Michael Carroll

> I wonder if this is something that we should put into rclcpp rather than here. That is, it seems to me that all ROS 2 binaries would benefit from...

No reasons against the change, likely just a lack of someone contributing the feature. Please open a PR and we would be happy to iterate on it.

There isn't a technical reason that this cannot be supported, it has just been largely outside of the scope of our current project workload. The simplest approach to trying the...

@akhileshmoghe Can you go ahead and rebase this on current `ros2` branch and I'll kick off CI, thanks.

Thanks for the report. If you could get the exact versions, that would help me figure out where the fix is most appropriate.

> To clarify, I am not suggesting providing ROS2 as an executable, but asking for the ability to turn applications that use ROS2 into standalone applications that do not need...

I think this brings up kind of a larger question. By IEEE-754 definition `nan != nan`, so if we have a message with a single float data field, both set...

I'm going to bring this up at our next team meeting, because I think you aren't wrong, but want to make sure that we think through the potential impacts.

> I can not use "ros2 run" or "ros2 launch" to run my packages, including ros2 demo(demo_nodes_cpp, turtlesim and so on). It is very strange, because I can still use...

> Another drawback of the current implementation is, that we need to hold a shared_ptr to the waitable while we call rcl::wait in order to make sure that the rcl...