SAN FRANCISCO -- Here's the difference between Silicon Valley and Wall Street: The Valley is invested in the next big thing, but the Street invests when the thing is big.
That's it. That's all there is to it.
But here's the problem: Somewhere in that gap, a lot of investors -- indeed, a lot of Wall Streeters -- get confused. They start to buy the hype. They start to think the next big thing is close enough to big. Some companies will even tell the Street that the next big thing has already arrived. That kind of premature speculation has led to some stock disasters, and there is no better example of that than
, the well-hyped programming language created by
Java has received an extraordinary amount of hype, and, indeed, it has probably been adopted faster than any other programming language in computer history. Ken Berger, a veteran programmer and president of
Interactive Web Concepts
, a privately held Internet design firm, still gets giddy when he talks of how Java swept the Valley. "I think we were all blown away by Java because it was such a simpler way of doing object-oriented programming than C++," he says. "It was so much easier to write with; you didn't have to worry about memory management or pointers -- pointers were a pain in the !#@$% under C++. And the idea that you could 'write once and run anywhere' seemed like a dream."
The bubble hasn't burst yet on that dream -- just this week,
announced a whole host of new Java projects. But here's the dirty secret: the handful of publicly traded companies which bought into the Java dream have seen their product schedules and share prices blow up.
have all blown quarters by missing crucial product ship dates, each citing problems incorporating Java. Indeed, the success of Java isn't based on the success of the network computer (as today's
Wall Street Journal
would have you believe). Just getting Java working in client-based software is proving to be enough of a struggle.
"Java hasn't really happened," says Greg Vogel a
software analyst. "Java is a good technology in theory, but the performance isn't there yet." Vogel points to the experience of
, the Toronto-based software firm that essentially bet the farm on Java. "Corel had a very ambitious plan to put out
and all these other applications on Java. But they ended up aborting that whole plan because Java was just too slow."
Oracle's blowup in the December quarter might even be a greater example. At the time, most of the press blamed the company's earnings shortfall on the Asian financial crisis -- a pithy three-word cliche the mainstream press could understand. But, actually, the company said it was hit by a double-whammy -- Asia
"Java was simply harder to incorporate than we expected," says CEO
. "We thought it would be a relatively quick process, but we found that we had to do a lot more tweaking of the code than we'd planned for. We could sit here and say, 'Gee this wouldn't have happened if Sun only distributed a better version of Java.' But, really, it's our fault. We screwed up. The reason we didn't deliver on the quarter was our screw-up and Asia. We can mask a serious problem, but we can't mask two serious problems."
Of course, Ellison's admission, which came earlier this year, came about only because most of the issues are history, and, of course, a new-and-improved quarter is in the books.
But similarly serious problems have slowed the rollouts of @Home's newest set-top boxes and Netscape's latest version of
. So what's next? Vogel says that Java success continues to elude client applications. " I'm very skeptical of any company that wants to create Java-based desktop apps. But for the server, Java makes more sense."
There is, however, one place where Java is working, according to Vogel. That's on the server side, where Java is running back-office applications with ease. "Java is a great language for talking to legacy systems. Private companies like
(recently acquired by Netscape), they all run Java on the server that the user never interacts with." Vogel expects a handful of these companies to go public in the fall of 1998 or early 1999.
But, until then, it seems like big Java announcements are a chance for savvy investors to take a break -- and let the fools rush in.