oreilly.comSafari Books Online.Conferences.


What Is Free Software
Pages: 1, 2, 3

Is It Free or Open Source?

As more and more people got involved with free software, the global pool of free programs expanded quickly. Because their source code was open to inspection by anyone, and because they could take advantage of massive parallelism in testing and debugging, many of these programs--particularly programming and networking tools--were of very high quality, and they gradually became part of the internet's basic infrastructure.

At the same time, the free software movement kept getting tripped up by an unfortunate linguistic coincidence: in English, the word free means both "costing no money" and "having liberty." In practice, free software satisfies both definitions: because you can always find someone to share a copy with you, the price of all copies is driven to zero by simple market dynamics. But for the FSF, the second definition, liberty, was the important one. After all, it's possible to have software that is available for no charge yet whose license prohibits redistribution or modification. Such software would not be "free" in the vital "freedom" sense. The FSF kept reminding everyone, "It's free as in freedom, not as in beer," but newcomers to free software still were regularly confused, because the language fails to distinguish carefully between low prices and liberty.

This recurring confusion frustrated a lot of people, and in 1998 a group of programmers came up with open source as a replacement for free software, creating the Open Source Initiative (OSI) to promote the new term. At first it wasn't clear that a schism had opened up in the free software world. But gradually it became obvious that the OSI was advocating a change not merely in terminology but also in attitude. As they saw it, the constant talk about freedom was off-putting to the corporate world, which was slowly beginning to wake up to the advantages of running and supporting free software. The OSI's position was: keep the same licenses, keep the same collaborative practices, but lose the talk of freedom and ideology; that would allow the movement to enter the mainstream and get many more participants and resources.

The OSI may have been correct. At least, it is undeniable that the term open source has caught on very effectively in the corporate world. How much of this is due to a desire to avoid talking about freedom, and how much comes from a marketing sense that the word free is fatally ambiguous, is impossible to say. But under the name "open source," an influx of for-profit dollars has changed the landscape of free software remarkably. Most major free software projects now have corporate backing of one form or another, and many of these companies have become quite adept at working with the volunteer developers who participate in the projects, and the users who report bugs and suggest new features.

The schism, such as it is, is an odd one, representing a profound philosophical disagreement that nonetheless has few practical consequences beyond terminology. Some people still say "free software" exclusively, because they want to remind people that freedom is the important thing. Some say "open source," either because they're not taking a position on the freedom question or because they find "free" too easily misunderstood. Some use the two terms interchangeably. (I'm in that camp, though in this article I've stuck to "free software" to match the title.) Any license that permits the basic freedoms necessary to allow unrestricted development is considered both a free software license and an open source license, and those freedoms, rather than the word used to describe them, seem to be what programmers look for when deciding whether to participate in a project.

The Future of Free Software

Those freedoms are also at the core of the business case for free software, even when the word freedom is not used. Free software offers a promise that few proprietary products can match: the promise that no one can take away from you that which you have invested time in learning and maintaining. When a corporation deploys a piece of software, even just internally, the corporation is making an investment. Employees will have to be trained; various internal processes will have to be adjusted to fit the software; documentation will have to be updated. Over time, important parts of the corporate infrastructure may grow to depend on the software's presence.

The more dependent a corporation is on a piece of software, the more it is at the mercy of the supplier of that software. When the software is proprietary, it means the corporation is at the mercy of the business decisions of a single software manufacturer. In theory, of course, that manufacturer doesn't want to disappoint its customers; it wants to keep them happy. But it might disappoint them anyway, by going bankrupt or by being bought by another company that makes a competing product. Even when those events are unlikely (say, when the supplier is Microsoft), the supplier might still disappoint its customers by failing to do the right usage research, by dropping backward compatibility in order to push customers along an upgrade path that they're not ready for, or by refusing to implement protocols that make it easy to interoperate with competing products.

A free software project, on the other hand, cannot be unilaterally shut down or taken down a wrong path. This doesn't mean that free software authors never make bad decisions; sometimes they do, just like any programmers. But the risk for users is much smaller, because a user who cares enough can simply take out the changes she doesn't like and put in the ones she does. When that user is a corporation, this is not just an abstract ideal but also a realizable practice. A corporation with a decent IT department can spend the effort necessary to customize software to its requirements, even if that means modifying the source code. It's a short step from there to contributing the changes back to others who use the software, and it's usually in the interests of everyone who uses the software to share their changes in common, since that reduces the maintenance burden on any single participant. The fact that no one, not even the original supplier, has the right to remove the software from circulation means that everyone is guaranteed that the effort they have put in so far can never be taken away.

Because of these freedoms and their practical consequences, I expect free software's share of desktop and office installations to continue growing rapidly, and for its current dominance in servers to solidify even further. But as a movement, it has implications beyond just the software we run on our computers. Free software's success, even at this early stage, calls into question some of the fundamental premises of intellectual property. If people will produce complex works of software without the monopoly control given by traditional copyright, will they do the same for books, songs, and movies? Custom tells us that copyright was designed to subsidize creation; but the vitality of the free software scene today, so strongly intertwined with the spread of the internet, hints that copyright may really have served to subsidize distribution, and that as distribution costs go to zero, people will start cooperating on works other than software. The degree to which this will come to pass is still an open question, but the free software movement has, at least, made it a question no one can ignore.

This article is released under a free copyright, the Creative Commons Attribution-ShareAlike 2.5 License.

Karl Fogel co-founded Cyclic Software in 1995, a company offering commercial CVS support. He is the author of Open Source Development with CVS (published by Coriolis), now in its third edition via Paraglyph Press. He has also written Producing Open Source Software, from O'Reilly Media.

Return to

Sponsored by: