Timer.getFPGATimestamp() vs System.currentTimeMillis()

[fovea1959]

coming from a desktop/server Java background, I’ve always used System.currentTimeMillis() for elapsed time calcs, and we use it on our bot consistently.

Just noticed Timer.getFPGATimestamp(), which strikes me as probably being better than System.currentTimeMillis() (though the former might be using the latter for all I know).

Anyone have any insight as too whether the two differ significantly? Don’t want to do the rework if not significantly better (though will start with Timer.getFPGATimestamp() next year!)

[Ether]

@fovea1959

coming from a desktop/server Java background, I’ve always used System.currentTimeMillis() for elapsed time calcs, and we use it on our bot consistently.

Just noticed Timer.getFPGATimestamp(), which strikes me as probably being better than System.currentTimeMillis() (though the former might be using the latter for all I know).

Anyone have any insight as too whether the two differ significantly? Don’t want to do the rework if not significantly better (though will start with Timer.getFPGATimestamp() next year!)

This thread has a discussion starting at post20 that might be of interest:

Velocity PID Programming: Terms dependent on time

[mmaunu]

One difference is the resolution. Timer’s method has a resolution of microseconds instead of currentTimeMillis()'s millisecond resolution.

[fovea1959]

thanks. I missed the difference in resolution. Not important enough to refit existing code (esp since Java TimerTasks only schedule down to the millisecond), but good to know.