CHANGELOG

Changelog for package clearpath_tests

Fix a bug where the e-stop test could cause the test node to crash
Add a 2s delay between clearing the e-stop and moving the wheels to allow CAN reconnections as needed
Remove CAN device ID count for platforms that don’t use CANopen
Lower the time threshold for linear drive test to 0.75 (from 0.8)
Change the default serial device for Jackal
Contributors: Chris Iverach-Brereton

Allow symlinks to the device handle
Fix serial MCU tests
Ensure the firmware is 2.3 or higher to allow fan control; otherwise just skip the fan test
Add support for lateral driving test for omni platforms (#5)
Re-enable the fan tests (#4)
Remove rotation test (#3)
Contributors: Chris Iverach-Brereton

Rewrite rotation test (#2) * Overhaul the logic of the rotation to use an accumulator instead of relying on the latest odom data. Factor in the rate of the EKF + the odometry twist instead of just looking at positional data. On 00005 we’re still overshooting slightly, but it’s now within the margin of error
Append the can interface to the node name to suppress duplicate node name warning
Contributors: Chris Iverach-Brereton

Move the confirmation about the lights being in the normal state before we call start()
Log the hardware ID and firmware version reported by the MCU status topic
Move the TF listener implementation to its own file
Add tests, refactor & reformat to address errors they caught
Contributors: Chris Iverach-Brereton

Add an optional flag for the e-stop, remove key-switch test, add wireless e-stop as an optional component
Contributors: Chris Iverach-Brereton

Reduce the minimum duration for a rotation. Log possible false-positives during the rotation test. Print the calculated duration error for the rotation & drive tests
Log the version of clearpath_tests in the report
Increase the length of expected lynx messages to 5, cast the length to an integer before comparing it
Contributors: Chris Iverach-Brereton

Invert the angle of the lateral test
Add a mutex to prevent issues with reading & writing the current & previous orientations asynchronously; this sometimes causes false positives or false negatives during the test
Don’t fail if we get controller_manager rate errors
Add newline between average motor currents in report
Increase the allowed margin of error on the IMU test to 20% (from 10%)
Add an extra confirmation that the lights are in a controllable state before starting the test
Contributors: Chris Iverach-Brereton