Re: [OT] Dilemma about OS <-> Postgres interaction

On 18/06/04, Harald Fuchs (hf517(at)protecting(dot)net) wrote:
> In article <20040617232536(dot)GB17879(at)campbell-lange(dot)net>,
> Rory Campbell-Lange <rory(at)campbell-lange(dot)net> writes:
>
> > I should have mentioned that we need the messages sent very soon after
> > they have landed in the 'inbox'; otherwise cron would definitely be the
> > way to go -- including an @reboot line.
>
> This rules out a cronjob. How about writing a pl/perlu function which
> sends the mail and calling that function by an AFTER INSERT trigger?

I was trying to avoid having an untrusted language in the database.
After all, this may be the way to do, and I don't need to worry about
daemons or anything like that.

Even further off-topic:
How does a pl/perl function access perl modules? Is there a sensible way
of testing them?

Thanks
Rory
--
Rory Campbell-Lange
<rory(at)campbell-lange(dot)net>
<www.campbell-lange.net>