Oracle was the first top-tier IT vendor to announce it was putting its key product - the database - on Linux. The logic was simple: Linux freed Oracle from depending on a single company for operating system - that company was Microsoft.
Taking the baton from Sun Microsystems' co-founder and chairman Scott McNealy at JavaOne this week, Oracle's chief executive Larry Ellison has seen his opportunity for independence again.
This time, however, he may struggle to get his way, and - in trying - actually hurt one of Sun's most prized and widely adopted open-source projects.
Ever the showman, Ellison threw a curve ball at JavaOne: he revealed he's been secretly meeting Sun's product groups and has decided that he likes JavaFX as an interface architecture. Ellison, ahem, "encouraged" the OpenOffice group to quickly build libraries for the C++-based suite using JavaFX.
"We encourage the OpenOffice group to quickly build their version of a spread sheet or a word app using JavaFX," Ellison said.
Such an early declaration is rare in the history of Oracle's M&A, as the company never plays its hand on product strategy until after a deal's done. During the acquisition of BEA Systems, for example, Oracle would not say whether BEA's WebLogic would become its flagship app server at the expense of its native product until product due diligence had been done by the experts in Oracle's server and middleware product team.
This time, Ellison's either done his own due diligence or he's preempted the product experts. Timing is everything, though, and regardless of JavaFX's technical merits, so is politics.
Ellison's declaration for JavaFX comes as the OpenOffice project assesses what it wants the next version of the suite to literally look like. Project Renaissance has taken feedback on user-interface design proposals, with results expected to be presented.
In 2005, Sun's outgoing chief executive Jonathan Schwartz expected AJAX to play a part in the future of OpenOffice. Ellison clearly has different ideas. Embracing JavaFX, Ellison sympathized with Java programmers that must adapt to AJAX. It's difficult to see Ellison, a salesman first and a database guy second, being able to empathize genuinely with the toiling Java masses.
Likely, Ellison's backing for JavaFX has two reasons. First, it potentially lets users of Oracle's Java middleware re-use their existing Java skills and technologies - that's the pitch Sun's been giving publicly and is likely to have given Ellison behind closed doors. It seems he took the bait.
More importantly, JavaFX is a Sun-owned technology, which - if the acquisition goes through - makes JavaFX an Oracle-owned technology.
Unlike the rest of the Java, JavaFX has not been submitted to the Java Community Process (JCP). Sun has never explained why, it's just dodged the subject saying it still believes in the JCP, which is like saying you believe communism is a good idea but that it's just not for you.
That fact JavaFX will become an Oracle technology confers both independence on Oracle and furthers the over-riding goal of devising the top-to-bottom, Oracle-software-stack approach that it's building to the interface. Until now, Oracle's interface technology of choice has been JavaServer Faces, which is in the JCP.
However, there's a big challenge to Ellison's wholesale declaration for JavaFX and the fact it's a Sun- and Oracle-only technology.
Sun might be the principal backer of OpenOffice, but there are major contributors and adopters who will likely disagree for political and technical reasons with the task of re-writing OpenOffice in JavaFX. Java's long been a pissing contest with IBM, Sun, Oracle and BEA trying to rest some form of control over the platform, or subtly lock in their users via features.
Just because BEA and Sun are gone or going doesn't mean this contest will now stop. IBM, Novell, Red Hat and Google are all major contributors and none has expressed an interest in JavaFX. IBM and Google have, infact, been prime movers and supporters of AJAX. You should expect them to resist moving OpenOffice to JavaFX, a technology that's unproven, owned by Oracle, considered inferior by some experts and that would - as a result - take OpenOffice right outside of the developer mainstream.
What happens next will be critically important. If Ellison gets his way, then the suite will get re-written in JavaFX. That will not only hurt development but set back a suite that's been slow to close the gap on Microsoft's Office. Such a move will also be seen by open-source supporters as early proof that Oracle is putting its own corporate goals ahead of the community's when it comes to running Sun's open-source projects. That'll further spook people alredy concerned about the future of MySQL.
Of course, Ellison might win skeptics by putting JavaFX through the JCP. It's hard to see, though, how companies like IBM and Google will view JavaFX as anything other than a Sun and Oracle vehicle, in the way NetBeans - which is needed to build JavaFX - is a Sun-only vehicle.
Executives from Sun told press at JavaOne they could not comment on what will happen to Sun's open-source projects like GlassFish under Oracle because Oracle had not told them what it's planning. If there is a product group breathing slightly easy inside Sun, then it's presently the JavaFX. How long they continue to breath easy will depend on just how far Ellison can convince developers and partners to support it, and how long he remains convinced of its worth. ®