Re: What is T.61 (Was: Chinese Character)

["Kurt D. Zeilenga" <Kurt@OpenLDAP.org>]

Villy Kruse <vek@pharmapartners.nl> wrote:
>I have long wondered what T.61 was as you see references to it in the
>open-ldap documentation.

Because it is the specified character set of LDAPv2 per RFC 1777.

>but nothing about what it is, let alone what the encoding is.

I try to add some references.  I am sure RFC 1777 provides such.

>Guess when we convert to UTF8 (8 bit encoding of unicode) we can forget
>about T.61.

You can forget about T.61 if and when you forget about LDAPv2.

However, if you are deploying software which supports LDAPv2 and
LDAPv3, you need to address T.61 vs UTF-8 encoded ISO 10646
charset issues.  You will find that some LDAP implementations
(correctly) use T.61 for LDAPv2 and other implementations that
(incorrectly) use UTF-8 encoded ISO 10646 characters for LDAPv2.
Mixing implementations which use different charsets for LDAPv2
in a deployment would reak havoc on your directory.  If you
expect to use LDAPv2 clients as part of a global directory service,
you must deploy LDAPv2 clients that use T.61 as required by
the specifications.

In OpenLDAP 2.0, we hope to correctly provide T.61 for LDAPv2
while providing UTF-8 encoded ISO 10646 for LDAPv3.  However,
doing so is a royal pain.  In the end we may follow the
UTF-8 encoded ISO 10646 for both LDAPv2 and LDAPv3 approach,
but recommend that installers disable LDAPv2 if they have any
LDAPv2 clients which expect T.61 or when deploying OpenLDAP
as part of a global directory service.

Kurt