Launch files to open an OpenNI2 device and load all nodelets to convert raw depth/RGB/IR streams to depth images, disparity images, and (registered) point clouds.
I'm going to merge this as a temporary fix, but I think I can fix a few things in the drivers later this week to just leave both on (which would allow you to dynamically reconfigure depth_registration)
This way the defaults for depth processing are set apropriately for both hardware and software registration.