On 5/29/2014 3:09 AM, Murray S. Kucherawy wrote:
On Thu, May 29, 2014 at 12:06 AM, John R Levine <jo...@taugh.com <mailto:jo...@taugh.com>> wrote: By the way, to return to the original point, it still seems vanishingly unlikely to me that if we invented per-sender whitelists that the two mail providers would implement them. Has anyone tried asking them? I'm not sure what value I should put in all this (ahem) third-party speculation about their intentions or what they care about.
Good point. Why don't you ask them? We need positive endorsement for 3rd party semantics which we don't have. I've implemented ATPS for ADSP and it works. Update the extension for DMARC and I'm sure Yahoo can manage 30K records if they decide to use it. That shouldn't be a problem and it will be a growth thing, not adding 30K records in one shot. It will be more manageable. A major consideration is that not all domains are YAHOOs so it they won't need the same scale, not even close.
But you see, thats been the problem with all this all along -- picking who this or that protocol, DKIM itself, will be using them and leveraging its value via a policy layer. We can't do that. The protocol modeling must fit all. That doesn't mean it works for all but from a mail integration and engineering standpoint, it has to apply to all -- small and large. It has to make sense at all scales. The danger was to miss this with the large who will have a higher impact when assumed they won't be using it.
-- HLS _______________________________________________ dmarc mailing list dmarc@ietf.org https://www.ietf.org/mailman/listinfo/dmarc
