GoesM

I've forwarded this issue to the Robotics stack exchange stack: [question 106008](https://robotics.stackexchange.com/questions/106008/ros2nav2user-misconfiguration-of-parameters-may-cause-instantaneous-crashs). Thanks for your feedback. 🙂

I used to discuss the relevant issues of msg-filter with the repository administrators of rclcpp. Currently, my understanding of this code architecture is that ROS primarily addresses problems caused by...

> I believe such changes would be beneficial for users of open-source code who may not be very familiar with software development. My personal viewpoint is that, instead of abruptly...

> If anybody wants to take on these changes and make a pull request, I'll be happy to review it. So happy for hearing that ! and so sorry for...

Sure, it's similar to #694, just issue about `humble` while seems to be fixed in `iron` and `main`

well, it seems that such issue has been fixed in `ros2` branch : https://github.com/SteveMacenski/slam_toolbox/blob/a3442d2f6824ff058fab0cb0a635e7a454294855/src/slam_toolbox_common.cpp#L240-L260 but no such code to reset these pointers in both `iron` and `humble` branch.

Shall I backport by `cherry-pick` ? 🤔 I think the past commit for this issue would change `humble` a lot: https://github.com/SteveMacenski/slam_toolbox/commit/de13553d9f92e3773da4b2ff6eebd1977b9337f6

There'are still changes for plugins, I would add soon.

It seems that in the test, some dyn_param_handler instances might not be registered, which can lead to related errors during resource release. I am currently making adjustments.

The latest CI test failed in `nav2_behavior_tree` , which seems not related to my change (I think) ?