​For context, my input is based on my involvement in two primary sectors: 
furthering effective interoperable specification development, and as a 
representative of a reasonably large email security gateway that handles a 
large volume of email.

Combining my experience in both sectors, my responses to the two points are 
embedded below:

> Claim 1. ARC is actively deployed and utilized

That is undoubtedly true.  We see evidence of ARC headers every day (primarily 
coming from a single source — by a LARGE margin).  The problem is that, from 
what we can tell, only the last hop signature appears to be utilized to any 
measurable extent with all earlier signed hops being ignored.

So… yes… ARC has been deployed and is used… but not for what it was intended.  
And any continued focus, time, money spent on it is actively draining resources 
from fulfilling the theory / hypothesis upon which we based the original ARC 
experiment.

> Claim 2. No interoperable replacement is yet published

This is arguable based upon the definition of “published”.  The DKIM2 draft 
specifications that includes ARC-like functionality has been published via the 
DKIM Working Group.  While the proposed specifications and associated drafts 
haven’t been published as Standards Track RFCs, I am aware of at least three 
interoperable implementations of the current drafts in various phases of 
completion.  Various interoperability events (some of which have been conducted 
under the auspices of the IETF) have already taken place.

And once again… for those of us without infinite resources to spend supporting 
experiments that have run their course at the same time as implementing the 
learnings from the experiment… moving the ARC experiment to “Historic” status 
will significantly free up resources to support the interoperable specification 
development process as well as increase the speed in bringing said technologies 
to market (ultimately improving security for us all).

In short… leaving the ARC RFC as-is continues to feed ambiguity into the system 
and is ultimately slowing everything down.

I hope that helps clarify the situation from my perspective and helps tick the 
boxes in the consensus conversation initiated by Barry.

Thanks,
Trent


From: Barry Leiba <[email protected]>
Date: Saturday, July 18, 2026 at 1:34 PM
To: DMARC Working Group <[email protected]>
Subject: [dmarc-ietf] Re: Evaluating consensus on the idea of making ARC 
"Historic"

This Message Is From an Untrusted Sender
You have not previously corresponded with this sender.
Report 
Suspicious<https://us-phishalarm-ewt.proofpoint.com/EWT/v1/ORgEfCBsr282Fw!YLP3Cs7RtFzudvuYvj_6nXV29St8gN_HUycrqQzIq6e1-AtmFI6Ejwftk5grM7zKDlZB3_4AL9wUWhR7W2v5wdkt8IcGkwafntePsQGRZsr2LZLWg7qtZMhTg9Bg$>


The objection collection period is over, and we seem to have clear
consensus on Seth's objection, among those objecting:

> I think we should NOT make ARC historic, yet.
>
> 1. ARC is actively deployed and utilized
> 2. No interoperable replacement is yet published
>
> I acknowledge ARC is not as useful as we'd hoped, and we've learned a good 
> deal which has been
> woven into DKIM2. I do not at all object to ultimately making the move to 
> historic.
>
> I simply believe DKIM2 should obsolete / make ARC historic when it publishes, 
> so there's a clear upgrade
> path. i.e. when we have a new protocol that works and interoperates, that's 
> the culmination of the
> experiment and when the bit flips.

Point 1 supposes that there's deployment of both applying ARC seals
*and* checking them; it would be good to have some real data on the
latter -- are they being widely checked and do those who are checking
them find the results useful?

Point 2 simply asks that we defer a decision on setting ARC aside
until after DKIM 2 is ready and we have some validation of it.

The next step is for those who still want to proceed with making ARC
Historic *now*... to respond to these objections and either accept the
suggestion to wait, or to address reasons we should not.

Just as I would like to see data about point 1, I would also like to
see data from those who reject point 1, to the extent that we have
any.  "No it's not," won't, by itself, be a sufficient addressing of
the objection.  That probably means that we need to hear from some of
the major email services.

The discussion of this starts now, and let's give it two weeks *after*
the IETF meeting is over -- so let's say we give it until 7 August and
then see where we are.  Again, please keep the discussion focused and
concise.  It'll be much easier for everyone to follow along if we make
our points concisely.

Barry, as chair

_______________________________________________
dmarc mailing list -- [email protected]
To unsubscribe send an email to [email protected]

_______________________________________________
dmarc mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to