On Sat, Oct 19, 2002 at 11:51:50PM +0100, Alan J. Flavell wrote:
> > Actually it is a problem you don't want to bounce the message yourself
> > if you can avoid it.
>
> With respect, I think you're presenting the right argument but against
> the wrong targets. We all seem to agree that you don't want to bounce
> the message yourself.
Apologies if I misunderstood then.
> > The reason is that when the envelope from has been faked,
>
> Sure - but that's the responsibility of whatever offered you the mail,
> if you reject at SMTP time. It's not your job to second-guess what
> _they_ might do when _you_ reject it. (In an ideal world they
> wouldn't have allowed it to reach you in the first place...)
Right, but I don't make any expectations on servers who send me mail
anymore, unless I also admin the remote site :-)
> You _could_ accept it silently and drop it on the floor. That would
I am very much against dropping any mail on the floor. If a mail doesn't
reach the recipient and doesn't get bounced, it will not be my fault. My
job is to make sure the mail goes somewhere.
/dev/null is not an acceptable place for it.
(I could think of exceptions, but let's not go there)
> seem to offer some short-term benefits, but I don't think it would be
> proper, in the Great Scheme of Things. I'm real glad that we have a
> feasible way of rejecting stuff at SMTP time. If the sender then goes
> and does something stupid with our answer then I'd like them to fix
> that, sure - but it's not my responsibility.
We agree on that. If I refuse an unwanted mail from an a relay, there's
not much I can do if they can't bounce the mail. Ideally they'd have
checked the mail to and refused it from the source, but that's not my
problem anymore at that point.
Marc
--
"A mouse is a device used to point at the xterm you want to type in" - A.S.R.
Microsoft is to operating systems & security ....
.... what McDonalds is to gourmet cooking
Home page: http://marc.merlins.org/ | Finger marc_f@??? for PGP key
