On Tue 29/Sep/2020 05:45:40 +0200 Dave Crocker wrote:
On 9/28/2020 8:42 PM, Seth Blank wrote:
Are there any objections to recording consensus that [1], that the domain and
selector of the key used to evaluate the DMARC status MUST be included, and
[2] opening a ticket to discuss how appropriate keys should be determined?
From the start of DKIM, there's been a common assumption that the selector is
part of the identifier, to use in evaluation. This defeats it intended
administrative purpose as merely a way to distribute keys, for the same domain,
such as for rolling them over.
Or have I misunderstood the import of this issue here?
After the introduction of elliptic keys, knowing the selector is necessary to
track the propagation of the new technique.
Best
Ale
--
_______________________________________________
dmarc mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dmarc
