Re: UTF-8 / Unicode / Local conversions

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

At 10:47 AM 6/19/00 -0400, Mark C Smith wrote:
>"Kurt D. Zeilenga" wrote:
>> 
>> At 04:09 PM 6/16/00 -0600, Dave Steck wrote:
>> >Applications need a way to display data obtained through the LDAP C API's.
>> >
>> >To provide truly cross-directory and *cross-platform* API's, it seems there should be
>> >a standard set of C APIs to do UTF-8 <--> Unicode conversions.
>> >
>> >There should also be a standard set of API's to convert between either UTF-8 or Unicode and local strings?
>> >Does anyone know of such a standard or work underway?
>> >
>> 
>> I believe it appropriate for the IETF LDAPext WG to specify an
>> an C API to converting between the local charset, T.61, and UTF-8.
>>         local <-> T.61
>>         local <-> UTF-8
>>         UTF-8 <-> T.61
>> 
>> However, as previously discussed, this can be specified as
>> a (Standard Track) extension.
>
>In most environments, these functions are already provided.  I am not
>convinced that the LDAPEXT group should duplicate the work being done by
>OS and platform vendors.

I have similiar concerns as well.  My comment was more to the point
that the working group had, IMO, reached some consensus previously
that such features would be best handled as extensions to the
API.

Note that I added T.61 to my statement because the C API
supports both LDAPv2 and LDAPv3.  If an extension provides UTF-8
conversions, it should provide T.61 (and vice versa).  It's not a
matter of letting T.61 die, it's a matter of letting LDAPv2 die.
I do believe that inconsistent LDAPv2 charset handling by various
vendors (including OpenLDAP) will eventually kill LDAPv2.

Kurt