Meet Ubuntu Linux's new CEO (Q&A)

LONDON--Jane Silber has been chief executive of Canonical for 11 days.

But she's no outsider swooping in to take over Ubuntu Linux's corporate sponsor. She joined Canonical in June 2004, two months after previous CEO Mark Shuttleworth founded the company with a few programmers he recruited from the Debian Linux project on which Ubuntu is based.

Since then Canonical has grown to about 320 employees and has made Ubuntu a major presence in the world of Linux--version 10.04, one of the important "long-term support" versions that arrives every two years, is due in April. It's an unusually sustained effort to make Linux a force on desktop and laptop computers, and among Canonical's accomplishments is a mainstream foothold on Dell PCs.

What hasn't changed is the company's insistence on making the version of its software it gives away for free identical to the product it supports commercially--a move that still contrasts with Linux incumbent Red Hat. And another thing: six years on, Canonical still is not profitable.

Being in the red now doesn't mean that the company--funded in part by Shuttleworth's proceeds from selling his earlier company, Thawte Consulting, to VeriSign for $575 million in 2000--doesn't plan to be in the black. Canonical has three main businesses: selling server management services to companies using Ubuntu Linux; working with original equipment manufacturers (OEMs) such as Hewlett-Packard or processor companies that need help with Linux; and most recently, an Internet-based tool for buying and synchronizing music files and other personal data.

I sat down with Silber, who moved from chief operations officer to CEO on March 1, in the company's London headquarters. This is an edited transcript of our chat about desktop Linux, cloud computing, and the company's profit plans.

Q: When Canonical announced the change in management in January, you signaled that there wouldn't be a cataclysmic difference from the Shuttleworth CEO era.
Silber: This is very much a continuation of what we've been doing before. I think it's hard for people from the outside to see how Canonical has been run over the last couple years, but there has been a strong partnership. To the two of us, this feels like a natural evolution and shift in our responsibilities rather than some dramatic knife-edge change. My role is to lead and drive us to accomplish the same goals on the same strategies we've had over the last couple years.

So if little is changing, what was the reason?
Silber: Canonical is changing in its life cycle. We're maturing as an organization. We're six years old now. We are 320 people--of that order. We have a much more robust set of relationships with partners and customers and the open-source community, and the type of work we're doing is different now and needs a different type of leadership and focus. It's the type of focus I'd like to bring. Mark has gotten more interested in elements of product design and strategy, and he's gotten more focused there. What the organization needs now we believe fits more naturally into the new sets of roles and responsibilities than the old ones. It allows both of us to focus where our strengths and interests are and on what Canonical needs at this stage in its life.

Start-ups often change from the visionary founding leaders to new management that focuses more on execution and operations. Is this that transition for Canonical?
Silber: It's part of that. We are still a very visionary organization. The work we're doing is still very disruptive. Some of the work we're doing on cloud computing on the server side is visionary. We are still breaking the model, exploring the boundaries between commercial and community. An element of this was about a drive toward operational excellence--benefiting from the foundation we've built over the last five years.

Investors who might not see eye to eye with management often pressure start-ups, but Canonical has funding from Mark Shuttleworth.
Silber: We are a for-profit company. We have product goals and technical goals that have been the case in the past and will continue to be the case on my watch.

But is there more urgency about profit now?
Silber: There is a sense of great opportunity right now. When we started Ubuntu in year one, we didn't put a strong push on trying to sell Canonical services, not because we were not interested, but it's hard to build a business around selling services around an operating system that nobody is using. We knew we needed to gain a user base and momentum before we could sell services. That user base is now there. There is urgency and momentum around that at a level we hadn't necessarily seen in the first couple years.

I've heard for years that Linux on the desktop will catch on, and it's had some modest success among programmers and developing countries. Where is it popular, where will is going to be popular, and where are you going to make it popular?
Silber: Is this the "Is 2010 the year of the Linux desktop" question?

I'm not going to go that far. Mac OS X is not the market-leading operating system, but it's reasonably successful. You can have success that isn't 90 percent of the market.
Silber: Creating a platform to get vastly widespread consumer use takes time. Nothing in that platform-changing realm will happen overnight. I think there are signs of change. We notice a dramatic change even in dealings with OEMs. If you think about the hardware ecosystem--the process of developing new components that find their way to end-users' hands--changing the dynamics of that industry. Where their staff is trained, where they drive product management from where innovation happens. There's a subtle but really important change happening across that whole ecosystem.

We see companies now having operating system that five years ago you'd never think needed an operating system. There's still going to be a lot of change in that industry. Lots of people today are trying things. Some things are going to work, a lot are not, it's going to be very dynamic over the next couple years. But what we're seeing is the result of the opportunities that open source and Linux have provided. The opportunity for choice and for innovation is coming out. In these disruptive environments, there's opportunity, and we think Ubuntu is at the forefront of that.

It seems like Netbooks would be pretty high on that list.
Silber: Netbooks are high on the list. I was at Mobile World Congress in Barcelona a couple weeks ago walking around the floor. It's primarily a mobile conference, but similar to what you'd see at places like CES where most of the products you see being shown are Linux-based and coming from quarters that traditionally have accepted other people's software products and put them together. There's so much activity in the area of taking what has been a Linux desktop and spreading it across that spectrum of form factors, from desktops to laptops to Netbooks. I think Ubuntu plays across all that spectrum. We have a core with common technology. It allows us to span that spectrum efficiently.

It seems to me Netbooks got a lot more popular once Windows began to show up on them. People are familiar with Windows, and they have software they want to run. I see Linux-based Netbooks coming out Taiwan, but I'm not convinced they're succeeding in a big way.
Silber: I think broadly as a category they're succeeding in the marketplace. There's exploration going on in terms of where the sweet spot is. Some companies are trying to discover where their skill set is. Lots of companies think they can make an OS, and they don't have that DNA in the company to really do that. There's exploration and experimentation happening. It produces a lot of devices and projects which aren't going to have a lot of staying power.

Linux has been successful in the server market. That's where Red Hat made its business. What are you doing differently to crack the server market?
Silber: Our main opportunity is in the cloud, both as a guest OS and in the infrastructure-host OS piece. I'm sure you're familiar with our partnership with Eucalyptus to build Ubuntu Enterprise Cloud?

Basically an in-house version of Amazon Web Services.
Silber: Right. It allows a company to build its own private cloud behind a firewall. That plays to our strengths for a number of reasons, including the simple fact that Ubuntu instances are free. You want to scale up, you want to burst? That's very hard in a Red Hat model where you need a licensed subscription for each of those instances. There's a good match between the inherent characteristics of that sort of cloud computing and Ubuntu. We're seeing a lot of interest there.

Red Hat is a great company that is going to be around and has a good business model in certain areas. We are not going out trying to target Red Hat customers and convert them to Ubuntu. Our main opportunity is in a different area than the one where they traditionally play.

So if using Ubuntu is free, where does the revenue come in?
Silber: Support and management services. Those instances, whether they're cloud instances or virtual instances, need some management services. This is our Landscape product offering. Landscape comes in two forms. One is a software as a service that we host. We also have something called Landscape Dedicated Server, which is an on-site version. There's a slightly different pricing model for that. It's basically per-machine under management.

Your third big business is Ubuntu One. Where is the money coming in there?
Silber: That's our newest business unit. The core offering is the storage and syncing capability. That is a freemium-based model. A certain amount of storage is free, and there's a subscription for larger amounts of storage. With Ubuntu 10.04, we're introducing Ubuntu One music store, which is through a partner providing the digital content. It's a purchasing MP3 model.

One of the most fascinating things now facing desktop operating system companies is cloud computing. For Linux in particular, it seems a blessing and a curse. It gets around a lot of the application availability problems. Quicken is a great example. For years people would say, "Oh, there's no Quicken on Linux." With Mint, now there's Quicken on Linux--with a lot of qualifiers, but you get the idea. The curse is the operating system just becomes a piece of the stack, down there below, not something that the end users even is recognize is there necessarily. It's not as much an opportunity to sell in one way or another to end users. How do you see it?
I think ultimately it's a benefit. It certainly is a changing dynamic. I'm not sure how many people will go completely into the cloud in the near future. This is a pendulum that swings back and forth from local client apps to centralized apps somewhere else that may or may not be under your control. The sweet spot is in that middle ground. It's naive to think everything will always be running on your machine locally, but it's equally naive to think everything will move to the cloud. The challenge for us is playing to strength, finding that sweet spot for the Ubuntu user base, what they need and want, and providing an appropriate set of local applications and Web-based services.

I've seen the pendulum, too, with time-sharing and whatnot. But the Internet strikes me as profoundly different from the old days of running stuff on a server, the green-screen terminal days. It's just so pervasive and you must be connected to it for a bunch of routine things, even if you're not jumping into cloud computing whole hog. I don't know where that pendulum is going to end up, but it seems it's going to be a lot more toward the cloud.
Silber: It's not something we're fighting against. We're trying to leverage this opportunity rather than cower in a defensive corner. For instance, another 10.04 feature is an initiative we're calling Social from the Start, which is making desktop this social gateway to your social networks, to your online life, in a very seamless integrated manner. One of the key features is what term the Me menu, a menu in the top panel bar. With just one click, it'll drop down and let you post to Twitter, Identica, Facebook, in a very simple, elegantly integrated into the desktop interface. It's that merging of online worlds and your local desktop world we think is very interesting.

Linux has been persistently popular with software developers. There's a lot of benefit there to having serious local computing horsepower when you're compiling your code. What are some other examples of software running locally where Ubuntu can make a difference? You don't have Photoshop, you don't have Final Cut Pro, you don't have a large number of games.
Silber: A lot of things related to media are still very valuable in a local perspective. The notion of being able to have access to your content locally, even if you're using a cloud-based application to deal with that, is very compelling to people. There are environments still, while we are moving to pervasively connected online world, there are still instances where you want offline horsepower to do things as simple as editing documents.

You mentioned Photoshop. The open-source community has really powerful photo tools like the GIMP. Inkscape is a great app for illustrations. There are a number of strong client applications like that. There are a number of good Web-based implementations of tools like that as well.

When you look at your three big business, OEM, customer support, and Ubuntu One, what are the ones that are going to bring you into the black?
Silber: It's going to be a combination of those three. Our OEM relationships are going from strength to strength. Some of the work we've done in the last couple years in terms of aligning hardware ecosystems around Ubuntu, getting people to enable various components on Ubuntu, are really starting to pay off now. When somebody wants to put together a computer, the components they're selecting from all work with Ubuntu as the base Linux platform. That's taking off in a very significant way.

Online services is a newer business unit. We have a ways to go there. It's not as mature as our OEM services offering, and in some areas, we're still finding out what's going to be successful. With corporate services we have a solid base of enterprise users now, and I think the cloud in the next couple years is going to make that grow quite substantially.

Under your management, is the profitability push going to be stronger than it has been under Mark Shuttleworth?
Silber: Certainly driving us to profitability is one of my important goals. People might not give Mark credit for how much that was one of his priorities.

It was clear to me it was a priority. It just wasn't clear when it was going to happen.
Silber: We are closer to it now than we have been before. I'm determined to drive us there. But it has been a priority for Canonical all along. It is one of my focal areas as I take on this job.