From Ravi Mathur, "New functionality is the ability to automatically determine the

maximum traversal depth necessary to obtain an accurate estimate of
the minimum number of required cameras.  In addition, the user can
specify an absolute maximum traversal depth that will not be exceeded."

examples/osgdepthpartition/DepthPartitionNode.h

@@ -37,11 +37,19 @@ class CURRENT_CLASS : public osg::Group
 	inline osg::CameraNode::RenderOrder getRenderOrder() const
 	{ return _renderOrder; }
 
+	/** Set/get the maximum depth that the scene will be traversed to.
+	    Defaults to UINT_MAX. */
+	void setMaxTraversalDepth(unsigned int depth) 
+	{ _distAccumulator->setMaxDepth(depth); }
+
+	inline unsigned int getMaxTraversalDepth() const
+	{ return _distAccumulator->getMaxDepth(); }
+
 	/** Override update and cull traversals */
 	virtual void traverse(osg::NodeVisitor &nv);
 
 	/** Catch child management functions so the CameraNodes can be informed
-	    of new/removed children */
+	    of added or removed children. */
 	virtual bool addChild(osg::Node *child);
 	virtual bool insertChild(unsigned int index, osg::Node *child);
 	virtual bool removeChild(osg::Node *child);