Re: [dmarc-ietf] Ticket #55 - Clarify legal and privacy implications of failure reports

[Douglas Foster <[e-mail]>]

Sorry that my name was dragged into this.   The study in question was one
provided by Laura Atkins several months ago.   I forwarded it to Michael to
bring him up to speed.   To date, it is the only study submitted to this
group.  I was not trying to introduce new information or re-ignite this
fight.  The chairs and our charter call for us to proceed with DMARC as a
tool to control misuse of the From address, which is also what I want.

To that end, we want to make sender, intermediary, and forwarder interests
converge as much as possible.    Our strategy is for intermediaries to
provide sufficient information for an evaluator to fairly judge whether a
particular message is in his interests or contrary to them.   By my
analysis, we are not there yet.

The particular information that I seek to achieve this result is the
following:
(a) know whether forwarding action(s) have occurred during transit
(b) reconstruct the message identity state prior to the forward(s), so I
can evaluate the message based on both the intermediate and final message
states.

I cannot reconstruct prior state using most ARC sets, because ARC only
provides an identifier if that identifier was evaluated.  Even then, some
of the identifiers are provided in comments or not at all.   Evaluating the
message based on its prior state requires a complete set of identifiers.

DF