Comment: The Future of Java
On the 27th, Oracle presented its roadmap for the technologies it has acquired with its takeover of Sun Microsystems. Oracle took great pains to stress that it could now deliver a complete stack of integrated hardware and software, from the processor, operating system, language platform, to the database, middleware and business applications. After nine long months of fevered speculation, debate and scaremongering, Oracle made a visible effort to ease the IT world's collective mind. Besides promising a range of improvements, it pledged unwavering support for the community of Java developers and users. But are all these promises realistic, considering that Oracle, in one fell swoop, has purchased an entire portfolio of competing products?
Development of the Java Runtime
In his presentation, Thomas Kurian discussed the roadmap for the Java SE Runtime. Alongside many promises of optimisation, the two things that struck me were the disclosure that Hotspot and JRockit were to be merged, and Oracle's pledge to continue supporting all the leading operating systems. The consolidation of Oracle's two JMVs is the only reasonable way of guaranteeing continued development of the platform. Since the JVM specification will continue, the impact of the Hotspot/JRockit merger on planet Java will be negligible. The 'leading' operating systems in this context are, of course, the ubiquitous Windows and Linux systems, and the in-house Solaris Unix derivative.
Which application server for which application?
Later in his presentation, Kurian moved onto Oracle's two Java Application Servers. Under Oracle, GlassFish will continue to be supported as Java EE reference implementation and will keep to the same release model and road map. Meanwhile, WebLogic will continue as a strategic product for enterprise applications. A technology exchange is planned between the two servers. Because of the conceptual proximity between WebLogic and GlassFish (Admin Console, domain model, etc.) this shouldn't cause any problems. Unfortunately, more details cannot be gleaned from the presentation, which leaves room for speculation. It would be conceivable for GlassFish to serve as a 'community edition' of WebLogic, which would support Kurian's vision of GlassFish as a target server for smaller departmental applications. It is questionable just how much 'enterprise capacity' will remain in GlassFish. For example, for EE-reference and departmental applications it would be possible to renounce the GlassFish monitoring, which is weak compared to the WebLogic anyway. Even the clustering could be renounced in extreme cases. The problem is, that with so many strong open source competitors, there isn't much room in the server market.
With the integration of Sun products and technologies, Oracle has filled the spaces in its integrated hardware and software stack. The combination appears coherent and allows for optimal matching of all Sun/Oracle components. This will bring its advantages. Unless this is realised at the cost of the platform neutrality of Java SE / EE, it's difficult to object to this part of Oracle's plan. But it still remains to be seen how the integration of the competing middleware components could be realised.