The sunny climes of the antipodes attracted Oracle and Novell bosses Larry Ellison and Eric Schmidt to Australia and New Zealand over Christmas and the New Year. Ellison had the additional motive of sailing in the Sydney to Hobart yacht race. Schmidt was just taking a holiday. Randal Jackson caught up with both in Sydney.
What a difference a few days make. "Never in a thousand years," replied a shaken Oracle CEO Larry Ellison when asked after the tragic Sydney to Hobart offshore yacht race if he'd enter the event again. Six yachties lost their lives in the storm-ridden race, won by Ellison's New Zealand-built boat Sayonara, skippered by Kiwi Chris Dickson.
Just three days before the race began on Boxing Day, Ellison was waxing lyrical about yachting in general and Sayonara in particular at a press conference in Sydney. Indeed, he was showing considerably more passion about the forthcoming event than he was in fielding questions about Oracle.
"It's a great opportunity for me to get away from talking about the Internet and worrying about computing," he said. "I've never seen anyone quite like Chris Dickson — he's incredible. It's humbling.
"Humility is a good thing."
It's also a quality that is not often obvious in industry leaders like Ellison, who is one of the most outspoken.
He recalled to his Sydney audience, in a lead up to his now standard dig at Microsoft, that when he joined the industry 21 years ago you didn't compete against IBM; rather, you competed in the IBM environment.
"Five years ago IBM was in trouble because applications were taking so long to develop, and they were vulnerable to the PC.
"The centre of the industry now is not the mainframe but the PC.
"History is repeating itself. If you're a PC programmer in Silicon Valley now you can't get invited to a good party. It's not cool.
"The Internet is shoving the PC off centre stage." Internet computing, he says, costs one tenth that of PC computing, and is the future of computing. Internet computing is where every private network will mimic the Internet, and complexity will be centralised. The Internet model will change the world eco-nomy, he says. "We have gone down a dark road in computing. So severe is the problem — the complexity on the desktop — that the industry suffers not from an applications backlog but from a labour shortage.
"This year more than 50 cents of every dollar spent in the US will go on IT — and there's a labour shortage."
Ellison talks of irrationally distributed complexity.
"Our industry is so profoundly inefficient that we are consuming unreasonable amounts of capital and labour.
"The road ahead is clear. All computer networks will become intranets.
"We need to consolidate information and professionally manage it."
Ellison moved on the the Microsoft antitrust case, which he believes will dissolve what he calls the Microsoft monopoly regardless of what else happens. On Microsoft as a monopoly ...
"Think of Netscape as Little Nell and Microsoft tying her to the tracks and driving the train to kill her. The death of Netscape shows just how damaging the Microsoft assault has been.
"Microsoft is a monopoly. Oracle isn't, so there are different rules. Under US law you can't leverage your monopoly into another monopoly by giving things away.
"Microsoft can put Intel out of business whenever they feel like it. They ordered [Intel chairman] Andy Grove out of the software business."
Despite SQL Server, Ellison does not rate Microsoft as a database player. "If we took over 100% of their [database] business, we couldn't grow much at all."
A quarter of database sales are now Internet-enabled, Ellison says. He expects that by the year 2000 a move away from PCs to an Internet-centric computing model will become clear. "The Internet will be closer to the centre of the universe."
That said, the bubble of over-inflated US internet vendor stock prices is set to burst because the value is not there, he says.
Some Internet valuations are "inexplicable".
Microsoft is so strong that every company has to have a strategy to deal with it, says Novell CEO Eric Schmidt.
"We've worked hard to change our historical relationship with Microsoft. We don't want to go back to that tit for tat.
"The conventional wisdom is that monopolies dominate. But there's something wrong with conventional thinking. It's more likely the case that both [open source and monopoly] will work, one not exclusive of the other. They can co-exist for a very long time."
Novell, he says, was in turmoil when he joined last year as CEO after years at Sun as chief technology officer. That was mainly because of acquisitions that didn't work out.
"I put together a team, and the focus is now on network infrastructure products. Historically, we were the largest network solutions provider. We stumbled into the directory, and are now using that to solve people's management problems."
Schmidt says what he didn't know after joining Novell was that he would become a turnaround expert.
"When you're in a platform business, there's a tremendous amount of wealth created with your partners.
"I inherited a tremendous amount of cash — some say too much — and I needed to put that to work. The real reason for [my] investments is to accelerate the adoption of the platform."
Schmidt says he expects to make more investments.
"Our basic strategy is to try to work out how to work with partners and take advantage of the operating system."
He says Novell will bring out a "whole bunch" of directory products before Microsoft ships NT 5.0.
"Do you know when it will ship?" he asked journalists pointedly. "On assumption it ships, it will have an adoption curve not dissimilar to other Microsoft products."
Novell is already committed to fully inter-operate with Microsoft's Active Directory, he says.
"We've made this commitment to deal with the Microsoft marketing machine."
Novell is about to ship a Unix version of its directory on top of Solaris.
Both Novell's C-based and Java-based environments are doing well, Schmidt says. "Programmers doing Java are solving a different set of problems than those doing C.
"Java tends to be a client-server application, often at the front end."
He says Novell will endorse Jini, a look-up mechanism for finding things on the network.
"We are primarily a server sales company."
Schmidt describes Novell's relationship with Sun as obviously unique. "I know where all the bodies are buried.
"Now we have the directory running on Solaris, there's likely to be more alignment with Sun."
Most of Novell's forthcoming products will be in the ZEN camp, he says. "We're focused on getting more user information into directories.
"It's about security, identity and remote management. Security is the hard one."
Long-term growth for Novell will come from applications that use the directory, Schmidt says.
"The next big upgrade for Netware 5.0 is clustering at the high end. We'll spend the next year or two getting people converted from 3.0 and 4.0 to Netware 5.0.
"With Y2K, you have to be careful not to make architectural changes that will break anything."
Asked whether Novell was setting itself to compete with enteprise management frameworks such as IBM's Tivoli and Computer Associates' Unicenter, Schmidt says he doesn't want to go into that space. "I'm talking about very horizontal management."
Novell, he says, is definitely not for sale though the question had some legitimacy when the company was in trouble.
"When companies get in trouble, you have don't have to do much on the strategy side. You figure out what the customers want and ship it to them You're only viable with people's ideas if you can establish relevancy.
"There were surprises when I joined Novell, things I hadn't discussed or hadn't been disclosed.
"I've learned to be suspicious of great demos."
Schmidt says there is no temptation whatever to get back into the application server market, to anything that's a diversion from the architectural model with applications running on top.
The aggregate future for Novell, he says, should be in America and Asia but ... "I don't think in terms of geographies because products are so generic. I think in terms of establishing partnerships."
He says developing countries can't afford big iron and that, in some, Novell is bundling products at one low price.
"It's very hard to sell software at a loss."
The Asian downturn has meant $US50 million to $US80 million missing from the profit and loss ledger. "That's real money we'd like to come back. Everyone is struggling through this and I'm strongly in favour of America helping.
"It's important Asia comes back quickly. The conventional wisdom is that it will take 12 to 18 months."
Around 60% of Novell's total revenue still comes from the basic Netware platform, he says. "There is growth there because networks are growing. Because Netware is such a large componenet of the business, you would have to wind up with a $US100 million application business to make a dent in that."
Schmidt says he is spending almost as much time on technical issues as he was at Sun. "The complaint is that I'm spending too much time on the fun stuff."