What about using structures?
struct t_count {
int input_rcpt, accept_rcpt,....;
} counters;
That's the most elegant solution to me.
....
small note: well, if it was completely up to me, I would go for classes
and C++, but that's a different story :-)
Ondrej
Constantine A. Murenin wrote:
>
> On 28/07/2008, Emmanuel Dreyfus <manu@...
> <mailto:manu%40netbsd.org>> wrote:
> > On Mon, Jul 28, 2008 at 06:48:55AM +0100, Constantine A. Murenin wrote:
> > > Well, that's more of an intended behaviour -- we want to use rcptcount
> > > to blacklist spammers that send messages to possibly non-existing
> addresses,
> > > and not valid cases where the recipients are actually good to go,
> right? :)
> >
> >
> > Ok, so we want two counters: one for valid recipients, and one for
> > attempted recipient. I wonder if we should also have tempfailed and
> > rejected recipient count available as well. Opinions?
> >
> > It seems rcptcount will go deprecated and we will introduce new
> > counter names. In order to avoid longer lines, I thought about that:
> > ircount: input recipient count
> > arcount: accepted recipient count (like old rcptcount)
> > trcount: tempfail'ed recipient count
> > rrcount: rejected recipient count
> > nrcount: not accepted recipient count (trcount+rrcount)
>
> Sounds a bit too excessive. :) I think rcptcount, counting the number
> of RCPT statements, whether successful or not, should be quite enough.
> These "ircount" statements are not saying much by themselves, and More
> is Less (tm), so I'd rather prefer an 'rcptcount' keyword with the
> obvious semantics that is not meaningless for greylisting.
>
> C.
>
>Message
Re: [milter-greylist] unbreak 'rcptcount'
2008-07-28 by Ondrej Valousek
Attachments
- No local attachments were found for this message.