Thank you for your continued rationality.
I do not have a strong opinion on what the solution to this problem
should look like. However, I am concerned about some related problems.
Local mail parts are often derived from or are usernames.
Administrators like the ability to be able to assign a single name to an
end user for the purpose of user identity. It is my hope that whatever
conclusion is reached will ensure that the i18n mechanism will allow
this single single name to be compatible with existing naming
conventions utilized with Kerberos, SASL, X.509, etc. It would I
believe detrimental to everyone if the various protocols which must
manipulate user identities did not have a common way of representing them.
Description: S/MIME Cryptographic Signature