Will the circle be UnOpen?
Email This Post
The buzz in IT markets and the Open Source world is Red Hat’s purchase of JBoss. Many have predicted the end of Open Source as we know it, and claim the purchase is clearly documented as one of the signs of the Apocalypse.
It is not as bad as all that, but people are scrambling. My morning email included requests for help from some Open Source organizations who fear that the mindshare of Red Hat and JBoss, and financial might of Red Hat alone might forever exclude other J2EE servers.
Unlikely, though it does significantly change the strategic landscape.
Markets reward winners and not whiners. Red Hat has brand preference for Linux. JBoss is rapidly gaining brand preference for J2EE servers and related SOA functions. The market has been steadily feeding both organizations more money because IT believes committing to market leaders is safer. For commodity J2EE servers, RedBoss certainly has momentum.
But market opinion is a fickle thing, and is based in part on brand trust. Other marketing mavens have correctly noted that a brand is "a promise kept." The promise of Open Source is one that IT has embraced. The worst thing that could happen to RedBoss is if this fundamental promise is forsaken. Much as Google broke its promise of "do no evil" when they embraced the evil dictates of Red China, RedBoss could break the brand if they violate any of the precepts of the Open Source pact.
This seems unlikely. JBoss is licensed under the Lesser GNU Public License (LGPL), which in my reading means the community could fork JBoss at will. But LGPL has derivative licensing arrangements, and this untested scheme might provide a stealthily shyster the wiggle room to take some of JBoss out of the public domain and promptly cripple any attempted fork. RedBoss would be insane to try this . . . today. But no telling if riches alone would cause Raleigh to act like Redmond.
I expect to see a backlash from several fronts:
- The other J2EE server communities (namely Apache/Geronimo and JOnAS ) already have significant mindshare
- IBM, a subject of derision from elders at JBoss, has a vested interest in promoting solutions that undermine the (un)holy union
- Novell can be the White Knight, and their cozy relationship with IBM might make a winning trilogy
From a market standpoint though, it is still a horse race. The adoption rate for J2EE servers (compared to straight HTTP or LAMP servers) is small, the learning curve to productivity is high, and thus there is a lot of elbow room in the market. But like Linux before it, the open ground is on the low end and continued development and defacto distribution of a completely unencumbered J2EE server will eventually displace anything that may well be too Red Hat centric.
