On Mon, Jun 29, 2026, at 07:51, Hannah Stern wrote:
> Hi!
> 
> On 6/24/26 21:08, Richard Clayton wrote:
> > In message <178232634716.1360785.12363329888242618970@dt-datatracker-
> > f9b87776f-xzl65>, [email protected] writes
> >
> > > Internet-Draft draft-ietf-dkim-dkim2-spec-03.txt is now available. 
> > It is a
> > > work item of the Domain Keys Identified Mail (DKIM) WG of the IETF.
> >
> > >   Title:   DomainKeys Identified Mail Signatures v2 (DKIM2)
> > >   Authors: Richard Clayton
> > >            Wei Chuang
> > >            Bron Gondwana
> > >   Name:    draft-ietf-dkim-dkim2-spec-03.txt
> > >   Pages:   42
> > >   Dates:   2026-06-24
> >
> > the changes are (and these affect implementors, so if you own code you
> > will need to look at the changes)
> >
> > #1  introducing nd= as an alternative to mf= rt= for an imaginary hop.
> > This is essemtially the "forward" signing scheme that Tobias suggested
> >
> > viz instead of:
> >
> > i=1; d=example.com; [email protected] [email protected]; f=..flags..;
> > i=2; d=esp.tld; [email protected]; [email protected];f=..flags..;
> >
> > hide the imaginary hop (because the ESP is creating both signatures)
> >
> > i=1; d=example.com; nd=esp.tld; f=..flags..;
> > i=2; d=esp.tld; [email protected]; [email protected];f=..flags..;
> >
> > similarly for a mailbox provider that forwards for vanity domains
> >
> > i=1; d=example.com; [email protected]; [email protected];
> > i=2; d=vanity.com; nd=mailbox1.tld;
> > i=3; d=mailbox1.tld; [email protected]; [email protected];
> >
> Questions:
> 
> - On verification, is it allowed that the latest signature (highest i) 
> has nd?

If you're receiving that message over the wire, no... unless you have a 
contractural relationship with the previous hop.  The point is to say "we're 
managing the handover between domains directly".  If you get something with nd= 
over the wire, it's breaking the replay protections.

> - Is it allowed to have more than one nd in a row? (I.e. i=4 has 
> nd=<domain of i=5>, i=5 has nd=<domain of i=6>)?

Yes, no reason why not to.  Again, so long as all these domains have 
contractural agreements.

> - How is the mf/rt chain checked in the presence of nd? Just skip the nd 
> hop? I.e. if i=4 has mf/rt, i=5 has nd, i=6 has mf=rt, do I check i=6 
> against i=4 using the usual rules?

No, you never skip over a hop.  You check:

i=4 rt=foo@X
i=5 d=X; nd=Y
i=6 d=Y; mf=bar@Y

Those are the invariants which must pass.

> > #2  Sorting out DSNs
> >
> > There are issues when a DSN flows back along the outgoing path and that
> > outgoing path visits the same system more than once (mail does that sort
> > of thing from time to time). This is fixed by insisting that DSNs always
> > have the header fields present as the email arrived at that hop (so
> > remove evidence of forwarding from the DNS that came in and is being
> > propagated). Also forbid "null recipes" for headers (which do not seem
> > to have a use case)
> 
> This means a) all forwarders _have_ to implement what's needed to 
> generate the recipes, b) security solutions can't remove my "Virus: 
> <base64 blob of malware>" field irreversibly.

If there's a virus in the header, then you're right - security solutions can't 
remove it.  All they can do is bounce it; which seems reasonable.

> > #3 Add the experimental Delivered-To: header field as an exception to
> > signing. Also the exceptions are now in a section of their own rather
> > than being buried away.
> 
> Stylistically, a simple list of ignored fields would be helpful instead 
> of digging the list out of prose.
> 
> Rationale could still be provided, e.g. as notes/footnotes to the simple 
> list/table.

Agree, this seems sensible.  I'm in favour of a nice sorted list of named 
header fields.

> > #4 Added "feedhere" flag so that privacy preserving forwarders can deal
> > with feedback reports (recall -- which are not defined here) do not
> > reveal where email was forwarded.
> 
> There's not even a mechanism to give a target for feedback reports (even 
> with not-yet-specified content format).

Yeah, we need to do this!

Bron.

--
  Bron Gondwana, CEO, Fastmail Pty Ltd / Fastmail US LLC
  [email protected]

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

Reply via email to