BusinessWeek reported yesterday that three core open source companies--JBoss (middleware), Zend (PHP) and Sleepcat (database) --are in Oracle's crosshairs. The omnivorous Oracle recently made waves by acquiring InnoDB, an open source storage engine for database competitor MySQL. It's not surprising. Larry Ellision has said the company would embrace open source (more like bear hug) and make it work to Oracle's advantage. It part of the Fusion strategy--annexing territory (customers and code) like Alexander the Great.
It's kind of a shell game. Instead of licenses for software, you pay a subscription fee for service and support but not for the software. The difference is you can get open source code for free, and get support from anyone. However, most enterprises will choose to pay a reputable firms for support. According to the BusinessWeek article, Oracle will do about $7 billion in subscription sales, with a 90 percent profit margin, this year. In any case, Oracle scoops up the revenue and increases its market share by acquiring the assets, which consist of customers, a mature open source distribution and key programmers and community members.
With a super complete stack of open and closed source enterprise software, Oracle becomes the one stop shop. Oracle can offer a la carte and full suite pricing (similar to Sun, which is open sourcing all of its code), and use its long reach into corporations and and extensive sales and support infractructure to move customers from the free and open source plan to the paid and open source plan. Importantly for Oracle, it's a hedge against the rise of open source, which is going to have profound impact on Oracle's licensed software business. How long can Oracle hang with its proprietary database? The open source databases like MySQL and Postgres have millions of downloads, but not billions in revenue. Oracle and IBM now have free, limited versions of their proprietary databases. The entire software industry is coming to the understanding the volume counts, and that free is a key growth lever. But, increasing open source will trump closed source, and Oracle isn't blind to that point. Ellison and company seem to be hatching a strategy that stacks the deck in their favor.
this overlooks the fact that you can't buy open source communities. It just doesn't work. Oracle can't even buy the technology from JBoss, arguably the most important of the troika, since it's LGPL and they don't require copyright assignations back to the company. (That said, JBoss does the bulk of its own development.)
So, Oracle can knock itself out "consolidating" the open source commercial world, but let's not kid ourselves. The only thing they get from it is a press release and 15 minutes of fame. None of these companies will be highly accretive to them.
They're hoping to buy community, but in this they'll fail. This is a big win for IBM and Geronimo, and for MySQL (though Sleepycat isn't a direct competitor - Sleepycat is an embedded database, with strong read technology but not ideal for transactions, which is where MySQL increasingly shines).
Matt suggests that the open source communities targeted will not surrender willingly into the arms of Ellison's army. But, Oracle will do whatever it can to appease the communities--if any of these deals go down. When you consider that most of the core development happens within the targeted open source companies, keeping the programmers happy is the key. Based on being one of the only games in town and its track record with its closed source roll up, I wouldn't count Oracle out.