It should be reversed: poke username -- path/to/command arg1 arg2 The command would fork, execute the child, wait for the child to finish, then send out its n

It should be reversed: poke username -- path/to/command arg1 arg2 The command would fork, execute the child, wait for the child to finish, then send out its notification. This would also allow flags like: poke --on-failure @koolba -- path/to/foo Or even handling the child taking too long: poke --on-lagging 20m @koolba -- path/to/something-that-might-hang