Mistaken Premises

All page numbers are from the PDF link sent to the class.

In The Performativity of Code: Software and Cultures of Circulation, Adrian Mackenzie attempts to describe the role of the Linux kernel as a cultural object. He walks through much of the history of Linux, and examines how different aspects of its development may have affected how it developed into the cultural force it is today.

While he makes many valid, interesting points, much of his analysis is flawed due to inaccurate or incomplete background research.

One major component of Mackenzie's analysis of Linux revolves around its licensing scheme, under the so-called "copy-left" (as opposed to copyright) GPL. Mackenzie cites how the GPL "prevents the software itself being sold," (Mackenzie 6) which is a common misinterpretation of the GPL. The GNU Project, the group responsible for the GPL, has a page which addresses this:

"Free software is a matter of liberty, not price..."

Mackenzie repeatedly alludes to the fact that there is no money necessarily involved in acquisition and use of Linux, but I question what portion of his conclusions are affected by the fact that working on free software challenges only notions of accessibility, not necessarily economics.

Mackenzie also ascribes great weight to the fact that Linux started out testing some i386-specific hardware (18), and while he does not explicitly state it, he implies strongly that Linux was the first to do so. However, Microsoft of all companies was selling its own version of UNIX - Xenix - which ran on the i386 processor, 5 years before Linux was conceived. So it was not Linux which brought UNIX to commodity hardware. Linux certainly made it more affordable, but was not the first to make UNIX on a widespread hardware platform a new possibility.

Similarly, Mackenzie also stresses how Linux's plethora of differently branded distributions differentiate it(11). He alludes to UNIX having a similar past when he mentions how the lack of official support for UNIX "forced users to share with one another" (16). But he neglects to explore the depth of this influence, as not only did UNIX users share with each other, but UNIX implementations splintered in various ways, similar to the divides between various Linux distributions. The most significant fork eventually became its own operating system, Berkeley UNIX (also simply called BSD), which has since evolved into a number of others: notably FreeBSD, OpenBSD, NetBSD. Reading the names alone hints at the philosophical underpinnings of the branches. It has often in fact been theorized that were it not for a lawsuit against Berkeley over BSD (around the same time as Linux's inception), BSD would have held the place currently possessed by Linux: http://en.wikipedia.org/wiki/USL_v._BSDi

Despite these shortcomings, Mackenzie accurately detects the cultural icon that Linux (or more generally openly available kernel code) represents. I have a copy of a book, Lion's Commentary on UNIX 6th Edition, which contains the full source code and explanation for 6th Edition UNIX. The book was originally a pair of manuscripts produced at one university, and then circulated for years via photocopy, and it is a fond memory for many experienced kernel hackers. I'll bring it to class tomorrow. It's interesting to me to see things similar to this (verbal readings of Linux source), which I simply absorbed in the past, drawn out as significant trends in a larger culture.