you for helping us make KDE software even better for everyone! |