On March 22, 2022 2:47:37 AM UTC, John Levine <[email protected]> wrote:
>I took a look and I think we're pretty close. I have two related
>suggestions.
>
>It is confusing that the default for psd is psd=n but an explicit
>psd=n isn't the same as no psd. So I suggest allowing threee values
>for psd=y/n/u with "u" for unspecified being the default.
>
>In the three numbered steps in sec 4.8 it says first to look for any psd=n,
>then any for psd=y. Consider this somewhat contrived case:
>
>x.a.b.c.d.e v=DMARC1;
>b.c.d.e v=DMARC1; psd=y
>c.d.e (nothing)
>d.e v=DMARC1; psd=n
>e (nothing)
>
>The current text says the org domain would be d.e but I think we want
>it to be a.b.c.d.e.
>
>The fix is simple, combine the first two steps so you stop at the
>longest label with psd=y OR psd=n, and it's that label if psd=n or
>the one below if it's psd=y. Arcane sort of special case, if the
>original domain has a psd=y/n, that's the org domain and no tree walk
>is needed.
Thanks. I agree that's the right answer. I'll take another shot at it.
I had considered just saying there's no default, but I guess using "u" works
too if you think that's better.
Scott K
_______________________________________________
dmarc mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dmarc
