FOSS is an abbreviation for Free and Open Source Software. In other words, FOSS is software whose source code is openly available. People can install and even modify FOSS as they please, so long as they follow a few basic requirements listed in the license. This arrangement makes FOSS the opposite of proprietary software, and one of the most original developments in the history of IT.
FOSS is a combination of two terms, free software and open source. Both free software and open source refer to software that is licensed in the same way, but the separate terms imply a difference in the reasons for the licensing.
For most free software supporters, the licensing is a way to ensure software freedom, or the ability of users to control their computers and their contents. By contrast, for most open source supporters, the licensing is a way to improve the quality of software. The open source argument is that, because the source code is available, bugs will be more easily discovered — or, as Eric S. Raymond put it, “with enough eyes, all bugs are shallow.”
However, this distinction is not as clear cut as it seems. To start with, some open source advocates such as Linus Torvalds have been known to argue that quality software is simply the means of ensuring software freedom. For another, the first people to use the open source label deliberately distanced themselves from free software, a decision that has left lasting animosities.
And finally, just to confuse the issue, some people insist that the terms are completely interchangeable. Certainly, individual and corporate memberships in the Free Software Foundation and the Open Source Initiative often overlap. Unofficially, “open source” tends to be used in a business context, but even that generality sometimes fails.
FOSS is used by those who want to avoid any arguments that using either free software or open source might provoke. Sometimes, because free software is often taken to mean shareware or software that is available at no cost, FLOSS is used instead, standing for Free/Libre and Open Source Software to emphasize that “free” is being used strictly in a social or political sense.
The use of FOSS or FLOSS is justified by the fact that, despite the often violent arguments, free software and open source software developers and users have far more in common than they have differences. They share the same history, and often share the same software licenses, community and ideals as well — to say nothing of the same failures and successes as the mainstream IT world discovers them.
Although FOSS and its component terms are less than thirty years old, the concept behind them is almost as old as computing itself. In fact, FOSS could be considered a specialized example of the academic belief that a free exchange of ideas is mutually beneficial.
Throughout the first few decades of computing, sharing and modifying source code was a common practice. For example, the SHARE user group was founded in 1955 to exchange technical information and source code modifications for IBM mainframes. Code was similarly exchanged by users of the IBM 7090/7094, and continues to be shared by retro-computing hobbyists. However, this practice began changing in the 1970s as computers became commercial products.
The origin of modern FOSS is usually traced to the early 1980s, when Richard M. Stallman encountered proprietary practices in his work. Stallman’s moral outrage over the change and its affect on his professional work caused him to create the GNU Project in 1983, with the intention of building a free operating system. In 1985, Stallman founded the Free Software Foundation to advance his goal.
The GNU Project and Free Software Foundation continued their work through the 1980s, releasing utilities for a free operating system such as GNU Emacs, a text editor that Stallman has personally maintained off and on ever since, as well as the GNU General Public License, the most commonly used license in FOSS.
By the early 1990s, the only major piece needed to achieve the goal of a free operating system was the kernel. This piece was supplied when Linus Torvalds, a Finnish computing student, released his Linux kernel under the GNU General Public License. Today, the resulting operating system is often called Linux, although free software supporters prefer to call it GNU/Linux to recognize the fact that the result was a joint effort. A project called the Hurd continues to this day to develop a separate GNU kernel, but is no longer a priority for the majority involved in FOSS.
The 1990s also saw the emergence of dozens of other FOSS projects, including Apache, Samba, Mozilla, and The GIMP and other standards of modern FOSS computing. Many of the projects that emerged in this period made fundamental contributions to building and popularizing the Internet, giving rise to the often-repeated claim that FOSS built and runs the Internet.
In 1998, a group that included John “maddog” Hall, Larry Augustin, Michael Tiemann, and Eric Raymond met in Palo Alto, California. According to Tiemann, “The conferees decided it was time to dump the moralizing and confrontational attitude that had been associated with ‘free software’ in the past and sell the idea strictly on the same pragmatic, business-case grounds that had motivated Netscape” to recently release its own source code. The term that the group started publicizing was “open source.”
The change of terms may have worked as intended, helping business to consider FOSS seriously. Or perhaps FOSS had simply reached a level of complexity where it could no longer be dismissed as a fringe idea and was starting to rival its proprietary counterparts. At any rate, during the Dot-com craze of 1998-2000, FOSS became the basis of countless start-ups. The same era also saw the first FOSS-based companies going public, such as Red Hat and VA Linux. Many established companies, such as IBM, Oracle, and Sun Microsystems, also began to participate in FOSS.
FOSS was briefly discredited by the Dot-com collapse, but continued to gain in popularity in both business and personal computing during the first decade of the millennium. Once developed largely by volunteers working in their spare time, today FOSS is so widely seen as an IT asset that 75% of the contributions to the Linux kernel are made by developers who are paid by corporations. Today, even Microsoft, whose executives used to claim that FOSS was un-American and a cancer sees benefits in at least paying lip service to FOSS.
These licenses have two chief concerns: specifying the rights of people to use and modify the software, and explaining how modified works based upon the software can be distributed — whether under the same license or a different one. Almost all have the minimal requirement that modified works must include a notice crediting the writers of the original software.
These lists of licenses overlap considerably, and, if a license does not appear on both, the chances are that the omission is an oversight. Conversely, if a license appears on both lists, you can be confident that it is a FOSS license.
An important distinction in FOSS licenses is between copyleft and non-copyleft licenses. Unlike a conventional copyright license, which is designed to restrict use, a copyleft license encourages the use of the software by others, so long as they offer their modifications under the same license.
By contrast, a non-copyleft license allows those who use the software to re-license under any terms they want. This general provision makes non-copyleft licenses popular in businesses, but makes these licenses second-best from the viewpoint of free software advocates.
The most popular copyleft license is the GNU General Public License (GPL). According to Black Duck Software’s tracking of FOSS licenses in September 2009, GPL in its various forms accounts for just under 64% of all FOSS licenses. That is no small percentage, considering that dozens of licenses are available.
The second version of the GPL (GPL2) is the most popular variant, with over 48% of the market. It remains the license of choice for many projects, including the Linux kernel.
However, in 2007, after extensive consultation with the FOSS community, the Free Software Foundation released the third version of the GPL (GPL3) to discourage new ways to circumvent the intents of the license, such as Digital Rights Management, patent protection and “Tivoization” (hardware that blocks the use of modified code). To date, GPL3 remains less popular than GPL2, with its use being strongest among free software advocates.
Another popular GPL variant is the Lesser GNU General Public License (LGPL). Also available in versions 2 & 3, the LGPL differs from the GPL in that it allows the linking of software licensed under it to non-GPL software. Originally designed for use with programming libraries, the LGPL is also a popular choice for dual-licensed software by commercial companies. By contrast, the Free Software Foundation considers the LGPL a compromise that should be used sparingly.
With the rise of cloud computing, the Free Software Foundation has also released the GNU Affero General Public License (AGPL). Its major difference from GPL3 is that it contains a requirement that if you offer software offered as a service, you must make its source code available. So far, this license is little used.
At the other extreme from the GPL family of licenses are permissive, non-copyleft licenses. The most widely known of these licenses are the Expat or MIT License, and the BSD-style Licenses. These licenses differ from any GPL variant in that they place no restrictions on the licensing for modified works. Consequently, they are short and easy to understand.
Other free licenses fall somewhere between the extremes of the GPL and Expat / BSD-style licenses.
Most free licenses have never been legally tested, but several — notably the GPL — have been successfully used as the basis for legal action by the Software Freedom Law Center on behalf of its clients. In addition, the Free Software Foundation maintains a licensing and compliance lab that informs users of violations, and assists them with coming into compliance. These activities tend to strengthen the legal status of FOSS licenses, especially since — despite the occasional complaints — even the strictest of them grant users far more rights in the software than proprietary licenses have ever done.
Community values are not part of the formal definition of FOSS or any of its aspects. All the same, several values and assumptions have been associated with FOSS for so long that not to mention them would be a distortion. Many of these values were first articulated fully by Eric S. Raymond in The Cathedral and the Bazaar, as well as his other writings.
FOSS development frequently challenges conventional ideas about human motivation, competition, and the sources of innovation. For instance, the experience of countless projects suggest that recognition for work well-done can be a major motivation, encouraging even paid developers to efforts far beyond their job descriptions.
Similarly, companies involved in FOSS discover that, in certain circumstances, co-operation with direct rivals is mutually beneficial. The Linux Foundation exists precisely for this reason. If you look at its list of corporate members, you will see that it includes many companies that compete in the same markets, such as Red Hat and Novell and Hewlett-Packard and Dell.
In addition, there is a striking similarity between Raymond’s statement that “Every good work of software starts by scratching a developer’s personal itch” and leading academic studies about the sources of innovations. In particular, the norms of the FOSS community parallel closely Erich von Hippel’s discovery that users, not corporations, can be a major source of technological innovation. In fact, when von Hippel discovered the FOSS community, he immediately identified it as a major proof of his research, much of which had been done before the start of the GNU Project.
Within computing, FOSS is most closely identified with the distributed development model — the co-operation of individuals in different locations. Teams working over a distance are not new, but, FOSS, arising at the same time as the Internet, has taken distributed development to previously unheard-of heights, developing methods for submitting modifications, evaluating them, and even testing them without most of the contributors ever meeting each other. Admittedly, many FOSS projects see the value of face-to-face meetings, either for intense developer sprints or large scale conferences such as GNOME’s GUADEC, but long-distance collaboration continues to be the norm in most FOSS projects.
Other values traditionally associated with the FOSS community include support for open standards and formats, non-hierarchical organizations, an emphasis on quality, and an accompanying disregard for deadlines. In some FOSS circles, too, a hatred of Microsoft might be said to form of negative identity — a means of identifying who belongs by what they reject and dislike.
Originating as a reaction to the rise of software as a product, FOSS has not always interacted smoothly with business. FOSS supporters often regard businesses that become involved in FOSS as opportunists that are eager to benefit from community-based development while contributing as little as they can. For their part, business executives often find the FOSS community overly-idealistic and its practices as hostile to business. Both views have some justification, especially since the Dot-com era.
Yet despite this traditional antagonism, business and FOSS have become increasingly close allies in the last decade, with many companies or executives becoming heavily involved in the community.
Today, FOSS appeals to venture capitalists and business executives in several ways. It is now seen as a source of disruptive technologies, potentially risky investments, but ones that, if successful, can mean a greater return on investment.
Just as importantly, by using FOSS, startup companies can reduce development time and bring products to market sooner than if they have to do all their development themselves. According to longtime FOSS venture capitalist Larry Augustin, the average time for a return on an investment is five to seven years, while for FOSS-based companies, it can be as little as three or four years. As As Lisa Lambert of Intel Capital summarized, “Speed to market, speed to revenue.”
Yet another advantage of FOSS is that it not only allow companies to compete with smaller staffs, but also to exploit market niches that would be too expensive to be worth developing by proprietary means. For instance, developing a generalized social networking site to compete directly with Facebook would likely seem so risky that few would try or invest in such an effort. However, thanks to FOSS, a more specialized social networking site might have a chance of being profitable.
Still another sign of business’ growing acceptance of FOSS is the idea of crowd-sourcing — the attempt to involve customers in product development and promotion. Yet, although obviously based on the FOSS development model, crowd-sourcing is rarely as successful as FOSS development. Unlike FOSS development, crowd-sourcing does not de-centralize activities, since the sponsoring companies remain firmly in control. Nor is the credit for ideas always rewarded to the same degree that it is in FOSS. All the same, the attempt to emulate indicates just how much fascination FOSS holds in some business circles.
One of the most obvious signs of FOSS’s influence is the number of movements that have adopted its principles to their own concerns. Some of these other movements are natural expansions of FOSS, such as the the OpenBios Project, whose goal is to extend the idea of FOSS to the firmware used on computer motherboards and other devices so that all the software used on a computer is politically free.
Similarly, the term “open hardware movement” is sometimes used to describe various groups who apply FOSS principles to hardware design and manufacture. For instance, Bug Labs sells a variety of modular hardware appliances, and encourages purchases to exchange ideas and plans about how to use them. CEO Peter Semmelhack describes Bug Lab’s goal as “to create a platform that allows anyone to build any project they want by snapping it together like Lego.”
One of the earliest and most successful FOSS off-shoots is the free culture movement, which (so far as it can be credited to a single person) is usually attributed to Harvard law professor Lawrence Lessig. Lessig’s books such as The Future of Ideas and Free Culture provide the philosophical underpinnings for the free culture movement.
Lessig is also one of the founders of Creative Commons, the central organization of the movement. Creative Commons is best known for its series of licenses that allows artist to declare the conditions under which their work can be reused.
Another major off-shoot of FOSS is the open access movement in academia. Concerned about the monopoly in textbook publishing and the growing cost of journal subscriptions at a time when library budgets are being continually reduced, members of the open access movement have been working for the last decade to encourage the public availability of academic research.
In the United States, a public face of open access is the Alliance for Taxpayer Access, which is particularly concerned about access to research that is publicly funded. Yet, if anything, the movement is even more popular in developing nations, where academic institutions cannot afford the journal subscriptions needed to foster first rate academic programs.
Although such movements have different goals, what they share with each other and with FOSS is a belief in free access to information and a de-centralized structure that challenges the existing order or way of doing things.
FOSS requires no registration, and statistics about its usage are scattered between projects. Consequently, exactly how prevalent FOSS is can only be estimated.
In some areas, FOSS seems to have become become a dominant force. As of May 2010, Netcraft estimates that Apache has a 55% share of the server market — a figure 10-15% lower than the ones given over the last decade, but still over twice that of Microsoft. Similarly, Mozilla estimates its Firefox browser to have about a 30% market share, largely at the expense of Microsoft’s Internet Explorer.
FOSS is also recognized as being a major player in the embedded market. Although figures are hard to come by, Android’s capture of 28% of the mobile market in less than six months gives some indication of FOSS’ position in the market.
The only area in which FOSS is generally considered to be weak is on the consumer desktop. In 2009, NetApplication announced GNU/Linux’s market share as slightly over 1%. If so, then, the FOSS operating system might be considered a victim of its own success behind the scenes — when GNU/Linux is so skilled at reacting with proprietary operating systems like Windows, perhaps the incentive to change desktops is reduced.
However, this figure has been criticized as being inaccurate and as being based too much on North American usage. Taking into account adoption by developing countries like Brazil, other estimates range from 1-5% to a high of 8-12%. The trouble is that all estimates require so many assumptions that their subjectivity quickly becomes obvious.
That said, few people are betting against FOSS in the near future. In 2008, the Gartner Group estimated that, by 2013, 90% of businesses will use FOSS either directly or in embedded devices. By the same year, IDC predicted, FOSS adoption among businesses would grow by 22% each year to reach $8.1 billion.
As Matthew Aslett pointed out, the greatest indication that FOSS has become an accepted part of the world’s IT infrastructure is that most analysts no longer dismiss it or see it as a transitory phenomenon. Instead, most analysts predicted a mixed infrastructure in which FOSS and proprietary software battle for market share on almost equal terms.
Currently, the largest challenge to FOSS appears to be cloud computing. In the past, FOSS’s major challenges have been efforts to reduce its effectiveness, such as patents. What is different about cloud computing is that it shares one of FOSS’s main advantages: the ability to deliver services to users at no cost. But, unlike FOSS, cloud services remain in the control of their owner, which might make them more appealing to executives educated with traditional business models.
Still, even this advantage might do nothing more than continue to stall FOSS on the desktop and ensure that predictions that a coming year will be “The Year of the Linux Desktop” continue to be a running joke. Issues of availability and privacy may be enough to ensure that cloud computing remains far less important than analysts are presently predicting. In the end, cloud computing may be just another player in a mixed market, exactly like FOSS.
Behind the scenes in the corporate world and in the community at large, FOSS is sure to remain a major feature of the IT landscape for years to come. Even in the unlikely event that the use of FOSS does not continue to grow, it seems likely to remain a major part of the mix of computing technologies, as well as a major academic and social influence.