Sun promotes Java aggressively, arguing it will allow programmers to write software that runs on any computer as well as countless other electronics devices. But how exactly does Java benefit the Palo Alto, California, computing powerhouse?
Put simply, the proliferation of Java and its cousin Jini is intended to fuel spending on Sun hardware and services, executives and analysts say.
Java is the "WD-40" that will let people plug billions of devices into a global network, and "Sun has terrific products and services for building that world," said Jon Kannegaard, general manager of the Java platform at Sun.
"That's probably the No. 1 way Sun makes money on Java," he said in an interview.
Jini, which taps Java to make printers, cameras, light switches, and other devices available on the Internet, will accelerate Java's proliferation, added Jim Mitchell, vice president of architecture and technology for Java at Sun.
"Jini will help spread Java in the world of small devices," he said. "It acts primarily as an accelerator."
To date, however, neither scenario has come to pass.
Though popular on the Net, Java is otherwise something less than universal, and lies at the center of a licensing lawsuit Sun has brought against software giant Microsoft. Sun claims Microsoft has been trying to usurp control of Sun's technology by writing proprietary "extensions," or programming commands, that work only within a Microsoft environment. True or not, the specter is hardly promising for Sun's revenue hopes.
But Sun hopes Java will lead customers to other Sun products, said Sally Cusack, an analyst at International Data Corporation.
"They haven't made a lot of money on [licensing] Java as a language," she said. "Most of their revenue is in hardware. Obviously, Sun's real hope is to sell a lot of Solaris," Sun's version of the Unix operating system, which runs on systems based on Sun or Intel processors.
Selling Java products
Besides licensing Java for use in commercial products--which Kannegaard concedes is not "big money"--there also are less direct ways of making money off Java.
Like its competitors, Sun sells products that use Java technology. In fact, in October, Sun's acquisition of NetDynamics gave the company a Java-based "application server," a high-end software product that businesses use to connect Web-based client computers to "back-end" systems such as databases.
One potential pitfall in selling NetDynamics products is that it puts Sun in competition with rivals who fear Sun may favor its own Java products. But so far that concern hasn't been realized, according to Steve Garone of IDC.
Sun says its Java division maintains a level playing field by preventing other divisions from getting sneak previews of new Java code. Instead, the company says, Java's appeal will produce a rising tide of computing products and services that lifts Sun and everyone else.
Sun's NetDynamics acquisition will likely pay off, Garone said. "We don't think that Sun, until recently, had a strong ability to make money off of Java," he noted.
Jini: The next cash cow?
And then there's Jini, which could produce royalty revenue for Sun--if it catches on.
Jini provides a way for Java-enabled devices to "announce" themselves on a network without being connected through a computer. Mitchell describes it as a thin layer added on top of Java that allows "universal plugability."
Where Java is finding use in large corporations, Jini "probably will reach its first market acceptance in the consumer space," such as within set-top boxes or other electronic devices, Mitchell predicted.
Sun hopes there will be billions and billions of Jini-enabled devices, each one leading to a royalty payment to Sun. Some companies, such as BizTone, believe Jini will show up in big business as well as a way to offload corporate computing tasks to outside companies.
But the more solid business proposition is that Jini devices will mean more usage of Java, and therefore more need for servers, which Sun happens to sell, Mitchell said.
The ill-fated Java chip
But sales of another Java product, the Java chip, haven't gone that well.
The chip, named picoJava, was Sun's initial effort to make a microprocessor that could run Java programs on a "native" basis (that is, without the need for a Java virtual machine, or JVM, to translate computer instructions). picoJava was supposed to speed Java's slow performance and make it possible to run Java programs on small equipment that doesn't have enough space for the more elaborate hardware Java sometimes needs.
Just as Sun brought the first chips to light, however, the company announced it wouldn't be manufacturing its own Java chips. Instead, Sun refocused the picoJava effort into a licensing arrangement wherein Sun helps other chipmakers incorporate Java chip technology into their own products, said Fadi Azhari, group marketing manager for picoJava.
Under the new model, Sun will reap rewards from both up-front fees and royalties Sun gets when licensees sell products using the Java chip technology, Azhari said. At present, several companies have licensed the technology, including IBM, Fujitsu, Siemens, and NEC.
The first products are expected to emerge in the second half of 1999, Azhari said.
But Jim Turley of MicroDesign Resources isn't so bullish on the Java chip's prospects, noting that licensees have been close-mouthed about what their finished products will be. And since running Java programs directly on a chip is awkward, he said, designing chips for the task will be doubly hard.
Sun's Java chip was delayed by about six months, Azhari acknowledged, but the Java chip effort is alive and well. "We didn't drop the Java chip program. We looked where we could focus the resources and know-how.
"They [the licensees] get to make chips, and we're not competing with them. It just made a whole lot more sense," he added.
Regardless the revenue Sun earns from licensing Java chips, it fits into the overall Java revenue plan, which hinges on spreading Java everywhere. "Sun wants to propagate Java technology as far as it can," IDC's Cusack said.