Almost one week has passed since BEA revealed its intentions to become the Switzerland of Services Oriented Architectures (SOAs). And, as with the Apple/Intel announcement, it usually takes about that long for news to sink in and for the dots between several takes on such deals to be connected.
My last blog on BEA's Aqualogic announcement had the headline "BEA reinvents itself as the Switzerland of SOA". In hindsight, the headline presumed that the reinvention has already taken place and was therefore a success. Already, though, I'm revising that assessment. Not only is this a bold step for BEA, it remains to be seen whether the strategy will actually work for a company which has, at the hands of IBM and open source J2EE outfits (like JBOSS), had its lunch eaten in recent years. Earlier today, in response to News.com editor Martin LaMonica's story on the rising popularity of the open source LAMP stack (Linux-Apache-MySQL-PHP/Perl/Python), I caught up with LaMonica via instant messenger. After I pitched him on a story about the open source LAMJ stack (Linux-Apache-MySQL-J2EE) -- now fully legitimized by virtue of the way open source Java application servers like JBOSS and JOnAS are finally getting their J2EE compliance certificates -- the discussion took a "So what did you think about the BEA announcement turn?"
In his initial response to the announcement, LaMonica blogged "BEA certainly isn't the first to pursue this notion. But the company has a lot riding on the success of AquaLogic.... The other reasons BEA needs a winner with AquaLogic are fairly obvious: the company needs a revenue boost."
Indeed, the company needs a revenue boost. But in the context of what's happening in the open source space and this announcement -- a revenue boost may only scratch the surface of the re-invigoration it really needs. Although its portfolio stretches beyond its J2EE application server WebLogic, BEA has largely been a one-trick pony -- with most of its success being built on the back of the mothership WebLogic as well as off the ecosystem it created. The "Switzerland" theme BEA has used to underscore the Aqualogic announcement is in many ways an acknowledgement that the WebLogic business has plateaued for BEA. In fact, given trends in the application server market, continued reliance on WebLogic as the mothership would probably have sunk the company over the long run.
Like every other part of the software industry, open source is taking its toll on commercial software providers. The aforementioned LAMP and LAMJ stacks (both open source) have always been something of a threat to software providers that specialize in dynamic, Web-facing, business process-oriented application infrastructures (eg: application servers, services oriented architectures, etc.). But now, as evidenced by LaMonica's story, LAMP is hotter than ever and, I'd argue, LAMJ is just as hot if not poised for even more stardom (than LAMP). One trend that both stacks having going for them is the rise of open source stack certifiers like SpikeSource and SourceLabs. As outfits like these continue to crop up, and continue to build credibility by hiring executives like Kim Polese and Bruce Perens, corporate America's comfort level (knowing that good help is around the corner) with these open source stacks will continue to improve. This can only mean bad news for companies like BEA that are so heavily dependent on commercial rivals to LAMP and LAMJ like WebLogic.
The writing is on the wall. I can see it. LaMonica can see it. And BEA can see it. And BEA's response, couched in the auspices of an exciting New York City announcement, is proof that the company realizes that its next move must be a dramatic one. The drama, of course, is that for the first time in the company's history, it is shifting from one of orthodoxy to one of agnosticism. In years past, for example, when the question of whether BEA and Borland (the WebLogic developer's tool provider of choice) should merge, one explanation for why that didn't make sense was that BEA's devotion to Java and Borland's neutrality as a toolmaker for both Java and Windows made the two cultural misfits.
But now, out of the blue, BEA is forsaking its orthodoxy for neutrality. Whereas before it was a premier ecosystem provider, it is switching gears in order to also mooch off the ecosystems of others (other J2EE providers as well as Microsoft with .NET). Once a host, now a parasite and with plenty of other parasites looking to ride the coattails of high-flying ecosystems like WebSphere, LAMP and LAMJ, BEA has way more competition on its hands -- and a much more difficult communications challenge ahead of itself -- than it did as a primarily one-trick pony in the J2EE application server space.
As good as the AquaLogic components are -- perhaps just as much best-of-breed for what they do as is WebLogic for what it does -- my sense is that there's much more drama ahead for this former darling of Wall Street. Up next for BEA is probably some sort of open source announcement, perhaps one that will insert it into the LAMJ fray before JBOSS does any more significant damage. In my interview with BEA VP of Marketing Bill Roth (see my last blog), Roth dropped a clue that some sort of open source-related announcement will hit the airwaves around the end of this month, about the time that Sun's annual Java love-fest -- JavaOne -- takes place in San Francisco. Beyond that, however, now that the company has shifted from host to parasite (well, to be fair, hybrid), the cultural barrier to a merger with Borland has been completely dropped. For the two companies to continue to survive the threat of players like IBM, JBOSS, Eclipse, LAMP, LAMJ, and Microsoft (with .NET and Visual Studio), perhaps now is the time that they'll realize that they can be stronger together than they can be apart.