X

Open source: One agnostic's view

Jack MacCrisken says the polarizing argument between conventional private software developers and open-source advocates misses the bigger point.

4 min read
The response of Bruce Perens and nine other notables from the world of open-source software to Microsoft Senior VP Craig Mundie's recent speech will no doubt puzzle those who haven't paid much attention to the subject.

For while the two documents (Read them here: Perens, Mundie) reveal some clear differences of opinion about the role of open source in software development, the group's hostility to conventional ideas about property rights makes it very difficult to define a common ground for debate.

Here's my take on what Mundie is saying:

Protection of innovation through patents and copyrights is a cornerstone to economic development. Thus, while a variety of intellectual property ownership arrangements between private and public entities may be fruitful in furthering technological progress, the relationships created by the General Public License (GPL) concept are highly problematic.

For one thing, GPL denies the primacy of the profit motive in innovative activity. Good software can--indeed has--emerged in an environment in which innovators have no legal claim to the fruits of their own labor. But such cooperation cannot be relied upon to sustain the pace of technological change that is so critical to economic growth. Funding of most businesses requires profits--whether self-funded, venture capital-funded, or Wall Street-funded.

For another, because of its viral nature, GPL cannot coexist easily with private initiative: When GPL code is incorporated in conventionally developed software, the latter loses its property protection. While there may be ways to create barriers against the virus, the emphasis is on "may"--and the resulting uncertainty is a serious deterrent to cooperative software development.

Perens and the other signers of the "Free Software Leaders Stand Together" manifesto counter this with a grab bag of assertions. First they claim there is indeed a conventional business logic to open-source development: It "reduce(s) the cost of software development and maintenance by distributing it among many collaborators."

That is surely true in some circumstances. But in the process of reducing costs, the provisions of the GPL variant on open source undermine individuals' incentives to put in the time and effort. The success of Linux proves that developers don't always need the explicit incentive of a salary or a royalty check to do good work. But it takes a major leap of faith to believe that the complex incentives leading first-rate programmers to participate in GPL projects would serve as well as property rights in driving the development of the software that consumers value most.

Ah, Perens explains, consumers are sure to gain from open source because "it means that the users control the software they use."

Indeed.

One of the appeals of open source is that sophisticated consumers (like the IT departments of large corporations) have the option to modify software to fit their own needs and to fix bugs on their own. But the vast majority of software users are in no position to muck around with source code. And in any case, it is possible to give consumers access to code without saddling the software with GPL. Even the dreaded Microsoft grants such access to some consumers.

Go back for a moment to the issue of the economic viability of enterprises trafficking in what Perens calls free software. He says that the "success of software companies like Red Hat and the benefits to vendors such as Dell Computer and IBM demonstrate that Free Software is not at all incompatible with business."

Major hardware vendors may be able to make money by supporting open-source software on their own machines. And they risk relatively little by investing modest amounts in optimizing the performance of GPL software on proprietary hardware. On the other hand, it is far from clear that companies like Red Hat, whose primary business is developing and supporting free software, will ever figure out a way to break even from ancillary revenue sources.

Red Hat--and a host of other companies identified with Linux--certainly convinced Wall Street that profits were forthcoming: They raised billions of dollars in capital during the go-go years of the late 1990s. But according to the company's latest filing with the Securities and Exchange Commission, "We have not demonstrated the success of our open-source business model, which gives our customers the right to freely copy and distribute our software. No other company has built a successful open-source business."

The most frustrating aspect of the group manifesto is the assumption that if you're not part of their solution, you're part of the problem--that one must choose between Free Software and the Dark Side, as represented by "traditional" software companies such as Microsoft. Open-source software can come in many forms, with varying rights to modify the source code and varying degrees of intellectual property protection. Ironically the Perens group alludes with approval to the success of the Apache Web server, which is developed and maintained cooperatively under a BSD-type open-source license that does not have the viral properties of GPL.

Moreover, as Mundie would probably be the first to admit, the surprising rise of open-source suggests that conventional private software developers have a lot to learn from alternative organizational models. While Perens reflexively dismisses Microsoft's "shared source" initiative as "look but don't touch--and we control everything," it seems like a plausible, good-faith effort at incorporating the benefits of collaboration in software development without undermining the utility of well-defined intellectual property rights.

For some members of the software development community, Free Software is a religion. For the rest of us, the sometime successes of GPL software in an economic culture that puts so much stock in material gain is simply--well, not so simply--a phenomenon worth notice. One can only hope the twain shall someday meet.