Migrating to Java 5.0 — Reality or wishful thinking?

I’m probably not the only one with the newly released Java 5.0 warming up on my disk. The question now and in the coming days is whether it’s worth moving to it. It’s less about Java itself than the support for it. Can we afford to move to 5.0, and what do we get from it?

I think we should distinguish two perspectives. The first is development itself. Do we have tools that can make use of the platform’s new features? I’m thinking primarily of development environments.

I wouldn’t be far off in saying most IDEs don’t fully support everything introduced in 5.0. My own IDE, Eclipse, is an example. The release notes for 3.1 M2 and 3.1 M2 make it clear that work on full support is just beginning. The NetBeans IDE may be a little further along, specifically version 4.0 Beta 2. Commercial environments like JBuilder and IntelliJ IDEA boast full support, but do they really have it?

Leaving IDEs aside, there are the other tools and frameworks we use. The first signs of support are starting to appear. Tomcat 5.5, for instance, proudly supports 5.0. But these really are exceptions we can set aside for now.

The second perspective is actually deploying products on 5.0. We don’t yet have performance test results comparing 5.0 with the 1.4.x or 1.3.x series. Reliability and security are important too. Too little time has passed for 5.0, or more precisely the JRE, to be properly tested outside Sun’s labs.

Version 5.0 is a turning point and will take hold, but from my perspective it’s too early to migrate for development or deployment. There’s no point pulling numbers out of thin air, but personally I’d wait at least for the first update, 5.0 Update 1, then decide based on where things stand.

Related articles

• Tigers and Mustangs and Dolphins, Oh My!