I realize the current code may be a bit hostile, but it's worth noting that you can most definitely monitor a dead process: you will just get the :DOWN message right when you monitor it and you'll get reason :noproc.

iex> pid = spawn fn -> :ok end
#PID<0.85.0>
iex> Process.monitor(pid) # "pid" is long dead here
#Reference<0.0.2.1188>
iex> flush()
{:DOWN, #Reference<0.0.2.1188>, :process, #PID<0.85.0>, :noproc}
:ok

Quoting the Erlang documentation:

Monitors are fired when the monitored process or port terminates, does not exist at the moment of creation, or if the connection to it is lost.

Yup, you should have no race but the proposed code is more readable, so thank you!

❤️ 💚 💙 💛 💜