Open Source is not an IT Solution but a Development Model

James M Whitehurst

What do you think about virtualisation?

Virtualisation is the future of the operating system because it is not just virtualisation to abstract computing cycles. Conceptually, when you start saying that cycles are abstracted away from how they are delivered, it opens a world of possibilities. So, we have multiple efforts underway to address that. But it starts with application mobility and for applications to truly move they need to be certified and supported by the internet service provider.

We have some key technologies around virtualisation and there is huge investment in management tools and in security. We are executing against the vision ‘true application mobility’ abstracting away from the data centre. It is cheaper to source computing cycles; the question is how you make those applicable at a larger level.

Virtualisation needs a very homogeneous architecture to be effective. How do you react to that?

That is relatively true because when you start thinking of things like application mobility, it’s great to say you have this big pool of resources. But I don’t know if my application can run in this pool or moved over here. So, one of the key positive things is that we are going forward with this massive base, which is the largest base of certified applications.

We have been very careful to architect our solutions. We are working on two layers of abstraction. One is at the hardware level, a combination of the operating system and virtualisation which makes that possible. I agree with Microsoft that virtualisation has to be a component of the operating system; it cannot be standalone. Now, splitting it and saying that we are going to have 80 per cent of the operating system with 20 per cent that touches the applications, could be desperate. So, again, that’s why we see it as a key component. We then have a second layer of abstraction and that’s at the application server.

We want to go all the way back conceptually in our business model; we are all about providing value by being open. Abstraction at various layers actually allows much more flexibility.

In this context, where would you fit open source? Also, what is the purpose sought to be achieved by open source technologies?

Open source is not an IT solution but a development model. We believe in a far superior development model to a typical proprietary development model. We are operating at a couple of different levels. One is making open source consumable to the enterprise by taking phenomenal open source projects that are all about innovation. That’s the beauty of open source innovation and we put those together into an enterprise-plus product. We make open source a consumable IT enterprise. And we take that innovation once every couple of years, both in the operating system and in the middleware for executing other projects.

We have a very customer focused model and I believe we are successful because of our own business model innovation. Step one of that is enterprise grade software to run mission-critical applications. Step two is defining that in the broader sense of architecture. These can truly reinvent IT architectures and help fully realise the value of this superior development model.

What is likely to be the market for this? 

I believe that we are at a fundamental inflection point in IT architecture.  It is much cheaper to have massive grids of computing with thinner clients. The questions were at the level of architectural change. IT hardware/software services is a trillion dollar business. If you look at Red Hat Linux, it’s the second most prevalent operating system after Microsoft Windows. If you look at enterprise operating system, we have a solid double digit share of server operating systems. According to IDC, the server operating system market was $21 billion in 2007. We are little over $500 million. Obviously, we had services and learning services in Jboss so not all that $500 million was Linux. So, open source fundamentally rips huge cost out.

Some days we get depressed that we have only 2 per cent operating system revenues even though we have a large footprint. We are one of the most profitable public software companies in terms of cash. We can charge dramatically less than our competitors, bring a lot more value to our customers and still be comparatively more profitable.

Is your existing channel going to address this? 

I think there is a massive opportunity out there for system integrators to really understand this and take a lead. We are not trying to be another IBM. That’s not our business model. Our core competence is building community use of open source. We don’t want to have a multi-billion dollar services organisation.

Because if you look at open source and what we are doing as a development model, we are enabling a new architectural paradigm which offers flexibility and huge cost advantages. Companies have to move to that and that in today’s times is a huge, multi-trillion dollar opportunity

Would you say that this is going to create opportunities for Indian companies because India clearly has a very major market share?

I think it is a huge opportunity for a couple of reasons. The Indian government has been probably one of, if not the most, enlightened governments around the world in understanding the power of open source — not just interns of the lower cost but also in terms of knowledge transfer, skills transfer, the flexibility it provides. We have worked very hard here to build large community.

It’s our third largest market in terms of contributors to Fedora, which is our Community Linux. We have more Red Hat certified engineers in India with deeper knowledge base and understanding than any other country in the world – more than even the United States. 

One thing I found breathtaking is the size and the scope of the IT projects that are being and will be undertaken.

While we are a fairly enlightened open source kind of a country, of late, there has been a concerted movement by the government to go in for a local version of open sourcing what they call BOSS (Bharat Open Source Software)?

I don’t characterise ‘free’ as a threat. ‘Free’ is a compliment to what we do. We believe this very deeply as a company, in contrast to many other open source companies and this is why I think we are profitable and others aren’t. We fully buy into the open source model. We are successful because of open source, not in spite of open source. Why is open source so powerful, and it is around economics of abundance. That’s how the models were built.

So, once you fundamentally believe and buy into that you have a business model. We take this huge, massive global movement called ‘Linux’ in one of our products and we say we make that enterprise revving. CentOS, which is very well known, works closely with the community. Open source does not say that you have to publish your source code. It says you have to agree to a general public licence (GPL), you have to agree to pass on the source code to your customers if they ask for it.

We fundamentally believe as more people use it, it becomes a more important part of infrastructure. They will also understand the value that we provide. Open source is a development model, it’s not a miracle implementation model.

If you look at the Indian context, CentOS is not happening. Your comments.

Given the millions of dollars that we spend every year certifying hardware and software, I find it hard to imagine that it can ultimately be offered for free.

When it comes to really run mission critical applications, I think when people look at it, they realise the millions of dollars of research and development and deep technical relationships that are required across so many different hardware vendors. In the end, when people run applications that require hardware enablement, they think Red Hat.