LDAP has become the application developers' Rosetta stone for the new generation of directory-enabled applications.
With Microsoft Corp., Netscape Communications Corp. and Novell Inc. each touting the benefits of their respective directory service implementations, coding a directory-independent application can be a daunting task. That's where the Lightweight Directory Access Protocol fits in. Rapidly becoming the common ground to which each directory service vendor has sworn allegiance, it's also the ticket to cross-directory compatibility (see related story, "LDAP is key to unlocking doors to global directories").
LDAP is extensible, allowing developers to alter a directory's structure for application-specific functions. For example, a scheduling program linked into a directory requires data fields that are not covered by the current LDAP Version 2.0 (RFC 1777 and 1778) specification. Solving this issue today requires nonstandard extensions to LDAP that poke holes in its promise of cross-directory compatibility. In fact, the present LDAP specification could look more like Swiss cheese as developers conceive of new applications that exploit the directory in mixed environments.
Because LDAP is a nonproprietary specification managed by the Internet Engineering Task Force, proposed changes to it are thoroughly debated before becoming part of the specification, a tedious process. Unable to wait for their proposals to become standards, some developers are pushing ahead with applications that will depend on an extended version of LDAP that has yet to be blessed by the IETF.
Closely involved with LDAP from the beginning (1993), the University of Michigan, in Ann Arbor, published in December a new draft, dubbed LDAP 3.3, for public review and comment--the early steps in the standardization procedure. LDAP 3.3 is stuffed with many of the functions the programming gurus have been clamoring for. Two examples are new directory replication and gateway implementations, which will relieve several interoperability concerns.
In other words, the "lightweight" protocol is going to have to put on a few pounds, but don't expect a full-fledged binge. Vendors are adding their own extensions to their directory service offerings, promising their directories will be translatable with the help of LDAP 3.3. At the same time, they are actively lobbying the IETF to make their definitions part of future specifications.
The vendors essentially hope to have it both ways. Only time will tell whether LDAP 3.3 will be the next Rosetta stone for translation of all directory service dialects.