(LinuxWorld) — This week is part two of a two-part series on comparing Windows to Linux for average-to-dummy users. This week I'll address one of the hot-button issues, installation. First, here's the simple conclusion of my previous column: When faced with the problem of using DHCP to make a connection to the Internet via cable modem or DSL, neither Linux nor Windows made the process as easy as it should have been. Windows had the edge in my case, but my problems with Linux were self-inflicted, since I choose to run the Debian unstable branch. The easiest way to connect was to hook up a dirt-cheap appliance. I used a SOHOWARE NBG800. It's a couple years old, and the latest product is NBG800/A. There are plenty of other dirt-cheap around you could use, as well, such as the Linksys Etherfast Cable and DSL router. The various routers run between $50-$100 depending on features and discounts.

The easy setup is only one benefit to using cheap routers instead of a direct connection. Another is that you can easily hook up multiple computers to your broadband connection. Perhaps the most important benefit is these routers are better firewalls than just about any average-to-dummy user could set up on Windows or Linux, and all the computers in the home or very small business benefit from the single cheap firewall. All of the units I've tried use network address translation (NAT, or IP Masquerading) to give your computers access to the Internet. That isolates your computers from the 'Net.

They're not perfect firewalls, to be sure. Some may still allow you to administer the router from the Internet by default, which is a big security hole if you don't change the default administrator password or if you choose a lousy new one. (I seem to recall one extreme case where it didn't matter because outsiders could even see your password by viewing the HTML source for the administrator page!) Nevertheless, most products have plugged these holes. The average-to-dummy Linux user is unlikely to configure a NAT firewall at all, let alone create an IP Masquerading gateway with IPchains. Something of a firewall is better than none.

The best installation available

All this goes to prove that the most desirable solution for average-to-dummy users has nothing to do with Windows or Linux. The most attractive solution is simply the path of least resistance, as long as there is an appreciable return on your investment. That's where almost everyone on the planet has it wrong when they wage a battle between Windows and Linux installation. There is no contest. Windows wins every time, easily, for the average-to-dummy user.

That's because the average-to-dummy user buys a PC with Windows pre-installed. The contest is not between installing Linux and Windows. For the most part, the contest is between purchasing a PC and running it as is versus purchasing a PC, wiping Windows off it and installing Linux in its place. At worst, pre-installed Windows requires only some configuration steps you may have to traverse when you first turn on the PC. Linux not only requires a comparatively lengthy installation session but some Linux install programs present the average-to-dummy user with extremely intimidating tasks such as partitioning a hard drive. Clearly, the path of least resistance with the biggest return is Windows -- so far.

Let me state for the benefit of my esteemed LinuxWorld colleague, Joe Barr, that this conclusion only has the average-to-dummy user in view. Joe's recent comparison of installations for Windows to Linux is appropriate for a corporate audience, since IT departments are weighing the advantages of migrating from the legacy Windows OS to Linux over updating to the latest version of Windows. In the latter case, one needs to compare how the two install fresh or over the existing OS, and Joe is doing an excellent job of that.

The obvious solution to this problem is to lobby to get Linux pre-installed on as many PCs as possible, right? Wrong. That would help, of course, but it's not looking far enough into the future.

Here's what matters — Xbox. The average-to-dummy market is rapidly changing. Heck, the market is rapidly changing even for those of us who pretend to be above average. I used to hate game consoles, but the latest games for these boxes are more inventive and exciting than PC games, and now they look attractive enough that I don't miss the high resolution graphics advantage of the PC. I also love my TiVo (the Linux-based appliance that records TV shows automatically, and lets you pause rewind as you watch live TV). I don't like one or the other because of the OS they run. I like them because they're simple, cheap, easy-to-use appliances. That's what average-to-dummies want. Their dollars are what Microsoft wants, which is why it makes the Xbox.

Would I consider building my own Linux-based TV recorder like TiVo? You betcha. I was thinking of contacting Russell Pavlicek about this, since he just went through the process (see his recent InfoWorld column, linked in resources). However, it's unlikely I'll ever get to that project. I just don't have the time and TiVo works well enough.

Even if I did create this box, how many people are going to that? How many hard-core games will continue building, tweaking, and downloading drivers for a gamers dream machine for $2,000 or more when it's so much easier and cheaper to plug in a $150 or $199 box? Some, maybe, but as HDTV grows, the advantages of the PC diminish. Even hard-core gamers are sure to defect eventually. Microsoft sees the trend and is back-pedaling on its idea of the expensive PC home media center because the benefits of a cheap appliance are so obvious.

We're not quite at the point where consoles will be the ultimate cross between a home entertainment box and PC replacement, but that day is coming soon. The PlayStation 2 and Xbox (and some new versions of the Game Cube) already play DVDs, so you don't need an additional DVD player. All these consoles need is a broadband connection (available for some consoles already) and good-enough software. Then there will be no reason for many average-to-dummy users to consider purchasing a PC. Heck, I can't wait until my kids can browse the Web and do email on a console so I won't have to worry about what damage they could do at my workstation.

The bridge to the future

Microsoft's edge in the average-to-dummy market is the Xbox, not PC pre-installs. Actually, Microsoft has two cards it has played in this market, but the most important one is the Xbox. It played the other some time ago when it bailed out SGI with some cash and took in return the patent on OpenGL. By doing so, Microsoft snuffed out the future of OpenGL as a gaming graphics API and ensured that DirectX would emerge victorious in the PC market. OpenGL is multi-platform, whereas DirectX is tied to the Win32 API and Xbox.

Since the PC market is saturated, it turned out to be a minor victory, although not entirely a hollow one. Those who flocked to DirectX reaped the benefit of doing so because they could apply their skills (along with many of their Win32 API skills) to the Xbox. The battle is no longer between DirectX and OpenGL but DirectX and everything else, including OpenGL and the APIs for Nintendo Game Cube and the Sony PlayStation 2. Here Microsoft has the advantage of offering what is basically a single API for both PC games and the Xbox console.

If you really want to eliminate Windows in the average-to-dummy user market, here's the way to do it. Build a bridge from Alaska to Russia using Xbox units. This defeats Microsoft in two ways. Microsoft only makes money on the Xbox games or the tools companies use to create them. Microsoft loses money on every Xbox it sells. The more Xbox units you buy, the further you drive Microsoft into debt. As long as you publish a picture of the bridge, Microsoft can't win the favor of game companies with the huge sales figures you created by purchasing so many Xbox units.

The answers

All right, so that's not very realistic. Neither is the Linux-for-the-Xbox idea, which is gaining a little popularity right now. At least this won't fly for average-to-dummy users until it is possible to run Linux on an Xbox without hardware modifications. Average-to-dummy users don't want to violate warranties or solder anything.

This is not going to be an easy market for Linux to capture, if it is possible at all. Assuming it is something you feel is desirable, perhaps you can email me your suggestions on how Linux could take over the average-to-dummy world, or share them in the Talkback section below. I've got a couple ideas of my own. I'll share yours and mine in a future column. Until then, happy gaming and TiVo watching.