BEA was the first member of the Java community to produce a product that simplified programming for the notoriously complex J2EE platform, introducing WebLogic Workshop, in June 2002. Behind the product was a simple strategy: hire brainy ex-Microsoft programmers to bring visual development techniques to Java, notably the Controls architecture that packages code and drastically reduces the need for hand-coding.

BEA expected to assemble a community of one million developers within a year of launching Workshop, drawn from Visual Basic and Power Builder developers sick of .NET, and Java programmers struggling with J2EE. Two years, and another version of Workshop later, BEA is still not only below the magic million number, but also is extremely reluctant to say just how many programmers it actually has.

Turning to open source

One thing is certain: things aren’t going well, and BEA has now adopted a new approach to expanding the Workshop and WebLogic Platform community – open sourcing Workshop’s runtime framework. BEA hopes the open source community will pick up the Workshop framework APIs and tailor them to non-BEA platforms, notably the popular Apache Software Foundation’s Tomcat web server, and possibly even Eclipse.

Codenamed Beehive, the project has been submitted to and accepted by Apache. Open source projects are notoriously complex and costly to run – two reasons why BEA’s number-one Java rival IBM this year spun out Eclipse as an independent entity. By leaving the Beehive with Apache, Apache will meet the cost and expense of managing the open source project, and will ensure that Controls is tailored to one of the industry’s leading web servers.

Problems at BEA

Beehive is only the latest indication that things are off kilter inside BEA. The company positions itself as the world’s leading application infrastructure software company, yet unlike other infrastructure vendors, notably IBM and even partner Hewlett-Packard, BEA has until very recently been silent on Service Oriented Architectures. At BEA’s eWorld developer conference in May, executives told WebLogic developers to begin building SOAs now. However, the role of any platform provider should be to lead by articulating direction not to follow others.

The company also seems to have misread the market. During this year’s first fiscal quarter, BEA missed targets for new license business and revealed a sales reorganization had taken place to help focus BEA’s business on millions of small and medium-sized businesses and vertical sectors. BEA is now planning componentized versions of WebLogic to suit customers’ specific needs, so they don’t have to buy a full platform. IBM, however, has been following the SMB money for more than a year, creating a business selling to SMBs and verticals in recognition of the fact that enterprise customers’ IT budgets have still not loosened up.

The outlook

BEA co-founder and chief executive Alfred Chuang told press at eWorld that his goal is to grow BEA to a $3 billion company over the next five years. It became a $10 billion company in 10 years, and stands to capitalize on a strong business, powerful WebLogic franchise and convincing ability to bring out new technologies first ahead of competitors.

Furthermore, putting BEA’s recent quarter in perspective, Chuang correctly said that building a company is a lengthy story, which goes far beyond the inability to execute in one area during a quarter.

However, BEA could struggle in the near future unless it puts recent problems completely behind it. The days of the 1990s when growth was a simple equation are gone, and BEA faces declining market share in J2EE application servers at the hands of IBM.

BEA’s ability to grow during the next five years will be limited unless the company sharpens its long-range strategic focus, and rediscovers its ability to both pre-empt trends and cause others to follow in its footsteps.