On balance, an ACU would risk downloading illegal software rather than spend time learning Linux. Of course, the analysis above assumes that the ACU values Windows at least, if not more than, Linux, and it assumes that there is a lower cost in downloading and installing pirated Windows than in installing and learning Linux. Both of these assumptions most likely withstand deeper scrutiny.

On a more macroeconomic level, “having these pirates means that Linux’s installed base does not grow as much as it would have if piracy weren’t there.” That is, software piracy bolsters network effects that add value to Windows. According to Ghemawat and Casadesus-Masanell, software piracy captures who users would not have bought Windows in the first place because it's too expensive. Thus, piracy converts the deadweight loss into a valuable asset by increasing the strength the network. Even Bill Gates “openly concede[d] that tolerating piracy turned out to be Microsoft's best long-term strategy,” stating that "It's easier for our software to compete with Linux when there's piracy than when there's not . . . .”

Ironically, software piracy hurts free software by hampering its adoption. Thus, the question becomes: How can Free Software fight back? Three solutions come to mind. First, freedom fighters can help Microsoft monitor piracy and enforce copyright law. Unfortunately, this solution is unlikely to reduce piracy by any appreciable amount. By analogy, piracy remains rampant in the music and movie industries despite increased enforcement efforts. Second, freedom fighters can embed viruses into pirated software. The idea is to “poison the well” – raising the cost of using pirated Windows such that Linux becomes a beneficial alternative. In addition to being malicious, this solution is unlikely to inhibit ACUs, as they feel protected by their anti-virus application. The third, most plausible option is for developers to make great Free Software that attracts the ACU more so than pirated Windows. Though some believed that Linux could compete against Windows based on pricing, this model is wrong. Due to the effects of piracy, Linux must out compete Windows pound-for-pound. Although this process has begun, more work is needed to refine Linux so that ACUs see it as a welcome alternative to Windows.

• This point, that unlicensed copying increases the network effect for Windows and helps to exclude competitors, has been noticed before. But in order to decide whether it has any effect on free software, you have to decide whether the point of the movement is to expand the use of the software, or to expand the demand for technological freedom. A user of free software who merely thinks of it as no-cost software, and who neither knows nor cares that it is free as in freedom, is not necessarily "helping the revolution." The fate of Windows and the fate of freedom are not photographic opposites of one another. Right now, Mr Jobs is grabbing eyeballs for a slimier version of "unfreedom developed at freedom's expense" as Windows contracts, which isn't getting us anything even as it subsidizes Mr Jobs' kingdom of ridiculously overpriced hardware.

• In fact, the short- and even medium-term direction of the software industry is less important to us than the long-term equilibrium, which is for software made by sharing because anarchist production of zero marginal cost functional goods is inherently superior. So we no longer have to ask about the fate of the software industry in late Stage One; in Stage Two of our progress, the proprietary industry dies off gradually, with entrenched high-quality proprietary incumbents like IBM and SAP taking much longer to abandon the model than entrenched low-quality incumbents, even of overwhelming size, like Microsoft. In the end, what advances our revolution is that not even Microsoft can avoid paying to distribute freedom.

• So what this essay needed was to go past the relatively easy insight that Microsoft sustains network effects through tacit encouragement of piracy, to something closer to a real analysis of the legal and social outgrowths of the free software movement. I have tried, at longer length using two extensive illustrations, to set out a possible approach.