Re: ldap v2/v3 interoperability

[Harald Alvestrand <Harald.Alvestrand@maxware.no>]

At 12:39 26.08.98 -0400, Christopher Oliva wrote:
>Hi,
>
>For ldap v2 clients interoperating with ldap v2/v3 servers, how is character
>set translation to be handled? In the case where an entry containing Unicode
>characters (created by a v3 operation) matches a query from a v2 client,
>what happens if the Unicode character cannot be translated to T.61? There
>are several possibilities:
>
>1) server returns an error (which error?)
>2) server returns the Unicode values encoded in some way (what encoding?)
>3) server returns partial results terminated with an error/warning (what
>error?)

don't see any answers to this....

generally, moving stuff between charsets is painful, and made even more
painful by the disasters in LDAPv2 where some of the leading software vendors
decided not to follow the spec, but to do two different non-standard things...

The usual strategy is:

- Translate what you can (and make the target charset configurable)
- Put in some kind of marker for the rest
- Pray that the customer is satsified

(It's actually possible to encode an incredible amount of stuff in T.61,
including Greek, Chinese and Hebrew - the problem is that most implementations
don't have a snowball's chance of presenting it decently.)

Group: Is there a need for an "ldap v2/v3 coexistence document", or are
we just hoping that LDAPv2 will die quietly?

                Harald A

-- 
Harald Tveit Alvestrand, Maxware, Norway
Harald.Alvestrand@maxware.no