Re: Review Request 118167: Flush xcb connection after preparing KSelectionOwner::claim
On May 16, 2014, 3:38 p.m., Thomas Lübking wrote: a) main.cpp, void Application::start() -Xcb::sync(); owner-claim(m_replace, true); +Xcb::sync(); Sufficient? (though it warns about a QPA crash later) b) Qt bug (should flush once per eventcycle) or feature (keeps dirty flag and only flushes on internal need)? - can such flag be set explicitly? Martin Gräßlin wrote: concerning a): no, as it could hit other applications, too. b) I'm not sure whether that's in the scope of Qt. If one goes down towards the lower level one needs to take care of it. Thomas Lübking wrote: What I meant is: if this is indeed just a bug in Qt, it's probably pointless to fix it here. If it's a feature (and we've no access to it) it's oc. our duty to make unportable things work. Martin Gräßlin wrote: right, I thought about it and yes an xcb_flush() before going to sleep could make a few things easier. I'll see whether I can hack Qt for that next week. investigation result from looking at Qt - not a solution. Qt uses a thread to poll the xcb connection socket, so we cannot flush before it's going to block again. - Martin --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118167/#review58061 --- On May 16, 2014, 3 p.m., Martin Gräßlin wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118167/ --- (Updated May 16, 2014, 3 p.m.) Review request for KDE Frameworks and kwin. Bugs: 334858 https://bugs.kde.org/show_bug.cgi?id=334858 Repository: kwindowsystem Description --- Flush xcb connection after preparing KSelectionOwner::claim Without the flush nothing starts. This caused KWin to stall during a restart. BUG: 334858 Diffs - src/kselectionowner.cpp 90a0004046135e7fee61fe2ef865b75eb421ea99 Diff: https://git.reviewboard.kde.org/r/118167/diff/ Testing --- Thanks, Martin Gräßlin ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118167: Flush xcb connection after preparing KSelectionOwner::claim
On May 16, 2014, 3:38 p.m., Thomas Lübking wrote: a) main.cpp, void Application::start() -Xcb::sync(); owner-claim(m_replace, true); +Xcb::sync(); Sufficient? (though it warns about a QPA crash later) b) Qt bug (should flush once per eventcycle) or feature (keeps dirty flag and only flushes on internal need)? - can such flag be set explicitly? Martin Gräßlin wrote: concerning a): no, as it could hit other applications, too. b) I'm not sure whether that's in the scope of Qt. If one goes down towards the lower level one needs to take care of it. Thomas Lübking wrote: What I meant is: if this is indeed just a bug in Qt, it's probably pointless to fix it here. If it's a feature (and we've no access to it) it's oc. our duty to make unportable things work. Martin Gräßlin wrote: right, I thought about it and yes an xcb_flush() before going to sleep could make a few things easier. I'll see whether I can hack Qt for that next week. Martin Gräßlin wrote: investigation result from looking at Qt - not a solution. Qt uses a thread to poll the xcb connection socket, so we cannot flush before it's going to block again. change to Qt: https://codereview.qt-project.org/85654 - Martin --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118167/#review58061 --- On May 16, 2014, 3 p.m., Martin Gräßlin wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118167/ --- (Updated May 16, 2014, 3 p.m.) Review request for KDE Frameworks and kwin. Bugs: 334858 https://bugs.kde.org/show_bug.cgi?id=334858 Repository: kwindowsystem Description --- Flush xcb connection after preparing KSelectionOwner::claim Without the flush nothing starts. This caused KWin to stall during a restart. BUG: 334858 Diffs - src/kselectionowner.cpp 90a0004046135e7fee61fe2ef865b75eb421ea99 Diff: https://git.reviewboard.kde.org/r/118167/diff/ Testing --- Thanks, Martin Gräßlin ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Review Request 118197: Fix regression in NETRootInfo::sendPing
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118197/ --- Review request for KDE Frameworks, kwin and Eike Hein. Bugs: 335024 https://bugs.kde.org/show_bug.cgi?id=335024 Repository: kwindowsystem Description --- Fix regression in NETRootInfo::sendPing Client message for _NET_WM_PING has a message_type of WM_PROTOCOLS. BUG: 335024 Diffs - src/netwm.cpp a4186d6d7beca67361a2ba6fa6fd9723296eaef9 Diff: https://git.reviewboard.kde.org/r/118197/diff/ Testing --- Thanks, Martin Gräßlin ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118197: Fix regression in NETRootInfo::sendPing
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118197/#review58132 --- Ship it! Fix tested and confirmed. - Eike Hein On May 19, 2014, 7:52 a.m., Martin Gräßlin wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118197/ --- (Updated May 19, 2014, 7:52 a.m.) Review request for KDE Frameworks, kwin and Eike Hein. Bugs: 335024 https://bugs.kde.org/show_bug.cgi?id=335024 Repository: kwindowsystem Description --- Fix regression in NETRootInfo::sendPing Client message for _NET_WM_PING has a message_type of WM_PROTOCOLS. BUG: 335024 Diffs - src/netwm.cpp a4186d6d7beca67361a2ba6fa6fd9723296eaef9 Diff: https://git.reviewboard.kde.org/r/118197/diff/ Testing --- Thanks, Martin Gräßlin ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: GenerateExportHeader and clang
On 18/05/14 20:03, Ivan Čukić wrote: Ah, but what is clang's return value when there is an unrecognized option? It is 0. That is why I find the situation so strange. :) And I thought I was being clever :-) It turns out that GenerateExportHeader uses check_cxx_compiler_flag, which greps the output of the compiler for text like unknown .*option or unrecognized .*option. Given the *other* unknown option, it finds this output and concludes that -fvisibility=hidden is unsupported. This happens entirely independently of the CMAKE_CXX_VISIBILITY_PRESET and CMAKE_VISIBILITY_INLINES_HIDDEN options that actually set those flags on the compiler command line. So there are a couple of interacting issues here: check_cxx_compiler_flag is fragile in the face of unrelated unknown options, and the mechanism for deciding whether to add -fvisibility=hidden to the command line is different to the mechanism for deciding whether to put export symbols into the library. Alex ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Review Request 118200: [StatusNotifier] Don't set custom window flags on ContextMenu
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118200/ --- Review request for KDE Frameworks, Plasma and Marco Martin. Repository: knotifications Description --- [StatusNotifier] Don't set custom window flags on ContextMenu A QMenu has proper default flags with BypassWindowManagerHint. Thus all the setting of hints and force activate is not needed. Also it takes care of grabbing keyboard and pointer in this case resulting in the popup being closed when clicked outside. This fixes interaction problems with Klipper. Diffs - src/kstatusnotifieritemdbus_p.cpp 0073afacfc3c0a5b6164e0e777471e1d5ea3cb43 Diff: https://git.reviewboard.kde.org/r/118200/diff/ Testing --- Klipper properly grabs the pointer and clicking outside Klipper's context menu, ends it. Thanks, Martin Gräßlin ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118200: [StatusNotifier] Don't set custom window flags on ContextMenu
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118200/#review58139 --- looks fine. can you try it also on an item that uses dbusmenu for menus, like kmix? - Marco Martin On May 19, 2014, 10:56 a.m., Martin Gräßlin wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118200/ --- (Updated May 19, 2014, 10:56 a.m.) Review request for KDE Frameworks, Plasma and Marco Martin. Repository: knotifications Description --- [StatusNotifier] Don't set custom window flags on ContextMenu A QMenu has proper default flags with BypassWindowManagerHint. Thus all the setting of hints and force activate is not needed. Also it takes care of grabbing keyboard and pointer in this case resulting in the popup being closed when clicked outside. This fixes interaction problems with Klipper. Diffs - src/kstatusnotifieritemdbus_p.cpp 0073afacfc3c0a5b6164e0e777471e1d5ea3cb43 Diff: https://git.reviewboard.kde.org/r/118200/diff/ Testing --- Klipper properly grabs the pointer and clicking outside Klipper's context menu, ends it. Thanks, Martin Gräßlin ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118200: [StatusNotifier] Don't set custom window flags on ContextMenu
On May 19, 2014, 1:06 p.m., Marco Martin wrote: looks fine. can you try it also on an item that uses dbusmenu for menus, like kmix? I just tried kmix and it looks fine to me on both left and right click. Both menus behave like I would expect (override redirect, grabbing pointer/keyboard). - Martin --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118200/#review58139 --- On May 19, 2014, 12:56 p.m., Martin Gräßlin wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118200/ --- (Updated May 19, 2014, 12:56 p.m.) Review request for KDE Frameworks, Plasma and Marco Martin. Repository: knotifications Description --- [StatusNotifier] Don't set custom window flags on ContextMenu A QMenu has proper default flags with BypassWindowManagerHint. Thus all the setting of hints and force activate is not needed. Also it takes care of grabbing keyboard and pointer in this case resulting in the popup being closed when clicked outside. This fixes interaction problems with Klipper. Diffs - src/kstatusnotifieritemdbus_p.cpp 0073afacfc3c0a5b6164e0e777471e1d5ea3cb43 Diff: https://git.reviewboard.kde.org/r/118200/diff/ Testing --- Klipper properly grabs the pointer and clicking outside Klipper's context menu, ends it. Thanks, Martin Gräßlin ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118200: [StatusNotifier] Don't set custom window flags on ContextMenu
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118200/#review58142 --- Ship it! Ship It! - Marco Martin On May 19, 2014, 10:56 a.m., Martin Gräßlin wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118200/ --- (Updated May 19, 2014, 10:56 a.m.) Review request for KDE Frameworks, Plasma and Marco Martin. Repository: knotifications Description --- [StatusNotifier] Don't set custom window flags on ContextMenu A QMenu has proper default flags with BypassWindowManagerHint. Thus all the setting of hints and force activate is not needed. Also it takes care of grabbing keyboard and pointer in this case resulting in the popup being closed when clicked outside. This fixes interaction problems with Klipper. Diffs - src/kstatusnotifieritemdbus_p.cpp 0073afacfc3c0a5b6164e0e777471e1d5ea3cb43 Diff: https://git.reviewboard.kde.org/r/118200/diff/ Testing --- Klipper properly grabs the pointer and clicking outside Klipper's context menu, ends it. Thanks, Martin Gräßlin ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118200: [StatusNotifier] Don't set custom window flags on ContextMenu
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118200/#review58144 --- This review has been submitted with commit c5c34851dc410aa14663cf3d8d0ea14fd061d31b by Martin Gräßlin to branch master. - Commit Hook On May 19, 2014, 10:56 a.m., Martin Gräßlin wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118200/ --- (Updated May 19, 2014, 10:56 a.m.) Review request for KDE Frameworks, Plasma and Marco Martin. Repository: knotifications Description --- [StatusNotifier] Don't set custom window flags on ContextMenu A QMenu has proper default flags with BypassWindowManagerHint. Thus all the setting of hints and force activate is not needed. Also it takes care of grabbing keyboard and pointer in this case resulting in the popup being closed when clicked outside. This fixes interaction problems with Klipper. Diffs - src/kstatusnotifieritemdbus_p.cpp 0073afacfc3c0a5b6164e0e777471e1d5ea3cb43 Diff: https://git.reviewboard.kde.org/r/118200/diff/ Testing --- Klipper properly grabs the pointer and clicking outside Klipper's context menu, ends it. Thanks, Martin Gräßlin ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118197: Fix regression in NETRootInfo::sendPing
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118197/#review58151 --- Ship it! ... and is the only theoretically correct procedure as well. - Thomas Lübking On May 19, 2014, 7:52 a.m., Martin Gräßlin wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118197/ --- (Updated May 19, 2014, 7:52 a.m.) Review request for KDE Frameworks, kwin and Eike Hein. Bugs: 335024 https://bugs.kde.org/show_bug.cgi?id=335024 Repository: kwindowsystem Description --- Fix regression in NETRootInfo::sendPing Client message for _NET_WM_PING has a message_type of WM_PROTOCOLS. BUG: 335024 Diffs - src/netwm.cpp a4186d6d7beca67361a2ba6fa6fd9723296eaef9 Diff: https://git.reviewboard.kde.org/r/118197/diff/ Testing --- Thanks, Martin Gräßlin ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118102: Use all entries from a cert attribute instead of just using the first one
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118102/ --- (Updated May 19, 2014, 6:56 p.m.) Review request for KDE Frameworks. Changes --- Update as per suggestions Repository: kio Description --- Instead of just using the first value of a attribute, use all entries, seperated by a ',' Also drop KSslInfoDialog::launchConfig since it's not used internally anywhere ( atleast as far as I can tell ), not to mention that the slot won't work anyway since it calls a bogus KCM Diffs (updated) - src/widgets/ksslcertificatebox.cpp 576acd0 src/widgets/ksslinfodialog.h 5280242 src/widgets/ksslinfodialog.cpp 33001f4 Diff: https://git.reviewboard.kde.org/r/118102/diff/ Testing --- KCM SSL still works Thanks, Rohan Garg ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118116: Rename DBus interfaces to use new namespaces
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118116/ --- (Updated May 19, 2014, 6:57 p.m.) Review request for KDE Frameworks. Changes --- Some more renames, forgot to include these in the last diff. Repository: kio Description --- Rename DBus interfaces to use the new Qt namespace. Diffs (updated) - src/core/org.kde.KPasswdServer.xml fbb590e src/kpasswdserver/org.kde.KPasswdServer.xml fbb590e src/widgets/org.kde.kuiserver.xml 75ad12f Diff: https://git.reviewboard.kde.org/r/118116/diff/ Testing --- Everything still compiles, tests still pass Thanks, Rohan Garg ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118195: Look for kded modules without kded_ prefix first
---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/118195/#review58153
---
Idea sounds ok, just minor issues.
src/kded.cpp
https://git.reviewboard.kde.org/r/118195/#comment40422
!obj.contains('/') is slightly more readable
Why this test, anyway? I don't really understand the logic here.
src/kded.cpp
https://git.reviewboard.kde.org/r/118195/#comment40423
Won't that warn for every old style plugin?
It's ok to qDebug using old style naming, but a qWarning annoys the user
and the warning text makes it sound like the loading completely failed, even
though it will end up working.
I think the qWarning should be out of the loop, when module is null.
- David Faure
On May 18, 2014, 6:54 p.m., Alex Merry wrote:
---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/118195/
---
(Updated May 18, 2014, 6:54 p.m.)
Review request for KDE Frameworks and David Faure.
Repository: kded
Description
---
Look for kded modules without kded_ prefix first
The implicit kded_ prefix is weird, and contrary to what most other
kservice-based things do. It also means that kded modules cannot
(easily) be placed in a subdirectory of the plugin directory.
The old-style naming is still supported, for compatibility.
Diffs
-
src/kded.cpp 6a3c4935b81ad42324eddc756b24ca85450b8e83
Diff: https://git.reviewboard.kde.org/r/118195/diff/
Testing
---
Built, installed. Ran kded5, and made sure
`qdbus org.kde.kded5 /modules/kcookiejar`
still worked (loaded the module etc) with the old-style naming.
Performed the same test after removing the old kded_kcookiejar.so, and
installing a modified desktop file with the library set to
kf5/kded/kcookiejar, and the kcookiejar module installed as
PLUGIN_DIR/kf5/kded/kcookiejar.so (and restarting kded5, of course).
Thanks,
Alex Merry
___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118195: Look for kded modules without kded_ prefix first
On May 19, 2014, 1:41 p.m., David Faure wrote:
src/kded.cpp, line 373
https://git.reviewboard.kde.org/r/118195/diff/1/?file=273432#file273432line373
!obj.contains('/') is slightly more readable
Why this test, anyway? I don't really understand the logic here.
On the basis that there's no point looking for kded_kf5/foo.so, for example.
Although it made more sense with the code I originally wrote, which looked for
the kded_ prefix first.
- Alex
---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/118195/#review58153
---
On May 18, 2014, 6:54 p.m., Alex Merry wrote:
---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/118195/
---
(Updated May 18, 2014, 6:54 p.m.)
Review request for KDE Frameworks and David Faure.
Repository: kded
Description
---
Look for kded modules without kded_ prefix first
The implicit kded_ prefix is weird, and contrary to what most other
kservice-based things do. It also means that kded modules cannot
(easily) be placed in a subdirectory of the plugin directory.
The old-style naming is still supported, for compatibility.
Diffs
-
src/kded.cpp 6a3c4935b81ad42324eddc756b24ca85450b8e83
Diff: https://git.reviewboard.kde.org/r/118195/diff/
Testing
---
Built, installed. Ran kded5, and made sure
`qdbus org.kde.kded5 /modules/kcookiejar`
still worked (loaded the module etc) with the old-style naming.
Performed the same test after removing the old kded_kcookiejar.so, and
installing a modified desktop file with the library set to
kf5/kded/kcookiejar, and the kcookiejar module installed as
PLUGIN_DIR/kf5/kded/kcookiejar.so (and restarting kded5, of course).
Thanks,
Alex Merry
___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118195: Look for kded modules without kded_ prefix first
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118195/ --- (Updated May 19, 2014, 2:39 p.m.) Review request for KDE Frameworks and David Faure. Changes --- Address David's comments. Repository: kded Description --- Look for kded modules without kded_ prefix first The implicit kded_ prefix is weird, and contrary to what most other kservice-based things do. It also means that kded modules cannot (easily) be placed in a subdirectory of the plugin directory. The old-style naming is still supported, for compatibility. Diffs (updated) - src/kded.cpp 6a3c4935b81ad42324eddc756b24ca85450b8e83 Diff: https://git.reviewboard.kde.org/r/118195/diff/ Testing --- Built, installed. Ran kded5, and made sure `qdbus org.kde.kded5 /modules/kcookiejar` still worked (loaded the module etc) with the old-style naming. Performed the same test after removing the old kded_kcookiejar.so, and installing a modified desktop file with the library set to kf5/kded/kcookiejar, and the kcookiejar module installed as PLUGIN_DIR/kf5/kded/kcookiejar.so (and restarting kded5, of course). Thanks, Alex Merry ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Build failed in Jenkins: plasma-framework_master_qt5 » NoX11,LINBUILDER #416
See http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/416/changes Changes: [notmart] don't link agains KConfigWidgets -- [...truncated 371 lines...] [ 21%] Built target KF5Plasma_automoc [ 22%] Generating libplasma-theme-global.h, libplasma-theme-global.cpp [ 23%] Building CXX object src/platformstatus/CMakeFiles/kded_platformstatus.dir/kded_platformstatus_automoc.cpp.o [ 24%] Building CXX object src/declarativeimports/calendar/CMakeFiles/calendarplugin.dir/daysmodel.cpp.o [ 25%] Building CXX object src/declarativeimports/calendar/CMakeFiles/calendarplugin.dir/calendarplugin_automoc.cpp.o Generating moc_tooltipdialog.cpp Generating moc_corebindingsplugin.cpp Generating moc_datamodel.cpp Generating moc_datasource.cpp Generating moc_framesvgitem.cpp Generating moc_iconitem.cpp Generating moc_serviceoperationstatus.cpp Generating moc_svgitem.cpp Generating moc_tooltip.cpp Generating moc_units.cpp Generating moc_windowthumbnail.cpp [ 25%] Built target corebindingsplugin_automoc Linking CXX shared library libplatformcomponentsplugin.so Scanning dependencies of target KF5Plasma [ 26%] [ 26%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/plasma.cpp.o Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/pluginloader.cpp.o [ 26%] [ 27%] Built target platformcomponentsplugin [ 28%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/version.cpp.o [ 29%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/private/componentinstaller.cpp.o [ 29%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/applet.cpp.o Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/containment.cpp.o Linking CXX shared module kded_platformstatus.so [ 30%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/containmentactions.cpp.o Linking CXX shared library libcalendarplugin.so [ 30%] Built target kded_platformstatus [ 31%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/corona.cpp.o [ 31%] Built target calendarplugin [ 32%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/private/applet_p.cpp.o http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/applet.cpp: In member function ‘virtual void Plasma::Applet::save(KConfigGroup) const’: http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/applet.cpp:155:38: warning: ‘void KCoreConfigSkeleton::writeConfig()’ is deprecated (declared at /srv/jenkins/install/linux/x86_64/g++/kf5-qt5/frameworks/kconfig/inst/include/KF5/KConfigCore/kcoreconfigskeleton.h:1429) [-Wdeprecated-declarations] d-configLoader-writeConfig(); ^ http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/applet.cpp: In member function ‘virtual void Plasma::Applet::configChanged()’: http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/applet.cpp:652:41: warning: ‘void KCoreConfigSkeleton::readConfig()’ is deprecated (declared at /srv/jenkins/install/linux/x86_64/g++/kf5-qt5/frameworks/kconfig/inst/include/KF5/KConfigCore/kcoreconfigskeleton.h:1046) [-Wdeprecated-declarations] d-configLoader-readConfig(); ^ In file included from http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/containmentactions.cpp:24:0: http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/private/containmentactions_p.h: In constructor ‘Plasma::ContainmentActionsPrivate::ContainmentActionsPrivate(KService::Ptr, Plasma::ContainmentActions*)’: http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/private/containmentactions_p.h:33:22: warning: ‘KPluginInfo::KPluginInfo(KService::Ptr)’ is deprecated (declared at /srv/jenkins/install/linux/x86_64/g++/kf5-qt5/frameworks/kservice/inst/include/KF5/KService/kplugininfo.h:114) [-Wdeprecated-declarations] containment(0) ^ http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/containmentactions.cpp: At global scope: http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/containmentactions.cpp:49:1: warning: unused parameter ‘parentObject’ [-Wunused-parameter] ContainmentActions::ContainmentActions(QObject *parentObject, const QVariantList args) ^ [ 32%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/private/associatedapplicationmanager.cpp.o [ 33%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/private/containment_p.cpp.o [ 34%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/datacontainer.cpp.o [ 35%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/dataengine.cpp.o [ 35%] Building CXX object
Build failed in Jenkins: plasma-framework_master_qt5 » All,LINBUILDER #416
See http://build.kde.org/job/plasma-framework_master_qt5/Variation=All,label=LINBUILDER/416/changes Changes: [notmart] don't link agains KConfigWidgets -- [...truncated 370 lines...] Generating moc_effectwatcher_p.cpp [ 23%] [ 23%] Building CXX object src/declarativeimports/platformcomponents/CMakeFiles/platformcomponentsplugin.dir/platformcomponentsplugin_automoc.cpp.o Built target KF5Plasma_automoc [ 24%] Building CXX object src/platformstatus/CMakeFiles/kded_platformstatus.dir/platformstatusadaptor.cpp.o [ 24%] Building CXX object src/platformstatus/CMakeFiles/kded_platformstatus.dir/kded_platformstatus_automoc.cpp.o Generating moc_tooltipdialog.cpp Generating moc_corebindingsplugin.cpp Generating moc_datamodel.cpp Generating moc_datasource.cpp Generating moc_framesvgitem.cpp Generating moc_iconitem.cpp Generating moc_serviceoperationstatus.cpp Generating moc_svgitem.cpp Generating moc_tooltip.cpp Generating moc_units.cpp Generating moc_windowthumbnail.cpp [ 24%] Built target corebindingsplugin_automoc [ 25%] Generating libplasma-theme-global.h, libplasma-theme-global.cpp Linking CXX shared library libcalendarplugin.so Linking CXX shared library libplatformcomponentsplugin.so [ 25%] Built target calendarplugin Scanning dependencies of target KF5Plasma [ 26%] [ 27%] [ 27%] [ 28%] [ 29%] [ 29%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/plasma.cpp.o Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/private/componentinstaller.cpp.o [ 29%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/version.cpp.o Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/pluginloader.cpp.o Built target platformcomponentsplugin Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/applet.cpp.o Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/containment.cpp.o [ 30%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/containmentactions.cpp.o [ 31%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/corona.cpp.o Linking CXX shared module kded_platformstatus.so [ 31%] [ 32%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/private/applet_p.cpp.o Built target kded_platformstatus [ 32%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/private/associatedapplicationmanager.cpp.o In file included from http://build.kde.org/job/plasma-framework_master_qt5/Variation=All,label=LINBUILDER/ws/src/plasma/containmentactions.cpp:24:0: http://build.kde.org/job/plasma-framework_master_qt5/Variation=All,label=LINBUILDER/ws/src/plasma/private/containmentactions_p.h: In constructor ‘Plasma::ContainmentActionsPrivate::ContainmentActionsPrivate(KService::Ptr, Plasma::ContainmentActions*)’: http://build.kde.org/job/plasma-framework_master_qt5/Variation=All,label=LINBUILDER/ws/src/plasma/private/containmentactions_p.h:33:22: warning: ‘KPluginInfo::KPluginInfo(KService::Ptr)’ is deprecated (declared at /srv/jenkins/install/linux/x86_64/g++/kf5-qt5/frameworks/kservice/inst/include/KF5/KService/kplugininfo.h:114) [-Wdeprecated-declarations] containment(0) ^ http://build.kde.org/job/plasma-framework_master_qt5/Variation=All,label=LINBUILDER/ws/src/plasma/applet.cpp: In member function ‘virtual void Plasma::Applet::save(KConfigGroup) const’: http://build.kde.org/job/plasma-framework_master_qt5/Variation=All,label=LINBUILDER/ws/src/plasma/applet.cpp:155:38: warning: ‘void KCoreConfigSkeleton::writeConfig()’ is deprecated (declared at /srv/jenkins/install/linux/x86_64/g++/kf5-qt5/frameworks/kconfig/inst/include/KF5/KConfigCore/kcoreconfigskeleton.h:1429) [-Wdeprecated-declarations] d-configLoader-writeConfig(); ^ http://build.kde.org/job/plasma-framework_master_qt5/Variation=All,label=LINBUILDER/ws/src/plasma/applet.cpp: In member function ‘virtual void Plasma::Applet::configChanged()’: http://build.kde.org/job/plasma-framework_master_qt5/Variation=All,label=LINBUILDER/ws/src/plasma/applet.cpp:652:41: warning: ‘void KCoreConfigSkeleton::readConfig()’ is deprecated (declared at /srv/jenkins/install/linux/x86_64/g++/kf5-qt5/frameworks/kconfig/inst/include/KF5/KConfigCore/kcoreconfigskeleton.h:1046) [-Wdeprecated-declarations] d-configLoader-readConfig(); ^ [ 33%] http://build.kde.org/job/plasma-framework_master_qt5/Variation=All,label=LINBUILDER/ws/src/plasma/containmentactions.cpp: At global scope: http://build.kde.org/job/plasma-framework_master_qt5/Variation=All,label=LINBUILDER/ws/src/plasma/containmentactions.cpp:49:1: warning: unused parameter ‘parentObject’ [-Wunused-parameter] ContainmentActions::ContainmentActions(QObject *parentObject, const QVariantList args) ^ Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/private/containment_p.cpp.o [ 34%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/datacontainer.cpp.o
Re: Review Request 118195: Look for kded modules without kded_ prefix first
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118195/#review58164 --- Ship it! Perfect :) - David Faure On May 19, 2014, 2:39 p.m., Alex Merry wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118195/ --- (Updated May 19, 2014, 2:39 p.m.) Review request for KDE Frameworks and David Faure. Repository: kded Description --- Look for kded modules without kded_ prefix first The implicit kded_ prefix is weird, and contrary to what most other kservice-based things do. It also means that kded modules cannot (easily) be placed in a subdirectory of the plugin directory. The old-style naming is still supported, for compatibility. Diffs - src/kded.cpp 6a3c4935b81ad42324eddc756b24ca85450b8e83 Diff: https://git.reviewboard.kde.org/r/118195/diff/ Testing --- Built, installed. Ran kded5, and made sure `qdbus org.kde.kded5 /modules/kcookiejar` still worked (loaded the module etc) with the old-style naming. Performed the same test after removing the old kded_kcookiejar.so, and installing a modified desktop file with the library set to kf5/kded/kcookiejar, and the kcookiejar module installed as PLUGIN_DIR/kf5/kded/kcookiejar.so (and restarting kded5, of course). Thanks, Alex Merry ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 117695: change where dynamic replace tabs is performed
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/117695/#review58165 --- Looks good to me. Still, Christoph, since it touches the core editing parts, could you have a quick look? - Dominik Haumann On May 3, 2014, 12:57 a.m., Sven Brauch wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/117695/ --- (Updated May 3, 2014, 12:57 a.m.) Review request for KDE Frameworks and Christoph Cullmann. Repository: ktexteditor Description --- This makes typeChars handle replacing tabs by spaces, instead of insertText. The rationale is that insertText is often called programatically, and the caller should be able to rely on the text he requests to be inserted is actually inserted, and not changed on-the-fly. Examples for where the previous solution caused problems are KDevelop (the codegen) and kte-collaborative. I'm not sure what the code I removed was doing (heh). It looks like it is supposed to advance to the next indent level if the current indent level is odd, but that still works after removing it. The obvious user-visible change here is that tabs in pasted text will no longer be replaced. But since I always found this behaviour undesirable anyways, I did not bother to replicate it. I will instead wait for people to yell at me for removing it. ;) Diffs - autotests/src/katedocument_test.h cff61d0 autotests/src/katedocument_test.cpp f3eef18 src/document/katedocument.h 83cc031 src/document/katedocument.cpp 546d3e6 Diff: https://git.reviewboard.kde.org/r/117695/diff/ Testing --- Just some quick manual tests, it seems to still work as intended. Thanks, Sven Brauch ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Build failed in Jenkins: plasma-framework_master_qt5 » NoX11,LINBUILDER #417
See http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/417/changes Changes: [notmart] it's a QGuiApplication -- [...truncated 351 lines...] Generating moc_packagestructure.cpp Generating moc_associatedapplicationmanager_p.cpp Generating moc_datacontainer_p.cpp Generating private/moc_dataengineconsumer_p.cpp Generating moc_dataenginemanager_p.cpp Generating moc_packagejob_p.cpp Generating moc_packagejobthread_p.cpp Generating moc_storage_p.cpp Generating moc_storagethread_p.cpp Generating moc_theme_p.cpp Generating moc_appletscript.cpp Generating moc_dataenginescript.cpp Generating moc_scriptengine.cpp Generating moc_service.cpp Generating moc_servicejob.cpp Generating moc_svg.cpp Generating moc_theme.cpp Generating moc_plasma.cpp [ 22%] Built target KF5Plasma_automoc [ 23%] Building CXX object src/declarativeimports/calendar/CMakeFiles/calendarplugin.dir/calendarplugin_automoc.cpp.o [ 23%] Building CXX object src/platformstatus/CMakeFiles/kded_platformstatus.dir/platformstatusadaptor.cpp.o [ 24%] Building CXX object src/platformstatus/CMakeFiles/kded_platformstatus.dir/kded_platformstatus_automoc.cpp.o Generating moc_tooltipdialog.cpp Generating moc_corebindingsplugin.cpp Generating moc_datamodel.cpp Generating moc_datasource.cpp Generating moc_framesvgitem.cpp Generating moc_iconitem.cpp Generating moc_serviceoperationstatus.cpp Generating moc_svgitem.cpp Generating moc_tooltip.cpp Generating moc_units.cpp Generating moc_windowthumbnail.cpp [ 24%] Built target corebindingsplugin_automoc [ 25%] Generating libplasma-theme-global.h, libplasma-theme-global.cpp Linking CXX shared library libplatformcomponentsplugin.so Scanning dependencies of target KF5Plasma Linking CXX shared library libcalendarplugin.so [ 26%] [ 26%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/plasma.cpp.o [ 26%] [ 27%] Built target platformcomponentsplugin Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/pluginloader.cpp.o [ 28%] [ 29%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/version.cpp.o [ 29%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/private/componentinstaller.cpp.o Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/applet.cpp.o Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/containment.cpp.o [ 30%] [ 30%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/containmentactions.cpp.o Built target calendarplugin [ 31%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/corona.cpp.o [ 32%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/private/applet_p.cpp.o Linking CXX shared module kded_platformstatus.so [ 32%] Built target kded_platformstatus [ 32%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/private/associatedapplicationmanager.cpp.o [ 33%] Building CXX object src/plasma/CMakeFiles/KF5Plasma.dir/private/containment_p.cpp.o http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/pluginloader.cpp: In member function ‘Plasma::DataEngine* Plasma::PluginLoader::loadDataEngine(const QString)’: http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/pluginloader.cpp:256:63: warning: ‘KPluginInfo::KPluginInfo(KService::Ptr)’ is deprecated (declared at /srv/jenkins/install/linux/x86_64/g++/kf5-qt5/frameworks/kservice/inst/include/KF5/KService/kplugininfo.h:114) [-Wdeprecated-declarations] engine = new DataEngine(KPluginInfo(offers.first()), 0); ^ http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/pluginloader.cpp: In static member function ‘static KPluginInfo::List Plasma::PluginLoader::listEngineInfoByCategory(const QString, const QString)’: http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/pluginloader.cpp:308:44: warning: ‘static KPluginInfo::List KPluginInfo::fromServices(const List, const KConfigGroup)’ is deprecated (declared at /srv/jenkins/install/linux/x86_64/g++/kf5-qt5/frameworks/kservice/inst/include/KF5/KService/kplugininfo.h:181) [-Wdeprecated-declarations] return KPluginInfo::fromServices(offers); ^ http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/pluginloader.cpp: In member function ‘KPluginInfo::List Plasma::PluginLoader::listAppletInfo(const QString, const QString)’: http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/ws/src/plasma/pluginloader.cpp:506:44: warning: ‘static KPluginInfo::List KPluginInfo::fromServices(const List, const KConfigGroup)’ is deprecated (declared at /srv/jenkins/install/linux/x86_64/g++/kf5-qt5/frameworks/kservice/inst/include/KF5/KService/kplugininfo.h:181) [-Wdeprecated-declarations] return
Jenkins build is back to normal : plasma-framework_master_qt5 » NoX11,LINBUILDER #418
See http://build.kde.org/job/plasma-framework_master_qt5/Variation=NoX11,label=LINBUILDER/418/changes ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Jenkins build is back to normal : plasma-framework_master_qt5 » All,LINBUILDER #418
See http://build.kde.org/job/plasma-framework_master_qt5/Variation=All,label=LINBUILDER/418/changes ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118195: Look for kded modules without kded_ prefix first
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118195/ --- (Updated May 19, 2014, 5:02 p.m.) Status -- This change has been marked as submitted. Review request for KDE Frameworks and David Faure. Repository: kded Description --- Look for kded modules without kded_ prefix first The implicit kded_ prefix is weird, and contrary to what most other kservice-based things do. It also means that kded modules cannot (easily) be placed in a subdirectory of the plugin directory. The old-style naming is still supported, for compatibility. Diffs - src/kded.cpp 6a3c4935b81ad42324eddc756b24ca85450b8e83 Diff: https://git.reviewboard.kde.org/r/118195/diff/ Testing --- Built, installed. Ran kded5, and made sure `qdbus org.kde.kded5 /modules/kcookiejar` still worked (loaded the module etc) with the old-style naming. Performed the same test after removing the old kded_kcookiejar.so, and installing a modified desktop file with the library set to kf5/kded/kcookiejar, and the kcookiejar module installed as PLUGIN_DIR/kf5/kded/kcookiejar.so (and restarting kded5, of course). Thanks, Alex Merry ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118195: Look for kded modules without kded_ prefix first
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118195/#review58167 --- This review has been submitted with commit 17977dcd85442a4cf8f59593ab486ba813cbd0df by Alex Merry to branch master. - Commit Hook On May 19, 2014, 2:39 p.m., Alex Merry wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118195/ --- (Updated May 19, 2014, 2:39 p.m.) Review request for KDE Frameworks and David Faure. Repository: kded Description --- Look for kded modules without kded_ prefix first The implicit kded_ prefix is weird, and contrary to what most other kservice-based things do. It also means that kded modules cannot (easily) be placed in a subdirectory of the plugin directory. The old-style naming is still supported, for compatibility. Diffs - src/kded.cpp 6a3c4935b81ad42324eddc756b24ca85450b8e83 Diff: https://git.reviewboard.kde.org/r/118195/diff/ Testing --- Built, installed. Ran kded5, and made sure `qdbus org.kde.kded5 /modules/kcookiejar` still worked (loaded the module etc) with the old-style naming. Performed the same test after removing the old kded_kcookiejar.so, and installing a modified desktop file with the library set to kf5/kded/kcookiejar, and the kcookiejar module installed as PLUGIN_DIR/kf5/kded/kcookiejar.so (and restarting kded5, of course). Thanks, Alex Merry ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118192: Do not set the INSTALL_RPATH_USE_LINK_PATH and INSTALL_RPATH properties
---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/118192/
---
(Updated May 19, 2014, 5:06 p.m.)
Review request for KDE Frameworks and Alex Merry.
Changes
---
The generated plugins cannot be built without Qt5Designer's include directories
being set, so deal with that in the kf5designerplugin_add_plugin macro.
kf5designerplugin_generate_plugin now secretly takes a TARGET argument,
allowing those same include directories to be used when generating the moc part.
Summary (updated)
-
Do not set the INSTALL_RPATH_USE_LINK_PATH and INSTALL_RPATH properties
Repository: kdesignerplugin
Description
---
Do not set the INSTALL_RPATH_USE_LINK_PATH and INSTALL_RPATH properties
INSTALL_RPATH_USE_LINK_PATH already defaults to TRUE, and INSTALL_RPATH
already has a sensible default, because of the KDECMakeSettings module
from ECM.
This also adds a comment explaining why the other RPath properties are
set.
Replace the installed macros
The existing macro (kf5designerplugin_add_widget_files) did not work (it
tried to iterate of the name of the sources variable instead of
${ARGN}), and the API didn't really make sense anyway (you cannot
combine multiple plugin instance classes into a single plugin library,
so accepting multiple .widgets files does not make sense).
kf5designerplugin_add_widget_files is therefore replaced with
kf5designerplugin_generate_plugin (which does the same, but for a single
widget file), and we also have kf5designerplugin_add_plugin, which
replaces the usual add_library call, and automagically deals with
.widgets files and .qrc files it is passed.
The two designer plugins provided by KDesignerPlugin are then generated
using kf5designerplugin_add_plugin, so we can be sure it actually works.
Diffs (updated)
-
CMakeLists.txt c90208bc13a13b6dc8824c86ee102c5b42369331
KF5DesignerPluginMacros.cmake 43fde97f0e07a9d00d79f92bee1888bf8fcbf70c
src/CMakeLists.txt 2e7b9f0aa3a2ba5adc388961919a635b8ec6a2c2
Diff: https://git.reviewboard.kde.org/r/118192/diff/
Testing
---
Builds, installs, plugins seem to work in designer. I'm planning to write some
autotests as well.
Thanks,
Alex Merry
___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Loading Qt 5 translations
Hi Translations for frameworks are loaded automatically when they are used, but what about Qt 5 itself? Are those translations loaded and if not, how should they be loaded? From testing using German locale it looks like Qt 5 translations are not loaded. Most generic buttons like Ok, Close, Cancel and so on appear as untranslated. Looked at KWrite, Konsole and System Settings. Using same locale settings (LC_ALL and LANG set to de_DE.UTF-8 and LANGUAGE=de) Qt Designer appeared fully in German. Cheers, Lasse ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Framework Localization Policy
El Divendres, 16 de maig de 2014, a les 01:04:48, Aurélien Gâteau va escriure: Hi, Finally took the time to document framework localization. You can find the documentation here: http://community.kde.org/Frameworks/Frameworks_Localization_Policy . Feedback is welcome. Looks good to me. Cheers, Albert Aurélien ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 117695: change where dynamic replace tabs is performed
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/117695/#review58172 --- Ship it! I think that is a better way to handle it than the old code was :) insertText should not modify the text more than needed. - Christoph Cullmann On May 3, 2014, 12:57 a.m., Sven Brauch wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/117695/ --- (Updated May 3, 2014, 12:57 a.m.) Review request for KDE Frameworks and Christoph Cullmann. Repository: ktexteditor Description --- This makes typeChars handle replacing tabs by spaces, instead of insertText. The rationale is that insertText is often called programatically, and the caller should be able to rely on the text he requests to be inserted is actually inserted, and not changed on-the-fly. Examples for where the previous solution caused problems are KDevelop (the codegen) and kte-collaborative. I'm not sure what the code I removed was doing (heh). It looks like it is supposed to advance to the next indent level if the current indent level is odd, but that still works after removing it. The obvious user-visible change here is that tabs in pasted text will no longer be replaced. But since I always found this behaviour undesirable anyways, I did not bother to replicate it. I will instead wait for people to yell at me for removing it. ;) Diffs - autotests/src/katedocument_test.h cff61d0 autotests/src/katedocument_test.cpp f3eef18 src/document/katedocument.h 83cc031 src/document/katedocument.cpp 546d3e6 Diff: https://git.reviewboard.kde.org/r/117695/diff/ Testing --- Just some quick manual tests, it seems to still work as intended. Thanks, Sven Brauch ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118192: Fix/replace CMake macros provided by kdesignerplugin
---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/118192/
---
(Updated May 19, 2014, 9:57 p.m.)
Review request for KDE Frameworks and Alex Merry.
Changes
---
Fix the title of the RR.
Summary (updated)
-
Fix/replace CMake macros provided by kdesignerplugin
Repository: kdesignerplugin
Description
---
Do not set the INSTALL_RPATH_USE_LINK_PATH and INSTALL_RPATH properties
INSTALL_RPATH_USE_LINK_PATH already defaults to TRUE, and INSTALL_RPATH
already has a sensible default, because of the KDECMakeSettings module
from ECM.
This also adds a comment explaining why the other RPath properties are
set.
Replace the installed macros
The existing macro (kf5designerplugin_add_widget_files) did not work (it
tried to iterate of the name of the sources variable instead of
${ARGN}), and the API didn't really make sense anyway (you cannot
combine multiple plugin instance classes into a single plugin library,
so accepting multiple .widgets files does not make sense).
kf5designerplugin_add_widget_files is therefore replaced with
kf5designerplugin_generate_plugin (which does the same, but for a single
widget file), and we also have kf5designerplugin_add_plugin, which
replaces the usual add_library call, and automagically deals with
.widgets files and .qrc files it is passed.
The two designer plugins provided by KDesignerPlugin are then generated
using kf5designerplugin_add_plugin, so we can be sure it actually works.
Diffs
-
CMakeLists.txt c90208bc13a13b6dc8824c86ee102c5b42369331
KF5DesignerPluginMacros.cmake 43fde97f0e07a9d00d79f92bee1888bf8fcbf70c
src/CMakeLists.txt 2e7b9f0aa3a2ba5adc388961919a635b8ec6a2c2
Diff: https://git.reviewboard.kde.org/r/118192/diff/
Testing
---
Builds, installs, plugins seem to work in designer. I'm planning to write some
autotests as well.
Thanks,
Alex Merry
___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Review Request 118211: Add an autotest for QDesigner plugin generation
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118211/ --- Review request for KDE Frameworks. Repository: kdesignerplugin Description --- Add an autotest for QDesigner plugin generation Diffs - CMakeLists.txt c90208bc13a13b6dc8824c86ee102c5b42369331 autotests/CMakeLists.txt PRE-CREATION autotests/plugintest.cpp PRE-CREATION autotests/qpushbuttonview.png PRE-CREATION autotests/sth.png PRE-CREATION autotests/testplugin.qrc PRE-CREATION autotests/testplugin.widgets PRE-CREATION Diff: https://git.reviewboard.kde.org/r/118211/diff/ Testing --- The test passes. Thanks, Alex Merry ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 117016: Allow overriding DrKonqi lookup directories by PATH
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/117016/ --- (Updated May 20, 2014, 12:26 a.m.) Review request for KDE Frameworks. Changes --- Updated patch with LIBEXEC_PATH, as David suggested. I'm terribly sorry, I forgot about this completely. I hope it's not too late. Repository: kcrash Description --- Since KCrash is a framework that relies on DrKonqi binary being provided by a 3rd party software (kde-runtime), it should not make assumptions regarding location of the utility. This patch makes KCrash to look for drkonqi binary first in $PATH, then falling back to CMAKE_INSTALL_PREFIX/LIBEXEC_INSTALL_DIR. With this patch it's possible for distributions to ship KDE Frameworks in normal prefix (/usr), but have current snapshots of kde-runtime in /opt/kde5 for instance. Diffs (updated) - src/kcrash.cpp 87163cc Diff: https://git.reviewboard.kde.org/r/117016/diff/ Testing --- - Installed KCrash into /usr prefix - Installed drkonqi from kde-runtime master to /opt/kde5 prefix - started broken application - no could not find drkonqi warning anymore - crashed application, got drkonqi window Thanks, Dan Vrátil ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 117695: change where dynamic replace tabs is performed
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/117695/#review58180 --- This review has been submitted with commit f9d33a959dd03cd4f335532942e910119b44c8d6 by Sven Brauch to branch master. - Commit Hook On May 3, 2014, 12:57 a.m., Sven Brauch wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/117695/ --- (Updated May 3, 2014, 12:57 a.m.) Review request for KDE Frameworks and Christoph Cullmann. Repository: ktexteditor Description --- This makes typeChars handle replacing tabs by spaces, instead of insertText. The rationale is that insertText is often called programatically, and the caller should be able to rely on the text he requests to be inserted is actually inserted, and not changed on-the-fly. Examples for where the previous solution caused problems are KDevelop (the codegen) and kte-collaborative. I'm not sure what the code I removed was doing (heh). It looks like it is supposed to advance to the next indent level if the current indent level is odd, but that still works after removing it. The obvious user-visible change here is that tabs in pasted text will no longer be replaced. But since I always found this behaviour undesirable anyways, I did not bother to replicate it. I will instead wait for people to yell at me for removing it. ;) Diffs - autotests/src/katedocument_test.h cff61d0 autotests/src/katedocument_test.cpp f3eef18 src/document/katedocument.h 83cc031 src/document/katedocument.cpp 546d3e6 Diff: https://git.reviewboard.kde.org/r/117695/diff/ Testing --- Just some quick manual tests, it seems to still work as intended. Thanks, Sven Brauch ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 117695: change where dynamic replace tabs is performed
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/117695/ --- (Updated May 19, 2014, 10:36 p.m.) Status -- This change has been marked as submitted. Review request for KDE Frameworks and Christoph Cullmann. Repository: ktexteditor Description --- This makes typeChars handle replacing tabs by spaces, instead of insertText. The rationale is that insertText is often called programatically, and the caller should be able to rely on the text he requests to be inserted is actually inserted, and not changed on-the-fly. Examples for where the previous solution caused problems are KDevelop (the codegen) and kte-collaborative. I'm not sure what the code I removed was doing (heh). It looks like it is supposed to advance to the next indent level if the current indent level is odd, but that still works after removing it. The obvious user-visible change here is that tabs in pasted text will no longer be replaced. But since I always found this behaviour undesirable anyways, I did not bother to replicate it. I will instead wait for people to yell at me for removing it. ;) Diffs - autotests/src/katedocument_test.h cff61d0 autotests/src/katedocument_test.cpp f3eef18 src/document/katedocument.h 83cc031 src/document/katedocument.cpp 546d3e6 Diff: https://git.reviewboard.kde.org/r/117695/diff/ Testing --- Just some quick manual tests, it seems to still work as intended. Thanks, Sven Brauch ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118197: Fix regression in NETRootInfo::sendPing
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118197/ --- (Updated May 20, 2014, 5:27 a.m.) Status -- This change has been marked as submitted. Review request for KDE Frameworks, kwin and Eike Hein. Bugs: 335024 https://bugs.kde.org/show_bug.cgi?id=335024 Repository: kwindowsystem Description --- Fix regression in NETRootInfo::sendPing Client message for _NET_WM_PING has a message_type of WM_PROTOCOLS. BUG: 335024 Diffs - src/netwm.cpp a4186d6d7beca67361a2ba6fa6fd9723296eaef9 Diff: https://git.reviewboard.kde.org/r/118197/diff/ Testing --- Thanks, Martin Gräßlin ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Re: Review Request 118197: Fix regression in NETRootInfo::sendPing
--- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118197/#review58184 --- This review has been submitted with commit eb4f296767e5b3f17a4b9f8db33acc66581e3fe3 by Martin Gräßlin to branch master. - Commit Hook On May 19, 2014, 7:52 a.m., Martin Gräßlin wrote: --- This is an automatically generated e-mail. To reply, visit: https://git.reviewboard.kde.org/r/118197/ --- (Updated May 19, 2014, 7:52 a.m.) Review request for KDE Frameworks, kwin and Eike Hein. Bugs: 335024 https://bugs.kde.org/show_bug.cgi?id=335024 Repository: kwindowsystem Description --- Fix regression in NETRootInfo::sendPing Client message for _NET_WM_PING has a message_type of WM_PROTOCOLS. BUG: 335024 Diffs - src/netwm.cpp a4186d6d7beca67361a2ba6fa6fd9723296eaef9 Diff: https://git.reviewboard.kde.org/r/118197/diff/ Testing --- Thanks, Martin Gräßlin ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
Jenkins build became unstable: kwindowsystem_master_qt5 » All,LINBUILDER #65
See http://build.kde.org/job/kwindowsystem_master_qt5/Variation=All,label=LINBUILDER/65/changes ___ Kde-frameworks-devel mailing list Kde-frameworks-devel@kde.org https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
