Why should we use destroy_node() for Python nodes in ROS2?

edit

Hi everyone,

In the ROS2 examples using rclpy (for example the subscriber example), in the main function I always see the use of the rclpy destroy_node() method.

def main(args=None):
    rclpy.init(args=args)

    minimal_subscriber = MinimalSubscriber()

    rclpy.spin(minimal_subscriber)

    # Destroy the node explicitly
    # (optional - otherwise it will be done automatically
    # when the garbage collector destroys the node object)
    minimal_subscriber.destroy_node()
    rclpy.shutdown()

To me the explanation in the comment does not give much info.

I've tried to run the code without this line to see if something would happen, and I got no error so far.

So:

Why should we use destroy_node() before calling rclpy.shutdown()?

Can I expect any kind of issue if I don't explicitely destroy nodes (if I just shutdown ROS and exit after)?

Also, is this practice temporary, and will this line be removed from examples and tutorials in the future?

Thanks!