Re: [exim] Limit number of recipients

OK as the OP in this exchange I have been impressed by the depth of
knowledge show by the people on this list. The problem that made me post
originally is that we have some users who cut and paste a list of 1800+
addresses into the bcc line. I would rather they used the university
dedicated majordomo list server. What I would like to do is reject the
message with a permanent failure and an error saying something along the
lines of "too many recipients" nothing fancy like accept the first 100
or so.

There seems to be several ways of doing this, I was just wondering which
is the best. recipients_max_reject would seem to be the obvious choice
or am I missing something.

Anyhow thanks to those of you who have taken time to consider my problem
and offer answers, much appreciated.

Ron

On Thu, 2005-05-05 at 17:01 -0500, Edgar Lovecraft wrote:
> "Fred Viles"
> >
> ..[snip]...
> >
> > That's not all it does, it also 554's the DATA command. No recipient
> > gets the message, same as with the ACL approach. I think this is the
> > bit you and Tony are overlooking.
>
> I was overlooking that, I only skimmed the Docs on it when I went to
> re-read the spec before I originally posted, OOPS! Shame on me for that.
>
> >
> ..[snip]...
> >
> > (assuming you mean recipients_max+recipients_max_reject)
> > Right. And my question for you was: how is that different from
> > rejecting the DATA command based on $rcpt_count in an ACL?
>
> It's not, but again, I missed the 554 at DATA for recipients_max_reject
>
> >
> ..[snip]...
> >
> > But that's what the OP asked for. If you want to argue that what he
> > wants to do is a bad idea, that's fine. But it sounded like you were
> > just saying recipients_max+recipients_max_reject was a bad way to get
> > there, not that he shouldn't want to go.
>
> I am not arguing or disagreeing with what the OP wants to do, just that
> the first three or so posts back to him all delt with doing it in the
> ACL (which is more complicated) when there was no need to do it there.
>
> Now, even I can argue that there may be valid reasons to use an ACL over
> a global setting, as with ACL's it is easier to 'turn on/off' the
> desired policy for individual hosts/senders/recipients, etc. But that
> is an entirely different matter.
>
> Again, I jumped into this thread just to note that there is already a
> global configuration option(s) to do either.
>
> Either being temp fail over x amount or RCPT's, or perm fail x RCPT's,
> I only missed that the perm fail option also failed to ALL RCPT's
> not just thost after x number.
>
> > | It may be true that the OP does want to reject any message that has |
> > been sent to more than 100 recipients in the same transaction, in |
> > which case, Tony's suggestion does just that, and recipients_max_reject
> > | does not.
> >
> > If true, that's what I'm (still) missing. As I read the spec, it
> > *does*.
>
> My second time through, a touch slower this time, I read the same ;)
>
>
> --
>
> --EAL--
>
> --
>
>
>
--
Ron McKeating
Senior IT Services Specialist
Computing Services
Loughborough University
01509 222329