On Mon, 20 Oct 2014 09:51:20 -0400 Miles Fidelman <mfidel...@meetinghouse.net> wrote:
> Joe wrote: > > On Sun, 19 Oct 2014 23:17:28 +0200 > > lee <l...@yagibdah.de> wrote: > > > >> Tanstaafl <tansta...@libertytrek.org> writes: > >> > >>> On 10/17/2014 9:24 PM, lee <l...@yagibdah.de> wrote: > >>>> You do not accept messages you can not deliver unless you are > >>>> relaying them. > >>> Absolutely wrong, this rule fully applies to relays just as it > >>> does final destination servers. > >> I'm not sure what you mean. How will you know whether messages to > >> a particular destination address can be delivered before sending a > >> message to that address so that you can decide whether to accept a > >> message you're relaying to that address? > >> > >> > > I think it's generally an admonishment not to get involved in > > relaying. The point of relaying is that the original sender cannot > > directly reach the recipient's authoritative mail server, in which > > case it can't generally query for recipient validity. > > Relaying happens all the time - e.g., when an organization designates > a single mail gateway, that then distributes to department-level mail > systems. > Yes, but there's at least a fighting chance in this case that the organisation can configure the gateway server to verify recipients locally. The problems occur where there is no real mail admin, where a small company outsources its spam-cleaning, and nobody in the company even knows what a recipient list is, let alone that their spam-cleaner should have one, kept up to date. Many small and medium businesses collect their main domain-wide by POP3, giving no way of automating recipient verification. > And, in the corporate world, NDRs from down-stream servers are > commonplace. > Which is why the spammers love them. -- Joe -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20141020201722.65ee0...@jresid.jretrading.com
