tobold.org

correct • elegant • free

△ comp.mail.* △

◅ Return Receipt To and Re: Standard Mail Headers

Looking for doc on Content-Length ▻

Where are bounces supposed to go?

In article <2m6041$cb5@bosnia.pop.psu.edu>, David Barr <barr@pop.psu.edu>
wrote:
>   The someone else contends that a bounces should go to the
>address spelled out in section 4.4.4 of RFC 822.   Basically this is
>the address in the "Sender:" line, or failing that, the
>"From:"/"Reply-To:" lines.
>
>Who's right?  If me, what is this section of 822 talking about then?

You are indeed right (as Erik has explained).

I have a couple of points which may shed some extra light (or not :-).

1. Certain RFC-822-like protocols (notably the JNT mail protocol, aka
`greybook') do not carry the originator in a separate envelope, but rely
on being able to extract it from the message headers.  (I doubt that
it was the intention of RFC-822 to sanction this, though; greybook has
to extend RFC-822 (with `Via:' lines) so that the full return path is
available.)

2. By my reading, section 4.4.2 guarantees that `Sender:' (if present,
otherwise `From:') will be the same as the original envelope originator.
(The latter may of course be updated, for example if the message is
processed by a list expander.  In this case, greybook mailers have to
update `Sender:' too.  Yuck.)

Tim.
--
Tim Goodwin | "scanf()... usually does something almost but not
PIPEX Ltd   | completely unlike what you want" -- Chris Torek.

Original headers:

From: tim@pipex.net (Tim Goodwin)
Newsgroups: comp.mail.headers
Subject: Re: Where are bounces supposed to go?
Date: 16 Mar 1994 18:36:23 -0000
Organization: PIPEX Ltd, Cambridge, UK.
Message-ID: <2m7jj7$hfd@tank.pipex.net>
References: <2m6041$cb5@bosnia.pop.psu.edu>

△ comp.mail.* △

◅ Return Receipt To and Re: Standard Mail Headers

Looking for doc on Content-Length ▻