Comment by blankverse

So, in the worst case, for achieving basic code reloading in other erlang like systems(ex.- cloud haskell), we could let every process in the cluster crash and restart with the new version?