@henningkayser

Yes, let's remove getTargetWayPointIndex() since it isn't used. 👍 Will push a commit for that

Running the demo today, I get an error about the OMPL planning pipeline not loading. Sigh. I wonder if it's something to do with my workspace. (It's not just your branch, my branch has the same error.)

I'm on the latest commit of moveit_resources, ros2 branch.

Yes, let's remove getTargetWayPointIndex() since it isn't used. +1 Will push a commit for that

How are you supposed to get the target waypoint from the local solver if the TrajectoryOperator doesn't have a function to provide it? I thought the whole idea about this class is that it defines the logic what the next waypoint is based on the current state. But maybe there is another way to do the same thing.

Yes, let's remove getTargetWayPointIndex() since it isn't used. +1 Will push a commit for that

How are you supposed to get the target waypoint from the local solver if the TrajectoryOperator doesn't have a function to provide it? I thought the whole idea about this class is that it defines the logic what the next waypoint is based on the current state. But maybe there is another way to do the same thing.

Yes, it is done within the operator and the local solver receives a local trajectory through a call of

getLocalTrajectory(const moveit::core::RobotState& current_state, robot_trajectory::RobotTrajectory& local_trajectory)

The local trajectory for the solver could contain more than a single waypoint. This is why I think this function is only useful in local sampler which only forwards a single waypoint of the global trajectory. getTargetWaypointIndex() is implemented in the simple_sampler to disclose the internal member variable.:

size_t SimpleSampler::getTargetWayPointIndex()
{
  return next_waypoint_index_;
}

But it was not used.

getTargetWayPointIndex() is something I added just recently for detecting if the local planner was stuck for X iterations. But, I think I implemented it in the wrong place.