On 14 Dec 2007 at 18:41, Alex van den Bogaerdt <ietf-smtp(_at_)imc(_dot_)org>
On Fri, Dec 14, 2007 at 05:13:58PM +0000, Tony Finch wrote:
To stress: The first and second are very similar, but they are not the
same and I think the differences are important.
In which case we should continue to use the term "alias" for type 1, and
we should continue to discourage type 2a lists that don't change the
Indeed. owner-ietf-smtp(_at_)mail(_dot_)imc(_dot_)org can automatically
me should my email address become unreachable. Someone posting to
this list cannot, and thus has no use for bounces, out of office
messages, delivery notifications and what more.
I just asked mail.imc.org to expand ietf-smtp(_at_)imc(_dot_)org(_dot_) It
member, majordomo - the processor and the resubmitter.
In which case, why do DSNs for mail to this address show as delivery to a
mailing list (in the sendmail alias-of-more-than-one-destination sense)?
I know this is probably an implementation detail, but it seems to me that
sendmail ought to be delivering mail to a mailbox under the current 2821
reading of what an alias-based mailing list is. Perhaps some other
characteristic of an alias sets off a sendmail-specific behaviour (the
owner-x syntax?) but in this case it's programmatically annoying since
some directly wired tools given aliases from the mailer (like Ecartis, for
which addresses of lists seem to show up as mailboxes at most sendmail
sites running it) need the return path to be accurate the whole time. The
bouncer processor in Ecartis, for instance, simply can't work without a
valid "From x(_at_)example(_dot_)org date" line, but does the SMTP loopback
connection thing itself in order to prepare new submissions.
Sabahattin Gucukoglu <mail<at>sabahattin<dash>gucukoglu<dot>com>
Address harvesters, snag this: feedme(_at_)yamta(_dot_)org
Phone: +44 20 88008915
Mobile: +44 7986 053399