Moronic 'Keynote' Turns Slightly Nasty
Bad planning by Sun was eclipsed only by the bad manners of Ellison in the blessedly final keynote address of JavaOne
By Lori Piquet, Editor-in-chief
San Francisco (June 7, 2001)
—I have found myself wondering several times this week how Sun Microsystems—which has done a darned nice job planning the "little things" at JavaOne (extracurricular entertainment, most notably)—has managed to make a complete mockery of the venerable institution of the keynote address.
Perhaps Sun was hoping that an embarrassing and useless program today, featuring the chiefs of BEA and Oracle, would purge Monday's deadly dull session from attendees' memories.
Today's session was unabashed propaganda, with a smattering of unapologetic potshots from Oracle Chairman and CEO Larry Ellison.
More on that in a moment.
Keynote for Sale
We all know that those who spend the biggest money on show sponsorships tend to get very special dispensation when it comes to placement and mindshare with attendees. Platinum sponsorships come with the opportunity to put one's CEO in front of the audience for an hour, which seems reasonable. But my beef with Sun is that they have seemingly thrown out all notion of holding keynote speakers to standards of intelligence and technical relevance. Didn't it occur to anyone to ask the CEOs to address a given topic, submit an outline, refrain from hype, and restrain the marketers—or even address the interests of the audience, for that matter?
This morning began with Sun Executive Vice President for Software Pat Sueltz speaking about the Java Community Process and very nearly begging all in attendance to join it. The JCP is a worthwhile endeavor for any Java developer and I've seen much this week to further convince me of that, but Sueltz was painful to watch at times: stilted, clinging to the monitor prompts, and gesturing incongruously.
Sueltz introduced BEA Chairman and CEO Bill Coleman, who led off his presentation with a pointless, ridiculous video, whose script should be set afire—while strapped to the back of the person who wrote it. Coleman took a few minutes to speak on the recent progress of the J2EE platform in removing the biggest roadblocks to Java's success in the enterprise. Like nearly every other luminary at JavaOne this week, he affirmed the idea that Web services are the key to the future of e-business and Web development, in particular.
Then he launched into a stream of WebLogic propaganda, including nearly a dozen slides that extolled the WebLogic architecture. Coleman stressed BEA's adeptness at application integration, thanks to the announcement this week of its WebLogic Integration Server version 2.0, which provides WebLogic customers with integration adapters for major enterprise applications such as EAI without loads of custom programming or installation of proprietary adapters.
Larry Turns Up the Thermostat
Here's where things got interesting. Coleman was followed by Larry Ellison, who was unabashedly confrontational, self-serving, and competitive. In other words, it was Ellison at his notorious best.
Ellison took full advantage of his clean-up position by using the entire portion of his allotted time—and then some—to tell developers about the performance and scalability enhancements that Oracle 9iAS provides over BEA WebLogic and IBM WebSphere.
Twice Ellison described Oracle as having "leapfrogged" BEA in the marketplace, saying first that the two-week-old update release of 9iAS was the best J2EE implementation available and then demonstrated this claim with several slides that compared Oracle very favorably in Web page performance tests with BEA and IBM, particularly at higher numbers of connected users.
Though Ellison was equally dismissive of both BEA and IBM, there was clearly an air of discomfort surrounding this presentation, as it came on the heels of Coleman's own. Twice Ellison told those in attendance that he had been asked not to make the remarks he had prepared; they had been deemed impolite.
"It's considered impolite to give the facts about Java performance?" Ellison asked rhetorically. Ellison argued that Java is an open standard and that an open standard should include open communication about performance.
"You want Oracle, BEA, IBM, Sun, and others to run faster to make Java better," he said.
Oracle directly challenged BEA in a race to be the first vendor to implement the J2EE 1.3 standard.
Ellison's comments clearly were unnerving, even to him. Ellison declined to attend the post-keynote press conference, at which Coleman told reporters that he believed Ellison's data was made up and that he did not believe that the newly updated version of 9iAS was capable of the performance gains Ellison had claimed. Coleman openly suspected that Oracle's new version was the result of a deal struck with a small coding company called Orion.
Akamai CEO George Conrades, who was invited onstage by Ellison to discuss the caching technology that Akamai has provided to help speed Web page downloads for Oracle shops, made it a point to give Coleman—another Akamai client—a sympathetic wave.
Just a hunch, but I'm betting that Coleman may insist next year that he not be followed by Ellison.
If I had my way, they'd both be under gag orders.
In Other News...
Last night's Java Technology Achievement Awards, sponsored by Java Pro magazine, were notable for two reasons: Borland and IBM. The two companies swept the majority of categories, garnering 10 awards between them. The big difference: a large and boisterous contingent from Borland was on hand; IBM was absent. Borland took the overall award for Most Valuable Product for JBuilder 4.
And BEA took the award for Best Application Server. But perhaps Ellison hadn't heard.