Re: [ldapext] [Fwd: I-D ACTION:draft-hall-ldap-idn-00.txt]

["Eric A. Hall" <ehall@ehsco.com>]

on 7/2/2003 8:51 AM John McMeeking wrote:

> If we require storing UTF-8 IDNs, does this require that clients be
> able to convert the IDN to the ASCII compatible sequence?

Only if the client does some kind of active DC<->DNS mapping. In practice,
there are very few applications which do this right now (although more are
coming, and it would be a good idea to specify the use of IDNs sooner
rather than later). Of those few applications, most of them can be handled
through some of the workarounds described in the draft (eg, using
referrals to point the dc searchbase to an idc naming context).

> With respect to the schema, hmm...  I'm not too fond of the idea of
> having a dual set of objectclasses and attributes that applications
> must use.  But I recognize that changing an IA5 String attribute to
> Directory String may not be palletable to standards bodies, and
> requires work on the part of implementations (schema updates, at least,
> possibly data migration, other changes).

I modelled after the domainComponent draft but I'm not married to it.

> idc/inetDomainComponent - Could we drop the alias?  Aliases cause
> problems. Defining an alias, while at the same time saying that systems
> MUST NOT use the alias... seems a bit contradictory at best.

I was trying to be "complete" by allocating the long form of the name, but
I'm not married to it.

-- 
Eric A. Hall                                        http://www.ehsco.com/
Internet Core Protocols          http://www.oreilly.com/catalog/coreprot/


_______________________________________________
Ldapext mailing list
Ldapext@ietf.org
https://www1.ietf.org/mailman/listinfo/ldapext