Jim Whitehurst, president and CEO of Red Hat, the oldest and by far the most successful company whose business is based purely around open source, makes no bones about it: “Selling free software is hard,” he says. In fact, he goes further: “Open source is not a business model; it's a way to develop software.” 

By opening up the code and letting anyone suggest improvements, free software can tap into the best brains anywhere in the world, using the internet to create a virtual team. But it's a team that largely chooses itself, with roles apportioned according to ability, not corporate hierarchy. Despite this unconventional structure, it's a development method that leads to better software than the traditional top-down approaches, according to Whitehurst. 

That's certainly borne out by the recent top500 supercomputer listings, where 91 per cent of the world's fastest supercomputers run some form of GNU/Linux (Windows runs on 1%), the fact that the open source Apache Web server well over half of the Web – as it has for the last 10 years - and the near-parity in Europe of Firefox's market share with Microsoft's Internet Explorer.

But Whitehurst admits that free software's strengths also make it hard to implement in the enterprise. The fact that open source is not monolithic, but consists of hundreds of smaller pieces working together, each of them generally the product of independent projects, and that code is released at a dizzying rate, with rapid iterations every few months, makes it a CIO's nightmare in terms of control. But one person's problem is another company's business opportunity.

Whitehurst says that Red Hat's approach is based on the fact that “once every two or three years, we freeze the spec” of a core set of free software enterprise applications, including the main Linux operating system, and its own JBoss enterprise middleware. It then puts 40 to 60 engineers on the release to make sure that everything works well together.  One benefit of freezing the code in this way is that Red Hat can offer Service Level Agreements, and guarantee to support its product for years – thereby countering the old accusation that nobody stands behind open source.

Red Hat's Werner Knoblich on lock-in, Oracle/Sun, IBM, cloud, JBoss and public sector

Red Hat's business model “works by adding value on top of the code”, he says. “We sell subscriptions – not to bits, but to a stream of innovation.” Impressively, Red Hat has “never raised its prices” during the last eight years – although Whitehurst doesn't go so far as to guarantee that it never will. Unusually, he says that Red Hat's “biggest competitor is people stopping paying for code” -- not least because it works so well nowadays that companies feel more confident supporting it themselves, using the active online communities that are a key part of the open source ecosystem.

Not surprisingly for someone who joined Red Hat in December 2007 from outside the world of computing – he was Delta Airlines' COO, and before that Director of The Boston Consulting Group -- Whitehurst has some fresh views on the current IT landscape. For example, he says: “For every CIO in the world, Google is your competitor in terms of the expectations your constituents have”, noting that “the best technology experience happens at home.” One consequence is that a corporate IT project faces the daunting task of needing to match Google in terms of “richness, content, speed it can be delivered – and setting a cost ceiling.” As Whitehurst says, “big, monolithic software just doesn't work anymore – it's too slow, and too expensive.”

Comments