Terje Petersen wrote:
TP> When I hit reply I expect to be corresponding with the PRA.
MW> That would be funny, because the PRA should be b(_at_)b(_dot_)com,
MW> which is yourself!
TP> No! According to your example I am C(_at_)C(_dot_)com(_dot_)
TP> You have not addressed the issue.
In the original example b(_at_)b(_dot_)com and c(_at_)c(_dot_)com are the same person. b(_at_)b(_dot_)com
is an address that only exists to forward mail on to c(_at_)c(_dot_)com(_dot_) If you, as
c(_at_)c(_dot_)com sent a reply there, it would just get forwarded back to you.
This case is I would guess the main reason why the Sender and Resent-
headers exist.
There are a number of identities in an email. In 2822 there are From,
Reply-to, Sender, and Resent-From to name a few. In 2821 there is
currently the, HELO domain, the MAIL FROM and now the proposed
SUBMITTER. The reason for all of these is that there are legitimate
cases where these identities will differ. (SUBMITTER actually must be
the same as one of the 2822 headers chosen via the PRA algorithm, but it
is not gauranteed to always equal any one of them). The debate in this
group has not been about whether it should be possible to represent the
fact that lots of people may touch and claim some level of
responsibility for an email between the end points. It has been about
which of those identities to check and how.
Robert