John Leslie wrote:
Frank Ellermann <nobody(_at_)xyzzy(_dot_)claranet(_dot_)de> wrote:
In 2.1.3 you say "bounce handling address". That's a neologism
for a concept that already has more than enough names.
I have sympathy for that name. I think we're heading towards
models where the RFC2821 MAIL FROM points to an MTA near the submission
server which will redirect DSNs in accordance with directives from
the originator -- after passing some validity tests.
We (Postini) do that extensively in our SMTP proxy server to manage
blowback. We aren't currently using a standard like BATV (we could, and
probably should), but the concept is certainly in production use now.
"Bounce" in the name is problemmatic because folks sometimes refer
to an error _during_ the SMTP connection as a "bounce". This, IMHO,
is just plain wrong; but perhaps we have already lost that battle...
It's amusing (albeit unhelpful) to note that, in those places where
Internet Mail and X.400 have the same features, they popularly use the
same language and terminology. Where we have trouble is features that
are unique to SMTP; different communities have created different
language to fill the gap. For example, at Postini, folks have turned
"500," "400," and "NDR" into verbs, like so:
"That setting will cause the server to 500 the message."
"We just four-hundreded the message, so it will retry."
"Exchange doesn't check addresses in SMTP, so it NDRs the message."
We actually use these with customers, because their complete lack of
ambiguity makes them immediately understood. Heaven help anyone who
attempted to codify these into a standard. :-(
I've also have heard "bounce" used for an SMTP reject (5xx return),
although that was among people who would never read a standard. I'm not
real worried about that case; and I've never heard the term "bounce
address" used ambiguously. Indeed, the reason I like it is because it
has always had 100% recognition even among people who'd never heard the