Ten years after we officially launched Java in May 1995, our dream of a ubiquitous software platform to power a networked world has actually come true. Today, some form of Java runs on 1.4 billion devices, and there are more than 4.5 million Java developers worldwide. Mobile applications like Java-based digital wallets generated more than $1.4 billion for the almost 100 mobile carriers who use Java worldwide in 2003.

Java was, quite simply, the right technology at the right time.

In the early 1990s, Java's architects at Sun anticipated a world in which a ubiquitous public network would connect devices of all kinds and let people collaborate on an unprecedented scale. That was only a few years before the Mosaic browser was released and the Web was born.

When we planned the launch of Java (then called "Oak"), our goal was ubiquity. We knew we had a powerful technology, but our challenge was finding the right platform on which to launch it. Initial forays in the nascent PDA and interactive TV markets proved premature, but our persistence paid off when we downloaded an early version of the Mosaic browser.

We realized that the World Wide Web was the ideal platform to launch Java. Exciting as it was, Mosaic displayed only static text and images. What was missing was interactivity: the ability to run a program in the Web page, see animations, and get a real-time response. So the team developed HotJava, the world's first interactive browser, so people could see animations, live stock quotes, sports scores, and other data come alive on the Web.

From the day we released the newly renamed Java - along with HotJava, the full spec, and the source code - developers embraced it.

Technically, Java broke through platform barriers. It freed developers from proprietary hardware, and let them write applications once for many different operating systems. It was flexible. As a language, Java was designed to be small enough to run even on low-powered mobile devices, but complete enough to support complex applications. And by using a virtual machine, Java could address security problems that had foiled previous attempts to create portable code.

In assessing Java's business potential, Sun's top executives realized the potential to encourage widespread adoption of this powerful technology through free distribution combined with innovative licensing terms. With the support of Eric Schmidt and Bill Joy, we put the full spec and source code for Java online, while Sun retained the licensing rights.

We were convinced that freely distributing the system to individual developers was the only viable path to ubiquity. Java was made freely available for download, which spurred thousands of software developers to build "applets," fueling Java's growth and adoption by showing off the potential of the Web.

When thousands of companies, from start-ups to major telcos and consumer electronics manufacturers, adopted Java to deploy new network-based services, its success was ensured.

Today, innovation in software is coming from another powerful phenomenon: open source development.

Java benefited greatly from shared learning and the collaborative development of hundreds of thousands of software developers. As an early stepping-stone in the new era of software design, Java showed what global, dynamic collaboration between individual developers could do. The current open source phenomenon shows the success of that approach: speedier deployments, dramatic cost savings, and often more reliable software systems.

The parallels are clear. In fact, Java's success derives from principles that are central to the growth of open source software:

First, the key to ubiquity is to make a technology freely available. Profits come from elsewhere: the value-added around the technology. Companies like MySQL and Red Hat have validated this model.

Second, technologies that allow greater independence from proprietary standards win. Java was an important step in liberating developers from proprietary hardware. Now, open source technologies are freeing enterprise IT from dependence on proprietary software.

Finally, developers are a technology's strength. The best thing you can do is to provide developers with useful tools and access to underlying code - and get out of the way.

When it was launched, Java empowered software developers to innovate and create a new vision for the Web. Now that open source development has become mainstream, a new period of software innovation has arrived, where the best technologies (not just the best-marketed ones) can actually win. And there's no going back.

Ten years later, and looking at the decade ahead, the future of software looks bright indeed.