On 2004-03-15 Edgar Lovecraft <exim-list@???> wrote:
> Andreas Metzler wrote:
> ..[snip]...
> > > Why?? All that really needs to be done to is to change HOW you are
> > > doing the sender callouts.
> > [...]
> > I think[1] you are missing the point. - Steinar is not concerned about
> > /his/ callouts but the callouts other machines will make to his exim
> > and these are not under his control.
> > cu andreas
> > [1] I am not 100% sure but I cannot see how greylisting can have an
> > effect on callouts /issued/ by this machine that s using greylisting.
> Okay, perhaps I am missing something in regards to the entire greylisting
> thing (and yes I have read the papers).
[...]
I think this is correct.
> Other than possibly delaying email for longer than necessary, this would
> not be a problem even from a callout setup,
Afaict delaying email for longer than necessary is the problem:
* Both host doing greylisting with callouts: delay is doubled.
* client is doing greylisting, server is doing callbacks: People on
the server suffer greylisting delays (because the callback is
greylisted) although they don't want it. Introducing greylisting on
client suddenly delays *sending* instead of *receiving*.
[...]
> As to rejecting at the start of the DATA session, exim does this now as
> long as ALL recipients have been either rejected, defered, or discarded
> BEFORE the data command is issued by the CLIENT:
> '503 valid RCPT command must precede DATA', and in a greylisting approach,
> are not ALL of the recipients for a given message defered, denied, or
> discarded??
[...]
This DATA rejection happens in *addition* to the RCPT rejection and
therefore does not help at all in this respect. - Steinar wanted to
*accept* to stop the RCPT double-delays.
cu andreas
--
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"
