| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
Hopefully this fixes the remaining issues.
|
|\ |
|
| |
| |
| |
| |
| | |
static_cast'ing this pointer type is not possible. You must call data()
or constData() to get the actual pointer to cast.
|
| |\ |
|
| | |
| | |
| | |
| | | |
Reviewed-by: Jason Barron
|
| | |\ |
|
| | | |
| | | |
| | | |
| | | | |
Reviewed-by: Jason Barron
|
| | |/
| |/|
| | |
| | | |
Reviewed-By: jbarron
|
| |\ \ |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
The C++ standard says that templates do not consider non-exported
operators, but most compilers ignore that rule. RVCT however, does not
so add the Q_STATIC_GLOBAL_OPERATOR macro here which expands to
'inline' for RVCT instead of 'static'.
Reviewed-by: Harald Fernengel
|
| | |\ \
| | | |/ |
|
| | | | |
|
| | | |
| | | |
| | | |
| | | | |
... instead of QScopedSharedPointer
|
| | | |
| | | |
| | | |
| | | | |
Getting us closer to removing QScopedSharedPointer
|
| | | |
| | | |
| | | |
| | | | |
Remove QScopedSharedPointer, this class will go soon.
|
| | | |
| | | |
| | | |
| | | | |
part of the task to get rid of QScopedSharedPointer
|
| | | |
| | | |
| | | |
| | | | |
Reviewed-by: Thiago
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
For some reason, gcc was not picking up the qSwap specialization.
Removing the explicit template call made it work.
Reviewed-by: Thiago
|
| | |\ \
| | | |/
| | |/|
| | | |
| | | |
| | | |
| | | |
| | | | |
Conflicts:
examples/painting/svgviewer/files/bubbles.svg
src/corelib/kernel/qobject.cpp
src/network/kernel/qhostinfo.cpp
tests/auto/qhostinfo/tst_qhostinfo.cpp
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
New output that tells if a QNetworkReply was retrieved by using
HTTP pipelining.
The demo browser now shows some (debug) output about usage of cache,
pipelining and SSL.
Reviewed-by: Prasanth
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
It's a long stretch, but it could happen that the compiler optimises
out the checking of the variable because it could not change
synchronously. In any case, it's best to set it to volatile to ensure
that all architectures reload properly (memory ordering semantics)
Reviewed-by: Alexis Menard
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
If Qt is used from within a native cocoa application. This patch
allowes the application to terminate if we have no current event
loops running. This should be safe in itself, and also handle the
case where NSApp run is called from outside Qt.
Task-number: 259928
Reviewed-by: msorvig
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Added support for from-by-animation and by-animation in the
animateTransform element. Updated bubbles.svg in the svgviewer example
to make it work like intended with the corrected animation code.
Task-number: Partially fixes 254784
Reviewed-by: Trond
|
| | | |
| | | |
| | | |
| | | | |
Reviewed-by: Trond
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
On animate transform if additive = "replace" then transform specified
with the element should be replaced with value of animateTransform. If
additive = "sum" , then apply the animateTransform on top of element
level transform.
Reviewed-by: Kim
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
When calling drawImage(), you expect the entire target rectangle to be
filled, but an SVG user agent preserves the source image's aspect ratio
by default. This fix explicitly turns off preservation of aspect ratio.
Task-number: 257052
Reviewed-by: Trond
|
| | | |
| | | |
| | | |
| | | | |
Reviewed-by: TrustMe
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Apparently setting WA_DontCreateNativeAncestors and WA_NativeWindow
attributes on widgets introduces painting artifacts. Temporary
removing these attributes until there is a proper fix (those
attributes will not be needed after the next gesture api improvement).
Reviewed-by: trustme
|
| | | |
| | | |
| | | |
| | | |
| | | | |
The code where the infinite recursion could appears should only be
called by code generated by very old version of moc
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Was introduced by commit 919b723
This function is not even supposed to be called anymore by anything
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Now we have 2 real implementations for listviews and iconviews
This is a big refactor but the behaviour is kept.
It also fixes a crash that could occur in setPositionForIndex
while in ListMode.
Acknowledged-By: ogoffart
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Since mouse tracking is always enabled on all QCocoaView-s, we are
getting mouseMoved events for both parent and child (if mouse is over
the child). In such cases the mouseMoved events are ignored for the
parent view.
We are using the native NSCursor stack for setting the override cursor.
The current implementation for changeOverrideCursor is modified to keep
this stack in sync with Qt's internal list.
Task-number: 258173
Reviewed-by: Morten Sorvig
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Those @$#! ifndefs.
Reviewed-by: TrustMe
|
| | | |
| | | |
| | | |
| | | | |
Reviewed-by: TrustMe
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
GL 2 engine on embedded.
Worked in release because we don't really use any internals in QDataBuffer
that sets the internal size correctly.
Reviewed-by: Samuel
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Build was broken when Qt3 support stuff was not enabled
Reviewed-by: TrustMe
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Make some functions static that are not used anywhere but in the
current file. Others that are used, add the declaration to the _p.h to
ensure we don't forget about them.
Finally, there's no need to enable debugging code if it's not used
anywhere.
Reviewed-by: TrustMe
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
With the series of commits ending in ff1280178, I made QUrl::toAce
fail if the parameter is not a well-formed hostname (i.e., if it
violates STD3). IPv6 hostnames do, so we can't preemptively run
ToACE.
Instead, delay running ToACE until we've tried literal matching.
Reviewed-by: TrustMe
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
scene bottom-up.
When adding n items, the depth was computed n² times. Adding lazy computation
fixes this performance issue.
Reviewed-by: Andreas
|
| | | |
| | | |
| | | |
| | | | |
Reviewed-by: Thierry
|
| | | |
| | | |
| | | |
| | | | |
Reviewed-by: ariya
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
HTTP Pipelining should improve the performance of HTTP requests
for high latency network links. Since some servers/proxies could
have problems with it, it is disabled by default.
Set the HttpPipeliningAllowed attribute of a QNetworkRequest
to enable it for that request.
Reviewed-by: Thiago
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
We will use this for auto testing.
Reviewed-by: Thiago
|
| | | |
| | | |
| | | |
| | | | |
Reviewed-by: Thiago
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
When sending XdndLeave events, we should use the effectiveWinId() of
the drag source (just like we do when sending XdndEnter and
XdndPosition events).
Task-number: 255559
|
| | | |
| | | |
| | | |
| | | | |
Otherwise this codec would not be available.
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
... by not allocating space for slots in the vector.
Before, the vector uses the signal index as index. The problem is that
the slots and signal are mixed in the same index space.
We solve the problem by having a different index space for the signal in
the connectionlists vector. All we need to do is to add the information
about the number of signals in the moc.
Also, we are not connecting to cloned signal but only to the orginial
ones. For example, destroyed(QObject * = 0) would generate two signal,
we now only connect to the first one.
This also improve a little bit the performence while activating signals
since it removed one call to indexOfMethod.
Reviewed-by: Brad
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
be handled by connected views/proxies. Otherwise, the views/proxies may
end up in an invalid state.
Task: 227718
Reviewed-By: Olivier Goffart
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This includes docs on the default QWidget::event() behavior, how to
use touch with QAbstractScrollArea subclasses, how the propagation and
grouping works, as well as some caveats.
Reviewed-by: David Boddie
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
QXmlStreamReader and QXmlStreamWriter can be used conveniently without
subclassing, which the example now demonstrates.
Reviewed-by: mae
Reviewed-by: David Boddie
|