ietf-822
[Top] [All Lists]

Re: DSN redesign

2002-05-07 22:37:53

At 00:37 08/05/2002 +0000, D. J. Bernstein wrote:
Eric A. Hall writes:
> Secondarily, some of the recipient-specific
> fields can't be implemented as shared header fields

Have one field for each recipient, naming the recipient:

   Notice-Requested-Upon-Delivery-To: ehall(_at_)ehsco(_dot_)com

What about BCCs? Also, I think the "guaranteed" 'some response' of DSN is good.

OTOH, it wouldn't be impossible to extend DSN (EDSN?) to make it 'wrap' a DSN request into the header the first time it sends back a 'Relayed' notification. Then when a EDSN-aware MTA receives a wrapped DSN header it re-creates the DSN envelope information

So, you have something like:

MUA     (DSN aware)     - adds DSN envelope info
MTA1    (DSN aware)     - receives, and passes on DSN info
MTA2 (EDSN aware) - receives DSN info, sees that MTA3 doesn't know DSN so creates a 'DSN-Info:' header with the DSN information encoded into it (not difficult to come up with a schema for doing this). Also sends back a 'relayed' notification with text indicating that 'DSN may be resumed at a later hop' MTA3 (not DSN aware) - receives message with DSN-Info header. Since it's not DSN aware it just ignores the DSN-Info header and passes it on MTA4 (EDSN aware) - receives message with DSN-Info header. Recreates the DSN info in the envelope and removes DSN-Info header. (Optionally sends back a notification saying that this has happened) MDA (DSN aware) - receives message with DSN info in header and acts on it, sending back success notifications.

Comments?


Paul                            VPOP3 - Internet Email Server/Gateway
paul(_at_)pscs(_dot_)co(_dot_)uk                        http://www.pscs.co.uk/



<Prev in Thread] Current Thread [Next in Thread>