| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
The change makes QTextCursor not extend the selection when inserting
characters at the end of the selection, provided that the cursor
position is at the end.
Reviewed-by: Simon Hausmann
Reviewed-by: Roberto Raggi
|
|
|
|
|
|
|
| |
When using style sheets, QSlider::handle couldn't be customized if
QSlider::groove wasn't. QComboBox::down-arrow:focus wasn't properly drawn.
Reviewed-by: Olivier
|
|
|
|
|
|
| |
complex widgets.
Reviewed-by: Olivier
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We need a way to support converion to and from multiple native pixmap
types from multiple pixmap backends. Instead of adding more virtual
functions to QPixmapData, make the existing one more generic but pass
an opaque pointer and a type and do some internal casting.
Currently this function is Symbian only, but could easily be extended
to work on other platforms.
Reviewed-by: Aleksandar Babic
Reviewed-by: Jani Hautakangas
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Since we only get one pointer event at a time, we need to keep a list of
all known touch points in QApplicationPrivate (otherwise the QTouchEvent
won't contain enough points). The QApplication machinery can handle
having inactive touch-points in the list, so at the moment we don't
clear the list.
We treat PointerNumber zero as the primary touch point, and only send
regular mouse events for that pointer, never for the others.
Reviewed-by: Jason Barron
|
|
|
|
|
|
|
|
| |
Turning this attribute ends up calling
RWindow::EnabledAdvancedPointers(), which tells Symbian to send us
multiple pointer events with extended info.
Reviewed-by: Jason Barron
|
|
|
|
| |
Reviewed-by: TrustMe
|
|
|
|
|
|
|
| |
The QTextCodec::codecForHtml used to expect http-equiv attribute before
the charset attribute in the meta header, which is not always the case.
Reviewed-by: Simon Hausmann
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This introduces an undocumented "phonon_compat" subdir and also adds
it to the INCLUDEPATH when QT += phonon is specified. With this, these
styles of #includes should be supported:
#include <Phonon>
#include <Phonon/Phonon>
#include <phonon>
#include <phonon/phonon>
#include <Phonon/MediaObject>
#include <phonon/mediaobject.h>
Still need to check if the headers get installed during "make
install". I couldn't find where in our code that is done.
BT: yes
Reviewed-By: Trust Me
|
| |
|
|
|
|
|
|
|
|
| |
The "fast path" will write bad premlutiplied pixels into the
backbuffer so the blending will be broken, plus that the window
does not get mouse events for alpha=0 pixels.
Reviewed-by: Marius Storm-Olsen
|
|
|
|
|
|
|
|
|
| |
Try not to hold an unused 4k QByteArray. This may introduce a
performance regression e.g. for QNetworkReply in exchange for lower
memory usage.
We will see if this is really the case.
Reviewed-by: Thiago
|
|
|
|
|
|
|
|
| |
This fixes a crash that occured because aborted requests were
not properly removed from the channel.alreadyPipelinedRequests.
Task-number: QTBUG-4507
Reviewed-by: Peter Hartmann
|
| |
|
|
|
|
|
|
| |
On X11 We have to wait we get the size change from the window manager
Reviewed-by: Jason Barron
|
|
|
|
|
|
| |
We skip parts of watchDirectory for the same reason.
Reviewed-by: thartman
|
|
|
|
|
|
|
| |
This is a maintenance headache. Just reference the canonical supported
platforms list instead.
Reviewed-by: Trust Me
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| | |
While refactoring gtkstyle in 4.6, the "0" .so suffix was dropped.
This patch restores the 4.5 behavior to look for *.so.0 since the *.so
is only available if you install the gtk developer packages.
Reviewed-by: jbache
|
| |
| |
| |
| |
| |
| |
| |
| | |
Large enough max heap for script tests is now defined by default
for all tests in qttest_p4.prf, so removed the statements from
individual script tests.
Reviewed-by: Janne Anttila
|
| | |
|
| |\ |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Many people have requested this. And since its protected it does not
really pollute the API a lot.
Task: QTBUG-3524
Reviewed-by: mbm
|
| | |
| | |
| | |
| | | |
Reviewed-by: TrustMe
|
| | |
| | |
| | |
| | | |
Reviewed-by: TrustMe
|
| | |
| | |
| | |
| | | |
Reviewed-by: TrustMe
|
| | |
| | |
| | |
| | | |
Reviewed-by: Trust Me
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Dependency on windows style, without including the header file
Needed static casts to compare QWidget* with QComboBox* at least for
the Nokia x86 compiler.
Reviewed-by: axis
|
| |/
| |
| |
| |
| |
| | |
Task-number: QTBUG-3211
Task-number: 236875
Reviewed-by: Alex
|
|/
|
|
|
|
|
|
| |
These files were previously added to source packages by the packaging
system. There's no reason why these files shouldn't be available to
repo users too.
Reviewed-by: Trust Me
|
|\ |
|
| |
| |
| |
| |
| |
| |
| | |
The test goes in only almost 30seconds now.
But i still have few failures
Reviewed-by: Denis
|
| |
| |
| |
| | |
ok, so this is a feature ... a 4-line-one ...
|
| |
| |
| |
| | |
Reviewed-By: Simon Hausmann
|
| |
| |
| |
| |
| |
| |
| |
| | |
When the script class is set to 0, we need to remove the script
object's delegate entirely, because the delegate requires a non-0
script class in order to be useful.
Reviewed-by: Olivier Goffart
|
| |
| |
| |
| | |
Reviewed-by: TrustMe
|
| | |
|
| | |
|
| |
| |
| |
| | |
Reviewed-by: Eskil
|
| |
| |
| |
| |
| |
| | |
Also, don't depend on the existence of ~/.qt/
Reviewed-by: Olivier Goffart
|
| |
| |
| |
| |
| |
| |
| |
| | |
The test of ea0c0dfa64f5c614bfd728f323d9b6731ee4898c is not valid
for the change introduced in that commit. QCoreApplication
does not use the event loop of Cocoa.
This is the same test in QApplication.
|
| |
| |
| |
| |
| | |
The test is failing when Qt is compiled with Qt 3 support. Qt 3
support send events for the insertion of the new child.
|
| |
| |
| |
| |
| |
| | |
On WinCE GetIconInfo returns a 0 hotspot for the system icons, thus
we cannot use this value to determine width / height of the icon.
Instead, we use the icon's bitmap to get this information.
|
| |
| |
| |
| | |
Reviewed-by: thartman
|
| |
| |
| |
| |
| |
| |
| |
| | |
CFG_PATH evaluates as "udeb", not "UDEB" when building the debug-armv5
target. So changed the FLM to check for that.
Also changed the variable redefinition to ifeq/else/endif.
Reviewed-by: Iain
|
| |
| |
| |
| | |
Reviewed-By: TrustMe
|
| |
| |
| |
| |
| | |
On Mac OS X, the directory /bin is hidden because it does not have
the "Visible" attribute in the filesystem.
|