Ryan's critique

From CyberOne Wiki
Jump to navigation Jump to search

Saied Pinto's project, “An Inconvenient Choice,” is aimed at convincing Windows “power users” to switch to Linux. He argues, primarily, that an investment of time and effort in switching to Linux would produce a continuous stream of future payoff in the form of efficiency and enjoyability. His blog posts are structured as a narrative chronicling his switching experience, in an attempt to make himself into an example with which readers can identify. This provides him with an ideal window through which to address the rational, as well as irrational, concerns of the prospective switcher. Although “An Inconvenient Choice” deals indelicately with some of the finer points of Linux adoption, it effectively allays the most important fears likely possessed by members of his target audience.

The narrative format is a good one, and not uncommon in Linux advocacy. However, its major problem is that prospective converts often have trouble identifying with Linux advocates – even the more moderate ones – because they are assumed to be computer scientists or at least geeks. Saied went out of his way in his first post to make clear that he does not have a computer science background aside from autonomous fiddling. Indeed, Saied is the only Linux user I know who is not either in possession of or currently pursuing a degree in computer science, and I am responsible for a good deal of Linux proselytization myself.1 This makes him an excellent protagonist for such a narrative.

However, the narrative was at times marred by metanarrative that, in my opinion, detracted from the persuasiveness of the piece as a whole. He states that he is not going to “give you cost-benefit analysis or tell you in strictly logical terms why you should switch to Linux.”2 However, he does eventually use cost-benefit analysis and straightforward argumentation, which readers may see as inconsistent with this statement. Furthermore, he states that he will not be “presenting a bullet-point list of arguments for and against switching to Linux,”3 but his future posts contain inset bullet-point-like (they are even prefaced with asterisks) statements – which I think add greatly to the effectiveness of those posts. While a statement differentiating this piece from other arguments for Linux adoption is probably good, it should be more carefully worded so as not to significantly limit his argumentative options in future posts.

He also “doth protest too much” at times. For example, he seems defensive about his lack of a computer science degree – which I consider his greatest asset in this endeavor. He says, “It’s easy, and perhaps understandable, to dismiss what I say on the basis that I’m in law school rather than pursuing a CS PhD.” This, I believe, quite untrue; the audience is less likely to dismiss him because he has the same relevant education as them – none. His entire second post ,”Birth of a Nerd,” seems to be focused on defending his technical prowess in a way that seems to fundamentally undercut his argument.

His substantive argument for Linux adoption consists primarily of an argument that Linux's customizability and a adaptability make it more useful than Windows, several less-important affirmative points, and a number of defenses against potential counterpoints. The customizability theme runs throughout all four initial pages. He describes the limits of customizability he ran up against on Windows that precipitated his switch, the process of learning how to customize Linux, and the benefits he currently reaps from his skills and customizations. In “Birth of a Nerd,”4 he describes his learning process, beginning as a small child using a Windows 3.1 computer with MS-DOS and proceeding through Windows 95, 98, and XP. Then, in “A Choice is Made,”5 he describes his experience switching to and becoming educated about Linux. “An Inconvenient Choice”6 is focused on his ongoing experiences as a Linux user, both negative and positive.

One of the problems with Linux is its reliance on the “command prompt,” a completely-textual way of interacting with the computer. People are generally afraid of and incompetent with the command prompt, which is why the familiar graphical user interface (GUI) was invented. While it is

not true that Linux is just a command prompt or even primarily command-prompt-based, it is unreasonable to expect someone who runs their own Linux computer to be able to completely avoid the command prompt. Saied deals with this problem adeptly by downplaying both the importance and the difficulty of the command prompt. He very effectively uses the familiar (to many, anyway, in his target audience) “win” command, used to load Windows 3.1 from DOS, as a way of making the command prompt seem less scary. He also differentiates the Linux command prompt from the Windows command prompt: the former is nearly powerful enough to write desktop applications, while the latter makes it difficult even to automate system backups. He uses this as an opportunity to return to his overarching theme of customizability: by investing the time necessary to write shell scripts, the user will save time repeatedly thereafter, with the time saved eventually greatly outstripping the time spent. The customizability benefit is interestingly dual to one of the major counterpoints he deals with: that Linux presents an overabundance of choice. Choice in Linux is pervasive, and those who think it is overly so are certainly not unreasonable. For example, just a few days ago I replaced my audio manager – that's the thing that provides the volume control and a few other features – with an alternative one. There are at least 4 major audio managers for Linux (and a handful of ones without much adoption), while both Mac OS and Windows don't even contemplate giving users a choice in that arena. This is one clear example of a very minor OS feature presenting a potentially-bewildering array of candidates. However, this creates a competitive market: I switched sound managers because the new one lets me adjust the volume of different programs separately (for instance, muting a game while keeping my music playing, or vice versa) and send sounds to speakers across the network. Neither of those features was present in the default sound manager, and neither is present (to my knowledge) in Windows or Mac OS. The desirability of choice is one of the most philosophical and deep arguments in the Linux community. Saied presents several similar examples, as well as more macroscopic choices – including, most importantly, the choice of a distribution. Both he and I use Ubuntu Linux, which is also widely regarded as the easiest Linux for new users. However, there are hundreds of Linux distributions with varying degrees of speciality (ranging from distributions designed to run on tiny, thumb-sized computers to those designed to power Google's massive databases). As a proponent of Linux adoption, this choice is difficult to deal with: on the one hand, finding the perfect distribution could well make adoption easy, but, on the other hand, finding that distribution – if it exists – could take hours of research. The bundling-based antitrust suits against Microsoft make it very clear that defaults can be largely determinative; this is because users don't want to take the effort to find better tools. Saied's stated audience, power users, may be more willing than most to embrace difficult choices, but even for the most open-minded and dedicated prospective adopter, the choice can be overwhelming. Saied gives a concrete suggestion, Ubuntu, – which I think is far superior to merely explaining the details of the choice – while also giving an overview of the choice to be made. Most users do not like making decisions. The other perpetual obstacle to Linux adoption is one it shares (to a lesser extent in the last couple years) with Mac OS: software compatibility. Although there is a project, called “Wine”,7 to add support for Windows programs to Linux, its compatibility inevitably lags months to years behind Windows' own support for such programs. For example, Microsoft Word 2007 is not supported, but there is reasonable support for Microsoft Word 2003. Likewise, Diablo II, an extremely popular game released in 2000,8 runs as well on Linux as it does on Windows, but support for recent titles ranges from full to none. Saied accurately and critically acknowledges that application support in Linux is a serious problem, saying, “The apps [(applications)] are understandably a major component for many, if not all people, myself included. That’s why we’re using the computer in the first place, right?”9 He implies that he would acknowledge that someone may be forced to stay on Windows because Linux does not run mission-critical applications.

However, he then coolly segues into an argument that turns the debate on its head and connects with the ongoing discussion about choice. “I now have a stable of excellent open source apps that do more than I could have imagined on Windows . . . .,”10 he says. Indeed, while there are admittedly some important Windows applications without Linux support, there are also many Linux applications without Windows support. Moreover, on Linux installing most applications is significantly easier and more secure than on Windows, due to the use of so-called “package repositories.” On my Ubuntu system, the package manager lists 23137 available packages, all of which are indexed and searchable, and all of which can be automatically downloaded and installed. Thus, package managers are a rather elegant solution to Linux's dual problems of choice and application support: first, they introduce a larger amount of choice than is present on Windows, which helps make up for the lack of certain Windows software; then, they make the installation process so much easier that even the expanded decision is easier to make than it would be on Windows. Saied makes an interesting gambit with regard to this point: he says, “Curiosity is crucial. Sometimes you must be willing to sacrifice efficiency in order to satisfy your curiosity.”11 Readers who see efficiency as paramount – and who, perhaps, even ascribe to the philosophy that “curiosity killed the cat” – will be significantly off-put by this comment; it will at least be a blow to the empathetic qualities of the argument. However, it will strike a resounding note with another set of readers – those who value learning for its own sake or who see it as a wise investment. It may even make readers in the latter set get a feeling of superiority by trying Linux, while simultaneously making it much more difficult to themselves to argue against at least trying Linux. This is especially true since, as Saied notes, “[t]aking Linux for a test run costs very little in terms of time and out-of-pocket costs. A Linux LiveCD gives you a great sneak-peek at Linux without installing anything onto your computer. If you don’t like it, eject the CD, reboot, and you’re back to normal.” I think that even without further additions, Saied's site could very well embolden some curious readers to attempt the switch.

1 However, I do not generally suggest Linux to non-programmers, due to the pervasiveness of the programmer mentality in the Linux system and the incompleteness of the graphical toolset that would allow users to avoid those subsystems with that mentality. 2 http://blogs.law.harvard.edu/inconvenientchoice/2007/12/03/a-letter-of-introduction/ 3 Id. 4 http://blogs.law.harvard.edu/inconvenientchoice/2007/12/03/birth-of-a-nerd/ 5 http://blogs.law.harvard.edu/inconvenientchoice/2007/12/03/a-choice-is-made/ 6 http://blogs.law.harvard.edu/inconvenientchoice/2007/12/04/part-iii-an-inconvenient-choice/ 7 A recursive acronym for “Wine Is Not an Emulator,” http://www.winehq.org/ 8 See http://en.wikipedia.org/wiki/Diablo_II 9 Comment #3, http://blogs.law.harvard.edu/inconvenientchoice/2007/12/03/a-letter-of-introduction/ 10 Comment #3, http://blogs.law.harvard.edu/inconvenientchoice/2007/12/03/a-letter-of-introduction/ 11 http://blogs.law.harvard.edu/inconvenientchoice/2007/12/03/birth-of-a-nerd/