Basically if we don't hard kill the container on both timeout and a wait error we can get in state where it's pretty difficult to kill the supervisor actor from the parent since the root perms get dropped in the parent after the child is spawned...
This also just changes the error raised when a nasty log is detected to instead be the more explicit ApplicationLogError which better describes what/why we're raising.
Hopefully the final fix for this 😂.
Basically if we don't hard kill the container on both timeout and a wait error we can get in state where it's pretty difficult to kill the supervisor actor from the parent since the root perms get dropped in the parent after the child is spawned...
This also just changes the error raised when a nasty log is detected to instead be the more explicit
ApplicationLogErrorwhich better describes what/why we're raising.