In the short history of software, there has been no more important customer than the government, especially in the form of various defense and intelligence agencies. The government sector’s importance continues even today and the posture that government adopts toward open source software can accelerate or retard its success. But just as open source software can be dramatically affected by government policies, the government itself can benefit or be hurt by these same policies.

What’s in it for the Government?

Recent studies have indicated that large organizations can save as much as 30 per cent by adopting open source software solutions because of the reduced licensing costs and also because of needing fewer (though sometimes individually more expensive) support staff. Other studies suggest that control, access, and flexibility are among the key benefits actively pursued by government organizations, especially agencies responsible for security. Outside the US, independence from foreign vendors also has been important in promoting open source software.

Most Governments are proceeding cautiously

Governments are mindful that much of their infrastructure is based upon legacy and proprietary foundations. Their approach has been two-fold: one overt and the other subtle, even insidious.

At an overt policy and political level, many governments have recommended that open source software alternatives be considered only in addition to proprietary solutions. However, at another level, which is reminiscent of the way in which TCP/IP and the Internet evolved, many government organizations are developing and using open source solutions anyway, especially for content and processing services. The January 2003 Mitre Report, “Use of Free and Open Source Software (FOSS) in the US Department of Defense”, demonstrates that many agencies have become dependent on various open source solutions such as the Domain Name System (DNS). According to this report, there has been significant penetration of open source software solutions at all levels of the US Department of Defense. Just as was with the early TCP/IP systems, many practical solutions can be implemented using open source software without having to undertake complex and often prohibitive procurement procedures. It is the irony of Christensen logic that such informal adoption can lay down a new infrastructure and advance the capabilities of the open source software solutions. Certain mature solutions such as DNS, Web services, and, increasingly, security and database services, have already reached a critical threshold of utility and even customer dependency within the government. Despite reluctance to openly embrace non-proprietary software, governments will continue to directly benefit from their informal use of and contribution to open source software.

Government’s Role

As the facilitator of local jobs and economy, government’s embrace of open source software can lead to the creation of high-skill IT jobs. Further, investment in the local software service industry produces an economic multiplier effect. This effect is enhanced by an egalitarian technology transfer shared non-discriminately by all the participants in the open source community and within the various open source software projects. Technology is not bottled up and hidden from the very groups that need access to it to generate new service jobs and business opportunities. Open technology transfer also leads to highly effective educational opportunities in the form of case studies of complex and successful projects and technologies.

Government should encourage healthy diversity!

If there is one lesson in the unforeseen global success of the Internet and the multi-decade triumph of TCP/IP over OSI, governments can neither create nor destroy successful technologies. They can, however, promote the welfare of a healthy and competitive software and technology industry and in turn benefit from increased revenues. Open source software forms an important part of a healthy and growing specialized services industry, especially internationally.

The evidence is mounting that open source software can create jobs, promote technology transfer, provide cost-effective software solutions, and enhance educational opportunities. The clear message to all the governments is that by encouraging the positive forces of open source software and technologies, they and the communities they represent will begin to reap the benefits.

Silicon Valley is known for innovation. Until recently the Valley has had an insatiable desire for hardware and software engineers and for visionary entrepreneurs who could craft this engineering talent into high profile product and market successes. Many Indians succeeded fabulously at all levels of engineering and business ventures in this nirvana of possibilities and optimism. Now that one of the highest status symbols in the Valley has become today just having a good-old-fashioned “job”, what can be more timely than for the Valley to introduce us to the next wave of imagineering.

April’s Embedded Systems Conference in San Francisco marked the signposts in this new era of opportunity. The signposts read “power”, “pervasiveness”, and “affordability”.

Power

Specialized embedded design enables the automation of both new processes and previously analogue processes. The Embedded Systems Conference demonstrated in many shapes and sizes that devices targeted for specific purposes have highlighted vast new levels of automation capability: from automobile control and robotics to communications and entertainment. Specialized digital devices are both easier to apply and more capable than their out-of-date analogue cousins. Further, many specialized hardware devices are now being replaced by more powerful generic platforms. The overwhelming trend suggests that as generic hardware becomes more powerful, specialized devices can be manufactured simply by building the right embedded software. Software, with good development tools, has become the key to producing powerful solutions cost-effectively.

As this show has demonstrated over the past few years, an inexpensive but comprehensive tool chain has been the value proposition of Linux since it began to be a significant player in the embedded devices arena. In addition to excellent intrinsic development tools, specialized tool sets are available from a variety of vendors including Lynuxworks, MontaVista, TimeSys, and IBM.

Pervasiveness

As more and more processes of everyday life become automated, specific purpose devices have become nearly universal. Modern automobiles typically have dozens of smart components using sensors and computers to control fuel injection and emissions, perform diagnostics, regulate temperature, etc. Once designed, cheap embedded devices can appear anywhere needed. Examples include self-configuring networks of smart sensors, process control nodes, factory automation, media gateways, VoIP, power management, controllers, and digital information devices.

But the automation of everyday life is not always cheap: for about a million Yen you can acquire an “almost human” companion based on sophisticated embedded technologies. Mitsubishi and MontaVista Software demonstrated a Linux-based human-size household robot that was a favorite of visitors to the conference this year.

Affordability

Economy of scale, level of integration and specificity of design combine to enable dramatic cost reductions. The vast majority of embedded devices can be produced with few parts, in simple steps and in high quantities so that costs are minimized. Although not all embedded devices will be cheap (as in the companion robot), most simple and widely adopted devices, such as entertainment systems, will remain inexpensive.

The Opportunity

The vast majority of new opportunities for Linux are not for the desktop, although the desktop remains a fertile battleground. Nor are these new opportunities for the server, although Linux will continue to sweep through all levels of server systems from personal servers to large-scale multi-processor and clustered servers. Instead, the overwhelming opportunities for Linux are in this new embedded device and applications space.

Likewise the business opportunities in this new era are for addressing the automation of all aspects of everyday life: computational devices will pervade everyday items from the car to the toaster and everyday activities from money and banking to medicine and travel. The opportunity for the new entrepreneur is to work new levels of automation into the fabric of everyday activities and to capitalize on the potentially large scale distribution of services and products.