January 06, 2005
The Command Line in 2004:
An updated and reader-annotated version of the original essay, "In the Beginning...was the Command Line" by author Neal Stephenson. Full text of the original essay, with extra notes applicable to today's world added by reader, Garrett Birkel.
Written in 1999, the original essay became a semi-bible for the open-source community. Explaining the basic differences between the traditional big software companies and free software, concepts addressed in this essay have continued to play out. However, that was 1999, and this is now. Technology has advanced (somewhat) and the big companies are smarter (somewhat). This annotated version of the original text covers happenings in the last five years. For original essay only, go here to download full text. Also Neal Stephenson's (awfully designed) website, here.
-
Damn. I forgot to mention the Monkeys.
-
Interesting reading. And the most impressive use of an extended metaphor I've found in a long while... Yet aren't car-driving monkeys a strange idea ?
-
The annotations aren't particularly interesting or insightful, IMO. To be fair, I didn't find the original essay particularly interesting or insightful either when I read it in 99. To compare operating systems to cars misses a crucial difference: one is merely a tool (or more precisely a tool schema, as we only care about its instantiations), while the other is a lot more than a tool. Every car in fact contains an instance of an operating system.
-
Holy shit, that guy's clueless. And boring.
-
And biased.
-
Really? The original essay has had a huge impact upon the development of open source. This provided a general direction for the disparate users to work towards, and in turn has made things better for us all. Even if you don't use Mozilla or Linux, the development of an open source community has greatly benefited the greater community of the internet through devloping better standards/protocols for us all. I agree that most of the annotation is a little facile, but I do like the monkey/chauffeur metaphor. It does sum up the differences between the various OSes pretty well. A car may not be a computer, but for the purposes of metaphor, its good enuff.
-
Great for the open source community. It means they're not easily bored. That's a great quality.
-
Has it really had as big an impact as you claim? I would think that Eric Raymond's The Cathedral and the Bazaar has had at least as much and probably more influence on the open-source movement. In fact, history disagrees with you. If you consider free software (note: larger scope than simply "open-source"), then the origins of the community are much earlier. Richard Stallman founded The FSF in 1985. Linus Torvalds released the first versions of Linux in the early nineties. (I have been using Linux every day since around 1995, and using GNU software and contributing to various FSF projects since at least 1989.) The Mozilla foundation has existed in some form since 1998 when Netscape opened the source of their browser. Stephenson wrote the essay in 1999, when free software was already an established movement. One might argue that the essay popularized some aspects of the free software movement, but to call it a "huge impact" is giving it too much credit.
-
It might come to some pleasing conclusions, and people might like Neal Stephenson, but I've been aware and following the open source movement since 1997, and I had never read it or seen anyone make the claim that it changed anything. It's full of half-truths, failed metaphors, and often just plain wrong. The sociology is bunk, the computer science is mostly worthless, the political history is boring paraphrase of Allan Bloom. It's just some guy who knows a little stuff about relatively mainstream OSes, but anyone who has done some reading will aknowledge that he doesn't know what he's talking about.
-
True. The Cathedral and the Bazaar has been far more influential. True. The free software movement/open source community has been around for quite a while. how does 'biggish impact' sound?
-
I don't know. The inaccuracies really irritated this Programming Language/OS geek. Reading up on Multics, ITS, VMS and Xerox PARC/Smalltalk should be required before you're allowed to write self-styled 'influential' essays. And before you call elisp 'beautiful', know what Scheme and Common Lisp are. And ideally ML, Haskell and Forth.
-
So I think Stephenson's essay is of very little worth. But you can't deny he's an influencial sci-fi writer popular amongst geeks. So some open-source programmers/sci-fi fans probably ignored/could not spot the flaws and got some motivation that some cool writer guy they admire likes open source/free software.
-
I agree that it has had some impact, but it has been of the stroking rather than stoking kind. And I disagree with Richer too. Stephenson's essay has only consumer PCs as its subject, not mainframe operating systems, and he talks more about usability than technical merit. Insofar as this narrow goal is concerned, and keeping in mind that his primary audience is almost certainly not technically inclined, I think his metaphors are okay. (The OS-as-car metaphor is certainly not his invention.) I do agree about the mistakes, however. Emacs-LISP is unique among contemporary LISP dialects in not having lexical closures. This not only forces programmers to engage in any number of name-mangling hacks, but also constrains the garbage-collector into essentially a mark-and-sweep style. It is therefore a worst-of-breed. On the other hand, if Stephenson thinks elisp is beautiful then he is in good company: RMS has staunchly refused to allow lexical closures to sully his LISP dialect.
-
I don't know. He makes sweeping generalization on OSes, including Unix. If one is to make those, he should mention Unix is not The One True Operating System.