Evan Martin wrote:
> Like I said in my original post, I understand the workaround. I just
> think that:
>
> 1) The workaround requires extra work for each developer (or at least
> each client application) using PostgreSQL, while a fix in PostgreSQL
> would solve this once and for all.
For a developer it should not be a problem to write an
auxiliary function that kills the connections and drops the
database.
> 2) The workaround requires superuser privileges, which I don't think
> should be required to drop your own database.
You said that such a feature would be useful for developers.
Developers usually have superuser privileges.
PostgreSQL 9.3 will probably allow you to pg_terminate_backend()
your own sessions even if you are not a superuser, at least there
is such a patch in the queue.
I don't think that saving a few keystrokes for lazy developers
is a good enough reason for such a thing in core.
Yours,
Laurenz Albe