[Top] [All Lists]

Re: email-arch: bounce vs... return?

2007-10-15 06:38:04

On Mon, Oct 15, 2007 at 07:43:36AM -0400, John Leslie wrote:

   "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...

"Bounce" is indeed problematic, as folks sometimes refer to it
as an email which is received and then sent back.

Older RFCs mentioned in 2821 don't mention "bounce".

RFC2821 has this to say:

   Delivery SMTP systems MAY
   reject ("bounce") such messages rather than deliver them.

Here "bounce" is an alternative word for "reject".  In all fairness,
the text suggests this is done after receiving the body (i.e. some
5xx code after DATA... <crlf>.<crlf>).

   Servers MAY reject or bounce messages ...

It isn't clear if this means "choose between reject or bounce", or
rather "reject ("bounce")...".  Codes are 550 or 551, so my money
is on "reject ("bounce")" again.

   Similarly, the discussion in section 3.4 applies to the
   use of reply codes 251 and 551 with VRFY (and EXPN) to indicate
   addresses that are recognized but that would be forwarded or bounced
   were mail received for them.

I read this as 251->forward, 551->bounce.

And then there's the use of bounce in e.g. light.  Light isn't received
and then retransmitted when it bounces.  It is not entering the material
on which it bounces.  So, why would mail enter the MTA?