On Thu, Sep 7, 2017 at 10:11 AM, Seth Blank <[email protected]> wrote:
> I'll go over this in more detail and post substantive comments sometime in
> the next day or so, but at first glance, a crucial change in 5.1 was missed
> and the draft language still makes a normative change to 7601 ("data SHOULD
> be added to the normal A-R content") to include data the WG consensus was
> should not be in the A-R (like smtp.client_id), instead of adding this data
> to the AAR where the WG consensus was that it does belong.
>
The way those fields are added seems wrong to me. RFC7601 spells all of
that out in terms of distinct (but related) ptypes and properties, which
are separate things. Adding "header.ds" as a single thing conflicts with
the way A-R is defined. At a minimum this should reflect that, though it
also brings up the fact that A-R has a registry for these; are we planning
to ignore, replace, clone, or augment that registry for this purpose?
-MSK
_______________________________________________
dmarc mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dmarc
