Being on the right track means that the directory is finally front and center with early adopters, and Novell has a better directory solution than anyone else in the market. Period. With years of experience in the field and millions of happy installed users, Novell will be looking in the rear-view mirror at its directory competition for some time.

Schmidt's challenges over the past three years have been monumental. NetWare's core design was outdated, and the company had missed not only the Internet bandwagon, but the application revolution as well. Third-party developers, once the Novell silver bullet, were nowhere to be seen, and its loyal corporate following, while still strong, was at the department level, not at the top levels of management. With IT becoming central to business success, departmental interest was giving way to top-down strategic planning. NetWare and virtually all the other Novell products--with the possible exception of GroupWise--were considered tactical, not strategic. The products served the needs of internal IT organizations--a cost center in every company. For Novell to return to the greatness it had known, the corporate-executive gang had to embrace the company, and its products had to fill strategic needs--where the money is made.

Building products that serve a profit center brings a whole new level of attention and demand. It also brings a willingness to write bigger checks, and that's where Novell wants to go. So Novell's latest road map to success looks something like this:

1. Take Novell's incredible caching know-how out of the operating system and make it serve the needs of the content-hungry Internet. Directory-enable the products along the way.
2. Combine the existing products centered on NetWare and put them into a business unit that can be run as a cash cow. Slow development but continue to directory-enable all products.
3. Create a professional-services support group to help customers build complex, profit-oriented, directory-based applications for the Web.
4. Most important, take the directory technology out of NetWare, make it platform independent, robust and scalable--and make it the centerpiece of a strategy for integrating every aspect of a corporate network, internally as well as externally, into a single coherent system.
5. Execute, execute, execute: products, marketing and services.

In May, Novell announced a restructuring that it hopes will deliver on this road map. If the first four steps seem reasonable, then it all comes down to the fifth step. Novell's track record on execution of new business initiatives has been one of remarkable failure, and so the question becomes: Can Novell do it now, and if it does, will customers trust a Novell solution? With its restructuring, Novell's business groups are now Net Directory, Net Management and Net Content. We'll look at each of these groups and its mission in detail. (Novell also has a professional-services group, which we will not explore in depth.)

Net Directory

As melodramatic as it sounds, the future of Novell lies with the success of the company's directory product. The good news for Novell is that its NDS, NDS Corporate Edition and eDirectory offerings (different product bundles centered on NDS) are far more mature than any other network directory on the market (read our evaluation of NDS). Novell claims its directory supports up to a billion objects, and that's a good order of magnitude more than any other product. What's also good news for Novell is that while competition is limited, there is growing market demand. Even better, the company is finally delivering on its promises of SSO (single sign-on) authentication and the directory as a resource for applications.

In some sense, there has been demand for directory services as long as there have been networked applications. IT managers know the pain of managing many naming systems: NOS, e-mail, remote access, human-resources software--the list is virtually endless. SSO has been a pipe dream for years, but it's been thwarted by parochial interests of vendors, none of whom is willing to be dependent on other vendors for this critical service. That's changing as Unix systems support PAMs (Pluggable Authentication Modules), and applications, such as PeopleSoft's business apps, support directory add-ins. Customer value is finally outweighing parochial interest.

But the real game is on the sales and support side. Web-based apps driving sales or tracking and serving customers are the silver bullet that has made solving the directory problem a real priority. One might say that Novell got lucky, inasmuch as it had a directory solution searching for a problem. And problems--or, rather, "opportunities"--it has found.

Novell prides itself on providing a best-of-breed solution. If storing a billion objects is the goal, then Novell no doubt has that solution. But in a truly heterogeneous environment with Linux, Microsoft Windows NT and Unix (so far, Sun Microsystems Solaris is supported) servers, the management and development tools are desperately lacking, though the directory is solid. On the big-picture end, metadirectory tools are lacking. On the other, more mundane end of the equation, installation and administration tools vary from marginally acceptable on NetWare to being a complete mess on Linux.

In the cost-center, internal world of IT departments, having marginal tools is an annoyance. People who have built complex directories are accustomed to encountering occasional glitches that can be solved only by having Novell tech support dial in and "do some magic" to solve the problems. Annoying, but tolerable. In the profit-center world of taking, tracking and managing customers, it had better be possible to do the magic in-house. Development and management tools need to be robust, and organizations need to be able to fully understand the operations of systems that sit at the heart of their income-generating businesses. Novell isn't used to this; that's the myth of One Net.

A newspaper publisher serves as a good example. A publisher owns two kinds of printing presses. One, a laser printer or copier, is a cost-center device and is generally serviced by whoever makes the thing. The other press prints newsprint, and in-house staff better be able to fix 99.9 percent of any problems that might develop with that press.