'mytrackd' configuration file (trackdrc)

'mytrackd' operates depending on it's config file. Config file, specified with "-f" option or default one, "trackdrc", get read on program startup and should consist of separate lines of following:

{token} {cave_index} {parameters_depend_on_token}

Each position in line must be separated with at least one space or tabulation character. Case does matter only to {name@host} (see bellow). Empty lines and comments starting with '#' symbol are also allowed.

{token} is one of 'sensor', 'button', 'valuator' or 'calibrator'

{cave_index} is zero-based index of CAVE structure that this line defining. For example, sensor 1 ... will describe CAVE sensor 1, i.e. wand sensor, and button 2 ... will describe third CAVE controller button. In case of 'calibrator' this index denote CAVE sensor to attach calibration filter to.

'sensor', 'button' and 'valuator' tokens share same parameters like:

{sensor/button/valuator} {cave_index} {name@host} {remote_index}

{name@host} specify name of device that provides data and host on which it resides, for example 'tracker2@localhost'

{remote_index} denotes which exactly sensor of tracker, button or joystick channel serve data. This index is zero-based. For example, Polhemus Fastrak tracker supports up to four sensors each of them reports independently.

Definition of 'sensor' requires two more parameters, {transmitter_setup} and {sensor_setup}. Both can take values 'off', 'local' or 'remote' and determine the way this sensor setup should be done. 'off' means no setup will be made, 'local' and 'remote' mean use of tracker config file, local or remote respectively (for more information please consult dedicated section). {transmitter_setup} involves configuration information about tracker's transmitter position and orientation (such information will be applied to ALL sensors), while {sensor_setup} involves information about this particular sensor (it will be applied only to this sensor). Keep in mind that this way you only specify where sensor just defined can get a configuration. Actual setup information is stored in separate local or remote (i.e., residing on 'host' machine in VRPN server folder) file, 'vrpn_tracker.cfg' or can be borrowed from CAVELib configuration. So, line to define a sensor should looks like this:

sensor {cave_index} {name@host} {remote_index} {transmitter_setup} {sensor_setup}

'calibrator' must goes AFTER definition of sensor it going to attach to and requires special definition like:

calibrator {cave_index} {dso_name} {parameters...}

{dso_name} defines name of calibration DSO to attach to specified (by {cave_index}) sensor. DSO will be searched in usual way, using LD_LIBRARY_PATH environment variable.

{parameters...} is list of parameters that will be passed to DSO without any affecting.

Example of 'mytrackd' configuration file:

# wand sensor (1) obtains data from sensor 3 of 'fastrack' tracker
# resides on 'tracker_host' machine, ignores tracker transmitter
# setup and use local one for sensor itself
sensor 1 fastrack@tracker_host 3 off local
# head sensor (0) gets data from sensor 1 and use remote config files
sensor 0 fastrack@tracker_host 1 remote remote
# 'trackdCalibrator.so' calibration DSO attached to sensor 0
calibrator 0 trackdCalibrator.so FIT 2 test.dat DEGREES
# two valuators mapped to joystick with channels swapped
# i.e 0(X channel)->1 and 1(Y channel)->0
valuator 1 wand@controll_host 0
valuator 0 wand@controll_host 1
# Second wand device button is mapped to second CAVE button
button 1 wand@controll_host 1