Marc Andreessen has redefined the adjective meteoric, at least when used to describe rising careers in the software industry. Two years ago, the 22-year-old was working for US$6.85 an hour as an undergraduate employee of the National Center for Supercomputing Applications at the University of Illinois. Today, as the vice president of technology at Netscape Communications Corporation, his net worth exceeds $50 million, based on the shares of Netscape stock he holds. Andreessen is enormously assured for someone so young. His answers, honed during two weeks of road shows for Netscape investors, burst out of him at a rapid-fire pace. Chip Bayers met Andreessen at the company's Mountain View, California, offices shortly before Netscape's spectacular August stock market début. Follow-up exchanges took place via e-mail.
Wired: Are we going to have two Nets any time soon - a low-bandwidth version like the one we're used to and a separate, high-bandwidth version that's graphics and video intensive?
Andreessen: No. Their interconnection is too valuable - there'll be high bandwidth, but it'll be integrated into the other environments, like everything else has been.
Does the Web browser become something like an operating system?
No, it becomes a new type of platform. It doesn't try to do most of the things an operating system does. Instead of trying to deal with keyboards, mice, memory, CPUs, and disk drives, it deals with databases and files that people want to secure - transactions and things like that. We're going to make it possible for people to be able to plug in anything they want.
Doesn't that become too monolithic a structure?
That's the great thing - it's not monolithic. The platform is this environment into which you can snap new things. Instead of having a world where linking and embedding - through OLE or whichever acronym you choose to use - connect things together, the Web environment can encapsulate all of those, surround the whole thing.
Because it allows plug-ins, you're calling Netscape 2.0 a platform rather than a browser in the company's product descriptions. But Photoshop has some pretty sophisticated plug-ins, and it's only an application. Is Netscape 2.0 a platform or an application?
It's a little of both. It's an application in that it runs on top of what is traditionally thought of as an operating system - like Windows or Unix - but it's a platform in that people can build applications on it. We use the term live online applications for the types of applications that people build on our platform - online because the applications are network-centric and distributed, live because they're highly interactive with users and with data retrieved in real time from databases and other sources over the network. Live objects is our term for things like Java applets and inlined viewers embedded in HTML documents. So, a live online application is built using HTML as a framework. That gives developers great flexibility in linking together people and information over networks.
Our platform is also operating-system independent; you can run the client on Windows 3.1, Windows 95, Windows NT, Unix - any of 12 flavours - or Mac; the server can run on Windows NT, Unix, and in the near future, Windows 95.
We don't use the term browser because we think it's pretty clear that Netscape Navigator 2.0 is far more than just a browser. On the one hand it's a suite: it does e-mail, threaded discussion groups, ftp, gopher, chat, et cetera. On the other hand, it's a platform: it allows people to build these live online applications on top of it.
Netscape and certain content providers have leapt ahead of the common standards in their new versions. Are standards pointless?
No, standards are a good thing. It's easy to get wrapped up in the academics and debates, but the issue is that things have to interoperate on a network. I think the devil's in the details. Are the standards created from the top down, like interactive TV, or from the bottom up? Is there room for innovation in the process? If there is, then the process is going to be successful. The reason TCP/IP was successful is that people just went ahead and did it. It interoperated; it worked. OSI was created from the top down. Nobody used it. I think there are a lot of similar lessons in the 3-D world.
But no one wants to be the guy who buys the Sony Betamax player and then can't get any tapes for it.
In an environment where there's lots of innovation, consumers have a lot of different products to choose from. Is that good or bad? The fact is, to fix that problem, you'd have to freeze innovation. You'd have to say, "OK, everyone, stop."
Except it's not only software; now we're dealing with the type of Net connection available: America Online, CompuServe, or The Microsoft Network.
The tremendously exciting thing that's happening is that all of the online services are basically becoming TCP/IP services. Their users are going to be able to run whatever software they want - it may be ours or somebody else's, but all these services realise this. Microsoft is realising that proprietary services are going to be a dead end.
The Net, more than anything else, is a platform for entrepreneurial activities - a free-market economy in its truest sense. It's a level playing field where people can do anything they want.
Since this is an entrepreneurial environment with a level playing field, who do you see as your biggest competitor right now? Microsoft and its new browser?
We have a wide range of competitors entering this space, competing with one part of what we do. The Microsoft browser is basically what we did with Mosaic - I'm glad to see they've caught up to what we did two years ago.
Well, if Microsoft isn't a direct competitor across the board, how about IBM? Is the Web - and is Netscape - a groupware alternative to IBM's Lotus Notes?
Groupware is a marketing term - it leads people to think about all the wrong things. Any company with a signiFIcant infrastructure is moving towards the Internet standard, TCP/IP, as its protocol.
Notes is funny because it has virtually no presence in the IP network space. And Notes is so expensive to deploy, it's so complex and monolithic. It's a very IBM-like product. It's going to be interesting to see if there is a future for it in these open network environments, where people are looking more for flexible systems.
So what is the future of HTML?
Instead of having Open Doc containers, or OLE containers, or whatever they're going to be called, you're going to have HTML documents containing all these different types of things: Java objects, Macromedia Director FIles, different types of audio formats, video formats, portable document formats - just about everything.
On the content creation side, your new products are directly challenging Microsoft. But instead of getting tools to create HTML, content providers seem to be getting separate sets of tools for MSML (Microsoft markup language) and NSML (Netscape markup language). Should you get together with Microsoft to agree on a standard?
Microsoft is introducing a new content format called BBML - Blackbird Markup Language.
Netscape regularly contributes proposals and ideas to the standards forums, as we are doing with all the new stuff in Navigator 2.0. We also fully support all standards, including HTML 3.0 tables and backdrops. Microsoft will choose whether or not to contribute its innovations to the standards process, just as it will choose whether to support existing standards - right now Netscape is clearly in the lead on both.
I think there's a misconception that all innovation must be originated inside standards bodies - that is, the standards body is supposed to start with a clean slate and design something from scratch. The Internet grew as people and companies made many proposals for many ways to do things. The standards bodies just played a shepherding and stabilisation role.
When you talk about Microsoft, you and co-founder Jim Clark sometimes sound like you're from Apple. If you cannot agree on standards with Microsoft, is a Netscape-controlled future for the Net a good alternative to the evil of one that's Microsoft-controlled?
I don't think it's a matter of good or evil - Microsoft is a competitor, and a smart one. Jim and I both think it's important to point out what Microsoft is doing in various areas, since they are very good at using FUD (fear, uncertainty, doubt) to attempt to paralyse the market - Blackbird is a perfect example; it's much less than it's cracked up to be, and people are taking it much more seriously than it deserves.
I don't see a Netscape-controlled future for the Net - it can't happen; the Internet is too dynamic and innovative. No single company will control everything. Our approach is pretty straightforward: we support open standards. Today we support a wider range of open standards to a greater level of fidelity than any other company in this space that I know of.
What are you going to do with all that money you're worth as a result of Netscape's public offering?
Well, for one thing, none of it counts, because it's all paper. It's going to be a while before any of it is available. Who knows if we'll even be here in two years?
Tell me about the atmosphere at Netscape on the day of the initial public offering. What was it like to have your personal stake go from $0 to $75 million to roughly $50 million in the course of a few hours?
It was business as normal, for the simple reason that it's all just play money unless we have a profitable business over the long term. I was surprised at the market's reception of our stock, but it wasn't something we had a lot of control over; we did the road-show presentations for institutional investors in which we gave a fair overview of the company's business, and the rest was out of our hands.
Are you working with the rest of the engineering staff on a day-to-day level? Are you working on code yourself?
I'm not working on code. But I spend a fair amount of time involved with the design of the product. Part of it is, Please add this feature, and, Please fix a bug. I save up most of the significant stuff I would like to see done differently, or would like to see added, and roll it into the next revision, the next plan.
Small private companies get their screw-ups buried in the back of the business pages. Two billion-dollar public companies get their mistakes splashed on the front page of The New York Times, as we saw in September with the story on the security hole in your encryption. Any lessons?
One is being more open to having outside experts investigate our security mechanisms. Another is to beef up quality assurance, particularly at the deep technical level. In the long run, what happened with our encryption is good for us - we, and the whole world, will know that we now have more secure products than we did before this news broke. So, from that standpoint, I'm happy hackers are concentrating on trying to break our products, since as a result we get to improve them.
Chip Bayers (firstname.lastname@example.org) is managing editor of HotWired and co-author of the books NetGuide (Random House, 1994) and Where We Stand (Bantam, 1992).