On 23 Mar 2004 at 21:05, Alan J. Flavell wrote about
"Re: [Exim] Greylisting + multiple M":
| On Tue, 23 Mar 2004, Bruce Richardson wrote:
|...
| > If greylisting becomes at all common, won't spammers simply adjust
| > their programs to retry at least once?
|...
| But to go back to your question: they'd have to retry _and_
| guesstimate how long to wait before trying.
That wouldn't be an issue. All they have to do is implement retry
timeouts the same as a typical legitimate MTA. IOW, for greylisting
to become ineffective, spamware has to act like a well-behaved MTA.
The idea as I understand it is that that would make sending spam a
much more expensive operation, which would still be a beneficial
effect.
I'm not so sure it is a valid assumption, however. I think normal
MTA queue management and retry behavior would indeed be very
burdensome for the spammers, but ISTM dedicated spamware wouldn't
need to do that. A technique like simply running through their
address list twice in two hour chunks would do the job.
- Fred
