HI Riccardo, I have been back tackling the outstanding OS work today and have a chance to build and test your callback test. I can recreate the issues with the crash and double calling.
So far I've fixed a bug in the osgGA/EventVisitor that was causing the crash and have replaced a couple of dynamic_cast<> in EventHandler with more efficient as*() calls, and currently investigating the double calling issue. I hope to get a fix for this soon, once I do I'll check in these fixes/refinements to the event callback handling. FYI, the lack of the bool being passed back by the callback methods has different meaning depending upon the context. An event handler that is handling a single event it returns true if that event is handled, then the calling code should then set the EventHandled to true on that event. For a method handling all the events assigned to an EventVisitor it will need to do this task itself, and it's return value is not intended to be used in this way. The duality is an unfortunate consequence of retaining backwards compatibility as the various underlying OSG classes have evolved to become more flexible. To fix it would require breaking lots of OSG user code so I've taken the decision to do what I can to offer the new functionality without breaking backwards compatibility and just have had to accept it's not as clean as it would have been if we have started from scratch with the callback API. Robert. _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
