Re: Are international characters allowed in email addresses?

Patrik, (non-AD hat noted)

> You at least need to either agree on a charset to use, or negotiation of
> charset. Compare MIME spec with current IDN discussions. And, 8 bits are
> not enough today for many charsets (including UTF-8).

Or use unknown-8bit (RFC 1428 section 3).

I honestly don't know how to compare anything useful with "current IDN
discussions", the list is all over the place, anything other than ACE
is as popular as djb-advocacy on namedroppers, with my-dog-is-bigger
behavior the norm, and periodically a "technical advisor" blows by with
helpful notes of the form "it-can't-be-solved-use-ldap".

Thank heavens we didn't decide we simply had to add "+" to the set of
characters we use for names, or authoritative root would be divided into
monotheists (of the negative, or "hypenated" form) and +/- dualists by
sundown.

I don't get the point of your final sentence.

Eric Brunner+/-Williams
architect, utf8-based code-set independence, solaris, 1995
implementor, utf8-based code-set independence, hpux, 1996

Quoting myself (headers), before Keith's commentary on my comment.
> The discussion thus far has employed terms like "address" and "domain".
> 
> In general, sendmail (a MTA with a modest deployment) is 8-bit encoding 
> indifferent. The only interesting issue is can sendmail handle 8-bit headers?
> 
> Currently, the answer is "no".
> 
> The IETF has not yet scheduled a transition plan to 8-bit clean header 
> processing, nor an EHLO extension specification and a way to do 8->7 bit
> encoding when sending to a system that didn't announce the extension.