An interesting thing is about to happen to home computing—the “Desktop” that GNU/Linux never seems able to liberate from proprietary Windows may be just about to become irrelevant. Three independent, ultra-low-end computing platforms are being released—platforms that, like the first “desktop PCs” will be mostly owned by people who’ve never owned computers before. Every one of them will run GNU/Linux!

Considering that GNU/Linux and other free operating systems are already a major force on the high end of servers, mainframes, and supercomputers and that GNU/Linux has begun to become extremely popular in embedded computers because of its portability and low cost—the once almighty desktop is now bracketed on the top and bottom by newer, GNU/Linux-based systems. Microsoft’s core domain is starting to look kind of puny and obsolete.

Embedded gadgets or the new PC?

When the first “personal computers” came out, they were pretty primitive systems. The very first Apple to go to market, for example, was not a full “personal computer”, but more like what today we would call a “console”—it kept costs down by using your television set instead of a dedicated monitor. That’s useful, because monitors tend to be the most expensive parts of modern desktop computers. Also, the monitor cost is not nearly so elastic as other parts of the computer—there are plenty of technological measures you can apply to get the cost of a computer PC’s electronics from $500 to $50, but if the monitor is going to cost a minimum of $300 anyway, then why bother?

Of course monitor technology is improving and getting cheaper at the same time. The monitor on my son’s Nintendo DS is a wonderful thing to behold, when your first computer was a TRS-80 Model I! Evolved from this end, developed for pricey “smart phone” and cheaper portable game consoles, this is one solution for the problem: make a new computer as just a really smart version of existing embedded technology.

One Laptop Per Child

Give that arrangement “laptop” styling, and you have the One Laptop Per Child (OLPC) project’s “$100 Laptop”, which is currently in production design in Taiwan, by Quanta, and will likely be delivering its first units towards the end of this year.

The OLPC laptop design makes sense, because it’s targeted at kids that most likely don’t even have TVs (blessed are the meek—for they shall inherit participatory culture before they are addicted to consumerism?), which means that the computer must be a complete package. Also, since many of the locations of interest are rural, developing world locations, a number of other design features make the machines relatively proof against real world problems like poor network connectivity and harsh environments (well actually, that’s just because they’ll be taken care of by kids!).

The target number of sales of the OLPC laptop is staggering. I think maybe you have to be a person who regularly deals with numbers to appreciate this. Adding another zero doesn’t look like much on paper, but it makes a lot of difference, nevertheless. One hundred million, is a very large number of computers, and while it still falls far short of the total number of kids on this planet, it’s a pretty good start. For the flip side of that comparison, it will mean just about doubling Quanta’s total output, for one product. That’s why the low price tag is possible: the designers are applying a quantity of scale that has not been common in the rapidly changing PC marketplace. This is a case of “good enough” being the enemy of “best”: in the higher end of the PC marketplace in the developed world, there is a perception that everyone wants the best, top-of-the-line, newest thing available, and that they will always be willing to pay a (virtually fixed) price to get that. It’s what the marketplace delivers, and we do buy them, so they must know something.

On the other hand, I have often personally wanted to see the priorities change. For example, I wouldn’t mind a laptop that was circa 1997 in speed, graphics, memory, and weight, if it would run for 12-24 hours on battery and cost less than $500. But the industry has never delivered one. Instead, they keep making thinner and lighter laptops that are meant to be as good or better than top-of-the-line desktop systems, but they still have lame battery life of only 2-4 hours on a good day, and they cost twice as much as I’ve ever spent on building my own desktop system. This is why I don’t have a laptop—and why I personally would pay $300 for an OLPC-style laptop if Quanta decides to market one to us poor, deprived “developed world” folks (in fact, I have spent about that much on an “obsolete” laptop in the past, but it still has the battery-life problem. And yes, of course, I could buy a Nokia 770, which is likely similar in need—but it’s also pretty similar to the OLPC laptop in design, isn’t it?).

Reds waving a Red Flag

Red Flag Linux, that is. A company (if that’s the proper word) inside of mainland China is producing a new computer that’s a lot like new game console systems, or a lot like that old Apple (or the old TRS-80 Color Computer, which is what I learned to program on), in that it provides a CPU and keyboard, and uses the television set for a monitor. They plan to sell it for about $150 a unit, making it definite competition for the $100 laptop for the folks who do have access to televisions (and dismissing, IMHO, the concerns that the laptops would just wind up on the black market: there will be good, legal alternatives).

This new system is also interesting because it’s based on a processor designed by a Chinese research group. I’m not sure if the design is truly “open hardware” in the world at large, but the Chinese motivation in using it, is clearly the same motivation that free-licensed hardware provides: they will not be beholden to US chip manufacturers in setting their bottom line, their production needs, or what they need their hardware to do.

They’ll comply with new Chinese laws to provide legitimate operating systems on each of these pre-installed systems by installing Red Flag Linux, of course. What else would they use? The new systems won’t be running Intel architecture, so portability is a major factor.

Playstation’s new “super computer”

Sony CEO Ken Kutaragi was definitely getting a little over-excited when he started billing the upcoming new Cell-based Playstation 3 as a “super computer”, but he made it quite clear that the new console will be a computer, not just a game console, and it will have an operating system. Oh, it will run games, and in full, living, DRM-locked color (I know, it would be nice, if implausible, to report that Sony would jump on the free software bandwagon, but, hey, baby steps are still good—at least they are ditching region-coding, which I think is a good sign), but it will also have a 60 GB hard drive, and at least the ability to run GNU/Linux from that drive (the details are still fuzzy, but it sure would make business sense to pre-install some GNU/Linux distro on that hard drive).

Speculation about the new system is rampant, and it won’t come out until this Fall (it was delayed), so you must take a bit of what I’m saying with a grain of salt, but I am basing this on unofficial statements from Sony, reported on a number of fairly reputable game news sites. However, the apparent cost will be in the neighborhood of $400 for this new console. It’s a big step up from the PS/2, which you can get now for about $150, but it’s still awfully cheap compared to desktop computers.

Between the people who will buy it only because it’s a game console, and the people who are smart enough to figure out that $400 is a great deal for an actual computer, this machine is going to find its way into a lot of homes in the developed world (especially Japan and the USA). Many of those will be homes that were never in the “desktop market”.

Switching is hard to do

“So?”, you might be inclined to ask. Well, Microsoft does not hold onto its dominance of the desktop market because people have looked at Windows, Mac OS X, and GNU/Linux, and decided that Windows is a better product. They use Windows because it’s what they’ve always used. If they remember that far back, they used it because it came from the same place that MS-DOS did. It was a straight upgrade. For many, it was what came already installed on their computer, and they may not even feel competent to change that, even if they wanted to try something else.

In other words, the resistance to GNU/Linux and other free software in the desktop world is not that it is somehow “not ready for the desktop” or “not as polished as Windows”, though you will still hear both claims.

No, the real problem is that people are conservative and lazy: they use the tool they have, instead of searching for a better one.

So, what will the world look like, when there’s a huge class of people for whom it is GNU/Linux that has that status?

The independent development of these new very low end systems, and the fact that they have all settled (pretty easily) on using GNU/Linux, makes it seem likely to me that this is just the leading edge of a new wave in the commoditization of computing.