Re: [Exim] Wish list (I think) regarding sender verify callo…

Top Page
Delete this message
Reply to this message
Author: Alan J. Flavell
Date:  
To: Tony Finch
CC: Exim users list
Subject: Re: [Exim] Wish list (I think) regarding sender verify callout.
On Tue, 1 Oct 2002, Tony Finch wrote:

> "Alan J. Flavell" <flavell@???> wrote:


> >Unless there's an overwhelming reason to accept mail from hosts that
> >are known to behave that way, I'd vote for putting them into a
> >blacklist with a special message.
>
> Why not just let the callout do the job?


It's a fair question.

I already felt the need for a stanza to deal with mail domains whose
MX hosts either persistently did not respond, or which cause some kind
of persistent "temporary" error, leading to days of repeat attempts
until the offering host finally gives up. So that was available
already.

Now, fair comment: if we do a callback to verify the envelope sender,
and the callback gets a raspberry when we say MAIL FROM:<> , then
we're going to react by giving the offered mail a 5xx and say that
sender verification failed. Maybe that's all that could be expected
of us...

But somehow I wanted them to get a different kind of response, in the
hope that someone, sometimes, might take the hint that it's their
server which is misbehaving, rather than there necessarily being
anything wrong with the email addresses that are involved. Which was
why I was deliberately sending it to that error section I mentioned
above.

> >On the other hand, there's no point in listing a domain for callbacks
> >if they accept any RCPT TO that you give them, and only reject bad
> >ones later.
>
> Actually there is: it means that they have to deal with the double
> bounce not you.


I think we must be misunderstanding each other here. I was just
saying there's no point in asking a particular server the question "is
this address OK?" if we know that _that_ particular server will
always answer "yes, it's OK". So we can leave that server off our
callbacks list.

all the best