[Imap-protocol] Modified UTF-7: justice delayed, but justice served... ; -)

[Mark Crispin]

I just spent the past several hours implementing support routines for 
Modified UTF-7, with the intention that Alpine will support multinational 
IMAP mailbox names.

I always knew that the design of Modified UTF-7 sucked.  It was certainly 
well-intentioned.  Punycode would have saved the day, but it didn't appear 
until years later.

The road to hell is paved with good intentions.

Not until today have I fully comprehended the sheer implementation 
insanity and hideous design of Modified UTF-7.  I shudder to think of all 
the people who have beaten their heads against a wall in implementing 
Modified UTF-7.

To those of you with dents in your wall: I'm sorry.  As ot today, there's 
a head-shaped dent my wall too.  You've been avenged... ;-)

This has been a learning experience for me.  I feel that we need to push 
ahead on UTF-8 mailbox names sooner rather than later.  I feel that it is 
far more important for the i18n document to make UTF-8 mailbox names 
possible than setting the language for error messages or translating 
namespaces.  The latter two are really l10n issues and belong in a 
separate IMAP-L10N document.

More specifically, I would move Section 3 of imap-i18n to a new imap-l10n, 
and upgrade the former section 5 to be a full mechanism rather than a 
discussion.  I suggest that the former section 5.1 of imap-i18n deprecate 
the LOGIN command entirely, and highlight that SASL names are Unicode.

As I suggest above, I feel that section 5.2 of imap-i18n is highly 
unsatisfactory for today and needs a real mechanism; the now-expired 
draft-ietf-eai-imap-utf8-01.txt was an option although I would prefer a 
simplified mechanism.

-- Mark --

http://staff.washington.edu/mrc
Science does not emerge from voting, party politics, or public debate.
Si vis pacem, para bellum.