'mytrackd' and VRCO 'trackd'

advantages

* Biggest one. 'mytrackd' build on top of FREE VRPN software.
* 'mytrackd' automatically grabs CAVELib configuration to determine keys for tracker and controller shared memory areas. Also, you can use CAVELib configuration files to setup head and wand CAVE sensors. This circumstance can cancel use of dedicated tracker config file in case you dealing with head and wand sensors only.
* VRPN provides record/playback embedded feature for data connections. This means, for instance, you can record reports sequence from specific sensor to further playback. Quite useful.

disadvantages

* 'mytrackd' needs for VRPN-based server running somewhere to obtain data for sensors, buttons and valuators over network connection (even in case 'mytrackd' and VRPN server running on same machine). Instead of this VRCO 'trackd' has embeded feature that allows data flow from tracking & controlling hardware directly to CAVELib shared memory. This means VRCO 'trackd' can be used alone, without 'trackdserver' in case of all hardware plugged into same machine used to run CAVELib application.
* 'mytrackd' and generic VRPN server are using primitive data collection behavior. Both applications implicitly poll all VRPN stuff they created by "mainloop" call for each class. Such polling is done in infinite loop with some sleeping in each cycle (amount of pause in milliseconds can be specified as command line parameter in both cases). As it seems, VRCO 'trackd' uses more advanced solution with light multiprocess design.