So is the normal usage path to start using osh for interactive use and then use osh in a bash compatible mode to write scripts that are less buggy but can still be run on bash?
Or is it more common to write scripts that tend to then require the osh runtime exist in whatever context they are deployed to?