I’ve told a story of how regulation works and of increasing regulability—of changes in the architecture of the Net that will better enable government’s control. These changes, I have argued, will emerge even if government does nothing. They are the by-product of changes made to enable e-commerce.

That was part one. In this part, I’ve upped the stakes. My aim has been to give a deeper account of the values built into a particular architecture of the Net, and thus a deeper understanding of the ways in which government might act to shape those values.

But now the story changes. I want to introduce a complication on this road to regulability. This complication promises (or threatens) to bring about an important change in the character of the Net and the feasibility of regulating it.

The complication is free software, or open source software or, more simply, open code.¹ Open code has actually been in existance since the beginning of the Internet, it's just more recently that it has come to the forefront of Internet technology. Put too simply, everything I have said about the regulability of behavior in cyberspace—or more specifically, about government’s ability to affect regulability in cyberspace—crucially depends on whether the application space of cyberspace is dominated by open code. To the extent that it is, government’s power is decreased; to the extent that it remains dominated by closed code, government’s power is preserved.² Open code, in other words, can be a check on state power.

This is a lot to convince you of in a single chapter—especially since the conclusion will seem to be an important reversal on much that I have argued so far. To see the point, we must back up and understand a bit more about the nature of the code space that government might regulate and the nature of the actors who might control that space.

Footnotes

¹ I want to sidestep the raging debate about whether to call this movement the free software movement, the open source software movement, or something altogether different. The reality is more important than the label. Activists will work out how best to claim the tradition. My aim is simply to understand the consequences of the struggle. Edit Delete

² By “closed code” I don’t mean anything conspiratorial. I mean simply code that does not reveal its source. A code writer might have many reasons for hiding the source—including economic survival, security, or embarrassment. The crux of my argument is that, regardless of the reason, closed code is not as easily modified as open code. Edit Delete

³ Hunt, TCP/IP: Network Administration, 1–22, 6, 8; Loshin, TCP/IP: Clearly Explained, 13–17. Edit Delete

⁴ There is no standard reference model for the TCP/IP layers. Hunt refers to the four layers as the “network access,” “internet,” “host-to-host transport,” and “application” layers; TCP/IP: Network Administration, 9. Loshin uses the terminology I follow in the text; TCP/IP: Clearly Explained, 13–17. Despite the different moniker, the functions performed in each of these layers are consistent. As with any protocol stack model, data are “passed down the stack when it is being sent to the network, and up the stack when it is being received from the network.” Each layer “has its own independent data structures,” with one layer “unaware of the data structures used by” other layers; Hunt, TCP/IP: Network Administration, 9. Edit Delete