It appears that Rob Sayre <sayrer@xxxxxxxxx> said: >> At a protocol level, DKIM operates just fine without knowing anything >> about SMTP at all. > >I'm not sure that's realistic, though. Of course it's realistic. If I wrote a message, added a DKIM signature, wrote it on a floppy disk, put it in an envelope and mailed it to you, you could trivially read the message and run it through a DKIM verifier, no SMTP in sight anywhere. The reference to SMTP in 6376 is saying in an elaborate way that an invalid DKIM signature is the same as no signature all. Even the new effort has a normative >reference to SMTP: >https://datatracker.ietf.org/doc/html/draft-gondwana-dkim2-motivation-02#name-a-chain-of-aligned-dkim2-si The new thing tentatively called DKIM2 not at all the same as the DKIM we're talking about. It is deeply unhelpful to try and drag it into this discission. R's, John -- last-call mailing list -- last-call@xxxxxxxx To unsubscribe send an email to last-call-leave@xxxxxxxx
