Re: I-D ACTION:draft-ietf-ldapext-cldap-00.txt

[mcs@netscape.com (Mark C Smith)]

Internet-Drafts@ietf.org wrote:
> 
> A New Internet-Draft is available from the on-line Internet-Drafts directories.
> This draft is a work item of the LDAP Extension Working Group of the IETF.
> 
>         Title           : Connection-less Lightweight Directory Access Protocol
>         Author(s)       : L. Johasson, R. Hedberg
>         Filename        : draft-ietf-ldapext-cldap-00.txt
>         Pages           : 10
>         Date            : 31-May-00

Section 3 says:

   ... Note that it is
   possible for a client to issue a modifying request (add, delete,
   moddn, modify) together with a control or an extended request which
   modifies the directory such that the response is too large to fit in
   a datagram which would make it impossible for the client to know if
   the requested operation was successful or not. For this reason
   servers implementing this protocol must respond with an error of
   unwillingToPerform(53) if such a request is received.

Does the above mean that a server must determine the size of a response
before it actually carries it out?  That might be tricky to implement.

Also, I'd prefer to see a more specific error introduced for this case
(or just use resultsTooLarge).  An unwillingToPerform error may be
returned for other reasons, and it would be nice to unambiguously tell
the client "This can't be done over UDP, please try again over TCP."  In
fact, I'd like to see the client's recommended behavior with respect to
error recovery and use of LDAP over TCP (when a CLDAP request fails)
spelled out in the draft.

It would also be valuable to provide a "scope" or "requirements"
section.  For example, I think a useful CLDAP protocol could omit all of
the update operations (keeping compare and search).

-- 
Mark Smith
Directory Product Development / iPlanet E-Commerce Solutions
My words are my own, not my employer's.            Got LDAP?