On Mon March 7 2005 17:13, willemien(_at_)amidatrust(_dot_)com wrote:
I specifically mentioned delivery notification because a
sender may receive such a notification from any of the
distinct message-handling components, and the significance
of such notifications (and the user's reaction) depends
on the component.
A that makes it much more clear but I still disagree.
section 2 is a (very) general discription of the mail architecture.
error messages nitty gritty belongs elsewhere.
Nitty gritty, yes. But notifications (success as well as
errors) are mentioned to illustrate the point; if I receive
a notification from a nearby MTA shortly after sending a
message (as has recently happened), it has quite a bit more
significance than a random notification from some random,
distant MTA (or anti-virus spamware, etc.).
Maybe it is better to create a seperate section on messages with subsections
- Delivery Status Notification
- Message Disposition Notification
- Message Filtering (SIEVE)
Filtering is an operation, not a message (while one thing
that SIEVE can do is sending "vacation"-like messages, that's
not its sole (or even primary) purpose).
And don't forget Message Tracking Status Notifications.
List expansion and forwarding are done by the MDA, (they depend on the
localpart of the email address)
OK, I'll buy that.
the MDA sends the message to an MTA and the MTA sends it to other domains.
Not if the alias/forwarding destination are local.