On Tue, Jul 10, 2018 at 1:24 PM, Jim Fenton <[email protected]> wrote:

> On 7/10/18 12:43 PM, Murray S. Kucherawy wrote:
>
>
> AMS is basically the same as DKIM-Signature, and so it covers body
> modifications.  When you verify the seal, you must also verify the latest
> AMS, which in turn means the seal is invalidated as soon as someone changes
> the content.
>
>
> Which goes back to my original question. If you need to check the AMS
> anyway, what additional purpose does the AS serve?
>

It's an interesting question.  I imagine it's similar in nature to the fact
that the crypto part of a DKIM-Signature will still pass for a message with
a modified body; you have to then check that "bh=" matches the current
content before you can truly assert that the signature is valid.

> I disagree.  If I can cause your MTA to crash because of oversized header
> fields, that's at least a denial of service attack.  If I can cause your
> filter to crash because of oversized header fields and your MTA fails open,
> I can bypass whatever protections the filter offers.
>
>
> In that case, I'm fine if this is expressed in terms of a security threat.
> But any ARC-specific nature to this threat is a marginal case; an attacker
> could do the same in most cases by just adding lots of garbage header
> fields.
>

True, but I think the nature of the beast here is that this is a
non-garbage addition to the email infrastructure that will certainly bloat
headers, and participants (and maybe even non-participants) need to be able
to deal with it.

OK, it's possible that this has been done. But the important question is,
> does this draft need to wait for 7602bis to be published?
>

This draft, as I recall, is trying to register some things into the
authentication method registries that the prose of 7601 didn't allow.  So,
yes.  I had thought that we planned to do them as a cluster anyway, and
that 7601bis was ready for WGLC ahead of this one, but I've forgotten where
we are with it now.

-MSK
_______________________________________________
dmarc mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dmarc

Reply via email to