Re: [ietf-smtp] Delivered-to and Return-path

[John C Klensin <[e-mail]>]

--On Friday, February 19, 2021 19:36 -0800 Dave Crocker
<[e-mail]> wrote:

> On 2/17/2021 9:55 PM, John C Klensin wrote:
>> If I correctly understand the intended use of Delivered-to and
>> the meaning of "added at the time of delivery", I believe that
>> the examples in Section 4 are in error.   The example does not
>> show a Return-path and I'm having trouble figuring out a case
>> in

> You appear to be adding a requirement for the example to be
> exhaustively complete.

No.  I am just suggesting that, given the obvious relationship
between one as a forward pointer and the other as a backward
pointer, both assigned at or after the time SMTP hands the
message off to something other than another SMTP server (acting
as such), an example that shows "Return-path:" as well as
"Delivered-to:" and "Received:" would be a bit more clear and
helpful.   Please note "suggesting" in the previous sentence,
which I do not believe to be a synonym for "a requirement".

> That isn't claimed or intended in the current example text. It
> merely seeks to show a sequence of Delivered-To fields, with a
> bit more context.

Then, as an alternative to the suggestion above, I would suggest
a sentence somewhere that makes it explicit that the examples
show only those header fields that are relevant to what you are
illustrating.

Dave, I am not the enemy.  Given that the header field is
already implemented and deployed, I favor having a spec go
forward that provides implementation and interpretation
guidance.  I am just trying to constructively point out a place
where the document might be made more clear for the benefit of
those who do not already understand how the field is used and
who are looking for implementation guidance.

     john