Comment by wolletd

Just adding `set -e` also exits the script when a subshell exits with non-zero error code. I'm not sure why I would leave `set -e` out in any shell script.