I woke up this morning to the news that Oracle, which on April 20 said it will purchase Sun for $7.4 Billion USD. This news came, just after IBM withdrew its offer earlier in the month.
One of the key assets of Sun, is the Java programming language. Larry Ellison, CEO of Oracle, has said Java is "the single most important software asset we have ever acquired.".
I’ve been involved with Java since it first came onto the scene, I think some thirteen years ago now. I can remember traveling to one of the earliest Java conferences in Melbourne, Australia and receiving a version of Solaris for Intel. I then spent a couple of days getting it working on my PC (it might have been a 386 or a Cyrix box, can’t remember now). But that shows the age, not just of myself, but of the Java language. It has had a long time to embed itself into Enterprise software development and has accrued some good and not so good legacy.
In Jan 2008, I wrote a blog entry "Java losing some of its mojo in enterprise development?", where I was questioning the relevance of Java moving forward. What’s more, I was hoping that Java would evolve not to compete against C# and .NET but because it makes logical sense to do so. I see now that with Oracle’s stewardship careful focus will be required to see how the Java Community Process unfolds and how R&D efforts will be directed. The pressing question in my mind is will Java’s evolution be tied purely to support ERPs moving forward?
On the positive side, I see Oracle taking the fight to Microsoft. Which can only be a good thing, given the current complacency in a number of organisations towards just using Microsoft based environments.
However, why did IBM pass up on this opportunity?
Java after all underpins the majority of its current software portfolio. Or does it?
In my 2008 prediction blog, I wrote at the end of 2007, that we would start to see the further evolution of language neutral SOA standards, such as SCA and SDO. This starts to open the potential for other languages besides Java to evolve in prominance in the middleware space. I’ve written about this evolution as I’ve seen the quiet rise in the background of the Composite Service Architecture specificaton through the OASIS Open CSA committee. In parallel I’ve seen the evolution of REST (Representational state transfer) based approaches, which are more readily understood by developers, than following conformity to standards developed by others.
Lets look at IBM’s Project Zero. The opening splash describes the project as "Delivering the best of agile Web 2.0 with PHP scripting, REST and Dojo in an integrated runtime and tooling package: IBM WebSphere sMash". The commercially supported product is called IBM WebSpere sMash. Did you notice that there is no mention of Java in this new IBM product?