More Trails to Blaze

This spring marks the 10th anniversary of the launch of the Mosaic Internet browser, the application that opened the door to the Web for many consumers and served as a wake-up call for Microsoft Corp. and many other companies. Mosaic also launched the meteoric rise of Netscape Communications Corp. and its founder, Marc Andreessen, who helped to create Mosaic. Andreessen went on to found Loudcloud Inc., which morphed into OpsWare Inc., a Sunnyvale, Calif., software company focused on automating data center provisioning, deployment and operations. eWeek Executive Managing Editor/Features Jeff Moad caught up with Andreessen this month to discuss boom-and-bust cycles in the IT industry, OpsWare and other issues.

eWeek: Youve said you believe what were seeing now among enterprises is a period of intense commoditization, which will be followed by investment in more ambitious applications. Why and when?

Andreessen: [One] reason I think this is a cyclical situation to be followed by what I think will be a very interesting recovery ... is that adoption of important new technologies follows an S curve. The analogy I draw is to the PC. From 1975, when it was invented, to 2000, when it really became mainstream when everybody had one, followed this kind of curve. ... Were only 10 years or so into the deployment of the Internet ... in sort of a serious form. That means theres another 15 years left to go in that generational shift. And, if historys any guide at all, were going through a period right now where there is certainly a big cleanup and a major psychological reset. ... But sometimes these things just take time.

eWeek: Is there a technology that will accelerate Internet adoption in the same way the 386 chip and Windows accelerated adoption of the PC?

Andreessen: Thats maybe the key question of all. For the PC, it was what took the PC above the hundred-million-unit mark. There were tens of millions of PCs in the 1980s, not hundreds of millions. ... The Internet, on the one hand, is being used by hundreds of millions of people but not yet 500 million or 700 million or a billion. What would you need to get there? It certainly helps a lot that PC price points are coming down. ... But more than anything, I think its just time. Its just people acclimating to the technology.

eWeek: Youre competing with large companies such as IBM and Sun that are, like OpsWare, focusing on utility computing, virtualization, autonomic computing—whatever you want to call it. Whats OpsWares advantage?

Andreessen: Several advantages. No. 1 is that were shipping, and theyre not. And that actually turns out to be a big advantage. We rarely go head-to-head against the big system vendors in a deal today, which is to say theyre just not present. Most of the time thats just because theyre not shipping. Even when they are shipping ... the customers all say the same thing. They say they dont want to buy one thing to run their Sun boxes, one thing to run their HP boxes, another thing to run their Linux boxes, another thing to run their IBM boxes. Theyre not going to buy five of these things. And thats just the OSes. So the customers all say cross-platform is necessary, important, critical. You gotta have it.

eWeek: What happens when you go up against IBM?

Andreessen: We ran into them at MetLife. We beat them at MetLife. At MetLife, they actually bid a $20 million Global Services contract to custom-implement them a utility computing software platform. And MetLife said, Wed rather buy a product. And they said, We can sell you this OpsWare thing. They did this without talking to us. But we were already talking to MetLife at that point. It was a great endorsement.

eWeek: If you look out two years, what will be the scope of problems that OpsWare is addressing?

Andreessen: What I would want us to be in two years is the leading provider of software for automation for utility computing. Leading in the same sense that Oracle or Veritas or BEA are in their markets, which is leading independent, cross-platform. ... We will continue to build out support for the breadth of technologies that we cover. Today we cover a lot of OSes, a lot of app servers, a lot of databases. In the future, it makes sense for us to cover a lot of business applications. ... It makes sense for us to extend our support for other kinds of computing devices, desktops potentially, mainframes maybe. Maybe load balancers, maybe firewalls, things like that.

eWeek: After acquiring Netscape, AOL was supposed to become the Microsoft killer. What happened?

Andreessen: Six months after they bought Netscape, [AOLs] market valuation went from $40 billion to $170 billion. It went up [by four times] in six months. That was really good for Netscape shareholders, but it meant that they immediately stopped caring about Netscape. If youre a $170 billion company and you werent six months ago, the question is not what do you do with this $10 billion thing you just bought [Netscape], the question is do you go buy Time Warner, Viacom, AT&T, the United Nations? Who do you go buy? Topic A at AOL board meetings at that time was a bar chart they had of the biggest companies in the world rank-ordered by market cap. They were on the list; therefore, anybody below them on the list could be bought. At one point, they had a draft plan to buy all of the top five newspapers in the country—The New York Times, The Washington Post, Los Angeles Times, The Wall Street Journal, The Boston Globe. [But] what they really wanted to do was to buy Verizon or somebody like that. But those guys were a little too big.

Then the dog caught the bus, and they bought Time Warner and, at that point, they didnt care about Netscape, and they increasingly didnt care about their own business. ... The AOL part of the business just wasnt being paid attention to by good-enough people.