Kissing the J2EE software business goodbye?

When these companies collide, Eric Knorr says HP's homegrown efforts to compete in J2EE and Web services may be part of the collateral damage.

commentary If Hewlett-Packard and Compaq combine, the merged company will immediately square off with IBM, in a showdown involving hardware, software, and services.

HP/Compaq will have hardware to spare and instantly become the third-largest services organization after IBM and EDS. But software? What software?

One of the secrets of IBM's success has been its WebSphere product line, which extends from its J2EE application server to its MQSeries middleware. Of course, IBM Global Services will deploy whatever software its customers want--but when it comes to integrating legacy systems or rolling out Web operations, WebSphere has given the company a powerful, profitable tool. By contrast, HP's amorphous middleware appears to be in disarray--while Compaq has no homegrown software to speak of at all.

Ironically, HP was an early leader in the Web services trend with its E-speak initiative, which fizzled due to proprietary technology and a lack of corporate commitment. In late 2000, HP acquired Bluestone, developer of Total-e-Server, a well-respected J2EE application server that never rose above a single-digit share of the market. Then, last fall, HP Application Server (HP-AS) 8.0 arrived, which HP now offers as a free download in a desperate attempt to gain market share.

If the HP-Compaq merger goes through, I wouldn't be too surprised if HP phases out its native application server business in favor of partnerships. The lead enjoyed by IBM, BEA, Sun, and Oracle just seems insurmountable at this point. BEA would be an obvious partner, since BEA WebLogic is engaged in the battle of its life against WebSphere. At the same time, both Compaq and HP have voiced strong support for Microsoft .Net--a realm IBM may be reluctant to explore deeply, since it competes with its J2EE-based WebSphere business.

IBM is a rarity: a hardware company that's equally strong in software and services. HP and Compaq have been pushing services for some time, but they have some serious catching up to do if they want to take on IBM (most of their service employees simply repair stuff or run outsourced computer operations rather than consult). Trying to build a J2EE software business at the same time sounds like a losing proposition--and buying a software company, as in HP's acquisition of Bluestone, hasn't worked too well either.

In the age-old choice of build, buy, or partner, partnering would make the most sense for HP/Compaq's customers, who could freely choose either a .Net or J2EE solution from the company as circumstances dictated. Of course, HP may instead cling to its native J2EE software and (like Oracle and IBM) push it as part of a package deal. Customers developing on that platform could end up locked into a niche solution with limited staying power.

Eric Knorr is a an award-winning freelance writer and consultant who frequently works with e-business consultancy Envivid Solutions in San Francisco. He is the founding editor of CNET's