| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
| |
Reviewed-by: Kent Hansen
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
rect is not drawn correctly.
Depending on the type of server, QAxClientSite::TranslateAccelerator()
needs to process the message differently. For ActiveQt based
in-process-servers, only normal Qt event handling is required. For
ActiveQt based out-of-process-servers, the message has to be forwarded
and Qt event handling needs to continue. For all other type of servers,
forward the message and stop Qt event processing.
Styles use the WA_KeyboardFocusChange attribute set on the window to
decide on drawing the focus rect. ActiveQt handles the VK_TAB key in the
QAxServerBase::TranslateAcceleratorW(), the attibute is now set when
focus is changed.
Task-number: 253763
Reviewed-by: Volker Hilsheimer
|
|
|
|
|
|
|
| |
QRasterPaintEngine assumes device does not change size. Therefore create
a new engine when resizing.
Reviewed-by: Tom
|
|
|
|
|
|
| |
Reading a help project was unacceptably slow with pattern matching.
Now we do it only for filenames that contain wildcard symbols.
Also, we cache the results of QDir::entryList() calls.
|
| |
|
|
|
|
|
| |
Broke on WinCE since QStateMachinePrivate is now using
Q_AUTOTEST_EXPORT.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This reverts commit 95560abfdd3a1755b69f238eff2954d164dff6a5.
Fix comboboxes used as editor.
isAncestor() stops at windows. we want to go tought the whole widget
chain, including parent windows
Use case: examples/itemviews/coloreditorfactory/coloreditorfactory
Task-number: 259034
|
|
|
|
|
|
|
|
|
|
|
|
| |
I'm not sure this is the proper fix, though. xlC 7 complains with:
"/usr/vacpp/include/new", line 97.10: 1540-1298 (I) "void *operator
new(size_t, void *) throw()" needs to be declared in the containing
scope to be found by name lookup.
Also add some calculations to the padding in the data structure.
Reviewed-by: Trust Me
|
|
|
|
| |
Needed for Declarative UI integration.
|
|
|
|
| |
Reviewed-by: kh
|
|
|
|
|
| |
QObject::sender() does not return const QObject*, so neither should
this API; it just forces you to const_cast for no good reason.
|
|
|
|
|
| |
Task-number: 258371
Reviewed-by: Trust Me
|
|
|
|
|
|
| |
See 3ae2cab9c8bd1790a00da2755ac036143a3a35f4 for another similar fix.
Reviewed-by: Trust Me
|
|
|
|
|
|
|
|
|
|
| |
There is no monotonic clock support on HP-UXi at all, and the
_POSIX_MONOTONIC_CLOCK macro is not defined at all (not even to
-1). We handle this in the event dispatcher, but not in
qcore_unix.cpp. Since the monotonic clock time code has moved, the
define of _POSIX_MONOTONIC_CLOCK to -1 should also move.
Reviewed-by: thiago
|
| |
|
|
|
|
| |
RevBy: denis
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
not C99-compliant.
C99 says vsnprintf returns >= `size' when not enough memory was
available. In practice, on Windows, glibc < 2.0.6, and probably plenty
of other places it returns -1 instead.
Reviewed-by: Rhys Weatherley
|
| |
| |
| |
| |
| |
| |
| |
| | |
Fix a few typos and add more descriptive documentation to the
class itself and its subclasses.
Reviewed-by: Volker Hilsheimer
Reviewed-by: Martin Smith
|
| | |
|
| |
| |
| |
| |
| | |
Explain the implications in the QWebHistoryItem documentation, and get rid
of the "group".
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Close button was enabled if the native window handle was created
implicitely (for example by the setWindowTitle() function). Moved the
code that enables of disables the close button to the place where we
create native window handle.
Task-number: 256360
Reviewed-by: trustme
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
On Windows we will add maximize button to the titlebar even if the
window has a fixed size if the user explicitely asked for it by
setting Qt::CustomizeWindowHint | Qt::WindowMaximizeButtonHint.
Task-number: 250188
Reviewed-by: Leonardo Sobral Cunha
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| | |
QWidget::setFixedSize() constrains the size of the widget, however there
was no way to remove those constraints, so setting the constraints to
QWIDGETSIZE_MAX will make the widget resizable again.
Reviewed-by: Leonardo Sobral Cunha
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
test wouldn't finish.
Running under valgrind and slow operating systems, the transfer rate is
only a couple of MB per second. As long as you can reach 3.2 MB/s now,
it should still work
Also, don't kill the thread, for any reason. That causes trouble.
Reviewed-by: Markus Goetz
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The delete reply; was introduced by a conflicted merge in the VxWorks
port. It should not have appeared.
But take this opportunity to convert the two other uses of
QNetworkReply* without lifetime management to QSharedPointer too.
Reviewed-by: Markus Goetz
|
| |
| |
| |
| |
| |
| |
| | |
Autotest: included
Merge-request: 378
Reviewed-by: Eskil Abrahamsen Blomfeldt <eblomfel@trolltech.com>
|
| | |
|
| |
| |
| |
| |
| |
| | |
"../shared/qm.cpp", line 556.45: 1540-0207 (S) No common type found for operands with type "const char [7]" and "QByteArray".
Reviewed-by: Trust Me
|
| |
| |
| |
| |
| |
| | |
made sure that we set the filter on the correct widget
RevBy: denis
|
| |
| |
| |
| | |
Reviewed-by: kh
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Text used to be formatted during the parsing of the SVG file, but
because the text can be referenced by a 'use' element, the text
formatting is not known at this point in time. Now, font attributes can
be inherited from 'use' elements, and the text is formatted each time
it is drawn.
Reviewed-by: Tor Arne
|
| |
| |
| |
| |
| | |
Task-number: 254040
Reviewed-by: Tor Arne
|
| |
| |
| |
| |
| |
| | |
Task: 254724
Reviewed-By: TrustMe
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Copy the first line from the texture to the destination
and from then on reference that copy reducing cache misses. Also
progressively copy larger sizes, reducing the amount of time spend
figuring out what to copy.
Merge-request: 371
Reviewed-by: Samuel Rødal <sroedal@trolltech.com>
|
| |
| |
| |
| | |
Reviewed-By: TrustMe
|
| |
| |
| |
| | |
Reviewed-By: Thomas Hartmann
|
| | |
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
loadGlyphs() has a big impact on text rendering speed on X11. Removing
the uneeded hash lookup makes this function go twice as fast, as it in
the common case (where the glyphs are already cached) the hash lookups are
100% of the running time of the function.
Reviewed-by: Trond
|
| |\ |
|
| | |
| | |
| | |
| | | |
Reviewed-By: TrustMe
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Squashed commit of the following:
commit fcf7e8cab339d0cf9f3f2a9756d7754c54c4d934
Author: Gunnar Sletta <gunnar@trolltech.com>
Date: Thu Jul 30 13:15:13 2009 +0200
note in the changes file...
commit 2c9c3880215988e6609c290a8e738b228736e601
Author: Gunnar Sletta <gunnar@trolltech.com>
Date: Thu Jul 30 12:51:42 2009 +0200
Don't leak native window HRGN handles
commit 6bb30d2075dd1d71a8a600d25f413a38af7f2f2c
Author: Gunnar Sletta <gunnar@trolltech.com>
Date: Thu Jul 30 11:09:22 2009 +0200
Moved qregion_wince.cpp -> qregion_win.cpp, platforms are identical now
commit 173fcc5baec73a198167985c6f777987e6015a71
Author: Gunnar Sletta <gunnar@trolltech.com>
Date: Thu Jul 30 09:42:06 2009 +0200
win32 calls on QRegion.handle() is no longer supported, use from HRGN
commit d7ddcce4ba29b70ed81f85274208b388a2bb9d4d
Author: Gunnar Sletta <gunnar@trolltech.com>
Date: Thu Jul 30 09:41:37 2009 +0200
Added convenience function to convert from HRGN to QRegion
commit 2fc53ac3d59a9c42bb4154fff7557610092b7946
Author: Gunnar Sletta <gunnar@trolltech.com>
Date: Wed Jul 29 09:28:10 2009 +0200
Kill qregion_win.cpp and use the unix code instead
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The reason is that the C++ compiler needs to add -I/usr/vacpp/include
before -I/usr/include, so our adding of -I/usr/include changes the order.
This causes a compilation error in the C++ header xlocinfo.h.
In any case, these checks done in qmake are unnecessary. If the OpenSSL
headers are in one of those include paths, they will be found on their
own already.
Reviewed-By: Bradley T. Hughes
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This reverts commit e181f8754bccb13cb59212f279e05c20cb76db62.
This code is used for compatibility issues for custom layouts that
dont delete their children, so its a false positive from coverity.
|
| | |
| | |
| | |
| | | |
Reviewed-by: ogoffart
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
1) fromScriptValue() needs to call qvariant_cast() on the data, in
order to be symmetric with toScriptValue().
2) use the overload of newFunction() that takes a prototype object,
so that the prototype.constructor and constructor.prototype
relationship is set up correctly; otherwise the instanceof operator
won't work.
|