Will Java and Jini be in everything?
April 26, 1999
by Clare Haney
(IDG) -- In the decade to come, computer users will no longer use their PCs as their main way of accessing the Internet, they will be tapping into the 'Net through their cell phones, their fridges, their televisions and their cars. They'll be surrounded by a wide range of Internet-related devices, many of them bearing the Jini logo.
This is part of Sun Microsystems Inc.'s vision according to the company's Chief Strategy Officer Bill Raduchel, who is also a member of Sun's executive committee. He heads up the vendor's planning and development efforts, including all mergers and acquisitions. It was Raduchel who signed on Sun's behalf the three-way agreement with America Online Inc. and its latest purchase Netscape Communications Corp.
Raduchel was in Hong Kong at the end of an eight-day visit to Asia where he met with Sun customers here, in Japan, the Philippines and Singapore, giving them Sun's perspective on the 'Net.
In a wide-ranging conversation with IDG News Service, he talked about how Sun's Java and Jini technologies are likely to develop and also gave his take on the year 2000 (Y2K) issue.
IDGNS: Was Scott McNealy (Sun's chairman, chief executive officer and president) serious when he said at the CeBIT show in March in Hanover that one of his goals is to put Java in every car made in Germany?
Raduchel: He was serious. Java will end up in all cars. It'll cost tens of U.S. dollars to connect to the 'Net in your car. That's why the 'Net will explode as the cost of access devices goes down. For example, Electrolux bringing out the first 'Net-enabled fridge. It's a novelty now, but in 10 years as costs go down, it'll make sense then. Even if you use it a few times a day, like a bulletin board, saying "Kids, I'm going to be home late, make dinner," it's worth it.
Both Microsoft At Work and Novell's NEST (earlier failed embedded systems initiatives) tried to go down too low. Java leaves most of the existing software stack to the device. The real different thing about Java is that it makes a layer you can put on many different operating systems.
Java and Jini will be in everything, but that doesn't mean that the user will ever know, 90 percent of them will never know. Everything is to do with economics. Technology doesn't drive change, economics drives change and that change in turn drives technology.
As we learn the economics of all devices below PCs, we discover that the PC model doesn't work very well. With the devices' very sophisticated engineering, it's not very easy to rip out one operating system and put in another. At the same time, all markets can't be different, there is a need to have a common layer so that you can use the same services such as stock tickers and e-mail notification in every device.
IDGNS: What about criticisms of Java? Is Sun working on a slimmed-down version of the Java Virtual Machine for use in embedded systems?
Raduchel: Our PJava (personal), EJava (embedded), which is a subset of P, and KJava (originally meant kernel) for use in the smallest devices, are all out in beta now.
Java has been criticized for being too big and for its performance. We're getting better on performance, with performance enhancements this year we'll largely eliminate the performance issue, then we have to work on removing the perception about Java's performance, that may take longer.
IDGNS: Given that Java partner Sony Corp. plans to make 100 million [M] of its next-generation PlayStations, is there a future for Java in games?
Raduchel: I think it will happen. Games are very graphics dependent and so have opted for optimal performance. With stand alone games, companies have traded off the cost of development for performance. Networked games will drive Java (adoption).
IDGNS: What do you see happening to the PC model over the next few years?
Raduchel: The most important 'Net access device in 2005 will be cell phones, PCs will the fourth, fifth or sixth most common device. Personal computing will be alive and well, but PCs won't be doing so well. In Japan already, you can sell cell phones as e-mail terminals because people are standing in lines in trains for one and a half hours.
Our Web tone idea is no different from the phone network -- you have mobile phones, home phones, fax machines and modems, they all use the same network. With Web tone, there will be lots of devices interacting with the 'Net. Perhaps, you'll use one to get e-mail notification, then go to another device to read your e-mail. We have to find a way to make lot technology easier to use. It's just too expensive now. Windows support is too expensive, people are stymied by it. Dedicated devices are simpler and you'll be able to afford many of them for e-mail, Web surfing and so on.
The 'Net is the ultimate communications medium. The big thing will be device-to-device communication over the 'Net. Think about repairs. It costs a lot of money anywhere in the world to have someone drive out to your house to repair your machines. Most machines know when they're wearing out and could send a "please schedule a replacement part message" to their makers. It would be US$20 to $30 versus $500 to $600 now spent on repairs.
IDGNS: How about Jini? Where does it fit in?
Raduchel: The true description of Jini is a spontaneous network with sets of Java programs. People will understand the beauty of Jini in Bluetooth (a wireless communications protocol) and HAVi (the home audio video interoperability specification.)
Consumers will never know about Jini, all they'll see will be the logo. It's all about making it easy to create local networks. For example, your cell phone belongs to 50 networks, it changes as you're moving along, when you're in a taxi, or if you're walking by your bank and want to do some banking. Jini eliminates the need for device drivers in the PC. Jini networks are self-healing, they never establish a permanent connection. According to the manufacturers at the Jini launch in January, Jini devices should be available by the end of next year.
IDGNS: What was your involvement in the Sun, AOL, Netscape deal?
Raduchel: I signed the contract. Scott (McNealy, Sun's chairman, CEO and president) went to AOL in 1994/1995 and talked to Steve Case (AOL Chairman and CEO), saying we should cooperate, but we had a hard time of it. At a December 1997 meeting, Steve asked Miles Gilburne (AOL's senior vice president, corporate development) and me to see about working together and we eventually focused on e-commerce. Miles suggested AOL acquiring Netscape and for Sun to partner with Netscape, bringing AOL's eyeballs together with Sun's servers.
IDGNS: What's Sun's take on speech recognition software? Both Microsoft Corp. and Intel Corp. are investing in that area.
Raduchel: We're looking at speech. It's an area where we've tended to rely on third parties, I'm not sure where we play in it, but speech is on our radar screen. Speech by itself is not a great advantage, it's not easier than touch screens.
The question is: What's user friendly? When I can say to a computer "Find the way home" and it can do it, then I'll really be interested. We're a long way from that, I can't do that yet from the keyboard and have the computer process it.
IDGNS: Would there be any technology in Asia Sun would be interested in?
Raduchel: We're a pretty polyglot company and are very open to ideas. For example, our workstations are built in Taiwan by Mitac. We understand Asia's potential, it will come back very strongly. Ten years ago, when I met with Asian users, there was a very different set of comments (from their U.S. peers), they were one to four years behind in IT thinking, that's not true today. The only big difference is that Asia lags behind the U.S. in terms of 'Net penetration.
The 'Net is such a leveling factor, it will drive the world to a smaller number of languages, perhaps five to ten languages. There'll be English, Japanese and Chinese, but will there still be French and German? There will be better automatic translation, but in a lot of business-to-business e-commerce, language won't be very important.
IDGNS: What do you think will happen on Jan. 1, 2000?
Raduchel: There will be a significant amount of hassle for people. It's a cold, not a heart attack. Small businesses will be hit, but lots of fixes are pretty trivial. We should have seen a lot of problems on Jan. 1, 1999, but we saw nothing.
Last fall, I was at a meeting where a very responsible person stood up and said "I think we have to worry about machine tools." When was the last time you checked a lathe for a date? Two-thirds of the people in the meeting room thought Y2K would be a major catastrophe. The former chief scientist of one of the top computer companies said we should all take US$10,000 in cash out of the bank in case of anything happening on Jan. 1, 2000.
I don't want to downplay the issue. I can't believe the doom scenario. A friend of mine, a CIO (chief information officer), as a joke, asked his CEO about finding funds to invest in the "Sept. 9, 1999 issue." It was a joke, but the CEO said, "How many millions do you need, we'll have to amend our SEC filing," and so on.
I know a top Web company that loves Y2K. They're planning to go and eat everyone's lunch, their competitors are spending millions of dollars on Y2K, while this company is spending millions on getting their rivals' businesses.
People talk about power plants' switches failing, switches don't look for a date. Electricity is electricity, it's not different after midnight. There will be a bunch of glitches. The thing we should be scared about is spreadsheets, people didn't follow the rules there. Y2K won't be catastrophic. The vast majority of problems will be cured by restarting the machine.
IDGNS: So, you'll be in the air then?
Raduchel: I have booked a flight for New Year's Day.
My home network: Look Ma, no wires
RELATED IDG.net STORIES:
Sun's Jini emerges from Java lamp
Sun Microsystems, Inc.
|Back to the top||
© 2001 Cable News Network. All Rights Reserved.|
Terms under which this service is provided to you.
Read our privacy guidelines.