Restart processes upon error about flock HOT 5 CLOSED

Just to be clear, are you suggesting that instead of daemonizing processes using nohup (as ZewoFlock and company currently do), Flock should integrate with systemd and/or related tools to automatically restart processes on failure?

If so, I think it's a smart idea and should be pursued in the server specific Flock packages (like ZewoFlock) rather than the general Flock package itself

from flock.

Why do you think so? Swift applications have a nasty tendency to crash with no way to recover (optional unwrap, index-out-of-bounds, fatalerror). I would say its pretty vital for a service like Flock that manages your application to restart it if that happens.

from flock.

My reasoning was essentially that, as of right now, Flock isn't actually in charge of starting any processes, be it through nohup or systemd; it delegates that work to ZewoFlock and the like. I could see value in centralizing this functionality in Flock though, as you're suggesting, since I would guess that regardless of Swift server library this functionality would be desired

from flock.

I decided to include systemd functionality within Flock itself, as you suggested, and then allow servers from within their own packages to take advantage of this functionality. VaporFlock and ZewoFlock both do so now

from flock.

Didn't see this until now - awesome! Thanks for being so responsive :)

from flock.