Commit Graph

10 Commits

Author SHA1 Message Date
Robert Osfield
dd996a3289 Introduced CMake option OSG_PROVIDE_READFILE option that defaults to ON, but when switched to OFF disables the building of the osgDB::read*File() methods,
forcing users to use osgDB::readRef*File() methods.  The later is preferable as it closes a potential threading bug when using paging databases in conjunction
with the osgDB::Registry Object Cache.  This threading bug occurs when one thread gets an object from the Cache via an osgDB::read*File() call where only
a pointer to the object is passed back, so taking a reference to the object is delayed till it gets reassigned to a ref_ptr<>, but at the same time another
thread calls a flush of the Object Cache deleting this object as it's referenceCount is now zero.  Using osgDB::readREf*File() makes sure the a ref_ptr<> is
passed back and the referenceCount never goes to zero.

To ensure the OSG builds when OSG_PROVIDE_READFILE is to OFF the many cases of osgDB::read*File() usage had to be replaced with a ref_ptr<> osgDB::readRef*File()
usage.  The avoid this change causing lots of other client code to be rewritten to handle the use of ref_ptr<> in place of C pointer I introduced a serious of
templte methods in various class to adapt ref_ptr<> to the underly C pointer to be passed to old OSG API's, example of this is found in include/osg/Group:

    bool addChild(Node* child); // old method which can only be used with a Node*

    tempalte<class T> bool addChild(const osg::ref_ptr<T>& child) { return addChild(child.get()); } // adapter template method

These changes together cover 149 modified files, so it's a large submission. This extent of changes are warrent to make use of the Object Cache
and multi-threaded loaded more robust.



git-svn-id: http://svn.openscenegraph.org/osg/OpenSceneGraph/trunk@15164 16af8721-9629-0410-8352-f15c8da7e697
2015-10-22 13:42:19 +00:00
Robert Osfield
6854d2968f From Jannik Heller, "Fix for Qt4 multi-threaded crash. This patch adds the AA_X11InitThreads flag to the QApplication setup, meaning QT will call XInitThreads for us. This also bumps the required QT version to 4.8+, because the AA_X11InitThreads flag was introduced in that version. If that is not acceptable, we could add a QT_VERSION >= 0x040800 guard, or manually call XInitThreads() instead. Let me know what you think."
git-svn-id: http://svn.openscenegraph.org/osg/OpenSceneGraph/trunk@14804 16af8721-9629-0410-8352-f15c8da7e697
2015-03-20 11:48:21 +00:00
Robert Osfield
30a62bc4ad From Jonathan Greig, "I was testing out the example and it works fine as a standalone example, but I found that it didn't work as expected when adding the class to some existing code I have to test it out. The osg widget was appearing but it didn't appear to be updating. After tweaking it to pass on the parent QWidget pointer from my code, it worked as expected. I added the WindowFlags to the attached file for good measure. If you look at the Qt Documentation for QWidget ( http://doc.qt.io/qt-5/qwidget.html#QWidget ), you'll see that is exactly where I got this from. When subclassing a widget as such, you should allow the user to pass in the parent at least, considering that has a major effect on how or if Qt may clean up the memory."
git-svn-id: http://svn.openscenegraph.org/osg/OpenSceneGraph/trunk@14718 16af8721-9629-0410-8352-f15c8da7e697
2015-02-25 19:19:37 +00:00
Robert Osfield
bc5575f83a From Kristofer Tingdahl, "I and my team have gone over the code again, and we feel that we are comfortable in our current proposal for change. It goes deeper than it did before, and I explain why:
There was code in the osgViewer/Viewer.cpp and osgViewer/CompositeViewer.cpp that transformed the Y-coordinates of an event. The code in the composite viewer did however miss the touch-data of the event. I thought that it should really be the GUIEventAdapter that should know about this, and hence I added the
GUIEventAdapter::setMouseYOrientationAndUpdateCoords which is re-computing the coordinates. First I simply added a boolean to the setMouseYOrientation function:

setMouseYOrientation( MouseYOrientation, bool updatecooreds=false );

but then the serializer complained.

This function is called from both the Viewer and the CompositeViewer. We have not tested from the viewer, but I cannot see it would not work from visual inspection.

The other change is in MultiTouchTrackballManipulator::handleMultiTouchDrag. I have removed the normalisation. The reason for that is that it normalised into screen coordinates from 0,0 to 1,1. The problem with that is that if you have a pinch event and you keep the distance say 300 pixels between your fingers, these 300 pixels represent 0.20 of the screen in the horizontal domain, but 0.3 of the screen in the vertical domain. A rotation of the pinch-fingers will hence result in a zoom in, as the normalised distance is changing between them.

A consequence of this is that I have changed the pan-code to use the same algorithm as the middle-mouse-pan.

The rest of it is very similar from previous revision, and there has been some fine-tuning here and there.

"
2014-04-24 17:14:54 +00:00
Robert Osfield
fd85542d8f Cleaned up the Qt find package, and made changed the default threading model to SingleThreaded when using Qt5 to avoid crash due to regresssion since Qt4. 2014-01-23 10:09:53 +00:00
Robert Osfield
a1bda7d39e Removed #if 0 #endif block 2013-07-19 10:11:42 +00:00
Robert Osfield
e7b86099d1 Fixed problem with the osgviewerQt code override the default View::Camera without setting it's default state on it's StateSet like View does by default. 2013-07-19 09:43:54 +00:00
Robert Osfield
bdfd18dc03 From Kristofer Tingdahl, with additions from Riccardo Corsi and Robert Milharcic, support for Qt5 build 2013-06-10 14:34:25 +00:00
Robert Osfield
d18250da59 Replaced .osg with .osgt file usage 2011-06-14 16:54:20 +00:00
Robert Osfield
b1d3763510 Renamed example from osgviewerQtContext to osgviewerQt. 2011-05-16 09:17:30 +00:00