Re: Required OIDs in DNs

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

Presently, we're exploring whether or not it would be possible
to use a registry approach instead of a static table approach.

If the registry approach doesn't work out, we might want to
consider other static table approaches... but, for now, I
rather focus on determining whether the registry approach
is workable.

While I'm going to duck your specific static table
suggestions, I do have a few general comments to make.


At 02:05 PM 2002-11-15, Hallvard B Furuseth wrote:
>As you may have realized, I've finally noticed that the LDAP standard is
>being updated:-) I suppose I'm beating a dead horse here, but could
>someone explain the points below or point me to an explanation?

>Why were DNs changed to require numeric OIDs instead of names for other
>attribute types than those in the table in [LDAPDN] paragraph 2.3? 

In my view, [LDAPDN] is not significantly different from
[RFC2253].  We've clarified that the "published table" is
the table is published in the specification.

Obviously, we have many DNs in use which don't conform to
this clarified specification.  And, so, the issue becomes,
how do we change [LDAPDN] so not to invalidate too many in
use DNs while continuing to preserve the DN properties
necessary to support Internet-scale directory service.

Kurt