Re: [dmarc-ietf] SPF doesn't accommodate third level .name domains?

[David Bustos <[e-mail]>]

Scott wrote:
> On May 30, 2022 9:50:05 PM UTC, David Bustos <[e-mail]> wrote:
> >Since I own david.bustos.name, someone forwards [e-mail] for me; I presume Verisign does.
> >
> >Lately I think email receivers have been quarantining my messages and I suspect the reason is SPF.  Specifically, no SPF record is published for bustos.name .  I asked Verisign to publish one and they declined.
> >
> >I'm thinking that SPF should be modified so that if a receiver finds no record and the domain ends with .name and the domain has two levels, it should look for the record on the third level domain.  Does that make sense?
>
> No.  My provider won't X, so we should special case the protocol doesn't scale.  It may be that .name is special to you, but there are a virtually infinite number of domains that are special to someone else.

So is your position that Verisign should publish an SPF record for bustos.name?  Don't you think they would say that doesn't scale because they would need to include an "include" for each third level domain in bustos.name?  Isn't changing one protocol more scalable than including N "includes" in the SPF record for each last.name domain?

.name was designed to be special by whoever regulated TLDs back in 2002, and that specialness was the main reason I bought the domain.  If the IETF position is that such addresses are second class citizens, then shouldn't that be expressed to Verisign so it can notify owners and registrars?

> Also, not really a DMARC question.

Is there an SPF working group?  The DMARC working group charter at https://datatracker.ietf.org/wg/dmarc/about/ includes

> 4. Related work
>
> Extensions to SPF/DKIM/DMARC that do not already fit under
> the charter of any existing working group can be considered for adoption
> by DMARC Working Group after consultation with the responsible AD.
> A prime example is draft-levine-appsarea-eaiauth, which provides EAI-related
> updates to DMARC and the protocols upon which it depends.
> Any such work needs to carefully consider interoperability implications.

I don't know much about DMARC, but it seems like it does not accommodate .name in the same way as SPF.  Is that true?  You guys are the experts, right?

Also, can you copy me on your replies?  Or should I subscribe to the list?


Thanks,
David