The answer to these demands and complexity is: simplicity. We've covered the issue before (see "Keep It Simple, Stupid"), but managing complexity is probably the biggest challenge IT shops will have to face in the coming year. In an increasingly complicated environment, development projects must be simpler and easier--or, at the very least, no more difficult--to design, produce, deploy and maintain. We've recognized the products and technologies that should give you expanded capabilities without making your life impossibly complex.

Middleware Technology

Sooner or later, you're going to have to move into the distributed world--and when you do, you'd better have a clear road map and be using an underlying architecture and technologies that offer a combination of robustness, flexibility and innovation. Over the past year, we've looked at the top technologies on which were--or will be--based the most innovative IT middleware development products.

Distributed computing is nothing new; CORBA-based systems have been around for long enough to be considered "legacy." But in today's world, multiple platforms run multiple languages from multiple locations, and all need to access information and business logic from the other. With so many vendors and IT systems integrated in one automated supply chain, an effective distributed architecture isn't just a luxury or appropriate for only the largest enterprises. It's mandatory for everyone. That means distributed computing isn't something only shops with large, specialized staff need to be able to handle. Solutions must be as simple as possible, yet afford the functionality to achieve success in the distributed world.

Last year, Enterprise JavaBeans (EJB) was a promising technology, but immature. A powerful buzzword, EJB in reality lacked key items, such as container-managed persistence for entity beans, and had no definitive reference implementation for its many features. This left vendors free to extend the spec in incompatible ways--which is exactly what hobbled CORBA interoperability. A year later, following the release of the 1.1 spec, EJB is maturing; it now has solid implementations, finally including most of what's needed, and a solid reference spec. Vendors of development tools and application servers alike are racing to include full EJB 1.1 support, and it looks as if EJB is continuing to gather steam. There's growing concern over who will control the Java specs, and how (or whether) standardization will occur. But, at least for now, EJB and the rest of the elements of the Java 2 Platform Enterprise Edition (J2EE) spec are leading the pack in middleware technology.

This year, Web application servers (and their Java-focused siblings, Java application servers) were stronger, more robust, easier to manage and deploy, and more full-featured than ever. But there's a dark cloud to go with the silver lining: Application servers face increased commoditization, and vendors that don't find a niche and differentiate their products will find themselves going the way of the dodo. Windows 2000 and COM+ 1.0 services now include many of the key features of Web app servers, albeit sometimes in a primitive form. But the writing is on the wall, and middle-tier vendors will have to read it carefully to survive.

Easy answers won't suffice. Once everyone's product has been hastily rebadged as an "e-business customer-relationship-management supply-chain storefront B2B XML server," the bottom-line question remains: "Who's building the most robust, most capable middle-tier application server?"

One option for differentiation is to morph into a services and custom-solutions company, as the Art Technology Group did early on with its Dynamo product. Some larger vendors, such as IBM Corp., are clearly at an advantage here in that they have the staff and other resources to tackle the divergent requirements of selling products and services without giving either short shrift. This year, expect many of the vendors of Web app servers to shift into service offerings.

Computing on the Half Shell

Starting this year, look for broader application development in the handheld-computing realm. In the past year, handheld devices broke out of their niche and began to be accepted within enterprises as tools for the masses. At the same time, IT departments purchased so many laptops that the portable units threaten to overtake their desktop counterparts. This all translates into increasing challenges for the IT managers and administrators who have to deploy, manage, synchronize, back up and maintain myriad mobile devices.

Meanwhile, computers keep shrinking. The Palm platform finally made the handheld more than just another cool tool for early adopters, technophiles and vertical markets (see "Will Success Spoil the Palm Platform?"). While it clearly had a variety of predecessors' shoulders on which to stand--including Apple Computer's Newton and PDAs from vertical-market vendors--the Palm was the first product to pick the right set of features, in the right form factor, at the right price. The Palm chose simplicity, battery life, stability and usability over "featuritis"--good choices at a good time. This market is sure to spawn new applications, thanks to a large and enthusiastic developer pool and well-timed third-party licenses.