Still in "ponder" mode on the possible Oracle acquisition of JBoss, and came across this interesting tidbit from the Enterprise Open Source Journal (worth a read, if for no other reason than to see Julie in "pensive mode :-). Mark Driver from Gartner writes [PDF download]:
"The [Oracle acquisition of InnoBase] has left many of us scratching our collective heads. On the surface, the motives seem obvious. Oracle, with a clear history of aggressive acquisitions, buys out a critical component of MySQL in order to kill off its commercial support channel. The problem is, we don't actually know the motive yet because Oracle has been mysteriously silent about the acquisition, citing "license negotiations" as the reason. By the time you read this, the issue may be resolved, but I fear it's a lose lose situation for Oracle, no matter what the outcome. Whether deserved or not, once the "stink" of being the bad guy sticks to Oracle, it will be virtually impossible to convince anyone in the open source community otherwise for many years to come.Bingo!
Not directly related to Oracle, per se, Mark also writes:
Mergers and acquisitions are a natural and healthy part of any free market economy. It's only natural then to see M&A activity as part of the open source landscape. However, many nuances in the business models that drive commercialInteresting stuff, Mark. Why aren't you speaking at OSBC? :-)
support for open source differ from traditional IT vendors. It's one thing when a company such as RedHat acquires Cygnus (as it did in 1999), but entirely another when an old-guard, traditional IT company gobbles up a smaller open source start-up. Are these acquisitions good or bad for open
source? The answer, I fear, is a bit of both.Companies buy other companies for many reasons, most generally to either acquire products and technology (including perhaps the human capital) or to convert a customer base. A large, established IT vendor might also buy a small Open Source Software (OSS) start-up in an attempt to gain instant credibility and influence within the open source community or simply to gain access to the "inner circles" of the OSS community. [Note: I've written before on whether one can effectively buy an open source community. Verdict? Not a chance.]
Larger vendors have the deep pockets to infuse a project with the capital and resources to accelerate adoption and critical mass. Moreover, large vendors can provide a level of clout and reduce the trepidation that more conservative users may have toward open source in general; after all, numerous IT organizations leverage multi-billion dollar vendors as safety blankets against risk.
However, these acquisitions can also be used as a weapon against open source. It's no secret that open source projects continue to mature in markets with direct competition against incumbent closed source solutions. Many large vendors might decide the best defense is a strong offense and simply attempt to kill off competing OSS vendors as they emerge.
For example, in May 2005, IBM acquired Gluecode Software, which specialized in supporting technology from the Apache Foundation. In particular, Gluecode was the principle source of engineering effort behind the Geronimo J2EE application server project. So, why then is IBM interested in a tiny software company focused on an OSS application server? After all, investments in Geronimo would certainly overlap with its own WebSphere product line. However, Geronimo is also a real threat to JBoss. IBM's strategy is a clear example of "the enemy of my enemy is my friend." I think IBM saw the writing on the wall, witnessing the pounding that BEA is taking from JBoss. Essentially, any OSS application server providing an alternative to JBoss is a good thing for IBM in the long run. On the upside, IBM provides a level of funding and support that Gluecode could never dream of as a struggling open source, pure play. This in turn will significantly help the Apache Foundation in driving Geronimo to critical mass much quicker. On the downside, the jury is still out on whether IBM will attempt to influence the direction of the Geronimo project and steer it away from potential competition with its own application server stack. In fact, IBM recently made WebSphere Application Server Community Edition available for free download. Two very interesting issues arise, however. First, WASCE is not open source-just free as in gratis. Second, it's not a full implementation of the Geronimo project. IBM hasn't included Geronimo's JBI-based ServiceMix Enterprise Service Bus (ESB); not surprising, since IBM doesn't support JBI.
For my part, I like the smell of IBM acquiring Gluecode much more than Oracle buying JBoss. IBM has legitimate open source street cred, and the strategy (Geronimo as the low end to Websphere) is a legitimate one, whether one agrees with it or not. But I have a hard time seeing JBoss thrive in Oracle's hands.
Posted by Matt Asay on February 7, 2006 04:18 PM
