Fix Servo logging frequency

[AndyZe]

A small fix to correct logger frequencies.

Prior to this, the messages only printed once.

main branch, message only prints once: (This is a gif, but nothing changes)

With this PR, prints every 3s:

[codecov]

Codecov Report

Merging #457 (7935ea2) into main (289437b) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main     #457   +/-   ##
=======================================
  Coverage   51.43%   51.43%           
=======================================
  Files         223      223           
  Lines       23339    23339           
=======================================
  Hits        12002    12002           
  Misses      11337    11337           

Impacted Files	Coverage Δ
moveit_ros/moveit_servo/src/servo_calcs.cpp	64.25% <ø> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 289437b...7935ea2. Read the comment docs.

[vatanaksoytezer]

This looks reasonable and good to me.

[griswaldbrooks]

Huh, what's the motivation to stop using std::chrono and count?

[AndyZe]

I just think 3000; // milliseconds is more readable than std::chrono::...whatever. I would have to google how to use std::chrono properly.

Will change if you feel strongly though.

[henningkayser]

How about 3000ms (https://en.cppreference.com/w/cpp/chrono/operator%22%22ms)?

[griswaldbrooks]

In general, I prefer expressiveness through types and not comments (though I like comments).
The issue becomes, it's immediately cast to a size_t which definitely doesn't enforce milliseconds.
I would prefer the use of chrono but won't be a stickler for it here because that type safety should really be propagated to point of use (which I guess is a `ros::Duration, so that should really be the type here.

@henningkayser he'd have to do constexpr size_t ROS_LOG_THROTTLE_PERIOD = (3000ms).count();

[henningkayser]

ROS_LOG_THROTTLE_PERIOD doesn't need to be a size_t, it could be an rclcpp::Duration that returns nanoseconds() to the logging call. I find it a little odd that the logging macro doesn't support the actual duration type.

[griswaldbrooks]

Absolutely, though I assumed increasing the magnitude of the change would have been out of scope for this PR. My guess is that ROS_INFO just wraps printf and so has a limited type set that it can support. ROS_INFO_STREAM uses the insertion operator so could print more interesting things.

[AndyZe]

I changed it to std::chrono::milliseconds(3000).count();