Re: [dmarc-ietf] Tree Walk impact

Thanks for confirming that your list of "problem domains" was not based on the tree walk design in the draft.

There's been no discussion of the how and the why of the design the design other than on the list.  I'm also mystified how it's possible you don't know.

Scott K

On October 9, 2023 10:19:56 PM UTC, Douglas Foster <dougfoster.emailstandards@gmail.com> wrote:
>Great question.   On Feb 23rd, I had this exchange which John settled:
>It appears that Douglas Foster  <dougfoster.emailstandards@gmail.com> said:
>
>>-=-=-=-=-=-
>>
>>I seem to have missed this redesign.   I thought the plan had always been
>>to take the top-most policy not flagged as PSD=Y.
>
>
>The current design has been in the draft since October, and we discussed
>it on this list at great length.
>
>
>R's,
>John
>
>Today, re-reading draft version 28, it says to use the top version.   I
>missed when stop-first was chosen "after discussion at great length", and I
>also seem to have missed the decision to switch back, presumably after
>equally vigorous discussion?
>
>Both approaches have problems.   Stop-at-last enables the walk to exit the
>current organization and stop on a private registry, for both alignment
>evaluation and for aggregate report transmission.   This is not a minor
>problem, even if it is arguably infrequent.
>
>Given that the problem with PSL is imperfect data, the solution is better
>data.   Instead we have chosen a heuristic, and consequently we can be
>certain of heuristic-induced harm.   We should give domain owners full
>control over their organizational boundary, and stop guessing.
>
>Doug Foster
>
>
>
>
>
>On Mon, Oct 9, 2023 at 9:00 AM Scott Kitterman <sklist@kitterman.com> wrote:
>
>> Where does it say to stop at the first policy found?
>>
>> Scott K
>>
>> On October 9, 2023 12:51:33 PM UTC, Douglas Foster <
>> dougfoster.emailstandards@gmail.com> wrote:
>> >Right, but we walk up from both domains separately, and each walk stops at
>> >the first policy found.  Since the two walks stop at different policies,
>> >they are presuned to be different organizations.
>> >
>> >Doug
>> >
>> >
>> >On Mon, Oct 9, 2023, 5:35 AM Alessandro Vesely <vesely@tana.it> wrote:
>> >
>> >> On Sun 08/Oct/2023 04:00:31 +0200 Douglas Foster wrote:
>> >> > Attached it is a spreadsheet with the problems from my data set.
>> >>
>> >> I see no blocking.  For example, the list shows From: bayer.com,
>> >> d=crm.bayer.com, the latter deemed blocking.  Both domains feature a
>> >> DMARC
>> >> record and (unsurprisingly) none has a psd= tag.
>> >>
>> >> According to the spec, one should look up:
>> >> _dmarc.bayer.com
>> >> _dmarc.com
>> >>
>> >> And then
>> >> _dmarc.crm.bayer.com
>> >> _dmarc.bayer.com
>> >> _dmarc.com
>> >>
>> >> The organizational domain is bayer.com and they are aligned.  No
>> blocking.
>> >>
>> >> Best
>> >> Ale
>> >> --
>> >>
>> >>
>> >>
>> >> _______________________________________________
>> >> dmarc mailing list
>> >> dmarc@ietf.org
>> >> https://www.ietf.org/mailman/listinfo/dmarc
>> >>
>>
>> _______________________________________________
>> dmarc mailing list
>> dmarc@ietf.org
>> https://www.ietf.org/mailman/listinfo/dmarc
>>