dhood

Hi, I don't have a repeatable example for this, but want to highlight that I've seen this issue in an rclpy executor as well (using Humble and Cyclone DDS): EDIT:...

yes, curiosity got the better of me, I'm a sucker for CI that needs troubleshooting :innocent: Perhaps explicitly removing the subscription from the waitset after it's been waited on, in...

FYI I think this has been closed by https://github.com/ros2/rclpy/pull/930! That has been released into Iron (not backported into Humble AFAICT)

@j-rivero @clalancette Was a new release of `urdfdom_headers` considered for bionic? Would it be too late now?

In your experience would it be covered by that policy? Essentially for machines with different locales, urdf parsing does not work, so it is high impact in the user-facing sense,...

Usually when we refer to node name uniqueness we are actually referring to "ns+node name" uniqueness. To clarify your statement: the remapping process is assuming there is only one `talker`...

As a note: `rclpy.logging` is not being imported automatically either (it could be imported manually but it also imports a C extension and FWIU that's to be avoided): https://github.com/ros2/demos/pull/210

I can confirm that remapping the topics as you suggested @clalancette does the job. This caught me by surprise, though - I wasn't expecting this feature to be missing, so...

Couldn't we rebase the repo+branch in the custom `ros2.repos` (e.g. dhood/* + feature) against the repo+branch in the default `ros2.repos` (i.e. ros2/* + master)? That is, mimic the process that...

yep, good point about the edge cases. I am thinking about this for an improved process for testing external contributors' PRs (so more likely a 1:1 mapping). But perhaps one...