So after I wrote favorably about a possible Sun acquisition by IBM, it's Larry Ellison and Oracle that actually consummate the deal. How about them apples? The technical fit-togethers are fairly obvious, but it's the market and business model issues that are going to be problematic. Looking at the major components of Sun's portfolio:
HardwareSun had been pushed into a niche building high-end servers at a low volume, not a sustainable position. IBM has other hardware lines, and has a long track record at being able to juggle their customer bases and arrange convergences or divergences to suit. Oracle has no other hardware business, and in its hands the business remains isolated, still beset by solutions built on arrays and clouds of lesser, but high volume servers. Oracle has software in its bones and that line of business will not be compromised to sustain a hardware line. Conclusion: Sun's server line dies a lingering death.
SolarisOracle has no existing operating system line, aside from its occasional desire to have no OS at all. Solaris' revenue opportunity is tied strictly to Sun's hardware, and there is no possibility to widen that footprint given the presence of Linux. Conclusion: Solaris dies a lingering death as well.
JavaIn spite of changing its stock symbol to JAVA, Sun never figured out how to directly extract revenue from its "dangerous watering hole". Oracle, Sun and IBM were of course all part of the 'Rebel Alliance' that pushed a combination of net-based services and browser-based clients as the antidote to Microsoft. Large portions of that vision are winning, and even the 'network computer' is coming back in other guise as 'netbooks'. Java failed as the client side part of the alternative platform, but found a solid niche in applications on the server side. That role is synergistic with Oracle's server side database and applications position, and it will now have a stronger voice in Java's direction. That begs the question that Java's users will now be asking: Will the semi-stable détente among Sun, IBM and other Java supporters remain stable when Oracle takes over the Sun position, or will Java become a strategic weapon between IBM and Oracle. Conclusion: Outcome unclear, but a few upside bets on Python and Ruby might be warranted.
MySQLHere is the heart of the matter. Oracle's mainstream database business has long been threatened by price and margin erosion, first from Microsoft and later by open source alternatives, particularly mySQL. IBM faced the same threat to DB2 (and other parts of its software business) and responded by shifting much of its revenue model to services, a business that remains fairly healthy today, in spite of recession (PDF). Oracle, too, has been working to establish a services model, but according to its 2008 10-K:
"Our services business, which represented 20%, 21% and 20% of our total revenues in fiscal 2008, 2007 and 2006, respectively, has significantly lower margins than our software business."But where has Oracle's strategic weight actually been applied? From the same filing:
Over the last four fiscal years, we have invested billions of dollars, including $9.4 billion in fiscal 2008, to acquire a number of complementary companies, products, services and technologies such as BEA Systems, Inc. in fiscal 2008, Hyperion Solutions Corporation in fiscal 2007, Siebel Systems, Inc. in fiscal 2006 and PeopleSoft, Inc. in fiscal 2005. We believe our acquisition program supports our long-term strategic direction...Oracle has tried to 'thicken' its application layer, moving from raw database to applications servers to fully configured solutions, and this effort has overbalanced its services initiatives.
The market conflict between high-margin, high-capabilty software and high-volume, commodity and OSS based solutions still exists, and Oracle has now brought that conflict in house by acquiring MySQL. If Oracle has become serious about shifting its strategic base from one side to the other, this is a good first move. The more usual outcome is that an acquisition which threatens a long established business model of the buyer is encysted and eliminated by the corporate immune system (remember Sun and Cobalt?) - to the detriment of the customers. MySQL is of course a strategic element of the famous OSS LAMP stack. If it is compromised, room will be left for an alternative, but its necessity would be a detriment to the open systems side.
(Update) To summarize: Solaris, SPARC and Sun's hardware are toast in the long run. To keep them alive Oracle would have to become competitive - simultaneously - with Intel, Microsoft, Dell and all their buddies, plus the Linux community. Ain't gonna happen.
The acquisitions of Java and MySQL are good for Oracle. What remains to be seen is whether they are good for the respective communities and existing users. Some prudent hedging is in order until Oracle's trajectory is clear.