Red Hat’s JBoss acquisition

I was pretty surprised today by the JBoss acquisition announcement. After some fumbling and missing huge opportinuties in the open source middleware space for a few years, Red Hat may finally be on the right track.

It’s a great strategic move, but it’s just mind-boggling why it didn’t happen like 3 years ago when JBoss had already established early dominance in the open source middleware space. Now Red Hat is paying up the ass for JBoss, and rightfully so. The JOnAS-based Red Hat Application Server was a valiant attempt at producing a subscription-based middleware solution with which Red Hat could tie together the entire application stack, but it seems that RHAS never really got off the ground. Don’t get me wrong, there are many insanely brilliant and talented software engineers at Red Hat, but surely management had to know that it didn’t have the middleware expertise needed to pull off its own application server business.

Of course it’s all about the juicy JBoss support contracts. It’s also about (finally) achieving an entirely Red Hat software stack, even though Red Hat *still* doesn’t have its own database product. It’ll have to buy another company to get that. Red Hat isn’t a software engineering company, it’s a software support company, and it’s lovely to see with the Sistina, Netscape Directory, and JBoss acquisitions that the company is starting to realize what it is and what it is not. “We know we cannot build it, so we must buy it” suddenly rings true.

I can’t wait to see JBoss AS running natively on libgcj et al. That will surely be sweet.