LDAPv2 DN (Was: LDAPDN problems, and changes since RFC 2253)

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

Though I didn't include it my previous Appendix B issue response,
it seems to be an Appendix B issue.

At 03:50 AM 7/23/2004, Hallvard B Furuseth wrote:
>>    - Replaced specification of additional requirements for LDAPv2
>>      implementations which also support LDAPv3 (RFC 2253, Section 4)
>>      with a statement (in Section 3) allowing recognition of
>>      alternative string representations.

This likely should be split into two statements as the allowance for
alternative string representations relates to other factors as well.

        - Removed specification of additional requirements for LDAPv2
          implementations which also support LDAPv3 (RFC 2253, Section
         4) as LDAPv2 is now Historic.
       - Allow recognition of alternative string representations.

and add
        + removed LDAPv2-only constructs
to the list of ABNF changes.

>That one is more far-reaching than it looks.

I hope the changes do have a far-reaching positive impact upon
interoperability.  RFC 2253 was a bit unclear as whether the
Section 4 requirements applied to, hence there may be some
disagreement as to how to reflect the changes made in Appendix
B.   Consensus of the WG was that they apply only to LDAPv2
implementations.  The removal of Section 4 certainly clarifies
that these requirements did not apply to LDAPv3 implementations.
I think that is adequately reflected in my reworded removal
statement.  However, if others feel an additional note would
be useful, I suggest adding:
        Consensus of the IETF was that the removed requirements
        do not apply to LDAPv3.

>It causes two changes to attribute value parsing:

As Section 4 (and the LDAPv2-only ABNF constructs) only applied when
"parsing a DN string generated by an LDAPv2 client", their removal
implies no change to LDAPv3.

I believe it reasonable to only describe changes to LDAPv3.
Changes to LDAPv2 are moot as 1) LDAPv2 is historic and
2) the Section detailing LDAPv2 requirements and other LDAPv2
aspects of RFC 2253 have been removed.

>Anyway, this means the new draft describes neither the "plain LDAPv3"
>nor the "LDAPv3 + LDAPv2" semantics of RFC 2253 in this respect.

The draft now specifies only the LDAPv2 DN string representation.
LDAPv2 is left to historic document.  "LDAPv3 + LDAPv3" semantics
is nonsense as only one version of the protocol is in use in
any one protocol session.

>I suggest that:

It's unclear to mean whether you are suggesting additions to the
Changes section or suggestion new changes be made to the
DN string representation.  Please clarify.  If the latter, please
provide justification for the change.

Kurt