Thursday, April 30

Sun of Oracle

Late last year I wrote a post entitled ‘Java may slay everything in its path - including Sun’ and asked the question: If Java is so universally good - it has been trying to prove itself for 13 years - then who enjoys and makes money working with it?

It appears that the only vendors who made any real money from Java are Oracle, BEA Systems (now Oracle) and IBM. As I've been tardy about posting to my blog there now seems little left to add to the well-documented saga of IBM choosing not to consummate a marriage with Sun and the man from Oracle saying “I do”. There is news today that Sun’s revenue plunged 20% last quarter and losses jumped to $201m – clearly a broken business model.

With Java finding a home at enterprise software vendor Oracle, whose sales force are notoriously weaned on six- and seven-figure deals, I have to believe that the door is now wide-open for Microsoft to dominate the SMB market. I have long believed that Java’s complexity and running costs have restricted its market opportunity and that Sun’s failure to find a viable business model (for this is what a sale at such a low valuation indicates to me) confirms my suspicion. Note that I am talking about commercial viability and not specifically whether the Java technology is good, bad or ugly.

I suspect that IBM i shops will be eyeing this transaction with some trepidation if they have bet the farm on a Java strategy. Larry Ellison has already talked about optimising it products for the Solaris platform and, by inheriting the MySQL database, Oracle has a new stick with which to beat IBM. I think it’s fair to say that Oracle and IBM are fierce competitors, far more so than either party’s current relationship with Sun. Java bigots will tell me that so much of the technology’s evolution is now influenced by the community process that a change of ownership will make little to no difference to Java’s future prospects. I think that viewpoint is naïve and woefully under-estimates Oracle's ability and appetite for making money.

It is a mantra of mine that to place bets on the future of a particular technology is to be certain of ‘losing big’ at least one time. That’s why I prefer to work with products that operate at a higher level of abstraction, thereby providing a degree of technology insurance rather than lock-in. Having choice is important at the operating system, development language, database and presentation layers because business needs and market conditions are always changing. So a rigid commitment to, say, DB2 + Java + Web carries risks that are not always communicated or considered.

I would not expect die-hard Java shops to down their tools just yet, but I urge those of you on the fence to reflect on this latest power play and consider the long-term implications for you and your organisation. If you are ready to step-off the treadmill of technology dependence then, as a reader of my blog, you hopefully know where to come first.