Perspective: The second coming of the network computer

Larry Ellison finally deserves at least a few kind words in this space. Not many, but at least a few.

The famously egomaniacal oracle of Redwood Shores has been the source of great copy over the years, running his mouth off so often and so outrageously that the Fourth Estate could hardly dream of getting a more delicious character to pick at.

But for all his bloviation and bluster, Oracle CEO Ellison remains a technologist and a serious influencer, and so when he talks, the industry does listen. He stirred up a controversy a year ago, when he said the enterprise software business was destined to consolidate and then gave weight to his words by launching an unwanted takeover bid for PeopleSoft.

The funny thing is that most of Ellison's prediction came true--though not in the way he thought it would.

We'll know soon whether the PeopleSoft gambit was the right bet. Friday marks the take-it-or-leave-it deadline Oracle established for PeopleSoft shareholders to tender their shares at $24. If they decline, then Oracle says the deal's off.

All this is a good walk-up to Oracle OpenWorld, the company's annual customer confab, which opens next month. It so happens we're also close to the anniversary of one of Ellison's other more controversial predictions. In a major address five years ago this month, Ellison tried to dust off his idea of the network computer.

The basic idea was that the PC as we know it was living on borrowed time. It inevitably would be replaced by a network device--small, inexpensive and untethered--that accessed data from a central network rather than from a hard drive. Updates and bug fixes would be automatic, and administrative costs would be just about zero.

Oracle's CEO was panned by many folks in the press and analyst communities for promoting a pie-in-the-sky vision of the future that oh, by the way, also would give Ellison an advantage against his archrival, Bill Gates. (No more operating systems, no more Microsoft, etc.)

Microsoft's still around and thriving, but the funny thing is that most of Ellison's prediction basically came true--though not in the way he thought it would.

When he started promoting the network computer in 1995 and 1996, bandwidth was scarce, computers were slow, and most machines were expensive. What's more, Ellison's idea was predicated on the notion that this evolution would occur with Java--though back then, Java barely worked.

The network computing revolution came and went, and most people never took notice.

Like so many other predictions, this one simply was ahead of its time. The network computing revolution came and went, and most people never took notice. But it did occur alongside the emergence of standard, Web-connected PCs and of server-based applications that dynamically generate HTML.

While Ellison was off giving speeches on the subject during the middle of the last decade, there was little that came close to qualifying as a real Web application. The world's now a lot different. The Web is used for a lot more than content like text, links and images. We use the Internet to do things--whether it be to send e-mail, research travel or tackle customer relationship management.

Ellison was only guessing at the time, but now software is developed and delivered into the browser, while desktop applications (or so-called client-server applications) are clearly on the wane. Networked applications are finally real, but they are called Web applications. It's Ellison's vision of the network computer by any other name.

Biography
Charles Cooper is CNET News.com's executive editor of commentary.

More Perspectives

I disagree

it is quite interesting to see that this vision about the death of PC App is so spread even in the high level people of Cnet.
But I think you are simply out of the way.

Why do you want to come back to the mainframe ?
Why do you want to put the control back to the network sysadmin that centralize and inhibit collective decentralized innovation ? Why do you think that the model of telco-network side control is so cool ?

I think ultimately that the real good question we need to ask about this incredible transition we have been living is this one :

Which is the most important revelation of the young Internet era : Google or Skype ?

I definitely it is the second one ?
What is your thoughts ?

[stephan6969]

network computers

Mr Ellison's predictions should be taken more seriously. He is a visionary in many respects. Though timelines are not prudent for him to expound, he sees what is and asks why. He sees what maybe and asks why not.

Food for thought

Web /= Internet

A part of the article says "The Web is used for a lot more than content like text, links and images. We use the Internet ...". This wording furthers the implication of what most of the general public thinks; that the Web and email is equivalent to the Internet, and that the Internet has nothing more to offer. If people's creative minds would get outside of the box of the Web, and start thinking of creative ways to use the Internet for other types of good things, the Inernet could become a much more and better thing than it is today.

IT's gonna be much more

Larry Ellision's prediction has a point. What we are seeing today is not even 10 percent of what web offers.

The highspeed net and wireless access are bound to change the picture drastically. I see 2 things surely happening. Timing? I'm not sure, it can be 5 yrs.
1. PC will stay for those who develop softwares.
2. Users will surely switch to Ellisions's version of pc.

The pc that Ellision is talking about is similar to Blackberries with extended capability.

What does a layman do with a pc? Surf the net and use Web applications.

When Bell invented phones, that is waht people used it for. Now we have mobile phones with extended capabilities. Same thing is happening to PCs.

Add to that convergence of communication technologies. Combine phones, mobile phones, TVs, Cable TV, Blackberries, PCs and internet. Reduce them in size and connect with wireless. Voila, here is your device for the future. It may have some applications which run on it, for other applications it will access Web Servers.

First step of convergence is here

Hi

I wrote that your phones, mobile phones, TVs, Cable TV, Blackberries, PCs and internet will combine. Reduce them in size and connect with wireless. I forgot to add camera.

Here is first step in the direction. A movie premiere being CellCast, for the first time.

http://timesofindia.indiatimes.com/articleshow/949481.cms

It's not long before your Cable TV is out and CNN watched on cell phone.

Regards
Himanshu

[David Arbogast]

Ellison is still wrong.

Not that his idea doesn't have merit, but it just won't happen. The assumption being made, is that Internet servers will have the power to handle thousands or millions of clients running all types of applications.... ummm... nope. Now, I'm not suggesting that power won't increase, but regardless how much it does increase, two machines will always have more power than one.

Networked devices today are growing, and are adapting more features... phones have cameras and do email... blackberry is going color and has calendars, etc... The simple fact, is that connected devices grow to incorporate more functionality. I would predict that network computers would grow and evolve into full-featured computers over time anyhow. Just like cell phones started without a screen, and now run multi-purpose operating systems.

A network device will always seem limited compared to other devices, and will never have the ability to take advantage of the latest technology. Who needs the latest technology? Ever seen a computer game? Can you imagine trying to run Halo on a PC with no power? Not going to happen. Even the MMRPG games that are Internet-based require tremendous horsepower from their clients.

Ellison has a neat idea... and within a business that can maintain complete control over the computing infrastructure, his idea has real potential for reducing costs and simplifying maintenance. But as a vision for the future of the entire Internet, it is ignorant and short-sighted. It is, in fact, like suggesting that people want less.

[ajbright]

30 years old.

So what Ellison was saying (and I remember laughing about it 10 years ago) is that the very thing the PC replaced, a dumb console that is driven by large mainframe style servers, will come full circle and replace the PC.

He seems to have more in common with the head of IBM, that ridiculed the notion that everyone would want their own, personal computer.

And what about home computers? Even if we had 1 gigabit internet connections, are you seriously telling me that people would be willing to rent all their productivity software and games via the internet, rather than having it immediately available on their own hard disks?

Server based applications may be of some, limited, use to large corporations, but in the modern world, the small to medium businesses that are able to compete with those same corporations, are not going to give up their independence.

[Ubber geek]

immediately available

http://www.analogstereo.com/infiniti_fx35_owners_manual.htm

Network PC applicability is very limited

There is a market for the network PC, just as there was for 3270 terminals, X terms and PCs running terminal emulators in the past. However, not for mobile users, not for home users and not for users with sophisticated graphics requirements.

The author made a mistake when he described these devices as untethered. They depend on the server pushing content down the wire. They only work in an "always connected" world, where the user needs minimal local resources.

I'll take a laptop or PDA with intelligence over a tethered, dumb device any day. So will the market.

[Jim Babcock]

Can the CEO of Oracle be one? Yes he can

Computing history (starting in the late 50's) and goes
like this: mainframes with Fortran/COBOL dominating;
followed by UNIX (You mean I can have my VERY
OWN piece of a computer?); then timesharing, then
PCs. This took about 30 years to happen. In the 80's,
PCs took off... decentralizing the industry.

But the Internet appeared in 1985- 1992. PCs were
networked (client-server syndrome). The next BIG thing
(after Mainframes; UNIX, PCs, Internet) is JAVA. It runs
anywhere, any machine, and is free.

Broadband bandwidth increases each year; (Internet II
is around the corner... soon Internet III will be
discovered).. Network devices with rich JAVA engines
enjoying the power of multi-GHZ speeds will make the
Network node machine envisioned by Ellison a reality.
JAVA IS a very high level programming language. It
does deliver large, complex applications on the Net,
now. It seems that every 10-15 years, we see an
evolutionary breakthrough; JAVA is this decades entry.

So, we come full circle; back to centralized control of
computing capacity.Downloading high quality JAVA
apps down to slim, super-fast PCs with minimum
storage will keep Intel in business but could see us
wave bye-bye to Microsoft...

It is just on the horizon, waiting for someone to
springboard it into a massive global system.

Look for it... it is ready.

Jim Babcock