> Does anyone know if the IETF is looking into
> this problem with an upgrade to the mail protocols?
The NOTARY RFCs (which have been approved and are just waiting on the
RFC editor to publish them) define:
+ a standard format for delivery status reports (delivery, nondelivery,
and "delay" reports)
+ standard, fairly precise codes for delivery errors
+ a means to request (via SMTP) which types of delivery status reports
should be sent
I sent the same analysis to the NOTARY list that I sent here;
I'm hoping that as MTA writers update their code to support NOTARY,
they will also improve error reporting.
The DRUMS working group (DRUMS = Detailed Update/Rewrite of Messaging
Standards) is updating RFC 822 and SMTP to clarify, fix things that are
broken, and make recommendations about the best ways to do things based
on experience. I would like to include some advice about error detection
and reporting here also; it's an agenda item for the WG meeting at next
week's IETF meeting in Dallas (but limited to 20 min discussion).
Both of these should help the situation somewhat.
On the other hand, the problems I see aren't so much in the protocols
as in the implementations, especially for mail gateways.