Author: James P. Roberts Date: To: exim-users, William Thompson Subject: Re: [Exim] Exim+SA: spamc non-0 exit when spamd is down = transport_filter panic
> > Indeed. Transport filters are not specific to the pipe transport.
You > > can have a transport filter for any transport. Transport filters are
> > assumed to be be things that normally work - they were originally
> > implemented for modifying messages on the fly.
> >
> > > Another solution might be to modify Exim to apply "temp_errors" to the > > > transport_filter command. Wouldn't that be best, for the same reasons > > > "temp_errors" is used for the main pipe "command"?
> >
> > No, because "temp_errors" is specific to the pipe transport, and
> > transport filters are not specific to the pipe transport.
> >
> > I suppose that in principle some kind of new generic option such as
> > transport_filter_temp_errors could be invented, but it seems a bit
> > over-the-top to me.
>
> How about just ignoring the filter if it failed? or maybe a
> "filter_failure_ok" option?
>
I think you should consider the suggestion someone else already posted;
that is, wrap the filter daemon in something to catch the error flag,
and pass it back to Exim with no error, as desired. Or patch the filter
itself to provide the desired error flag behavior. Trying to fix it in
Exim is not necessarily a good idea, since it is a multi-purpose
interface. I would recommend against "tuning" such an interface for a
particular application. Better to fix the application to conform to the
published interface.
Rather like asking the power company to add a 4th wire to every house
outlet, because you want to have an "on-off" feature, and your appliance
lacks a switch. Is it not better to put a switch between the wall
outlet and your appliance? (I exaggerate, of course, but the point is
valid, none the less).
