[RST-10727] Publish the start/stop status of the fixed-lag smoother

[svwilliams]

Now that it is possible to construct a fixed-lag smoother without immediately starting it, knowing the running state of the optimizer may be important for coordination with other nodes. Publish the start/stop state of the optimizer as a latched topic.

The status is currently being published as a std_msgs/Bool. I do wonder if I should add a dedicated message type to fuse_msgs and publish that instead? If we ever want to expand the status information in the future, having our own message type could make that easier. We would simply add new fields to the message, but we would continue to publish the same message type. Any thoughts?

[carlos-m159]

The status is currently being published as a std_msgs/Bool. I do wonder if I should add a dedicated message type to fuse_msgs and publish that instead?

My initial thoughts would be to keep it with std_msgs/Bool and change it if we ever require additional status info. For example, nodes that are toggled using std_msgs/Bool or are waiting on some signal to start up or stop: injecting a fuse_msgs dependency might be strange. But I am also fine with creating a custom message, and, for example, publish that continuously and not only at start and stop.

[carlos-m159]

Maybe move this to a function?
If we at any point store the state in a member, might will be easier to maintain.

inline void notifyStatus(const bool running)
{
  // Update status topic
  auto status = std_msgs::Bool();
  status.data = running;
  status_publisher_.publish(status);
}