Open source has been on an interesting journey over the past few years. How would you say it has developed?

The really fascinating thing for me about open source over the past few years is the way it has gone from being vendor-led to being more user-led. What you saw in the first decade of open source was a lot of work to create open alternatives to traditional software.

All that was great, but I would safely say the first decade of open source was about offering more plausible open alternatives to proprietary software. Then something fascinating happened about four or five years ago, and that was the first of the big web-scale companies. The cool thing about the web-scale companies is that they are large, technically sophisticated IT users that have both the financial and technical acumen to start driving their own roadmap.

After Google, Amazon, Facebook, LinkedIn and Twitter started building all of their infrastructure on open source, the next wave is all about innovation. Every major innovation in big data has come out of open source. Almost all of the new application models in mobile are open source. Most of the things happening in cloud are open source. And this is not because of vendors – it’s just a fundamental move from vendor-led innovation to user-led innovation.

>See also: How open source will power tomorrow’s tech unicorns

In big data and analytics, everything is happening in open source – even the proprietary vendors like Oracle and IBM are talking about how they’re getting involved in projects like Hadoop and Spark, and how they’re going to make this consumable.

It’s all fundamentally open-source software, and that’s a major change. When we talk to CIOs now, we’re not talking as much about how open source saves money – they just look at it as the best way to consume innovation because that’s where innovation is coming from.

How have you seen people’s perception of open source change over the years?

Firstly, the first wave of open source – Linux, JBoss and MySQL – have truly gone mainstream. 94% of the Fortune 500 are our customers, and 100% of large commercial banks, airlines and healthcare companies are using our software.

When I joined Red Hat eight years ago, much of what I talked about revolved around whether open source was safe, secure and reliable – now, I think everybody knows it’s all of those things. So over those intervening eight years, we’ve clearly checked that off.

The second major change is that CIOs are now looking at open source for where their innovation is going to come from. I talk to a lot of CIOs who basically say they’ve never really thought about it before but their purchasing and vendor management liked it so now they’ve got to understand it and develop an open-source strategy to keep ahead of competitors. That’s a fundamental change from where we were five years ago.

How does the culture of an open organisation differ from that of other businesses?

We’re moving from a world where most value is created by being more efficient. Traditional bureaucracy was really good at driving efficiencies, but we’re moving to a world where moving faster, meeting customers’ needs and innovation are becoming the primary ways that businesses create value.

Bureaucracy does not work well for innovation, so we started to think about working in a more horizontal network way. What that means is that an open organisation is one where leaders create the context for their people to do their best work. The traditional model of command and control is about deciding a strategy and then telling people what to do, so involves hierarchy and control functions. An open organisation is much more about enabling people.

Last year, you said that Red Hat is shifting from client-server to cloud computing. What was the thinking behind that move, and how has the transition gone so far?

I would say it’s cloud-mobile because cloud is the infrastructure and mobile is the end device – it’s hard to separate them because they’re so related. It’s going extremely well.

There are two components of our business. One is continuing to help modernise the workloads that enterprises have, to put them on a modern infrastructure that can run in virtualised, public cloud or private cloud deployment models.

If you look at our core Linux, it continues to grow in the mid to late teens. The second component of our business is our next-generation OpenStack. We’re the largest contributor to OpenStack and the largest enterprise provider of it. It’s going extremely well. We bought a company out of Ireland last year called FeedHenry to offer mobile services to developers – that’s been an extremely successful acquisition. Enterprises all over are excited that we’re offering the technology. So overall that transition is going well.

The good news for us is that because we were kind of born in the next generation our technologies are particularly relevant, so we don’t have that same legacy migration issue that companies like IBM, HP, Cisco, Oracle and Dell have faced. So, generally, our migration to cloud is all upside with very little downside.

What advantage does Red Hat have over other cloud infrastructure providers such as AWS, Microsoft and Google?

The value proposition that Red Hat traditionally had back in the days of Linux was to certify all of the hardware so you write your application once and it will run on anyone’s hardware.

To some extent we’re driving that same value proposition in the cloud – you write to a Red Hat infrastructure and you can run that application on-premise, on Amazon, on Google and, hopefully very soon, on Microsoft. We work with the major cloud providers so our infrastructure stack runs on the cloud. If you write your application using Amazon, Google or Microsoft’s software stack, you can never move it.

Our value proposition is that our software stack can run on any cloud. We disclosed last quarter that we’re already on over a $100 million run rate of just selling Linux on public cloud. So the proposition’s working really well – customers are concerned about cloud lock-in. We’re one of the antidotes to that – write to our stack and you can run it anywhere.

>See also: How open source can help businesses reclaim control of IT

The difference between us and the traditional IBMs, HPs, etc. is that we’re not competing against the clouds – the clouds are actually a deployment model for us and therefore they’re partners with us.

That’s one of the subtle but really important differences, and one of the reasons why cloud and cloud adoption is good for Red Hat, whereas it’s kind of bad for most traditional software companies.

OpenStack has been on an interesting journey since its inception five years ago. What challenges has it faced, and how have overall enterprise adoption rates developed?

I think the biggest challenge was that the expectation was ahead of the technology. One of the problems with OpenStack is that everybody was excited about it and wanted to be involved in it, so a lot of vendors started saying they were going to put out an OpenStack product before the software was ready.

If you think about how a traditional software company controls the release cycle of its own proprietary software, it would not put out a product until it was ready. The problem with OpenStack was that nobody wanted to be left behind, so people started putting out products when the software was still in alpha or beta stage. So I think there was frankly an 18-month or two-year period when vendors went out trying to sell products ahead of when they were ready.

Now, we have customers running production applications on OpenStack, the tools are touching up nicely, we’ve had great progress with our installer, and more hardware is getting certified to work on it.

So I think we’re really getting to the point now where we’re seeing enterprise adoption – it’s just two years after when people thought we’d see it. That’s because of the dynamics of having so many vendors involved around it.

We’re really starting to see adoption now in a material way, not just in enterprises, but also in telcos. So I’m feeling really good about where we are – I’ve been saying for years that major infrastructure technologies like this take a while.