Reason for using gpg --batch?

[adigitoleo (Leon)]

Hello,

In password-store.sh, some option flags are unconditionally added to the
gpg invocation. These are --batch and --use-agent. While I understand
the use of the latter, I'm not so clear on what the reason is for using
--batch. The gpg man page suggests this is intended for pgrogrammatic
use of gpg, to avoid blocking or waiting for interactive input. However,
pass is a CLI program that quite happily asks for input in many
situations.

My reason for asking this is because I have recently discovered that gpg
can use a 'loopback' mode whereby it accepts the PGP key passphrase on
the stdin of the terminal where pass was invoked. This is convenient,
because I mainly use pass from interactive terminals, and therefore
don't usually want nor need the PGP agent to resort to using a special
GUI or ncurses interface just to ask me for the passphrase.

It seems to me that the simple answer would be to use `pinentry-mode
loopback` in .gnupg/gpg.conf, however the gpg --batch option is not
compatible with this setting. For now, I am manually patching the shell
script to remove that flag, and have not yet encountered any problems,
but would like to know if that has other implications.

Cheers,
Leon