Anthony Campbell wrote:
> Maby thanks to everyone who has replied.
>
> 1 Eduardo:
>
> Not sure how to discover what the routers are doing. What command to
> use?
>
> But thanks for clearing up the part about procmail.
>
> 2. Bill
>
> I seemn to be getting into deep water here. I think it would be overkill
> for me to try to ditch procmail when it's working well in general, just
> in order to try to redirect the occasional email from one sender.
That part is dead-easy in Exim:
- a bespoke (hard-coded) router/transport set,
- a lookup / table-driven set,
or (in our case)
- PostgreSQL-driven. (most of our Exim is such - but I don't 'recommend' it,
as it is more flexibility than ordinarily needed, plus one more 'player' to
possibly fail (the RDBMS).
> Bouncing those from within mutt does work well, so I can do that, which
> isn't a big problem.
>
.. or any 'configurable' MUA. The one(s) in Mozilla/SeaMonkey/Thunderbird, for
example can do all manner of programmable stuff, and they are hardly alone. Nick
knoght's MR2/ICE for OS/2 et al was easily programmed to convert between and
among email, fax, voice-mail and so on.
Few of us need that these days, but Exim can do all that also.
> Icidentally, reading up on procmail I see that it will apparently
> forward mail, using "! user@???". Perhaps I should try that
> before going deeper into exim.
>
> Regards,
>
> Anthony
>
I suspect your use even of procmail may be 'overkill' - or at the least,
'parallel' to what you could do in Exim just as easily - so you may find moving
to an all-within-Exim solution less work than first appears.
One of the niceties is that you can easily take one situation at a time. Another
is the rich toolset in place for testing, debugging, logging, and analysis.
Arguably the best such toolset in the business, and well worth becoming (more)
familiar with.
YMMV, of course - BUT .. procmail or 'milter' use WITH Exim are relatively
uncommon situations largely because Exim just doesn't need the externals and -
over time - most of us who HAVE used them are too lazy to try to stay 'current'
with multiple sets of semantics and behaviours - more especially if they don't
need looked at but a few times a year and we have to retrain ourselves at each go.
Best,
Bill
