Beyond Java - Bruce Tate [0]
Bruce A. Tate
Editor
Mike Loukides
Copyright © 2009 O'Reilly Media, Inc.
O'Reilly Media
* * *
Preface
In March of 2005, I was shocked and honored to get the word that one of my books, Better, Faster, Lighter Java (O'Reilly), won a Jolt award for technical excellence. I talked about how Java? developers could buck standing conventions to build better applications, faster than before. That book will always have a special place in my heart. Yet, throughout the process, something was in the way, and I couldn't quite put my finger on it.
Around this time, one of my customers was building an application consisting of a complex database schema with a web-based user interface. We'd been using Spring and Hibernate with Web Work, a common stack for lightweight Java development, and we'd been generally pleased. Some things bugged us, though: the amount of repetition, the proliferation of XML configuration, and the pace of our changes. On a whim, we tried Ruby on Rails, a surprisingly productive and innovative framework that's sweeping quickly through non-Java communities, and is making some noise among Java architects, too. We were shocked with our productivity, and we moved the project to the new foundation.
Something clicked into place for me. For this kind of application, Java itself was in the way. Remove it from the equation, and I could reduce the amount of code by a factor of four, drive the XML down to one-tenth of what it was, and achieve stunning productivity, with good performance. Better still, the concepts in Better, Faster, Lighter Java still applied. For other projects, if I needed the community and tools that Java offered, I could use it instead. If I didn't need Java, I could take the principles in BFLJ to the extreme. A dam inside me broke, and this new book started pouring out. I had a message.
Months later, I found an audience, thousands of miles and 19 hours from home. I fidgeted nervously before the Java User's Group. I'd certainly addressed larger groups, but this trip was different. In this case, the Norwegian Java User's Group had paid my travel expenses to Oslo, not to sing the praises of Spring, or Hibernate, or agile development, but to call their baby ugly. It was hard for me. After writing the bestsellers, getting the Jolt, and building a thriving consulting practice in a down economy, I wanted the Java train to roll on, unstoppable, always building on an ever-strengthening foundation. I wanted Java to send my productivity through the roof, and for the impressive community and brain power to solve all the tough problems that Java faces today, but nothing lasts forever.
In the talk, I didn't pick an eventual winner. I laid out the reasons for Java's success, and then talked through its most serious problems. I showed some alternative languages and frameworks, as I saw them. Throughout the talk, I pointed out that conditions are ripe for an alternative to emerge. As I addressed the hospitable group, I answered questions and read faces. A few looked hostile, or hurt. Most others showed understanding, and a little fear. They understood my central thrust. For many of the most common problems that we solve with Java, some other frameworks in other languages can already do a better job. In some cases, the productivity discrepancy is wide enough to merit a serious look.
The talk, and the questions, went on way too long, but nobody left. They were surprisingly receptive. After the presentation, we went out to see some of Oslo. One of the hostile attendees cornered me for most of the night. The hard questions just wouldn't quit coming:
Why can't we improve Java to cover the shortcomings?
Do the other frameworks and languages that you presented have enough commercial backing?
What about distributed transactions, or web services, or XML support?
How can you find programmers, or train the ones you find?
These questions are real, and they show the tremendous barriers of entry against emerging languages. My questioner was a gentleman, but he could not completely