On 25 Aug 2005 at 17:07, Herb Martin wrote about
"RE: [exim] Anti Phishing Trick":
|...
| > [mailto:exim-users-bounces@exim.org] On Behalf Of Fred Viles
| > Sent: Thursday, August 25, 2005 1:31 PM
| > To: exim-users@???
| > Subject: RE: [exim] Anti Phishing Trick
|...
| > Which is why message bodies should not be included in DSNs.
I should have added "to external senders". Accept-then-full-bounce
to local MUA submissions is reasonable and appropriate.
| BUT enough of the body NEEDS to be sent to make the rejection
| useful in most cases.
Interesting point. How much is enough, do you think? And is
"enough" for legitimate messages less than "too much" for spam?
I don't return bodies at all. The headers already include the
recpient, subject, and date. The sender in most cases will get the
DSN almost immediately after sending the failed message. That should
be sufficient to identify the message to the sender ISTM.
The other use for returning the body is so the sender can resend if
he didn't save a copy. But in that case he needs the whole body to
be useful.
It's a moot point for a lot of us, I guess, since we don't generate
DSNs to external senders but once in a blue moon. If then - OTTOMH I
can't think of a scenario that would cause my system to generate one.
- Fred
