Prophet of Oracle’s evolving future

Sitting in a sparse Japanese pavilion beside a private lake hidden among the wooded hills of northern California, the software industry's second-richest man seems fully in command of his domain.

This is Larry Ellison's personal Xanadu. Designed, he says, to resemble a Japanese village with structures from different parts of the country's history, it has taken five years to plan and 10 to build.

"There's a primal connection to running water, and cut, unfinished wood," says Mr Ellison, as he tries to explain why, since he was a young man, he has felt most at home in Japanese gardens.

Watching shrouds of cherry blossom blow from trees across the lake, he seems momentarily entranced. "It's a wonderful counterpoint to the rest of my life, which is not calm. I'm very intense." With a laugh, he recalls the ironic nickname given to him by one of the first programmers at Oracle, the database software company he founded in 1977: "Captain Mellow."

Yet, at 62, Mr Ellison admits that lately he has been slowing down. "My work week is down to 50 hours - I work less," he says. Safra Catz, a former investment banker who has become Mr Ellison's second-in-command, has taken on most of the day-to-day tasks at Oracle, leaving Mr Ellison freer to concentrate on engineering, which he says has always been his main focus.

Among other things, it has also left him a little more time to think. Recently, he says, he has been dwelling a lot on the potential - and the threat - of open-source software. He recites the questions he has been addressing: "Is it disruptive, is it a risk, how do we respond to it, can it hurt us, how can we take advantage of it?"

This sounds very different from the Larry Ellison of four years ago, when the technology industry was still in free-fall after the boom of the 1990s. In an interview then with the Financial Times, he argued that the early era of innovation in technology was over and that making software was becoming as boring, and mature, as any traditional metal-bashing industry. He also predicted a wave of mergers, as software companies came to realise that the go-go days were over.

Within months, Oracle had embarked on a round of acquisitions that shook the software world. They included a bruising hostile bid for PeopleSoft, which led Mr Ellison to fight and win a court battle against the Department of Justice's trust-busters, and the purchase of Siebel Systems.

Innovation in the software world, however, has not come to an end. The rise of the Linux operating system and other open-source software is just one of the new forces that threaten to shake software giants such as Oracle. Others include "software as a service" - the practice of running software on behalf of a customer and selling access to it under a monthly subscription fee.

Meanwhile, Silicon Valley is also buzzing again over a new style of lightweight software development, dubbed "Web 2.0", that lets small groups of developers conjure up the sort of products that were once within the reach only of big software companies.

Faced with this new upheaval in an industry that dates back barely three decades, Mr Ellison takes it in his stride. The demands of the big corporate systems built by companies such as Oracle will still force more mergers, he says. "To tackle the big jobs, consolidation will continue."

In fact, Mr Ellison has shown himself adept at anticipating some of these new forces. He is the largest shareholder in Netsuite, one of the biggest software-as-a-service companies, which is expected to seek a Wall Street listing later this year. He was also an early backer of another, Salesforce.com, that was set up by Mark Benioff, a flamboyant former Oracle executive.

Mention of Mr Benioff's success draws a sideswipe from Mr Ellison, who becomes characteristically abrasive in the face of former lieutenants who turn into rivals, including the former heads of PeopleSoft and Siebel Systems. "It wasn't his idea," he says tartly. "I would argue I started the first big software-as-a-service company."

Mr Ellison's answer to the new disruptive forces in software - embrace them - has been most apparent in his recent conversion to open source. By allowing anyone to use and adapt their code free of charge, and generally charging only for maintenance, open-source companies benefit from a low-cost approach to developing and distributing software that threatens to disrupt established software empires.

"We don't have to fight open source, we have to exploit open source," says Mr Ellison.

Last week, it looked as though his plans had suffered a setback. Oracle had been trying to buy JBoss, one of the most prominent of the new breed of open source companies: instead, it was bought by the leading Linux company Red Hat.

The disagreement, says Mr Ellison, was all about price. "JBoss wanted to sell the company to us. Clearly, if we wanted to buy JBoss, we’d have bought JBoss."

Without control of their own intellectual property, open-source companies are vulnerable to seeing their products simply absorbed by companies such as Oracle, says Mr Ellison. "The reason I have a hard time writing cheques for billions or hundreds of millions of dollars for things that are open source is that if we could do this, other people could do this too."

Those considerations seem also to have dissuaded Oracle from trying to buy one of the big Linux companies, at least for now. "I don't see how we could possibly buy Red Hat . . . I'm not going to spend $5bn, or $6bn, for something that can just be so completely wiped off the map," he says. Oracle looked at buying Novell, owner of SuSe, Europe's biggest Linux company, but would have faced the same risk, he adds.

Along with open source, Mr Ellison says he has been pushing Oracle further into software as a service by hosting software on behalf of customers. Indeed, mention of the new business model around these types of software, which rely on regular subscription payments rather than the upfront revenue associated with traditional software sales, touches a nerve: Mr Ellison has complained for some time that Wall Street has failed to give his company credit as it shifted towards a greater reliance on regular maintenance payments.

"On the one hand, people say open source and software as a service are really hot - on the other hand, all they look at is our new licence sales," he says. "It's the kind of absurdity you find in the world sometimes."

While some of the forces that have hit the industry may have been far more disruptive than Mr Ellison himself seemed to expect four years ago, other trends look like a return to software's roots. Oracle, like its arch-rival SAP, has been intent on building a complete "stack" of software, ranging from the database software on which its business was founded to the "middleware" that acts as the glue holding internet-based systems together and the applications that run on top.

Tying all the parts of the software stack together makes information systems more manageable, according to Mr Ellison. "It's all synchronised - that's the Microsoft approach, and that was historically the IBM approach for a very long time," he says.

The prospect of software once more coalescing into monolithic blocks sold by single companies is likely to leave customers feeling decidedly ambivalent. But buying from a single supplier would make life easier.

"Most of our big customers would rather have one phone number to call - the classic 'one throat to choke'," says Mr Ellison. If Oracle also sold a Linux system, "from the operating system up to the application, we're completely responsible. We test everything together, have one set of management tools."

Yet stacks of software like this make it harder for companies to choose between components from different suppliers, potentially reducing competition.

"Customers want integration, and they want choice. To some extent those things are at odds," says Mr Ellison. Oracle will be able to ease the concerns, he predicts, "as long as we do a good job and we don't price-gouge them. That would be self-defeating - we make plenty of money now."

Gazing out at the peaceful lake, with the clean smell of Douglas fir and cedar seeping from the rafters and walls of the Japanese pavilion, the scale of the riches at stake are readily apparent.

In spite of his air of perfect calm, though, Mr Ellison seems far too restless to think of retirement yet. "I love my job - I don't need the money," he says.