Andreessen assesses browser prospects

Marc Andreessen ponders the future of the Web browser

Ten years ago this month, a team of programmers at the University of Illinois' National Center for Supercomputing Applications loaded a program on an FTP server that could be accessed by almost anyone on the Internet. Mosaic was the culmination of work launched about six months earlier by NCSA researchers Marc Andreessen and Eric Bina. Within a year, there were millions of Mosaic users worldwide. Andreessen, the quintessential Internet programmer, is the board chairman of Opsware, formerly Loudcloud, a Sunnyvale, Calif., vendor of data center management software. He talked recently about the future, present and past of the Web browser with Network World Senior Editor John Cox.

What have been the main results of the invention of the Web browser?

The big result is that the browser makes it possible to create Web sites. In the early months, there was very little content to actually view. It wasn't like you had Time magazine online. But it also made it very easy for people to view what was there. We went to 10,000 Mosaic users quickly, and to 100,000 in about three or four months. And the number of Web sites [available to be viewed] tracked the same way. It was the network effect, a snowballing. And it still is [continuing]. The browser catalyzed all that in the first place.

OK, so the future of the Web browser is what?

After 10 years, it's still a user sitting in front of a Web browser viewing HTML services. It's not [about] XML, Java applets, the semantic Web. It's HTML and some JavaScript. The difference is that now it's 500 million people doing this, with something like 3 trillion or so HTML pages.

It's the same dynamic as TV. TV was invented in 1950. Today, we have 500 channels instead of three. But it's the same model, exactly as it was 50 years ago. Once these things get started, it's hard to slow them down.

What about the idea of the semantic Web, new tags that will let applications and computers automatically interact?

The semantic Web [means you have to] retag everything that's out there [in HTML content]. Uhmm, no. I don't think so. But if the browser isn't changing, the [original] architectural changes are still relevant.

How?

Before the browser, if a business had a software application, then that's what it would give to its employees, period. They never even conceived of exposing their applications to anyone else.

With the browser, all this changed. Amazon and eBay today have millions and millions of people using Amazon and eBay applications via the Web. Consumers log on to their PC and typically now are running applications on someone else's computer.

[So] you can get [Web] services, and transactions, and all this stuff. EBay has thousands of servers running very complex applications in a complex infrastructure to make this possible. But the users just see Web pages.

How will wireless technologies affect this model?

Wi-Fi [wireless LANs based on the IEEE 802.11 standard] will be the dominant form of the 'wireless Internet.'

Five years ago, my PC was linked to the Internet via a dial-up connection. I'd dial up my provider, log on, use my computer online for a while, then log off. That takes a lot of effort.

With Wi-Fi, on the other hand, as the way to access DSL or cable modems, every computer is on the network all the time. Your usage pattern changes at once: you use your computer on the network intermittently, dozens of times a day. You're using it therefore more often in more day-to-day activities. Wi-Fi makes this all much easier.

What about the wireless Internet based on cellular data services?

Marc AndreessenCell phone data is a complete market failure in the U.S. because the user experience is deeply inferior to what you do with a browser on the PC or even a handheld. Cell phone browsers based on [Wireless Application Protocol] just make you want to cry.

You're not a fan of converged devices?

There will be multiple devices [for each user], and wireless networking will enable that kind of diversity.

When I first went to a Kentucky Fried Chicken restaurant, they'd give you a spork - something that was half-spoon and half-fork. And I thought, 'why don't we have this at home?' And here's why: because what you actually want is to have a good fork and a good spoon. And the spork is neither.

The multipurpose device will always fail.

What's been Microsoft's contribution to the browser?

Undoubtedly, an overwhelming benefit has been the proliferation of the browser. Hundreds of millions of people now have access to it. Microsoft has been a very effective force. But it's also been a force in the eradication of the commercial [browser] market, and the elimination of any incentive to change the browser. It's not like they've changed a lot [in it].

What's hot in browser technology?

There's nothing emerging right now. Creativity stopped in 1997. Before that, there were huge numbers of changes: dynamic HTML, JavaScript, Java mail, plugins for security and other functions. And these were created by Netscape and many others.

What killed the creativity?

The browser market went away. There's no commercial incentive. It's all free. The browser today is basically what it was in 1998-99. The good news is that everyone knows what a browser is and what it does.

What about the open source browser project, Mozilla?

Mozilla could affect this. It has lots of developers and a fair amount of users. But, again, there's no commercial incentive to create a competitor [to Microsoft's Internet Explorer].

At the beginning, you posted the beta version of Mosaic on an FTP server, not even a Web browser, because the Web in effect didn't exist at the time. How did the idea of graphic user interface [GUI] for the Internet arise?

We were at the University of Illinois, at the NCSA. Originally, the idea was to create supercomputers at a central site, and then let researchers log onto them from around the U.S.

But that model quickly didn't make any sense, because we all had these powerful Unix workstations on our desks. So the focus was shifting toward enabling the Internet as a medium for research and science. We were part of the software development effort to do that.

We needed to put a full GUI front end, as people had come to expect on the desktop. Originally, it was a renegade project with no [official] permission. It turned out people really liked it.

It was a free program, on the share-everything Internet. Why make it into a commercial product, by creating Netscape?

We took Mosaic as far as we could in a research environment. [There,] we could do new things without worrying about return-on-investment requirements. But you couldn't hire support people for the users. The National Science Foundation doesn't pay for technical support people.

So we decided to start Netscape around this idea.

In hindsight, would you change anything in the first release?

Probably not. Although there was one feature that was temporary in Mosaic: the Back and Forward buttons. That never made a lot of sense to us. Back to what? Forward to what? We thought there would be a better way to navigate. But no one ever came up with one.

When we started Netscape, we added stuff, such as a safe way to do secure transactions. The first Mosaic release was for Unix, and we quickly changed to Windows. At Netscape, Windows became our priority.

Editors' Picks
Join the discussion
Be the first to comment on this article. Our Commenting Policies