JavaOne was ahead of the curve in more ways than one this year. Last week's developer conference was forced out of its usual June slot in San Francisco's Moscone Center, throwing the timing of vendor upgrades out of whack. Attendance was down, and the show floor's aisles were wider to conceal the drop in vendor booths.
Sun took advantage of the confusion to delay its launch of J2EE (Java 2 Enterprise Edition) release 1.4 until January of next year. Web services may have to wait until then to be fully integrated into the J2EE platform, but Sun will let others carry the ball forward for now. BEA's WebLogic Workshop (formerly code-named Cajun) moved from an unstable alpha premiere at the InfoWorld NextGen Web Services conference in January to a public beta a month later at BEA's eWorld.
Sun showed unusual restraint in its messaging, from a cordial debate over development tools between IBM, Forte, Borland, and BEA to Sun Chairman and CEO Scott McNealy's keynote. The tools panel was moderated by Sun Chief Technology Evangelist Simon Phipps, who in recent months has been sharply critical of IBM's Eclipse open-source initiative as a redundant competitor to Sun's NetBeans program.
But honoring his role as moderator, Phipps let IBM Websphere market manager Adrian Mitu off with a muted, "Does the world really need two open-source Java tools initiatives?" And Forte Tools Vice President Peter Young dropped hints of détente between the two open-source communities. McNealy's keynote was long on marketing puns -- Java Won, Java One of Two - and short on attacks on the Evil Empire to the north.
Of course, there were the obligatory references to Bill Gates & Company's lack of education. Scott reminded the crowd that he, unlike Gates, graduated from Harvard, and suggested that might be why Bill missed the class on sharing in the second semester of the third year. Then there was the dig about not having to send out a companywide e-mail on trustworthy computing.
But McNealy was self-deprecating, both in his speech and later at a press conference where he passed up the stage to roam, à la Jerry Springer, through the audience. Asked to justify how Java won, he uncharacteristically demurred, "It's just a play on words. What can I talk about? Security models in Java? I'm a golf major."
McNealy got even quieter when asked whether he had any evidence to support his keynote assertion that Microsoft was hijacking XML. You could see he wanted to leap all over this, but the best he could muster after what seemed like a two-minute pause was, "That's a fair question, and we'll get back to you on what we think they're doing."
Indeed it was a fair question, I pointed out, because he brought it up by referring to XML with a question mark in his presentation on open standards. Like a shrink sensing a patient on the verge of a breakthrough, I asked Scott why he was persisting in marginalizing XML's role in Sun's architecture.
The dam burst. "We don't. We don't. In fact, what I said was it's incredibly important. We've introduced a whole bunch of XML technologies. Heck, [Jon] Bosak as a Sun engineer started the whole XML maneuver, and we've been very much with it, for it." For a golf major, McNealy displayed an intimate grasp of Sun's XML strategy and history, briefly passing the baton to Java and XML chief Rich Green before interrupting again.
"Excuse me for being skeptical of Microsoft's intentions," McNealy said. "But they were contractually obligated to stay compatible and interoperable with Java, and they didn't. What contract do they [have]... with what organization -- that can go to the court system and bring justice to their efforts to take proprietary XML extensions and deliver them bundled with their Windows monopoly around the world?"
Fair question followed fair question. "Who's going to do what we did with Java to ensure write once read anywhere? Who's going to do that for XML?" McNealy asked. The thoughts tumbled out: "Given that it's a standards body, kind of -- which is why I put a big huge question. ... This would really be a problem for Sun if Microsoft hijacked XML."
Finally he got back to answering the original question. "So do we have evidence? Just absolute, undeniable, unstoppable, inevitable, absolutely consistent, never wavering behavior from Microsoft from the time they got started. But if you want more detail you'll have to go check in with them and see what they're doing and try to make sure that XML and XML tags only work with Microsoft. That's always been their strategy."
As much as we tend to be cynical about the problems of the seriously rich, I can't shake the notion that McNealy's sense of frustration is sincere. Sun feels shut out of the standards body merry-go-round. IBM and Microsoft grab the remaining chairs when the music stops on the Web Services Interoperability (WS-I) carousel, and IBM conveniently forgets to let Sun know about meetings of Web Services Interactive Applications (WSIA) group.
McNealy sees the Java Community Process as an acceptable compromise between the opposing requirements of rapid time to market and the slow deliberate blocking and tackling of the standards bodies. Microsoft's ownership of the client (desktop and now browser) and the pressure of .Net on the developer community have forced Sun to rely on the courts to even the playing field. That brings up another fair question: I know you can't fight City Hall, but can you rely on it?
Besides, while McNealy goes after Redmond and the dark side, Simon Phipps seems more worried about IBM. Maybe that's because he came to Sun after a stint as Big Blue's XML evangelist and knows what he's up against. That may be why the shy Bosak, who prefers the collegiate work of the standards bodies to smoke-filled rooms, was willing to sit down with us for an insider's look at the XML standards game. Stay tuned for more.