Murray S. Kucherawy writes: > As I read 5.3, it says you need to make sure what you sign is what > the verifier will receive. It seems to me a signer that gets 8-bit > header fields can RFC2047-ize them before signing, presuming the > MTA will make the same conversion before putting the signed message > out on the wire.
I'm not sure what you're trying to say here. If the outgoing MTA has SMTPUTF8 enabled, then it has the option of not encoding them and leaving them as "raw" UTF-8. But if is encounters an MTA that doesn't support SMTPUTF8, it must encode or bounce. In an SMTPUTF8-enabled environment, whether the header is encoded or not is nondeterminis- tic. So that presumption is invalid. The signer really has to force RFC 2047 encoding, and so may as well do it itself. _______________________________________________ dmarc mailing list [email protected] https://www.ietf.org/mailman/listinfo/dmarc
