> > I don't want to skip syntax errors - I do indeed want an error to
> > occur. However, I would rather have the error go to
> > postmaster@$domain rather than to the original sender.
>
> I haven't tried this, but what the code is supposed to do is to defer
> delivery on a syntax error. So no error message is sent to anybody.
> Or has this gone on so long that the address has timed out? Please
> confirm what is actually happening.
Yeah, that's what it is - it is indeed deferring, but because the
syntax error doesn't get fixed by whomever made the syntax error, it
eventually times out and bounces the error back to the original sender.
> > Is "errors_to" not applicable in this instance?
>
> No it isn't. I had never thought of it as being relevant; my assumption
> is that syntax errors are either skipped, or cause delivery to be
> deferred. In neither case is a bounce message generated.
>
> The manual tries to make this clear:
>
> If the router successfully handles an address, it may queue it for
> delivery or it may generate child addresses. In both cases, if there is a
> delivery problem during later processing, the resulting bounce message is
> sent to the address that results from expanding this string, ...
Indeed, reading it carefully this time shows that Exim is following spec.
Hmmm.
Not sure what to say at this point - it would be useful to _me_ to have
a way to control error control on final retry timeouts, but I don't know
if errors_to would be the right way to do it, or if its such an isolated
case that it isn't worth considering as a future feature.
Future feature? :)
Regards,
Dean A. Brooks
dean@???
