On Wed, 19 Nov 1997, Philip Hazel wrote:
> Remember that while you are sitting there dribbling out stuff slowly,
> you are tieing up *your* resources as well.
Yes, but just about every form of recipient-based spam blocking requires
extra effort/resources at the receivers' end. Table lookups are pretty
cheap, but the RBL proposal means you have to query another server
every time a connection is made (its maintainers specifically request
that RBL replies *not* be cached).
So spam-fighting almost always costs resources by the one(s) doing the
fighting -- our only question left is what degree of trying up local
resources is reasonable.
> You are keeping an SMTP
> connection open, thus reducing the number available for everyone else
> (if you have a limit set), and you have a process running which is
> taking up real and virtual memory and other process resources.
...resources that were used up at one time by my systems' non-consensual
relaying of spam. I'm not saying this approach is for everyone. I'm just
curious to know whether there's a significant body of people out there
who are *so* ticked off by spam that they're willing to make the leap
from passive resistance (merely rejecting spam when it happens) to
actively fighting it (using local resources to make spammers' lives
difficult).
Such resources are, after all, the admins' to waste. If there are cycles
and sockets to spare, why not spend them fighting spam?
> Exim sends its SMTP responses in single write() calls. I would not want
> to do otherwise. It would be silly to send them out one character at a
> time. Besides, they are typically not very many characters long.
Fair enough. What about a configurable response delay, then, such that the
full message displays together, but maybe X seconds after the request is
determined to be rejected (X being a number that is as long as possible
without causing a timeout).
Think of it. Even if we made every spam connection take just four seconds
longer to fail, we could add more than 110 hours to the time it takes to
complete a 100,000-address mailing.
- Evan
--
*** Exim information can be found at
http://www.exim.org/ ***
