Closed akashmod closed 1 year ago
Summary notes for any reviewers: This covers similar ground as PR #249 (5 years old and waiting for changes) and #310 (4 years old and VERY minimal).
I think whatever happens to the socketcan_bridge auto-recovery code, the ThreadedSocketCANInterface should be merged as a (potentially) separate PR as it fixes a seemingly missing piece from ThreadedSocketCANInterface recovery code (as I ran into in issue #467).
Also, shouldn't ThreadedSocketCANInterface virtual void recover
have a bool return type?
The virtual void recover should have a bool return type yes. That needs to be corrected.
Hello, Tank you very much!! This was giving me a big headache, because we actually have a device that has on/off using dead man button and had to restart the node in order to send new data. Been following #467 since then, this seems to solve the problem.
Is it possible to know approximately when could this be updated to noetic release?
Thank you again for the effort!
This is to solve issue 468. A ros timer is started to check the driver state after every specified duration and to reinitialize the driver for the node if the driver has recovered.