Re: Re: User-based white- and blacklists

Marc MERLIN wrote:

>>http://www.exim.org/pipermail/exim-users/Week-of-Mon-20030317/051275.html

> Tom, have you checked that remote MTAs will not be spooked out by getting a
> 45x after the xth rcpt to, and will all correctly go on to data and deliver
> to the receipients that were accepted?

 From what I have seen so far on my 2 systems, there doesn't seem to be 
a problem. A "proper" MTA should handle this situation well, because I 
could also think of "legit" 4xx reponses ("Mailbox over quota", 
"partition full", or similar).

> I'm a bit afraid that some broken MTAs may just see a 45x, and issue
> a quit instead of data for the ones that were accepted.

I can confirm that _MUAs_ can't handle this. So you must make sure to 
skip the procedure for typical MUA delivery paths (ASMTP, Allowed Relay 
Networks).

>>However, I find it so horrible ("il s'oppose meme la nature", as my
>>granny might have said)

It's the best you can get if you want both SMTP time rejection and 
per-user settings. I can assure it can't get any better, unless we 
change the fundamentals of SMTP :)

I hope I can post to the sa-exim list, since I'm not subscribed ... :)

/tom

-- 
Tom Kistner  <[email protected]>
ICQ 1501527  dcanthrax@efnet
http://duncanthrax.net