Re: [yam] Issue #10: RFC 5321 3.9: add tiny subsection

[S Moonesamy <[e-mail]>]

Hi Alessandro,
At 10:28 12-12-2009, Alessandro Vesely wrote:
>This might just be
>
>  3.9.0 Plain forwarding
>  Plain forwarding implies both the envelope originator and recipient
>  addresses fields are left intact. The message is transmitted to the
>  next hop.
>I agree some confusion may arise because of the "forwarding" versus 
>"forwarding or delivering" --which may be dealt with by skipping a 
>transaction-- but beware of the "forwarded or redistributed" in the 
>current text... I'd propose a replacement for the first two 
>paragraphs of section 3.9 as follows:
>
>  When a message is sent to a forwarded address (also called an
>  "exploder", or  a "pseudo-mailbox"), copies are forwarded to each
>  target mailbox in the expanded set. In case the resulting target
>  address corresponds to a local mailbox, servers MAY skip an SMTP
>  transaction and just deliver a copy of the message to the relevant
>  mailbox. When replacing an envelope recipient address with multiple
>  target addresses, servers SHOULD simply utilize the expanded set;
>  application of heuristics or other matching rules to eliminate some
>  addresses, such as that of the originator, is strongly discouraged.
>
>  Existing header fields (RFC 5322 [4]) MUST be left intact; in
>  particular, the "From" field is unaffected. Forwarding across
>  different domains may be frustrated by policies aimed at transport
>  level integrity (see Section 7.1). Hosts MAY go along adding header
>  fields, choosing an adequate expansion method, and using supported
>  extensions as required.
>
>Again, that text is just for the WG to consider the changes:
>
>* header fields, e.g. signatures, may be added on forwarding,
>* expansion does not necessarily imply multiple delivery,
>* forwarding may break recipient's policies,
>* terminology...

I'll add the above text to Issue #10.

>Finally, several mailers replace the 5322.To field. Shouldn't this 
>be mentioned? Or, why do we consider the header at all?

The following is an individual comment.  There is a discussion of 
changes by a SMTP server in Section 3.7.1, Section 3.7.4 and Section 
6.4.  Section 3.9 mentions a requirement for the message header 
section.  There is also some text in Section 7.2 about attempts to 
deduce relationships between the envelope and the header section and 
use them to alter the header section of the message for delivery.

Regards,
S. Moonesamy
YAM WG Secretary