Alberto Soragna

@wjwwood I opened a PR to fix the unit-test for the events executor: https://github.com/ros2/rclcpp/pull/2465 However, I'm confused by the expected behavior here. The test was not triggering the waitables (this...

Ok, I think that there was also a problem in the `is_ready` method of the `TestWaitable` ``` bool is_ready(rcl_wait_set_t * wait_set) override { (void)wait_set; return true; } ``` This shouldn't...

Reopening this, since the test is still disabled for the `StaticSingleThreadedExecutor`, although it could be fixed by https://github.com/ros2/rclcpp/pull/2142

We just discussed this in the Client Library Working Group (with @mjcarroll @jmachowinski) The behavior proposed in the PR (throwing an exception if there's a qos mismatch) seems dangerous as...

> Why not allow this this opt-in The "opt-in" would consist of a subscription option, e.g. `bool throw_on_qos_incompatible`. IMO that's redundant. You already can set a custom callback for handling...

* Linux [](http://ci.ros2.org/job/ci_linux/20719/) * Linux-aarch64 [](http://ci.ros2.org/job/ci_linux-aarch64/15028/) * Windows [](http://ci.ros2.org/job/ci_windows/21402/)

> Note, there is something fishy going on, as the CbgExecutor receives more messages than the normal one The likely explanation is that publishers/timers are running faster (thanks to the...

> The main point of this executor is that it has similar behavior in single and multithreaded mode. Note, this is not the case for the current executors. That it...

I agree that this is a bug, however I don't agree with your recommendation > Since I think it's undesirable to throw in the destructor, I would recommend to catch...

We identified type adaptation as the main culprit for the build time increase between Galactic and Humble. I would strongly recommend to profile the build time changes caused by this...