Communications of the ACM

Practical programmer

A Look at the Economics of Open Source

By Robert L. Glass
Communications of the ACM, February 2004, Vol. 47 No. 2, Pages 25-27
10.1145/966389.966409
Comments

In the previous instantiation of the Practical Programmer column (Nov. 2003), I looked at the sociopolitical aspects of the open source movement. In this column, I examine its economic aspects. Because open source is most prominently about building software products for no financial compensation, the economic aspects are perhaps the most important to analyze and understand.

The open source movement, where people perform their software development activities principally for the accolades of their peers, seems terribly noble, and in fact that nobility is undoubtedly part of its appeal.

It also seems faintly utopian, as we saw in the previous column. There have been many utopian movements in the past, where workers banded together to work for the satisfaction of a job well done and for the common good (and, once again, for the accolades of their peers). There are two interesting things about utopian movements: They typically begin in enormous enthusiasm and they end, usually a few decades later, in failure. What's the most common cause of utopian failure? The impractical nature of the economic model (which I discuss here) and political splintering (I discussed that—seemingly unlikely—possibility in the November column).

So, regarding that economic practicality issue, just how impractical is the open source movement? To date, the answer would appear to be there is no sign of the movement's collapse because it is impractical.

There is little definitive evidence regarding the size of the movement, but the frequency of its mention in the computing popular press would tend to suggest it is growing. Its advocates are also increasingly outspoken. And companies have sprung up that, while not making money on open source products (which would be a violation of at least the spirit of the movement), are making money on servicing those products (for example, Red Hat).

Then there is the issue of communism, an issue that is usually present in discussions about the problems of open source, although it rarely surfaces. There is a faint whiff of communism about the concept of working for no financial gain. Open source is certainly not about "from each according to his ability, to each according to his need," so that whiff is indeed faint. But the sense of nobility that open source proponents feel, in working for no financial gain, resonates with some of the other basic communist philosophies. And the open source proponents themselves can sometimes sound just like communists. One columnist recently spoke of anti-Linux forces as "capitalist interests," and later as "capitalist forces." He also claimed some anti-Linux people are behaving as "Microsoft lackeys." While opposing capitalism and using the word "lackeys" is not proof the author of that column is communist, the rhetoric he chose to use certainly reminds one of communist rhetoric. Whether communism is a good thing or not is, of course, a determination that must be made by the reader. But in this discussion of the practicality of the open source economic model, it is worth noting that the communist system is in considerable decline and disfavor in today's world.

It is particularly interesting that advocates of open source refer to "the cathedral and the bazaar" in their discussions of the movement and its alternatives [2]. In that view, open source represents the bazaar, a place where people freely trade their wares and skills, and the proprietary movement is represented by the cathedral, a bricks-and-mortar institution with little flexibility for change. I find that especially interesting because, when I first saw this particular analogy, I assumed open source would be the cathedral, a pristine and worshipful place, and proprietary software would be the bazaar, where products and money change hands, and there is a strong tendency toward working for profit. I suppose that those who invent analogies are entitled to make them work in any way they wish. But my own thinking about this pair of analogies is that the open source way of viewing it is fairly bizarre!

Does the open source economic model show any long-term promise of working? There is no evidence at present that it will not, but on the other hand it is worth noting that the analogies we can draw between it and other relevant economic models are primarily about models that eventually failed.

Back to Top

The Future...and the Past

Given the critical importance of the economic model to the open source movement, let's spend a little more time looking back at the economic past of open source, and then look ahead to its potential economic future.

First, let's look back. In [2], the author dates open source back to "the beginnings of the Internet, 30 years ago." That doesn't begin to cover open source's beginnings. It is important to realize that free and open software dates back to the origins of the computing field, as far back as the 1950s. Back then, all software was free, and most of it was open.

Software was free because it hadn't really occurred to anyone that it had value. The feeling back then was that computer hardware and software were inextricably intertwined, and so you bought the hardware and got the software thrown in free. And software was open because there was little reason for closing it—since it had no value in the marketplace, it didn't occur to most people in the field that viewing source code should be restricted. There were, in fact, a wide variety of thriving software bazaars, where software was available for the taking from user organizations like SHARE, and the highest accolade any programmer could receive was to have his or her software accepted for distribution in the SHARE library, from which it was available to anyone in the field.

Free and open software remained the rule into the mid-1960s, when anti-trust action against IBM by the U.S. Department of Justice first raised the issue of whether the so-called bundling of software with hardware constituted a restriction of trade. Eventually, IBM unbundled hardware and software, and—for the first time—it was possible to sell software in the open marketplace. However, IBM—it was widely believed at the time—deliberately underpriced its software to inhibit a marketplace for software, which might enable computer users to move beyond IBM products into those of other hardware vendors (who had historically not offered as much bundled software as IBM). Thus, even when software was no longer free and open, there was still not much of a marketplace for it.

It was a matter of another decade or so before the marketplace for software became significant. Until that happened, there was a plethora of small software companies, not making much money, and nothing like today's Microsoft or Computer Associates.

Whether all of that was a good thing or a bad thing can, of course, be considered an open question. But for those of us who lived through the era of software that was free and open because there were no alternatives, a return to the notion of free and open software feels like a huge regressive step. However, there aren't many of us old-timers around anymore, so I suppose this argument is of little interest to the issue of the future of open source as seen from the 21st century.

Because of all of that, let's return to a discussion of the economic future of open source. Advocates tend to make it sound like open source is without question the future of the software field, saying things like "the open-source age is inevitable," while chiding Microsoft (the primary putative enemy of open source, as we have already seen here) for sticking to the "buggy whip" proprietary approach [1].

Open source advocate Eric Raymond goes even further, saying things like "Windows 2000 will be either canceled or dead on arrival," "the proprietary Unix sector will almost completely collapse," "Linux will be in effective control of servers, data centers, ISPs, and the Internet," and—ultimately—"I expect the open source movement to have essentially won its point about software within the next three to five years." And then he proposes a blueprint for fighting the war that he sees necessary to make it so, with things like "co-opting the prestige media that serves the Fortune 500" (he names, for example, the New York Times and the Wall Street Journal), "educating hackers in guerrilla marketing techniques," and "enforcing purity with an open source certification mark" [2]. It is important to realize, of course, that we are well into his predicted 3–5 year future, and none of those things have happened. Clearly, open source zealots are going to have to readjust their timetables for the future, if not give up on it entirely.

There are other views of software's future, of course. Some of those views I have expressed previously, suggesting that open source, far from being software's future, may be a passing fad or a utopian-like dream. Others simply go on about their business ignoring open source, for the most part, participating in the proprietary software marketplace with only an occasional uneasy glance over their shoulders. Still others, looking for a safety play, are betting on both open source and proprietary software, developing products consistent with both approaches (ironically, IBM is one of those companies). But perhaps my most favorite view of the future of the software field comes from Derek Burney, CEO (at the time of this pronouncement) of Corel, a company that had elected to host its future tools development work on the open source Linux operating system. Responding to a question about developing open source versions of Corel's WordPerfect software suite, he said "We have no plans to do so." Then he added "In my opinion, open source makes sense for operating systems and nothing more."

The economic future of open source? It could range anywhere from "the inevitable future of the software field" to "it's only good for operating systems, nothing more" to "it's in all probability a passing fad."

Inevitable future or passing fad? No matter how you slice it, the notion of open source software has certainly livened up the software scene, circa 2003 and beyond.

Back to Top

Back to Top

Back to Top

The Digital Library is published by the Association for Computing Machinery. Copyright © 2004 ACM, Inc.

---

No entries found