The Next Big Thing

Of course, the whole premise of this tutorial is arrogant beyond belief. I'm making an incredible number of assumptions and drawing some aggressive conclusions based on little more than a couple of dozen interviews, a keen sense of intuition, and a few massive piles of circumstantial evidence.

Java may need nothing more than a little overhaul. Maybe the problem is in the massively complex libraries, and a few rewrites with some tweaks of the language would extend Java's leadership for 10 more years. Maybe the community's culture doesn't help define our libraries. The driving vendors may do an about-face and focus more on simplifying the 80% path instead of building yet another XML-obsessed framework. The JCP could suddenly start supporting the best existing frameworks based on experience instead of standardizing a good idea that was born in a committee.

Maybe Dion Almaer is right, and the big companies that drive this industry are not remotely interested in moving away from Java, and we'll all be saddled with Java for the foreseeable future.

Maybe Jason Hunter is right, and the next big thing won't be a programming language at all. Maybe Java's all we'll ever need, and we'll use that foundation to move up the abstraction ladder. Maybe Glenn and David are both right and there won't be one next big thing, but lots of next little things, and both metaprogramming and continuations will play a significant role.

I don't know the ultimate answers, so I've leaned on my mentors and peers. The interviews in this tutorial are the opinions of some of the people I respect the most. It's been an honor to share these few pages with them. I'm not ready to say that Java's dead, or that Ruby is next, or that continuation servers will reign supreme. I just know:

  • I'm hurting right now, and my customers are, too. It's getting harder and harder to teach my customers to satisfy themselves with Java.

  • Certain things, like baby-sitting a relational database with a web-based UI, should be easier in Java, after nearly 10 years of effort, but they're still cumbersome.

  • The same people that dozed in conversations about other languages two years ago seem to be paying attention now. My "Beyond Java" talks, at Java conferences, are continually packed.

As for me, my eyes are wide open. I've seen what the alternatives can do. In particular, Ruby on Rails lets me build reliable code fast, and put it in front of my customer with more confidence and frequency. I've got every reason to maintain the status quo. I did find that some of the alternatives are compelling, and make for a smooth transition.

A Charge to You

If you're a Java developer and this message is troubling you, that's natural. You've got good reasons to feel threatened with this challenge of your world view. You may feel even more unsettled when someone challenges the foundation of your livelihood. I'd encourage you to put this tutorial down and do some research of your own.

Look around. When James Duncan Davidson did, he found a language that responded to his needs for low-level user interface development. Stuart Halloway found a language that let his start-up move at the speed of his ideas. Dave Thomas found the foundation for an increasingly important publishing series. Glenn Vanderburg found languages friendlier to his beloved metaprogramming techniques.

If you decide to expand your horizons beyond Java, you may find that I'm right, and some of the alternatives I've explored here, or even some I didn't, unleash you. You'll be surfing the next wave that propels us forward.

If I'm wrong, Java will still be there for you; heck, even COBOL is still there for you. But to you, it won't be the same Java. Other languages will expand your horizons to other approaches, just as a wave of Java developers will bring our unique view of the world with us. If you spend some time in Smalltalk, you'll probably use Java's reflection more, you'll look for more opportunities to invert control by simulating code blocks, and you may well tone down your use of XML. (OK, I may have pushed things too far with that one.) If you explore continuation servers, you may look for a way to simulate that programming style in Java. If you explore Rails, you'll likely learn to pay more attention to defaults and convention. Hibernate, Spring, Struts, servlets, collections, and the JDO could all use these techniques.

Pick up your eyes by learning a language. Expand your mind to something a little more powerful, and a lot more dynamic. Warp your perspective to functional programming or continuations. Annoy your friends with a contrarian's view. Tell them that you don't think the world's flat. There's a whole universe out there, beyond Java.