LDAP and character encoding...

["Richard, Francois M" <Francois.M.Richard@usa.xerox.com>]

Hi, 

I hope this list is appropriate for the kind of general information about
LDAP I am looking for.
I understood (from RFC2251) that strings exchanged through LDAP protocol are
utf-8 encoded.
But all different samples of message I could find (like in RFC 2849) seem to
apply Base64 transfer encoding. For instance

# cn;lang-ja;phonetic:: <JapaneseCn_in_phonetic_representation_kana>
title;lang-ja;phonetic:: 44GI44GE44GO44KH44GG44G2IOOBtuOBoeOCh+OBhg==

So I am trying to test my understanding on this character encoding issue:
1- LDAP v3 requires utf-8 character encoding for strings
2- The LDIF format for exchanging data through LDAP v3 protocol requires the
utf-8 encoded bytes to be base64 transfer encoding before being sent over
the wire.

Is this correct?

Other related questions:
Is  the only benefit of using Base64 encoding the 7-bit clean aspect? 
Why does LDAP has to be 7-bit clean when other protocol(HTTP, IPP, ...) can
transmit utf-8 over the wire?
Is it possible that this restriction will be removed in the near future, to
get utf-8 over the wire?
Is there still existing implementation of LDPA v2 (based on IA5 or IRA)?

I would appreciate some help since reading between the lines of RFCs is
tedious...
Tanks