ietf-822
[Top] [All Lists]

Re: I-D Recommendations for Automatic Responses to Electronic Mail

2002-06-06 13:07:00

Keith Moore <moore(_at_)cs(_dot_)utk(_dot_)edu> writes:

one reason I didn't just insist that MAIL FROM be <> is that too many
mailers categorically reject MAIL FROM:<>.  yes, they're broken, but
there's so much of this kind of brokenness in the world that a feature
that depends on it not being broken is doomed to failure...

there are also things that rewrite <> to MAILER-DAEMON or some such,
presumably because mailers don't accept <>.

The other reason not to use <> is because if you use <>, the mail system
will generally believe that the message is a bounce, and if it then
bounces itself, it will send the message to the local postmaster as a
double-bounce.  Even worse, if the message is relayed through other hosts
before it bounces, it will often end up in the postmaster box of some
random mail system not run by the same person as the autoresponder.

For many autoresponders, no one cares or wants to know if the autoresponse
bounces; they want that bounce to silently disappear.

Now, one can send autoresponses from a system specially configured to
discard double-bounces, but one doesn't always have that degree of freedom
to bring up a separate system for ones autoresponder needs.

I use nobody@<host>.stanford.edu as an envelope sender for autoreplies,
generally, where nobody is configured to silently discard all mail.

-- 
Russ Allbery (rra(_at_)stanford(_dot_)edu)             
<http://www.eyrie.org/~eagle/>

<Prev in Thread] Current Thread [Next in Thread>