Re: [dmarc-ietf] Separating DMARC alignment validation from DMARC policy
In article you write: >On 7/28/2020 9:59 AM, Murray S. Kucherawy wrote: >> I think it was you that suggested having the determination of the >> Organizational Domain be its own distinct process. Would that fit into >> the first document, or should that also be something external? > >I'd wish for that to be entirely outside of DMARC, because the issue and >the mechanism are not specific to DMARC. Unfortunately, we all know where to find DBOUND. I suppose now that the smoke has somewhat cleared we might take another run at the PSL problem, trying harder not to boil the ocean. R's, John PS: https://github.com/jrlevine/bound ___ dmarc mailing list dmarc@ietf.org https://www.ietf.org/mailman/listinfo/dmarc
Re: [dmarc-ietf] Separating DMARC alignment validation from DMARC policy
On 7/28/2020 9:59 AM, Murray S. Kucherawy wrote: I think it was you that suggested having the determination of the Organizational Domain be its own distinct process. Would that fit into the first document, or should that also be something external? I'd wish for that to be entirely outside of DMARC, because the issue and the mechanism are not specific to DMARC. d/ -- Dave Crocker Brandenburg InternetWorking bbiw.net ___ dmarc mailing list dmarc@ietf.org https://www.ietf.org/mailman/listinfo/dmarc
Re: [dmarc-ietf] Separating DMARC alignment validation from DMARC policy
On Tue, Jul 28, 2020 at 6:53 AM Dave Crocker wrote: > I am not a fan of splitting documents only for an abstract sense of > cleanliness. There needs to be justification of improved documentation > 'cleanliness' and/or useful removal of fate-sharing -- if separate fates > are reasonably expected. > > > DMARC alignment validation is a basic, mechanical process that produces > a yes/no response. At that level, it's comparable to the nature of a > DKIM validation, except for the From: field domain. > > DMARC "policy" is fundamentally different. It's not that its mechanism > isn't "mechanical" but that its semantics produce more semantic and > operational challenges. > > > I think that could reasonably make it worth strongly separating them > into two different documents, no matter has small one of the documents > might be. > I think it was you that suggested having the determination of the Organizational Domain be its own distinct process. Would that fit into the first document, or should that also be something external? -MSK ___ dmarc mailing list dmarc@ietf.org https://www.ietf.org/mailman/listinfo/dmarc
Re: [dmarc-ietf] Separating DMARC alignment validation from DMARC policy
The MIME document split was, in hindsight, partly successful (separation of media types registration was a really good thing), partly unsuccessful (part five should not have been pulled out of part two), and partly a wash (the cost/benefit of separating parts one and two ended up pretty much equal IMO). The devil is always in the details. The DMARC alignment validation process appears free of policy considerations as currently specified, and is thus a candidate for separation. But are we sure it will remain so, and is the separation of a relatively small part of the specification worth it? Ned I am not a fan of splitting documents only for an abstract sense of cleanliness. There needs to be justification of improved documentation 'cleanliness' and/or useful removal of fate-sharing -- if separate fates are reasonably expected. DMARC alignment validation is a basic, mechanical process that produces a yes/no response. At that level, it's comparable to the nature of a DKIM validation, except for the From: field domain. DMARC "policy" is fundamentally different. It's not that its mechanism isn't "mechanical" but that its semantics produce more semantic and operational challenges. I think that could reasonably make it worth strongly separating them into two different documents, no matter has small one of the documents might be. d/ -- Dave Crocker Brandenburg InternetWorking bbiw.net ___ 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
[dmarc-ietf] Separating DMARC alignment validation from DMARC policy
I am not a fan of splitting documents only for an abstract sense of cleanliness. There needs to be justification of improved documentation 'cleanliness' and/or useful removal of fate-sharing -- if separate fates are reasonably expected. DMARC alignment validation is a basic, mechanical process that produces a yes/no response. At that level, it's comparable to the nature of a DKIM validation, except for the From: field domain. DMARC "policy" is fundamentally different. It's not that its mechanism isn't "mechanical" but that its semantics produce more semantic and operational challenges. I think that could reasonably make it worth strongly separating them into two different documents, no matter has small one of the documents might be. d/ -- Dave Crocker Brandenburg InternetWorking bbiw.net ___ dmarc mailing list dmarc@ietf.org https://www.ietf.org/mailman/listinfo/dmarc
