Closed nakane11 closed 1 year ago
What was the reason for needing the commit at https://github.com/jsk-ros-pkg/jsk_robot/commit/ffc10275694e046a1b01e1e3030da56dc47d6043 (or https://github.com/jsk-ros-pkg/jsk_robot/pull/1251/files) in the first place? Is there any issue if we revert it?" @Kanazawanaoaki
I'm sorry. I just remembered.
The reason I created https://github.com/jsk-ros-pkg/jsk_robot/pull/1251 first was because I wanted virtual_force_publisher
to be up and running by default in pr2.
The reason for this nodes killed problem is that the same node was added to different files in the two PRs https://github.com/jsk-ros-pkg/jsk_robot/pull/1251 (in pr2_bringup.launch
launch by robot
) and https://github.com/jsk-ros-pkg/jsk_robot/pull/1314 (in pr2.launch
launch by jsk-pr2-startup
), and we created https://github.com/jsk-ros-pkg/jsk_robot/pull/1314 without considering the possibility that https://github.com/jsk-ros-pkg/jsk_robot/pull/1251 would be merged and left it as it was.
(It seems that only https://github.com/jsk-ros-pkg/jsk_robot/pull/1314 had been included in the develop/pr2
branch, but both were included at the time of the noetic transition, etc., causing the problem.)
I don't remember why it was added to pr2.launch
, in the PR of the newly created one, but I think it was said that virtual_force_publisher
sometimes died, so it is better to use jsk-pr2-startup, which is easier to start up.
@nakane11 Therefore, if there is no particular reason, it may be better to delete virtual_force_publisher in pr2_bringup.launch and leave pr2.launch.
@Kanazawanaoaki Thank you for reviewing. I deleted virtual_force_publisher in pr2_bringup.launch and rebased the commits.
Thank you very much. I think it will be OK this time.
Currently both
pr2.launch
andpr2_bringup.launch
run virtual_force_publisher and the nodes are killed fornew node registered with same name
error. This PR makespr2.launch
stop launching the nodes.