| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
| |
Task-number: QT-3292
Reviewed-by: Lars Knoll
|
|
|
|
|
|
|
| |
q->inherits("QGLWidget") gives incorrect result,
so use QWidgetPrivate::isGLWidget variable instead.
Reviewed-by: Jason Barron
|
|
|
|
|
| |
Task-number: QT-2139
Reviewed-by: Gunnar Sletta
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously, the following sequence:
1. Widget is hidden
2. Widget partially revealed
3. Widget fully revealed
resulted in the reference count of the backing store owned by the
widget's window() being incremented twice.
This patch replaces the simple reference count with a QSet which
stores pointers to the native widgets which are descendents of
the backing store owner, and which are currently visible. The
sequence above therefore results in just a single insertion at
step (2), with step (3) having no effect on the backing store.
The QRefCountedWidgetBackingStore class has been renamed
QWidgetBackingStoreTracker to better reflect its purpose.
Task-number: QTBUG-12800
Task-number: QTBUG-12817
Reviewed-by: Jason Barron
|
|
|
|
|
|
|
|
| |
This fixes a crash when a static QWidget is destroyed after the QApplication
object.
Task-number: QTBUG-12025
Reviewed-by: Zeno Albisser
|
|
|
|
|
| |
Merge-request: 734
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Avkon dependencies can be configured out with the -no-s60 configure
flag, conversely Qt on Symbian will use Avkon if -s60 is configured.
These changes are intended to keep or introduce binary compatibility
between the s60 and no-s60 configurations. To do this, it has been
necessary to introduce stub equivalents of the CAknAppUi related
classes into the no-s60 configuration, and override all Avkon framework
virtual functions in the QS60Main... classes.
Other than that, these changes are mostly just correcting the use of
the Q_WS_S60 flag so that it only refers to Avkon dependencies.
Reviewed-by: Sami Merila
|
|\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1: (55 commits)
Fix symbian-abld build failure with bearer plugins
remove certificate bundle
make QSslSocket::systemCaCertificates() use system certs
Adjusted RegExp in QSslCertificate::fromPath()
Fix QSystemTrayIcon::supportsMessages() on Windows
Use NIM_SETVERSION to get the latest behavior
Cleanup obsolete stuff
doc: Added DITA XML generator
Fixed a broken merge.
Changed the way we detect touch screen on Windows.
Fixed error deploying qsymbianbearer.qtplugin on Symbian.
Revert "Attempt to fix build failure on Symbian."
Attempt to fix build failure on Symbian.
Replaced redundant "!contains" scopes in "contains(A, B) {...} !contains(A, B) {...}" constructions with "else"
Updated leading whitespace to make indentation more consistent
Updated project files so it is now possible to use "-system-zlib" configuration option on Windows and Symbian platforms.
make postgresql support sensitive to PSQL_LIBS
support for cross building Qt for MinGW (win32-g++) on Unix
use "win32-g++*" scope to match all MinGW makespecs
Demos: Fix compilation with namespace.
...
|
| |\ |
|
| | |\
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1: (29 commits)
I10n: Update German translations for 4.7.0
Fix a freetype link failure.
Fix configure test for DirectFB
Split Symbian bearer plugin into three platform specfic plugins
Doc: Fixed whitespace in the other configuration file for zh_CN.
Doc: Fixed whitespace in the Simplified Chinese doc configuration.
Unbreak Linux build when qendian.h is included before qglobal.h
Revert accidental commit of irrelevant stuff. Silly mondays..
Unbreak Linux build when qendian.h is included before qglobal.h
Fix QApplication/QWidget to really take ownership of input contexts
Added setting a hotspot on standard gestures
Fixed GestureOverride event delivery in GraphicsView.
add check-ts target to auto-asses translation completeness
qt_ja_JP.ts => qt_ja.ts
Fix incorrect \since tag
XQuery test suite (and others): remove p4 dependency
add docu about pixmaps being invalidated on qapp destruction
remove somewhat misleading warning about x11 pixmap leak
Disabled item view items use incorrect background color
Update the mkspec for linux-icc: don't use jump tables in shlibs
...
|
| | | |\
| | | | |
| | | | |
| | | | | |
4.7-staging1
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
* src/gui/kernel/[qapplication|qwidget].cpp,
tests/auto/qapplication/tst_[qapplication|qwidget].cpp (setInputContext):
The documentation for [QApplication|QWidget]::setInputContext claims that the
[QApplication|QWidget] instance would take ownership. This commit fixes the
setter to also reparent the input context. Furthermore, the crappy test for
this setter was improved.
Reviewed-by: Denis Dzyubenko <denis.dzyubenko@nokia.com>
|
| | |/ /
| |/| |
| | | |
| | | | |
Task-number: QTBUG-11391
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Problem was that we didn't take the children's effective mask
into account when checking whether a point was inside or not.
This commit is also an optimization since we no longer check the point
against the widget's rect twice. Furthermore, unnecessary
QRegion operations are completely avoided in the common case.
Auto tests included.
Task-number: QTBUG-7150
Reviewed-by: paul
|
| |/ /
|/| |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Setting it to LayoutAuto will be ignored on an
application level (only LeftToRight or RightToLeft
are valid).
On a widget level is does the same as calling unsetLayoutDirection().
Reviewed-by: Simon Hausmann
|
|\ \ \
| |_|/
|/| |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Conflicts:
src/3rdparty/webkit/VERSION
src/3rdparty/webkit/WebCore/ChangeLog
src/3rdparty/webkit/WebKit/qt/Api/qwebpage.cpp
src/3rdparty/webkit/WebKit/qt/ChangeLog
src/gui/painting/qpainter.cpp
src/gui/painting/qtextureglyphcache.cpp
src/opengl/gl2paintengineex/qpaintengineex_opengl2.cpp
src/opengl/gl2paintengineex/qtextureglyphcache_gl_p.h
src/s60installs/bwins/QtGuiu.def
src/s60installs/eabi/QtCoreu.def
src/s60installs/eabi/QtGuiu.def
src/s60installs/eabi/QtNetworku.def
src/s60installs/eabi/QtOpenVGu.def
tests/auto/qfontmetrics/tst_qfontmetrics.cpp
tools/linguist/lupdate/main.cpp
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
On Symbian, the top-level widget's backing store must be destroyed when
it is no longer required, in order to conserve memory. The criteria
for destroying the backing store is when neither the TLW nor any of
its native descendents (which share the backing store) are visible.
In order to implement this requirement, a count must be kept of the
number of native widgets which are using the TLW's backing store.
This patch provides the mechanism for maintaining this count, and for
destroying the backing store when the count is decremented to zero.
No calls to either the increment nor decrement functions are made,
however, by this code included in this patch; this code will be added
to only the Symbian backend by a subsequent patch.
Task-number: QTBUG-8697
Reviewed-by: Bjørn Erik Nilsen
Reviewed-by: Jason Barron
|
| |/
|/|
| |
| |
| | |
Merge-request: 535
Reviewed-by: Andreas Aardal Hanssen <andreas.aardal.hanssen@nokia.com>
|
|\ \
| | |
| | |
| | |
| | | |
Conflicts:
src/plugins/bearer/symbian/qnetworksession_impl.h
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Added a guard check to return if the given input context is the same is we
already have.
Explicitly mention in the doc that we take ownership of the given input context
object.
Task-number: QTBUG-10780
Reviewed-by: Thomas Zander
|
| | | |
|
|\ \ \
| |/ / |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The problem was that the fix for bug QTBUG-10519 introduced a new state
for saving the geometry. However there is an exception and that
exception comes into play when saving a maximized window. This fixes
that by making sure we only enter into that state when the window is not
currently maximized.
Reviewed-by: tbastian
|
| |\ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Mac: restoreGeometry()
[Regression] Build failure on Mac Carbon
doc: Added some \briefs to how-to docs.
Some 4.7 changes.
Updated changelog.
Cherry pick fix for MOBILITY-932 from Qt Mobility.
Cherry pick fix for MOBILITY-962 from Qt Mobility.
Cherry pick fix for MOBILITY-828 from Qt Mobility.
Bearer management changes from Qt Mobility.
Cherry pick fix for MOBILITY-828 from Qt Mobility.
Added snippet labels to QML Dial example.
Fixed documentation typo.
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
The problem here is the fact that when zooming (maximizing) a window,
we should update the starting position since the window is now starting
at 0,0. However when restoring the window should be placed in the
original position. The bug here occurs because we kept the original
position and the new size. So the window was drawn were it was using
the new size. This patch introduces a boolean flag that keeps track of
the maximized state. We are interested in knowing if the window was
maximized but not restored, so when saving the geometry we have a way
to know that we need to "move" the window to 0,0.
Task-number: QTBUG-10504
Reviewed-by: Richard Moe Gustavsen
|
| |\ \ \
| | |/ /
| |/| |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-2: (35 commits)
Avoid running animation when loopCount == 0
Set the pictures for the examples of multitouch
Compile fix on Windows.
Remove an unnecessary assert.
Documentation for the Pad Navigator Example.
New Pad Nagivator example implementation.
Skip definition of wintab functions in case of QT_NO_TABLETEVENT.
Made curve tesselation be dynamically adjusted based on transform.
Modified QPainter and QPixmap benchmarks to use raster pixmaps.
Stabilize tst_QWidgetAction::visibilityUpdate
Fix compiler warning in QT_REQUIRE_VERSION
Stabilize tst_QColumnView::parentCurrentIndex
Really fix tst_QDockWidget::taskQTBUG_9758_undockedGeometry on Linux
Fix typos in Elastic Nodes example documentation.
Made paint engine texture drawing work in GL ES 2 and updated docs.
Opt out of visual-config size checks with extension
Fix off-by-one in text layouts and widget size hints on Mac
Stabilize tst_QDockWidget::taskQTBUG_9758_undockedGeometry
Mark QFileDialog::Options as a Q_FLAGS
fix tst_QDockWidget::taskQTBUG_9758_undockedGeometry on Linux
...
|
| | |\ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-2: (24 commits)
Stabilize tst_QColumnView::parentCurrentIndex
Really fix tst_QDockWidget::taskQTBUG_9758_undockedGeometry on Linux
Fix typos in Elastic Nodes example documentation.
Made paint engine texture drawing work in GL ES 2 and updated docs.
Opt out of visual-config size checks with extension
Fix off-by-one in text layouts and widget size hints on Mac
Stabilize tst_QDockWidget::taskQTBUG_9758_undockedGeometry
Mark QFileDialog::Options as a Q_FLAGS
fix tst_QDockWidget::taskQTBUG_9758_undockedGeometry on Linux
Fixed scrolling bugs in widget graphics effect backend.
Fixed source pixmap bug in widget graphics effect backend.
Fix crash with stylesheet if widget change style in the changeEvent
Fix textdrawing under GL on N900.
Fixed bug in QIODevice::read after first reading 0 bytes.
Fix auto-test failure on Mac/Linux/QWS.
Some EGL implementations does not return a EGLNativeDisplayType
Fixed the sizing of the dock area with fixed size dock widgets
QSortFilterProxyModel: Warning or assert failure
Performance issue with QGraphicsItem::ItemClipsChildrenToShape.
Fixes crash in QGraphicsItem::mouseMove for untransformable items.
...
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The cache wasn't invalidated for widgets with a graphics effect inside a
scrolling QScrollArea, so for now we disable caching for widget source
pixmaps. Also, we can't clip the source pixmap to device coordinates
since there's no knowledge of which areas of the source pixmap the
effect uses to compute the visible pixels. See change fd30cc9fabe6fc023
for the graphics item effect backend fix.
Reviewed-by: Bjørn Erik Nilsen
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The source pixmap shouldn't include the window background, as the
graphics effect might then overwrite other widgets.
Reviewed-by: Jens Bache-Wiig
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The problem appeared with KLineEdit and its proxy style.
When setting a stylesheet, we create a QStyleSheetStyle in 9# and set it
to the widget, which receive a changeEvent, that will set again the style
so 3# will create a new QStyleSheetStyle with the new proxy. and set it.
in 2# we will dereference the 'old' QStyleSheetStyle (the one created in #9)
that will destroy it.
Then, later in #7, we will still access that style ('newStyle') and we crash
0# QStyleSheetStyle::~QStyleSheetStyle()
1# QStyleSheetStyle::deref()
2# QWidgetPrivate::setStyle_helper(QStyle*, bool, bool)
3# QWidget::setStyle(QStyle*) (qwidget.cpp:2523)
4# ChangeEventWidget::changeEvent(QEvent*)
[...]
6# QCoreApplication::sendEvent(QObject*, QEvent*)
7# QWidgetPrivate::setStyle_helper(QStyle*, bool, bool)
9# QWidget::setStyleSheet(QString const&) (qwidget.cpp:2470)
The solution is to change the order, and do not use 'newstyle' after
we sent the event.
The origStyle is now protected wy a QWeakPointer, but this is
just for safety reason and not related to the crash.
Reviewed-by: JBache
|
| |\ \ \ \
| | |/ / /
| | | | |
| | | | |
| | | | |
| | | | | |
Conflicts:
src/declarative/qml/qdeclarativeengine.cpp
tools/qdoc3/htmlgenerator.cpp
|
| | |/ /
| |/| |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Another fix for the same problem - we also need to be careful - when ungrabbing
a gesture for the recognizer that has already been destroyed and cleaning up
the QGesture object for it we need to make sure we know it is removed from the
obsolete gestures list so that we won't delete it again in the QGestureManager
detructor.
Task-number: QTBUG-9801
Reviewed-by: Thomas Zander
|
| |/ /
|/| |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
QRuntimeGraphicsSystem is a proxy graphics system which
can dynamically switch underlying graphics system on runtime.
For example, switch from hardware accelerated graphics system
to raster graphics system on low GPU memory situation. This
feature is currently supported on Symbian platform.
Task-number: QT-3276
Reviewed-by: Jason Barron
|
|/ /
| |
| |
| |
| |
| |
| |
| |
| |
| | |
QGraphicsProxyWidget.
In case of a complex widget embedded in a proxy, we need to update the
input method acceptance whenever the focus changes inside the proxy.
Task-number:QTBUG-8847
Reviewed-by:janarve
|
|\ \
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Fix compilation with QT_NO_TEXTSTREAM
Fix the compilation with QT_NO_TOOLTIP
Add unit tests for a number of QList methods.
Crash while dragging on Mac OS X (Cocoa)
Doc & menubar hidden for QtDemo after switching back from fullscreen
qdoc: Added breadcrumbs for class pages.
QTabWidget documentMode not working correctly on Mac
qdoc: Removed MacRefs for Xcode documentation stuff.
Slight QList documentation fix.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This feature was implemented in Carbon and was never ported to Cocoa.
The major problem is the fact that Cocoa draws a line under the titlebar,
regardless of what we say. The only way to avoid drawing that line is
by adding a native toolbar and ask it not to draw its baseline. If there
is not toolbar, as it happens in this case, there is no way to prevent
that line from being drawn. So instead we paint over that line and hope
for the best.
Task-number: QTBUG-8159
Reviewed-by: Richard Moe Gustavsen
|
|/ /
| |
| |
| |
| |
| |
| | |
This reverts commit 434d2312b63090cfd2ba7e0f3728b5b225bcaeec.
This patch was partially wrong and introduced some unwanted effects on
KDE and Windows. And it's probably not the righ solution.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
stuck anymore
When using this flag the child of the widget which is embedded is becoming
a top level QWidget. So a right click on a text edit trigger the context
menu to be top level. When creating this top level context menu we are
grabbing the mouse but we were never releasing it when the menu was hidden.
The patch check if the widget uses the Qt::BypassGraphicsProxyWidget and
ungrab the mouse. The patch also fix a bug when positioning the QMenu, it was
for the same reason.
Task-number:QTBUG-7254
Reviewed-by:brad
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The Qt::WA_Moved was not being set during the move event as notified
by the window manager. This is a behavior change for 4.7, but we think
it's more user friendly than the previous behavior.
Reviewed-by: mbm
Task-number: QTBUG-9991
|
| |
| |
| |
| | |
Reviewed-by: Martin Jones
|
|\ \
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* '4.7' of scm.dev.nokia.troll.no:qt/qt-qml: (147 commits)
doc fixes
Don't test XmlListModel examples on platforms without QtXmlPatterns
More error output for QML_IMPORT_TRACE=1
Fix declarative examples autotest, avoid using native separators
Doc: update Rectangle smooth painting screenshot.
Add some TextInput properties and methods
Remove GraphicsObjectContainer from the documentation.
import Qt 4.6 -> import Qt 4.7
Move documentation code to snippet.
Test fix.
Fix test
Fix test on QWS.
Fix PathView crash.
Make bindings dump more useful
Improve font value type documentation.
coding conventions
Doc fixes
Warning
Read Maemo orientation at startup
Exclude gestures from examples autotest
...
|
| | |
| | |
| | |
| | | |
Apply the ~QObject QDeclarativeData logic to ~QWidget
|
|\ \ \
| |/ /
|/| /
| |/
| |
| |
| | |
Conflicts:
qmake/generators/symbian/symmake.cpp
src/gui/image/qimage.cpp
src/openvg/qwindowsurface_vgegl.cpp
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This change includes a fix for QTBUG-4887 and other exception safety
problems found while testing it.
The QTBUG-4887 fix is to qimage.cpp. QImage doesn't throw exceptions on
failure like a proper class should, instead it tries to fail "nice".
What happens here is that setAlphaChannel would crash on OOM as after
the convertToFormat call, d could be NULL. This new version checks the
result of the conversion before using it.
The other fixes are all cases where exceptions were thrown from
destructors. I added code to the test app to help debug these cases,
and I fixed all the problems I found.
With these changes, tst_exceptionsafety_objects runs and passes on the
Symbian emulator.
Reviewed-by: Shane Kearns
|
| |
| |
| |
| |
| |
| |
| |
| | |
This should prevent window managers from ever sending a WM_TAKE_FOCUS
message to those windows (useful for IM windows like virtual keyboards,
notification banners, etc.)
Reviewed-by: bhughes
|
|\ \
| |/
| |
| |
| | |
Conflicts:
qmake/generators/symbian/symmake_abld.cpp
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This also fixed the issue where an input panel would not be requested
by the widget, because it thought it was being focused over and over,
instead of focused, and then clicked (focus followed by click is
default input panel popup behavior on Symbian).
AutoTest: Included and passed
RevBy: yoann
RevBy: Jan-Arve
|
|\ \
| |/
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Conflicts:
configure.exe
examples/multimedia/audioinput/audioinput.cpp
src/corelib/io/qfsfileengine.cpp
src/gui/egl/qegl_wince.cpp
src/gui/egl/qeglproperties.cpp
src/gui/egl/qeglproperties_p.h
src/gui/embedded/directfb.pri
src/gui/kernel/qapplication_win.cpp
src/gui/painting/qdrawutil.cpp
src/opengl/qgl_p.h
src/sql/drivers/odbc/qsql_odbc.cpp
src/sql/drivers/odbc/qsql_odbc.h
tests/auto/auto.pro
tests/auto/qgl/tst_qgl.cpp
translations/assistant_adp_ru.ts
|
| |
| |
| |
| | |
Reviewed-by: cduclos
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Before doing anything we need to make sure that we don't leave anything
in a non-consistent state. When hiding a widget we need to make sure that
no mouse_down events are active, because the mouse_up event will never be
received by a hidden widget or one of its descendants. The solution is
simple, before going through with this we check if there are any
mouse_down events in progress, if so we check if it is related to this
widget or not. If so, we just reset the mouse_down and then we continue.
In X11 and Windows we send a mouse_release event, however we don't do that
here because we were already ignoring that from before. I.e. Carbon did
not send the mouse release event, so we will not send the mouse release
event. There are two ways to interpret this:
1. If we don't send the mouse release event, the widget might get into an
inconsistent state, i.e. it might be waiting for a release event that
will never arrive.
2. If we send the mouse release event, then the widget might decide to
trigger an action that is not supposed to trigger because it is not visible.
Task-number: QTBUG-8604
Reviewed-by: denis
|