There was never any confirmation, however, that Oracle was seriously looking at JBoss, even though many in the industry seem to have assumed it was in the works. JBoss’s close ties to Microsoft and Oracle’s own substantial middleware investments would have made the two an unlikely pair. JBoss and Red Hat, on the other hand – now that makes sense.
Although Red Hat released its own application server about a year ago, it has nothing to compare with JBoss’s product. And as successful as it has been so far, JBoss is dwarfed by the brand recognition and industry relationships enjoyed by Red Hat. More than likely, many JBoss implementations are already happening in Red Hat environments, so the combination will only increase the opportunities for further integration and stability around a complete open source package.
Red Hat’s purchase assures JBoss users of the company’s future in a way that no acquisition by a proprietary vendor would have. A buyout by IBM or BEA (who were also rumoured to be interested) would have been interpreted as a gateway deal — a means to achieve credibility among certain kinds of developers, to gain access to a particularly interesting customer list and to feed off the momentum around open source in the way that, for example, Novell has. If what I’ve heard is true, however, JBoss had yet to reach its peak. It was not the platform that people deployed on so much as they experimented. Red Hat has the chance to bring a promising set of technologies to their maturity and reap the benefits without having to invest as much R&D in its own product, which I suspect will be killed off or quietly subsumed into the JBoss product.
The deal immediately catapults Red Hat into a more interesting position in the business software market, though not exactly in the same league as the closed source firms. IBM has supported Red Hat on a number of fronts, but its own purchase of Gluecode and its frenetic effort to market its middleware portfolios towards service oriented architecture projects means the Big Blue team won’t be losing much sleep over this merger. JBoss was also moving in a SOA direction, but Red Hat would probably have to pump a lot more money into developing something that would rival Big Blue’s line. The same is true with Oracle, whose Fusion middleware is already available and will evolve considerably once the integration with PeopleSoft and Siebel is complete. BEA might have used JBoss to foster greater innovation around its own product line, but it basically remains as vulnerable a target as ever.
If there’s a downside to Red Hat’s US$350 million purchase, it’s that it keeps JBoss cloistered in a part of the market it seemed destined to transcend. Certainly it will help to have a Linux distribution paired with strong open source middleware, but I’d rather see it flourishing amid products that are both open and closed. Without alienating its core audience, Red Hat needs to ensure that JBoss fulfills its potential to become a great application server, not just a great open source application server.