Murray S. Kucherawy wrote:
I think it's redundant to refer to a signed message, since that's what the
entire document is defining; if the message isn't signed, the document
doesn’t apply in the first place.
So the question to me is more like: Is an intermediary adding a signature
generating a new message?
If the answer is "yes", then no change is required.
If the answer is "no", then it seems to me the simplest path is to change
"identity claiming responsibility for introduction of a message into the mail
stream" to "identity claiming some responsibility for a message in the mail
stream".
Good question.
I believe its depends on the intermediary type and what constitutes a
new message. IMO, the mere act of stamping a message with a
DKIM-signature is not a new message. An altered message at the header
level, but still the same message.
Maybe another way to consider this is if its reasonable to suggest
there are theoretical three levels of DKIM responsibility?
Creator
Signer
Sender
My OP was raising the point the sender is a passive transport process,
blind of the payload content or DKIM signature information.
On the other hand, we have the viewpoint of the receiver and the
sender is generally viewed the initial responsible source for sending
the mail and yet from an MUA (user) it may the FROM: as a responsible
party.
So we have Z-Order responsibility chain of technologies:
SPF targets responsible senders
DKIM targets responsible signers
MUA targets responsible creators
I guess when UIs begin to show signer info, the DKIM trust theory (and
hope) is the user beginning to shift to viewing the trusted signer as
the responsible party for the message and perhaps being to turn a
blind eye on the creator - the Cat's Meow for Direct Marketers.
Personally, I think its safer to view the SDID as the responsible
identity for stamping a statement of authenticity and trust.
Responsibility is really in the eyes of the beholder - receiver, DKIM
verifier and user.
So maybe the text could be reduced to:
A single domain name that is the mandatory payload output of DKIM and
that refers to the identity claiming responsibility for signing the
message.
This would allow DKIM-BASE to remove itself from the subtle
differences in responsibility party claims.
--
HLS
_______________________________________________
NOTE WELL: This list operates according to
http://mipassoc.org/dkim/ietf-list-rules.html