On Wed, 23 Sep 1998, Evan Leibovitch wrote:
> As for exim relevance, is there *anything* being discussed at a high level
> about regarding a receipt-confirmation mechanism -- this one *could* be
> implemented at the MTA level if enough systems agreed on the format of
> the header.
My own personal belief that the best protocol for learning
whether your message has been read is putting the magic letters RSVP at
the bottom of it. This gives you some assurance that the message has got
into the recipient's brain and not just been displayed on a screen.
However, despite my belief,
I read the DSN RFCs. I even put a bit of code into Exim (currently
excluded by compile-time conditionals) to accept the data on incoming
SMTP calls, carry it around with an address, and transmit it on output.
However, I got totally bogged down in trying to understand how to use
the DSN data correctly. Sending a message when a delivery fails or is
delayed is easy - Exim does that already. All the other cases are
conceptually difficult. How does Exim know when a delivery is a "final"
delivery? The most difficult thing of all is the area of aliasing and
forwarding. The RFCs are full of weasel words and I concluded that it
just wasn't clear enough to attempt anything further. Others have agreed
with this position.
You can, of course, configure Exim to look for certain headers and do
things if they are found. For example, if you set up a transport like this:
xxx:
driver = autoreply
from = $local_part
to = $h_return-receipt-to
subject = message received
text = Message from <$sender_address> re $h_subject has reached my inbox.
you could drive it with a director of the form
yyy:
driver = localuser
condition = ${if !eq {$h_return-receipt-to:}{}{yes}{no}}
transport = xxx
unseen
It doesn't quite guarantee what it says, because delivery to the inbox
might get deferred. Individual users can do similar things in their
.forward files.
Personally, I *hate* it when I get an automatic response saying "Your
message has reached my inbox. I'll get to it as soon as I can." It's
just so much noise.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
--
*** Exim information can be found at
http://www.exim.org/ ***
