Friday, May 1, 2009

Java and Oracle, the real question.

So unless you've been living under a rock, you know that Oracle is buying Sun. And of course one of Larry Ellison's motivators to do so is to gain control of Java.

Feelings on how this will affect Java run the gamut from "Blissfully happy" to "The end is nigh".

The press seems to look at Java aspect of the acquisition regarding the product overlap and how little money Sun made off Java. They also seem to view Java as purely an enterprise play, This is an incredibly narrow view as makers of everything cell phone apps to desktop apps to Blu-ray applications will happily attest to.

Java is not a product in the traditional sense of the word, Java is more a community marketplace centered around the Java technology as opposed to a shrink wrapped product.

Java's real value IMHO is then in it's vast reach both in the general development landscape and the vast community of programmers, not so much in the IP or the source code that makes up the JVM.

My question is then if Oracle realizes this.

Community is something that Oracle has not traditionally had to deal with.

Oracle has always been a product company focused on generating healthy margins of all of it's products from a set of clients with deep pockets.

It is now in a position where it will need to support and grow a community of Java users which provide little or no tangible benefit to Oracle's bottom line.

There are many examples of this.

One such example is Glassfish which overlaps with Weblogic and seemingly has no place, however Oracle is now in charge of the JCP and as such has to provide reference implementations for the specifications somehow. Glassfish currently fulfills this role. They could maybe use Geronimo or even JBoss but in a space that will eventually commodotize, it might a bad idea to promote a competitor. Oracle may decide to cripple Glassfish purely to be used for reference but then the fact remains that they have to expend resources on a piece code which makes them no money in order to support the Java community.

Another is the fact that the Java community also consists of companies that compete directly with Oracle. Oracle as the Java shepard must now support these competitors: most notably IBM; the 800 pound Gorilla of the Java community. Strategically it's better for Oracle and IBM to jointly run Java even as competitors lest they both cede a great deal of the Enterprise market to Microsoft. Oracle must therefore keep IBM happy, potentially even spend money to keep them happy in order to preserve the status quo. It must also do the same for it's smaller competitors to expand the reach of Java as far as possible.

Finally Oracle now enters into strange new markets through Sun's acquisition. Oracle's never had any presence in places like the embedded market and now must do work for people who make such strange things as cell phone applications and software for set top boxes, many of whom will never even see an Oracle database or an Oracle App server. Oracle must now keep them happy potentially at margins which don't quite resemble what they make on 11g CALs together with the reality that there is little chance to sell other Oracle products.

No doubt Oracle's shareholders will want Larry Ellison to make good on Sun's purchase but to fit Java into the way Oracle traditionally does things is probably not going to work and might very well end up killing Java.

1 comment:

Unknown said...

I think Java is not going to die... Mainly because it is not dead by now! Sun was not performing good at all anyway. IBM and Java enthusiasts, such as Apache and Spring guys, mostly led it before...