Description
So, I see that y'all gave up on tty detection because of some obscure bug in #9035.
But can we have switches in docker exec/run and docker compose exec/run to turn on tty detection when we want to use it?
Having to deal with -i, -t, and -T is driving me insane, especially since the behavior of docker exec and docker compose exec is flagrantly inconsistent in this regard (though the inconsistency is a separate issue):
> echo 'SELECT 1;' | docker exec db psql -U postgres
# (no output)
> echo 'SELECT 1;' | docker compose exec db psql -U postgres
# (outputs "the input device is not a TTY" and exits)
> echo 'SELECT 1;' | docker exec -i db psql -U postgres
# (operates on piped input)
> echo 'SELECT 1;' | docker compose exec -i db psql -U postgres
# (outputs "the input device is not a TTY" and exits)
> echo 'SELECT 1;' | docker compose exec -T db psql -U postgres
# (operates on piped input)
Description
So, I see that y'all gave up on tty detection because of some obscure bug in #9035.
But can we have switches in
docker exec/runanddocker compose exec/runto turn on tty detection when we want to use it?Having to deal with
-i,-t, and-Tis driving me insane, especially since the behavior ofdocker execanddocker compose execis flagrantly inconsistent in this regard (though the inconsistency is a separate issue):