OK, I won't try it.  But I will ask a few questions about the motivation 
behind some of the practices mentioned.

On Wednesday 03 August 2005 14:39, Marc Sherman wrote:

> This has been discussed many times before on this list; that is very
> risky, as for every site that refuses your mail with a 550 on the RCPT,
> you'll see a dozen that either wait till the end of DATA to refuse
> (which you'll never reach on a callout), 

Why would you configure that way?  It doesn't reduce incoming traffic, 
which I thought was the goal of blacklisting.

> or accepts after DATA and then 
> dumps the message in a spam quarantine or even /dev/null -- possibly
> after learning it as spam in their bayesian filter, poisoning your
> attempts to resend a week later through the smarthost when your
> intended recipient calls you to ask where that email you said you sent
> last week was.

That seems like a strange way to train a Bayesian filter.  Shouldn't a 
Bayesian analyser be trained on manually tagged data (i.e. mails that a 
human has classified as good and bad)?

> We're not being belligerent when we say that what you want can't be
> done; we're being helpful, because we've seen all these options
> proposed and shot down multiple times in the past.

Understood.

-- 
## List details at http://www.exim.org/mailman/listinfo/exim-users 
## Exim details at http://www.exim.org/
## Please use the Wiki with this list - http://www.exim.org/eximwiki/

Reply via email to