faith-based operating systems

Stevn has a lengthy essay on the analogy between operating systems’ battle for the desktop and the schism between Christian Orthodox and Proestant sects. In a nutshell, Linux is Protestanism, and it’s very non-centralized nature is what keeps it from ataing critical mass on the desktop. It’s a fascinating analogy and one that might I might well apply in reverse to Islam at some later date (elsewhere).

One thing I note, is that Protestanism is dominant in Christianity, at least in Europe and the United States. It’s mostly in the third world that Catholicism is ascendant, especially Southern America. What does that mean for Linux? One thought is that even tough there is no common unifying distro, enough “turnkey” distros may arise that the majority of computer users one day are using Linux in some fashion.

Ultimately, I think the real barier to entry is not the OS but the applications. The primary obstacle facing Linux is really the lack of a serious contender to Micosoft Office, though now that MS has adopted an XML standard that may well change. Hardware periperals are another factor – once anyone can plug any mouse or USB thumbdrive bought off NewEgg or Amazon into Linux and It Just Works, Linux’ fortune will markedly improve.

It’s not about installing an OS – it’s about using it. To that end I don’t think that there’s really a two-year window as ESR suggests. If the various linux factions can unite, not on a distro but rather on a API or plug and play infrastructure, then Linux doesn’t need unity. It can win by sheer dint of numbers alone.

5 thoughts on “faith-based operating systems”

  1. The big reason that the analogy is imperfect is that religion is not subject to network effect to the extent that operating system choice is.

    But a different point that I thought about bringing up was that it’s true that Protestantism largely displaced Catholicism in Europe and North America, but Protestantism in its turn is dying. If you look at the numbers, the majority of Americans are still considered Christian, but the majority of those are “Easter Christians”, if they even attend church at all.

    The first has gone out for most Protestant sects. There are only a few which are growing; most are shrinking.

    That’s another reason the analogy is imperfect; I don’t see any reason to believ a similar failure portends for Linux.

  2. agreed. But I think my main point is really that network effect – while important – isn’t the sole arbiter of fitness for an OS. After all, one can look at Apple as the classic example – it has a smaller-scale network effect and can survive, because it’s solved teh application problem and the plugnplay problem. The former is simply that Microsoft has a version of Office for OSX. If not for that, Apple would have died long ago, like OS/2 and Be.

    I’m convinced that we will see distros popu up that solve the hardware problem. Any USB peripheral you buy should work flawlessly; thats a prerequisite. The application problem is going to become a nonissue once the DOC format is replaced by teh XML one. Here’s a good article on the wrangling going on there:

  3. *Plugs in my Microsoft Wheel Mouse Optical USB into my Ubuntu-linux laptop and proceeds to browse to this site. Also plugs in USB thumbdrive with a few episodes of Mushishi on it, a window pops up with the contents of the drive and I proceed to double-click on one to start watching it. Coincidentally, both pieces of hardware did come from NewEgg…*

    Really, I understand where you’re going with this, but you really need better examples than mice and thumbdrives, as those have very standard interfaces that have worked in Linux for many years now (Heck, I got a Canon camera that worked on my Linux laptop right out of the box without needing any driver installs or anything). Printers and modems would probably be better examples (though in both cases, it’s vendor driver support rather than any API issues with Linux. The USB part works flawlessly, it’s the non-standard vendor-specific communications that is the problem here).

    As a heavy user of Linux as a server and desktop, I truthfully couldn’t care less about having Linux become a mainstream standard on the desktop. If you want to use Linux as a desktop, great. If not, awesome. Then again, I’m one of the worst Linux advocates out there as I’m rather apathetic as to whether this works for others or not. I basically use Linux because I found it works best for my own needs (especially with working on files on remote servers via SSH), I can browse the net without issue, plus I enjoy the newer XGL/Compiz 3D desktops. Heck, I’m not sure if I want to see Linux become a mainstream desktop as I’m seeing that many of the hardcore geeks these days are gravitating towards the BSDs.

  4. Nick, you’re right that mice and thumbdrives might not be the best example 🙂 Then again, last week I needed to get some image files off of a GE-built MRI scanner that runs the latest RedHat and it was an excercise in pulling teeth. Philips Medical Systems scanners run on Windows and its never a problem. My point is that the ease of use has to translate across all Linux distros – some kind of universal API or functionality embedded in Linux itself.

    The point is that “most” hardware on “some” distros just isnt good enough. It needs to be “all” and “all”.

  5. “Philips Medical Systems scanners run on Windows…”

    Sigh. Exactly 15 years ago I was in the thick of porting device drivers from VAX to Alpha for the VMS-based Philips MRI scanners.

    Now I feel old…

Comments are closed.