Often lost amid the focus on software you don’t have to pay for—such as Linux and Eclipse—is any mention of the organization that started it all: the Free Software Foundation (FSF).1
FSF was founded in 1984 by Richard Stallman. Then a programmer at MIT’s Artificial Intelligence Laboratory, Stallman resigned to protest its restrictive copyright policy. He started the GNU (GNU’s Not Unix) Project,2 an effort to build a free Unix clone, and wrote the GNU (General Public License),3 which essentially said you could redistribute GNU code for free as long as you also gave away any modification you added—and the free software movement was on its way.
The GNU Project’s success can be measured by the widespread use of its tools from the outset by cutting-edge software professionals. The most notable of those offerings include the GNU C compiler, GNU Emacs editor, and GNU symbolic debugger, all of which Stallman, not coincidentally, is principal author.
Today, the GNU Project continues with sorely needed efforts such as Classpath, which seeks to build a set of essential libraries for supporting the Java language, and DotGNU, a multiproject plan to craft a complete replacement for Microsoft’s .NET strategy. Worthy work also surrounds GNU Glue, which aims to create a distributed groupware application framework based on emerging Internet standards such as XML and GNU Octal—a digital-music software effort.
Nevertheless, the GNU Project’s efforts remain less high-profile in the public mind than other open source work such as Eclipse. (No doubt the reader here thought of perhaps an even better known example than Eclipse: Linux. As Stallman points out, however, Linux is itself a GNU project more correctly referred to as GNU/Linux.) And FSF, despite its status as parent to the code-producing GNU operation, is in many ways a political organization. Stallman himself acknowledges as much. “We developed GNU starting in 1984 as a campaign for freedom, whose aim was to eliminate non-free software from our lives,” Stallman told me in an e-mail exchange.4
All of which raises the question of whether GNU will remain an important source of software for tool users. “I believe GNU is and will continue to be technically relevant,” David Sugar told me.5 Sugar, a 20-year free-software veteran, is the chair of the DotGNU screening committee and principal author of the GNU Bayonne telephony server. “The best examples of cutting-edge development in GNU that come to my mind are Portable.NET and DotGNU, where new and original concepts in virtual machine implementations and Web services continue to be experimented with.”
But, as with many FSF adherents, Sugar believes that the whole of the effort is more valuable than the sum of its individual projects. “If you are asking if a specific GNU package is better or worse than a [non-GNU] package, then I think you are asking the wrong question,” Sugar noted. “The technical value and relevancy of GNU is of GNU as a whole, not of the individual packages in isolation.”
According to Eben Moglen, a professor at Columbia Law School who is known as the FSF’s legal expert:6
“ On the subject of the GNU Projectrelevance, I would say that technical vitality is in general best measured by breadth of use. Many components of GNU have no meaningful competition; others compete against both close substitutes and industry analogues on very favorable terms. I see no sign that GNU Emacs, GCC [GNU C Compiler Collection], GDB (GNU Project Debugger), Glibc, bash, gzip, GNU tar, and other primary components of GNU have become technically irrelevant. Ongoing development activities may or may not result in equally significant future products; time will tell. Certain products, like GCC, are unlikely to move to the commercial-toolbed context: In compiler development, we are essentially the honest-broker managers of an industry consortium.
When I queried FSF founder Stallman as to whether FSF is still politically relevant now that we have widely available open-source toolbeds such as Eclipse, he replied that the question was “a non sequitur, like asking, ‘Now that we have a hospital, is the fire department obsolete?’”7
If Stallman sometimes seems to be an enigma wrapped inside a software zealot, he takes great pains to set precise parameters for any discussion of FSF. He emphasizes that FSF supports free code, but has nothing to do with open source. To some, this may seem confusing. But open source and GNU are not synonymous. Programs written by the GNU project are indeed “free,” and they’re covered by the GPL “Copyleft” licensing agreement, which allows them to be redistributed provided that the source code for any changes made by the user doing the redistributing are included. Open source programs do indeed tend to be offered under often strikingly similar terms, but they don’t come from the GNU folks, nor do they have the exact same license. Accordingly, open source software can perhaps best be thought of as an offshoot of GNU. According to Stallman, the open source folks are a “different group which was formed in 1998 specifically to reject the idealism of the Free Software Movement that I founded.”
With regard to Linux, which is the most successful spawn of FSF, Stallman is in something of a dilemma. On the one hand, Linux would seem to be a powerful case in point for free software. In the public mind, however, Linux is an “open source” project associated with Linus Torvalds, not FSF.
The result is that Stallman may seem to have a case of sour grapes when it comes to Linux, since he spends so much of his time harping on the nomenclature issues that have become a particular sticking point for him. “One common error is calling the whole [Linux] operating system Linux,”8 he wrote in response to my recent e-mail requesting an interview for this column. “The system is basically GNU; Linux is actually the kernel, one program in the system. When people call the whole system Linux, they give the system’s principal developer none of the credit. A related error is the idea that GNU is a collection of tools; GNU is an operating system.” (Readers should know that Stallman declined to participate in my interview when I was unable to agree to a lengthy list of definitional ground rules, including referring to Linux as “GNU/Linux” at all times—using Linux only if specifically referring to just the kernel.)
While Stallman bangs the drum for credit on Linux, FSF may find itself drawn into an incipient battle: that of the legal rights surrounding free and open-source software. The issue has moved to the fore recently. Specifically, SCO recently sued IBM for $1 billion, claiming that IBM misappropriated SCO Unix trade secrets by putting some Unix code owned by SCO into Linux. More generally, the lawsuits by the recording industry against file-sharing downloaders have heightened legal concerns throughout the software world.
Stallman has manned the barricades in the SCO battle, distributing a statement to news organizations charging that SCO is conducting a “smear campaign” against Linux. “Our community cannot be defeated by this,” he vowed.9
Columbia professor Moglen sees FSF as “indispensable” to the fight for software legal rights. “When Microsoft’s Craig Mundie accused FSF last June of ‘destroying the global software industry,’ he was paying the Monopoly’s [Moglen’s term] respects to the effectiveness with which we perform our other functions, which are directed at breaking the paradigm relationship of ownership to technological innovation,” Moglen noted.10
The Monopoly and other “owners” [Moglen’s quote marks] have temporarily deluded the world into thinking that the free exchange of ideas, which has so successfully characterized Western science since the sixteenth century, is an inappropriate basis for technological, as opposed to basic scientific, progress. Our goal is to provide a mixed theoretical and practical demonstration that freedom is the only morally satisfactory basis for innovation in software, and beyond software in other forms of human intellectual effort. Far from being irrelevant, we are more important every day.
Moglen sees pervasive digitalization as creating an economy where information has a “zero marginal cost.” Old-line software and entertainment companies are trying to hold on to the monopolies that existed when software was distributed on disks and movies weren’t available outside of theaters. If they succeed, he contends, “the result would be vast profit margins for them, at the expense of exclusion for those who cannot afford knowledge and culture” at artificially high prices and “who ought to have free access to music, art, learning, software, and bandwidth.”
1. The Free Software Foundation (in 19 languages): see http://www.fsf.org.
2. The GNU Project: see http://www.fsf.org/gnu/thegnuproject.html.
3. GNU Classpath: see http://www.gnu.org/software/classpath/classpath.html.
4. Stallman, R. Private e-mail exchange with author (Oct. 2003). For more on this explanation, see http://www.gnu.org/gnu/linux-and-gnu.html.
5. Sugar, D. Private e-mail exchange with author (Oct. 2003).
6. Moglen, E. Private e-mail exchange with author (Oct. 2003).
7. Stallman, R. See reference 4.
8. Stallman, R. See reference 4.
9. Stallman, R. GNU founder calls SCO Linux suit a smear campaign, EmbeddedWatch.com (June 26, 2003); see http://www.embeddedwatch.com/stallman.htm.
10. Moglen, E. See reference 6.
ALEXANDER WOLFE received his electrical engineering degree from Cooper Union in New York City. A science writer based in Forest Hills, New York, he has contributed to IEEE Spectrum, EE Times, Embedded Systems Programming, and Byte.com.
Originally published in Queue vol. 1, no. 9—
see this item in the ACM Digital Library
Marshall Kirk McKusick, George V. Neville-Neil - Thread Scheduling in FreeBSD 5.2
A busy system makes thousands of scheduling decisions per second, so the speed with which scheduling decisions are made is critical to the performance of the system as a whole. This article - excerpted from the forthcoming book, "The Design and Implementation of the FreeBSD Operating System" - uses the example of the open source FreeBSD system to help us understand thread scheduling. The original FreeBSD scheduler was designed in the 1980s for large uniprocessor systems. Although it continues to work well in that environment today, the new ULE scheduler was designed specifically to optimize multiprocessor and multithread environments. This article first studies the original FreeBSD scheduler, then describes the new ULE scheduler.
Bart Decrem - Desktop Linux: Where Art Thou?
Linux on the desktop has come a long way - and it's been a roller-coaster ride. At the height of the dot-com boom, around the time of Red Hat's initial public offering, people expected Linux to take off on the desktop in short order. A few years later, after the stock market crash and the failure of a couple of high-profile Linux companies, pundits were quick to proclaim the stillborn death of Linux on the desktop.
Jay Michaelson - There's No Such Thing as a Free (Software) Lunch
"The licenses for most software are designed to take away your freedom to share and change it. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change free software to make sure the software is free for all its users." So begins the GNU General Public License, or GPL, which has become the most widely used of open source software licenses. Freedom is the watchword; it's no coincidence that the organization that wrote the GPL is called the Free Software Foundation and that open source developers everywhere proclaim, "Information wants to be free."
David Ascher - Is Open Source Right for You?
The media often present open source software as a direct competitor to commercial software. This depiction, usually pitting David (Linux) against Goliath (Microsoft), makes for fun reading in the weekend paper. However, it mostly misses the point of what open source means to a development organization. In this article, I use the experiences of GizmoSoft (a fictitious software company) to present some perspectives on the impact of open source software usage in a software development shop.