Hi,
On Mon 11/May/2026 20:23:52 +0200 Trent Adams wrote:
I’ve (finally) caught up on the really insightful comments sent to the list
about the value enabled by ARC… specifically it seems that folks find
utility in the final hop (if not the full chain). In fact, I’m impressed
that so many spent their weekend time thinking about this. It indicates
quite an honor to the work. So… thanks for the time and energy being put
into very thoughtful comments.
While trying to consolidate the input into a coherent frame for moving
forward, it seems to me that there’s a lot of commonality in the comments.
For example, many have said some variation of “ARC is valuable for [X]
reason”, with a roughly equal sense that “the current solution for [X]
reason needs additional work.”
To that end, some folks seem to be indicating that we should do that
additional work here, within the DMARC WG.
What we should do within the DMARC WG, according to the current charter, is
complete the analysis of ARC failure, assuming that it did not adequately solve
the problem it was intended to address. The operational outcome is before
everybody’s eyes: mailing lists largely switched to From: munging instead.
Completing that analysis is especially important because DKIM2 appears to be
encountering the same obstacle.
Here’s my suggestion… if this group agrees that the current ARC experiment
(as defined by RFC 8617) doesn’t hit the exact target (but may still have
some value), I’d suggest that everyone who wants to see the learnings from
this experiment implemented… that they join the DKIM2 work. The
collaborators are keenly aware of what worked (and didn’t) in ARC and are
baking into the DKIM2 proposals the best of what worked (and trying to
correct what didn’t).
I lost part of my enthusiasm for DKIM2 when the WG decided to redefine
well-known specifications, such as the tag=value syntax, on the grounds that
they belong to the old world. That stance does not seem to indicate much
willingness to distinguish what was actually problematic from what was merely old.
I also suggested adopting the concepts expressed in
draft-kucherawy-dkim-transform, in order to trace differences that can be
trusted to preserve the semantic integrity of a message. Instead, the WG
appears to prefer fully general difference tracking, whereby a DKIM2 message
can substantially diverge from the original content while still claiming that
the author's domain signature remains valid.
DKIM2 addresses several important problems faced by large mailbox providers,
such as DKIM replay. However, it does not seem to address the mailing list
problem, and I do not think it is the best venue in which to analyze why ARC
failed to solve that problem.
… side note… many of the folks working on DKIM2 were also collaborators on
ARC (and many have also implemented RFC 8617 to some degree)… so, they
understand the situation and are motivated to see the intermediary chain
model reach its potential. I hope everyone here is plugged in there, too.
With that, here’s my request to this group… I think it’d help for folks to
consider the following when posting to the list:>
* Do you think that ARC has no value at all and should simply be buried?
No.
* Do you believe there's “something” of value in the ARC model (even if it
falls short)?
Yes.
* Do you have concrete proposals for improving upon RFC 8617?
Yes.
In my opinion, if you use that as a framing exercise, we’ll find out if (a)
ARC should simply be marked “Historic” and be done, (b) we should move the
conversation for improving upon the model to the DKIM WG (after marking ARC
as “Historic”), or… (c) we respond to the current DMARC recharter with “no”
and someone proposes rechartering (again) to revise RFC 8617 (hopefully
based on a specific I-D proposing an improvement).
How about (d): we conclude that domain reputation alone is not an adequate
mechanism for solving the mailing list problem, that it needs to be
complemented by an authorization protocol, and we commit a draft stating so.
Possible discussion of further rechartering could then proceed on a clearer
technical basis.
Basically… we just need to answer the current charter question about
whether or not we should mark RFC 8617 as “Historic" (along with any
commentary)… knowing that doesn’t kill the concept itself, as it’s already
being folded into the current DKIM2 proposals.
Good practice is to move a protocol to Historic status when it has actually
been superseded.
Best
Ale
--
_______________________________________________
dmarc mailing list -- [email protected]
To unsubscribe send an email to [email protected]