Daniel Edholm

Would it be an okay solution to use `systemd.path` to just monitor /dev/input for newly attached devices, and if the specified keyboard is not present then just exit the application...

I created https://github.com/donniebreve/touchcursor-linux/pull/52 as an initial attempt. It works locally for me, plugging the keyboard in and out several times. Also re-attaching the dock, with multiple (input) devices attached, works.

I'm new to Github PRs, but I seem unable to push additional changes to my opened PR. Either way, I have an additional commit coming with updated documentation (README.md) to...

I was using VSCode to authenticate against Github, which worked with the https clone I had done since it can handle OAuth and 2FA. However, it did not allow me...

Thanks for giving it a spin! :) From what you describe it certainly sounds like a restart will be necessary, as opposed to just trying to start the service anew....

I've not had much time to take this further lately. However, I did manage to get a watcher working in the end, and I **tried** to clean up my mess...

I made the solution look better, and probably be the more systemd way of doing things. The previous commit would also eventually break completely for me, since the restarter service,...

@donniebreve I really thought I had checked that the service was indeed *restarted* and not just started. Also, based on a comment (that I had accidentally left in the servce.path...

I took another look at it again, and I'll be damned if this isn't possible to solve with systemd! Not that I prefer it any which way, but it just...

> I tried these changes, but I am still stuck at the same issue. The service does not get restarted when a new keyboard is plugged in while the previous...