ast year was an important one for Web application servers. These midtier servers are strategic technologies that plug some major holes for companies that want to run real business applications over the Web. They connect Web applications with back-end systems, provide integrated development environments to ease application building and deployment, and offer robust architectures and low-level services that help make Web applications more reliable, scalable, and readily available.
The media and analysts were quick to tout the significance and value of this new class of software. This technology is instrumental in helping companies build and run real business applications over the Web, while leveraging their investments in existing applications and infrastructures. The technology also plays a key role in supporting critical E-commerce applications.
Vendors such as Allaire, Bluestone Software, Forté Software, Haht Software, Pervasive Software, SilverStream Software, and Vision Software introduced robust products that are gaining attention. Participation from major software vendors also legitimized the market: Netscape acquired Kiva Software, Sun Microsystems acquired NetDynamics, and Microsoft released products such as Visual InterDev and Windows NT.
Many predicted the Web application server market would go the way of the Web server market, with only a few major vendors left standing. Indeed, with performance and scalability approaching parity in many products, there are signs of commoditization.
But I don't think commoditization is imminent. The market is new enough to support more than two or three vendors--and more appear all the time. A year ago, there were maybe a half-dozen legitimate products in this market. Today, there are more than twice that number.
Web application servers are now available from independent vendors that specialize in this technology, from major vendors such as Apple, Microsoft, IBM, Oracle, and Sun, and from middleware vendors such as BEA Systems, Inprise, and Iona Technologies. Even IBM's Lotus unit has a Web application server group. Which is best? The jury's still out.
The independent vendors show few signs of going away quietly. Many of these companies are sufficiently funded, have experienced management and marketing teams, and offer strong products that are no longer first-generation. These vendors are also ensuring continued growth by going after different or smaller market segments and application types than the big vendors, who mostly pursue large-scale enterprise implementations.
What does this mean to companies that need Web application servers? It means you shouldn't overlook the smaller vendors. Many of these companies will still be thriving down the road because the market won't shake out much in the next six months. This may be the best time to buy.
It's more important to look at your company's resources and technical expertise. The process of building scalable Web applications still isn't simple. Different products use different approaches, and will require varying degrees of programming skill and expertise with advanced technologies and architectures.
It's also important to pick a product that can leverage your existing infrastructure, legacy systems, and line-of-business applications. Web application servers are supposed to connect these systems into higher-level applications available via browsers--hardly a trivial undertaking. Your Web application server needs to work with what you already have--it's not a replacement for your existing systems.
All of this means plenty of custom integration and development work--no matter which tools you choose. So expect to spend money on IT services. Letting business managers or marketing folks build the applications probably isn't the way to go.
Jeetu Patel is VP of research at Doculabs, a Chicago advisory firm that focuses on the Internet, intranets, electronic commerce, document management, and knowledge management. He can be reached at info@doculabs.com.