Open source's next chapter?

Time magazine considered her one of the most influential people in America in 1997.

But fate is a fickle friend, and Kim Polese then embarked on a celebrity roller-coaster ride that took her from being Web poster child to nearly anonymous in the blink of an eye.

Polese, founder of Marimba and formerly the public face of Java for Sun Microsystems, is now back in the limelight--this time as CEO of SpikeSource, an open-source services company catering to corporate customers. And just like during the early days of the Web, Polese believes she's at the front of something big.

Talking about the future of the industry with CNET News.com, she said the combination of readily available software components, on the one hand, and the Web for collaboration on the other will usher in a golden period for the software industry.

Q: Silicon Valley went through a tremendous slump the past few years, and of course there's the whole debate over whether IT matters. Do you think the best minds are still working here, or are other areas of technology more compelling?
A: Actually, I think that we are at the onset of a renaissance in the software industry. The reason is that the industry is maturing. But that's not a bad thing--that's a good thing.

Software is going everywhere in every conceivable appliance or object that we interact with in daily life. The total market size is increasing, not getting smaller and not shrinking. All of that together means renaissance, means growth, means driving in the next generation of the software business.

What do you make of people like Larry Ellison who predict that there will be massive consolidation in the software industry?
I agree with Larry that the industry as we know it today is consolidating and that a lot of start-ups are going away because

you cannot manage the cost, the overhead of the sales force, and all the things that you need to do to compete against a big company. But there is a new crop. There is a new generation of companies that are utilizing the Web to deliver their services and that are utilizing open-source software and creating innovations that take advantage of the commodity, of the abundance. And that's perhaps something that people have missed.

Comparing this era to the early days of the Web, do you think open source is going to have the same sort of buzz or hype that we saw back in the mid- to late 1990s?
Well, I don't expect to see that level of hype ever again in my life--at least in this industry. I think that was just the product of a pretty unique time, when the Web burst upon the scene, and people started thinking about where this could go. They jumped to the ultimate end goal and realized this thing is big. And then, of course, Silicon Valley became the new Hollywood, and we--Marimba and I--happened to be caught up in that.

Is the idea of SpikeSource to make it look like there's a commercial outfit like an IBM or Microsoft behind a set of open-source products?
Yeah. What is sort of interesting right now is that IT developers, architects and chief information officers are aggressively adopting open source. The problem has become how to manage the abundance. There are more than 85,000 different open-source projects today.

All the things that IT is used to, like support documentation, reliability, road maps--none of that exists for open source when you start moving beyond a single component. When you start talking about actually integrating the components into applications, there is no sort of product management for open source. That is where we see an opportunity.

IBM, Hewlett-Packard, Novell and smaller companies like JBoss are providing services around a packaged set of open-source products. Why start a new one?
Platform vendors absolutely would be natural candidates for doing this. In fact, as you said, they are beginning to talk about and provide some of these services. But ultimately, the larger objective of that company is in promoting and selling that platform.

Clearly, part of the reason open source took off is because the software is free. Does a services company still have a cost advantage, even when there aren't any software license costs?
I think there is a whole new pricing model developing in a way that's different from how software has been priced for the last 20-plus years.

With licenses, you pay a significant up-front cost and then the ongoing maintenance. Maintenance is becoming a bigger and bigger chunk of revenue for software companies. But it is the model of a bunch of money up front, and then customers are left to essentially try to figure out how to implement the software.

That model is starting to get a lot of push-back, because IT can now supply itself. We are seeing instance after instance of CIOs deciding to reject commercial software and instead adopt open source, in part because it is also getting better from a quality and reliability standpoint and a robustness standpoint. When that happens, the pricing model gets turned upside down as well.

How is a subscription model for assembling and packaging open-source components an improvement?
Layering addition stacks and addition services over time--that's not about taking a huge bite up front or socking IT with these inflated maintenance contracts. But instead it's providing them with what they need and no more. I believe that ultimately, every software component application is going to be available in some form of open-source software.

Does that mean that software becomes free?
Well, I am talking about looking out on the horizon, where this is all going. Certainly, there will be intellectual property which is unique and innovations which are unique and which companies and developers want to protect. That is how it should be, but over time, I think that more and more categories will begin to be commoditized with open-source software.

So as more software categories become easily available commodities, does the money in the industry go toward services like you are describing?

It's what happens to every industry when it becomes commoditized; innovation moves up to the next layer, basically. You look at a company like General Motors today, which increasingly has a pretty significant piece of its business from OnStar. So it is further moving up the stack in the automotive business, licensing the service to other car companies.

Does that mean that building and selling products is not as interesting as it once was? Why are seasoned entrepreneurs starting services-oriented companies like SpikeSource and SourceLabs?
Think about the building industry. Doc Searls is someone who talks about this extensively: "Do-it-yourself IT"--The fundamental building blocks, the concrete and the lumber, that is becoming widely available--and a lot of complexity just is inherent in putting those pieces together.

That's what's becoming automated now in the companies, like ours, that are getting into the business. I don't for a second believe that this means that commercial software or closed-source software is going away. But innovation will happen in new areas, just as it has in the construction industry.