Hi Robert, I totally agree on the general approach. Still, if you want to go for the change on the osgGA::CameraManipulator base class as a first step, I'd be glad to help with that.
Ricky On Tue, Mar 6, 2012 at 11:44, Robert Osfield <[email protected]>wrote: > Hi Riccardo, > > The current osgGA::CameraManipulator's are a bit of mess, the way they > just control the view matrix is one of their flaws. I merged a set of > sweeping changes from Jan Perciva back in May 2010 that added a new > inheritance hierarchy and more features, but in hindsight I feels > these have just made them more convoluted in design and > implementation. > > Rather than tweak the current CameraManilulator API I feel that a > wholesale review of osgGA and the CameraManipulator is long overdue. > We have new devices such as multi-touch and MS's Kinect, as well as > old ones like joysticks that are really not served properly by osgGA. > We really should try and tie these all together. Perhaps this should > be an objective for 3.4. > > In terms of osgGA::CameraManipulator base class I'd like to distil it > down to it's core functionality - allowing a manipulator to control a > Camera. Potentially we could introduce this change prior to the > wholesale revamp. > > Robert. > > On 5 March 2012 17:35, Riccardo Corsi <[email protected]> wrote: > > Hi Robert and All, > > > > in more than one occasion I've noticed that having access to the camera > > manipulator in use is not enough to have "full control" over the camera > > motion handling. > > > > An example is if you want to temporary disable the manip to set the > camera > > position explicitly (in this case you need to remove the manipulator from > > the viewer, not to get the position overwritten), > > or if you want to write a manip for an ortho camera, in which case you > need > > access to the camera's projection matrix as well. > > > > At the same time we need to preserve the loose coupling between the > camera > > and its manipulator. > > > > I think that a possible approach would be a virtual method in the base > > camera manip like: > > virtual void osgGA::CameraManipulator::UpdateCamera(osg::Camera* pCam) > > to be invoked by the viewer classes. > > Instead of asking the ViewMatrix to the manipulators, they would call > this > > method passing the reference camera and leaving the matrix assignments to > > the implementation. > > > > Do you think it's a reasonable approach? > > The porting of the existing manipulators to the new interface would be an > > easy task. > > > > Please give me some feedback, > > if you like the approach I can submit a patch. > > > > Thanks, > > Ricky > > > > _______________________________________________ > > osg-users mailing list > > [email protected] > > > http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org > > > _______________________________________________ > osg-users mailing list > [email protected] > http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org >
_______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
