Sun's newest company scrambles to hire staff, switch locations, release products, cut deals, create a business plan
JavaSoft, Sun Microsystem’s latest addition to its growing constellation of operating companies, recently moved to new offices in Cupertino, CA. The new digs are located directly across the street from Apple Computer’s R&D complex. And it must be noted that what would otherwise seem a purely mundane space-and-economics-driven change of locus was made all the more intriguing by the choice of address: JavaSoft has taken up residence in the building on De Anza Boulevard that once housed the main offices of the short-lived Taligent Inc.
For those whose memories are only as long as the lifespan of some recent Solaris releases, Taligent Inc. was for a very brief time a corporate entity that had the dubious distinction of being — simultaneously — a subsidiary of Apple Computer and IBM. The mission of this unholy alliance was to build an object-oriented operating system that would function equally well on IBM DOS-compatible and Apple/Macintosh operating systems. The miraculous new OS would be networked, multi-platform, and international. But as often happens to the best laid plans and most noble of intentions… well, suffice to say the corporation evaporated, the building was left vacant, and is now occupied by Sun’s newest corporate venture.
JavaSoft: Mission and metaculture
JavaSoft’s stated mission since its inception has been “…to develop, market, and support the Java technology and products based on it … JavaSoft develops applications, tools and systems platforms to further enhance Java as the programming standard for complex networks such as the Internet and corporate intranets.” When it comes to an actual business plan, however, JavaSoft has yet to demonstrate how giving away the golden eggs of Java source code will enable it to become a profitable operating unit.
Jon Kannegaard, JavaSoft’s chief operating officer, concurred that it is hard to imagine how a business might thrive by giving away the product.
“Amazing, isn’t it? But profitability in the short term isn’t what we’re after,” Kannegaard said. “What we’re after is ubiquity — what we want to have happen is to get an entirely new set of players competing above and below the line. The question of what the new applications are and who will develop them still remains.” Kannegaard was quick to point out that Sun is not now and does not intend to give Java away to commercial users to incorporate in their products. The published rate for licensing Java source code for commercial use is a 25,000 up-front fee plus per copy. This amount, however, is still considered below Sun’s cost, which means the company will lose money in the licensing business for the foreseeable future. But that’s not what the corporate strategy is about.
For the immediate future, JavaSoft plans to get the Java platform — the core components that enable software developers to build, compile and test Java applications, including the Java Applet Viewer; the Java Compiler; a prototype debugger; the Java Virtual Machine (JVM); and class libraries for graphics, audio, animation and networking — licensed to as many big companies for inclusion in their products as possible.
“Our focus after getting the platform built and licensed is to widen the scope of class libraries and build a better API. Then once that happens, Java programmers can start to do something really interesting,” Kannegaard said. “We have the best-of-breed players signing up to help. Companies like Macromedia and SGI are working on multimedia toolkits. And there are many more coming down the pike…
“Right now, JavaSoft is like a volunteer fire department. We’ll take anybody inside Sun or outside who wants to be involved. There are maybe 300 people with Java tee shirts. Of those, about 100 are on the payroll.”
When asked to comment about the departure of three members of the original Java team (Arthur Van Hoff and Sami Shaio, software engineers, and Kim Polese, Java’s senior product manager), Kannegaard said: “There’s a misconception that the Java team left to build Java applications. Three people left; 97 stayed. James Gosling [Java’s primogenitor, a Sun engineer and Fellow since Sun’s early days] is still here. Of course you always have mixed feelings when good people leave, but the fact is, they’ve gone off to become ISVs. There are certainly no people better qualified to develop Java applications. They’re talented. They may be the best Java programmers out there now. And all you can say is: go off and make Java applications and keep broadening the scope of its use.”
Giving away the razor so people will buy the blades
Industry analysts, while no less enthusiastic about the possibilities of the technology, recognize that the jury is still out when it comes to the viability of a Java business unit.
Donald DePalma, a senior analyst with the Software Strategy Service at Forrester Research, sees JavaSoft’s mission as necessarily falling into three main areas of focus. They must consolidate what Sun has done around Java and make sure that SunSoft and Java are at arm’s length; they must make sure that Java fever doesn’t abate and take advantage of all the free press that they can through marketing programs and consortia to keep Java’s presence as much at the forefront as it is now; and they must decrease Java’s association with Unix, making it as cross-platform as the architecture promises.
As for the technology itself, DePalma said: “Is it great? So far, it’s the only alternative to Microsoft’s vision of the world. It’s the best hope of everyone else … When you get away from the hype, you end up with a group who falls into the camp of `the enemy of my enemy is my friend…’ As a technology, it’s C++ stripped down, but it retains the benefits of the object paradigm. It makes use of distributed objects, simple inheritance, assembling components. It runs on both server and client, and has the ability to create a small footprint of interpreted code. Furthermore, it does just-in-time compiling … The problem is the phenomenal amount of hype. Dave Litwack, the president of Powersoft said they are working on an application called `Decaf’ — Java without the hype.”
An analyst at the Gartner Group who chose to remain anonymous commented that the purpose of JavaSoft is not and should not be to make money. “They are acting as a provider of Sun technology. The business won’t work if they think they’re going into a software venture where people make money from Java applications… That scenario doesn’t make sense for Sun. The purpose of JavaSoft is to build Java into an OS. How they’re going to make money isn’t clear. What is moving Java now is Netscape.”
This analyst is not convinced that starting a separate company was a wise business move, but at the same time he sees the extent to which Java was and continues to be tied to Sun as a problem.
“The whole appeal of Java is its platform independence. A Java-based OS is cool as long as everyone’s running it. What isn’t smart is to build an OS that everyone has to use in order to run Java applets. Then you’re back facing the same problem. Nothing is `platform independent’ if the platform has to be Java,” the Gartner Group analyst said.
As far as making money is concerned, the Gartner analyst envisions JavaSoft continuing to market developer’s tools, just-in-time compilers and so on. He saw the deal with Microsoft as “…a foregone conclusion.” Microsoft’s agreement with Sun to develop and maintain the reference implementation of Java for all Windows platforms (announced on March 12) was necessary for them as providers of technology in order to keep abreast of the tide. The point of pushing Java technology — for both Sun and Microsoft — is not that they will build applications and thereby create profit streams from Java products, but to change the technological playing field.
While Java has shown a great deal of promise, and certainly seems to be a technology that has arrived at the right moment in the technological space-time continuum, aside from a few notable and noteworthy business applications (see “Java in the Real World”, a story in JavaWorld magazine) its real merits as the basis of a much-talked about (but yet-to-be-realized) Internet OS, and the limits of its capabilities — particularly in the personal productivity area — have yet to be tested.
“In the past, with any new desktop technology, it wasn’t until a compelling personal productivity platform arrived that actually took advantage of the wire that anyone could tell its value,” said Kannegaard. “When you see it, like when CPM people first saw Visicalc, the first thing you think is `wow, we should have thought of that…’ The same has held true for every new platform. It took Lotus to prove the value of DOS, Excel to show the value of the Mac OS…”
For the home market, Kannegaard envisions Java as the first seed in the evolution of a group of applications that are not browsers. “There’s a lot you can do besides browse and click on hyperlinks. I see three categories coming: news; entertainment; and publishing media of various kinds. There will be a virtual newsstand where you can look at categories of magazines, for instance, and then choose the category you’re interested in — sports, for example — and then see the selections in that area.”
For businesses, Kannegaard sees Java as a medium that will initially improve the way in which internal applications are developed and employed. “Businesses have a great opportunity to take advantage of Web protocols — internally — using Java. Sales people can use an information retrieval system to get data from the corporate net directly. They won’t have to spend hours downloading everything to a laptop while they’re on the road.”
Kannegaard agrees with the assessment that the real merits of Java have yet to be proved. “Any great tool has to be abused,” Kannegaard said. “It has to be stretched to the boundaries of imagination, and used for things that the original designers never even thought of to become really remarkable.”
(Almost) controlled chaos: The corporation of the future
The corporate structure at JavaSoft as of this writing continues to be something of a mystery. Ruth Hennigar, general manager of Java products at JavaSoft, admitted that the corporate structure has not quite congealed. “We’re adding people as quickly as possible. We plan to have three basic units, though nothing has been officially announced: developer services, products, and a marketing organization. Alan [Baratz, JavaSoft’s new president] hasn’t done an org chart yet,” Hennigar said.
As for JavaSoft’s business plan, Hennigar envisions the company will continue to do more of what they have been doing in the months since the first alpha release of Java: license Java source code, work on the Java Virtual Machine, add to the class libraries, perfect the Java Developer’s Kit, add features to the HotJava browser, and make OEM licensing deals.
In addition, Hennigar sees the development of the Java API and Java developer services as key to the success of the business and seemed to think that there was money-making potential for JavaSoft in the area of application development. “We need to take advantage of the multi-platform API,” Hennigar said. “We need to look at the Microsoft model… They make money on the OS but make a whole lot more money on applications.” Hennigar felt that Microsoft would benefit from Windows implementations of Java immensely. “They [Microsoft] have access to the premier Windows experts,” Hennigar said. “We want to make sure the Windows implementation is the best it can be. All their changes go back to customers so that they get the best possible Java applications for Windows.”
Hennigar viewed the departure of Van Hoff, Shaio, and Polese as a booster for Java business. “The whole point is to have as many people as possible out there building Java applets. We want to press the envelope. If there are no good Java applications out there, no one will come to the party.”
Although there is no official road map for JavaSoft as yet (Baratz’s office has promised an announcement several times during the last ten weeks, and backpeddaled each time the date approached). Hennigar claims that there will be an “official rollout plan” eventually. At the moment, the members of the organization “…have their heads down working….”
The hundred-person staff is expected to grow over the next six months to 200. Maybe after that, there will be time to figure out what is really happening, and where JavaSoft expects to go from here. As of this writing, JavaSoft’s president, Alan Baratz, who came to Sun from Delphi, an on-line service provider, and before that worked for IBM, was unavailable to comment.
A kind of religious fervor — more than the profit motive — seems to be driving JavaSoft at the moment. Both Kannegaard and Hennigar claim not to be troubled by the coincidence of their new location. Neither believe in portents, haunted buildings, or bad vibes. Said Hennigar: “I was at Apple during the Taligent era and watched the whole thing being mismanaged and fall apart… We’ve got a completely different thing going on here….”
