On Wed, Jun 27, 2018 at 12:52 PM, Jeremy Harris <[email protected]> wrote:

> On 06/25/2018 05:09 PM, Dave Crocker wrote:
> > On 6/25/2018 6:03 AM, Jeremy Harris wrote:
> >> On 06/25/2018 04:44 PM, Dave Crocker wrote:
> >>> If the creator of the information does not have a reliable way of
> >>> knowing what the receiver of it will do with it -- I mean
> interpretation
> >>> of the information, not follow-on policy decision-making -- then it
> >>> isn't a standard.
> >>
> >> On that view, the spec should be either
> >>
> >>    "any tag not specified here, in any of the three ARC headers,
> >>    invalidates the ARC set"
> >>
> >> or
> >>
> >>    "any tag not specified here, in any of the three ARC headers,
> >>    MUST be ignored".
> >
> >
> > While there are some operational environments where one of these two
> > choices really is appropriate -- the need for extremely rigorous control
> > over the entire range of operational choice is... extreme -- this isn't
> > one of them.
>
> Then I'm unclear what you were asking for with
>
> > Anything that comes close to 'do whatever you want with this
> information' demonstrates NON-standardization.
>
> Perhaps you could suggest some text for the spec?
> <https://www.ietf.org/mailman/listinfo/dmarc>
>

We already tried to walk that line in the spec. In general, any tag that
does not have a defined interpretation (firstly in the ARC spec, then
falling back to the DKIM spec) gets ignored with certain exceptions, such
as the "h" tag in an AS header.

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

Reply via email to