Linux is increasingly hard to ignore, but it's only one part of the open source software (OSS) debate.

OSS allows businesses to create standardised working practices, and tailored 'ecosystem' models, so that vertical markets can collaborate in a way that hasn't been possible before.

It effectively creates an 'Internet of Things', where many different devices will collaborate and transact together.

Supporting this requires non-differentiated standardised software which performs many more tasks than the relatively simple content management software handled by the OSS that created the world wide web.

The basic idea behind OSS is simple. When programmers can read, redistribute and modify the source code for a piece of software in an 'open' environment, the software can evolve rapidly and more people can use it as a 'standard'.

As more people use the code, it becomes richer and evolves to become the choice of new users.

Does this altruistic approach work? Yes: it was the model that powered the explosive growth of the web. If you're used to the slow pace of conventional software, the speed of development that OSS supports is astonishing.

For these reasons, OSS is likely to become the normal approach for many web services roles. The quality is usually high because the creative process includes strong peer pressure, urging excellence.

Many professional programmers work on OSS code in their own time, finding it highly rewarding compared with the commercial pressures of their day jobs.

The link with Linux is important. It was the first, and almost only, choice for web servers when the internet exploded.

Most hardware vendors now support a Linux version and, as it becomes more mainstream, it will be used with OSS in generic areas of an enterprise where the need for standardised 'utility' software is based on cost.

It is important to distinguish it from commercial software where value lies in the complex logic, often around a business function that has been developed to provide 'world class' capabilities to enterprises.

It is equally important to understand the licence model. Commercial licences are based on having a lease to use software, whereas OSS licences deal with its reproduction and modification.

Market ecosystems already exist in many industries, where value lies either in exchanging data or in access to pooled data.

These ecosystems are different from internal enterprises, and are based on protocols more than on extensive transaction logic.

Web services are designed to enrich the interchanges, much as TCP/IP was the starting point for the development of steadily richer internet services.

The ability to interact and collaborate within the market ecosystem will result in an increased need for commercial software, offering new forms of business logic.

The Internet of Things is already emerging, usually under the generic description of 'mobility'. One example is radio frequency identification (RFID) tags.

In the retail and logistics market, RFID can create a 'pool' of data that tracks loads or goods. Members of the 'ecosystem', such as shippers, receivers, carriers and retailers, will all require access to elements of the data pool.

The pool belongs to everyone, so it needs standardised ways to interact, collaborate and eventually automate transactions. OSS is likely to prove the most effective mechanism.

Higher level functionality between individual partners, evoked on the basis of the information provided, may well continue to be the province of commercial software.

OSS isn't a magic bullet and cannot be used for every type of software development project. Where conditions are right, its use promises benefits in functionality, cost and time.

But it still needs to be well managed, with attention paid to licensing, and it may have to be integrated with existing software.

Andy Mulholland is global chief technology officer at Cap Gemini Ernst & Young.