Re: Re: [COMMITTERS] pgsql: Don't override arguments set via options with positional argumen

On tis, 2012-04-17 at 19:19 -0400, Andrew Dunstan wrote:
> It was discussed. I think the previous behaviour is a bug. It can't be
> sane to be allowed to do:
>
> initdb -D foo bar

It's debatable whether it should be allowed. I don't see anything wrong
with it. After all, later arguments usually override earlier arguments,
and non-option arguments notionally come after option arguments. Also,
if this should be disallowed, would you also disallow initdb -D foo -D
bar?

But what I think is worse is that the "bar" is silently ignored now. If
you think that this is an error, it should produce an error.

My vote is to revert this altogether and leave it be. In the
alternative, make it an error.