Maybe bug? Out of sync last_angles (SDK version 1.6.10)

[asantanna-applied]

I was having a problem with set_servo_angle() where it looked like the robot was just doing something random. I tracked down the problem to last_used_angles being uninitialized. I fixed my SDK with a call to _sync() as seen below. I probably should have forked the repo and all that but this was so tiny a change I figured it wasn't worth it ... sorry ... :-)

• Andre

xarm/x3/xarm.py (line 312)

   def set_servo_angle(self, servo_id=None, angle=None, speed=None, mvacc=None, mvtime=None,
                        relative=False, is_radian=None, wait=False, timeout=None, radius=None, **kwargs):
        assert ((servo_id is None or servo_id == 8) and isinstance(angle, Iterable)) \
            or (1 <= servo_id <= 7 and angle is not None and not isinstance(angle, Iterable)), \
            'param servo_id or angle error'

        # ALS: line below fixes my problem
        self._sync()

        last_used_angle = self._last_angles.copy()
        last_used_joint_speed = self._last_joint_speed
        last_used_joint_acc = self._last_joint_acc
        is_radian = self._default_is_radian if is_radian is None else is_radian

[vimior]

Thank you, _sync will synchronize the reported position to the recorded value sent last time, but if the last motion command is non-blocking, then you want to perform relative motion at this time (or you did not specify all joint values (or X/y/ z/roll/pitch/yaw value)) will be problematic. The value of last_used_angles will be updated synchronously when the robot arm is in a normal state when it is reported for the first time

[asantanna-applied]

This occurred after several commands in normal state. I think it was the first joint command though. What should I do to make sure values are correct before issuing the joint command? Doing a relative command of all zeros forces resync?