Adopting Linux

Everyone has been blogging about Linux recently but we are approaching an important time for Linux. People need to know about Linux before Longhorn comes out, if it is to steal any microsoft marketshare Imagine if, when Longhorn comes out, it had the same problems that XP did when it first came out (aka a security nightmare). Lets imagine that Gateway picks up Ubuntu and offers it as an alternative. If people are sufficiently scared from Longhorn, have heard of Linux, and see a major distributor offering it pre-installed (on cheaper computers than Longhorn is offered), they might pick it up. This would be huge. Others would follow in offering Linux alternatives, there are plenty of user friendly distributions (Mandriva, Ubuntu, Suse), and Longhorn would see, maybe 20% penetration onto new desktops. That would be nice.

Linux does have problems. The relative difficulty of installing packages (when it doesn’t show up in default deb repo), difficulty installing the OS itself (in that you need to know how things work at least superficially), dependency problems, hardware support comes late if at all (only now are wireless drivers approaching ubiquitous) et cetera. There are others.

Microsoft makes a tight end-user product. Put the cd in the drive and, apart from some minor partitioning questions, it installs and works. You want to install something? Find a .msi or .exe file, double click on it, and ‘next’ though the EULA. Done.

That being said I don’t like Windows all that much. It is overpriced and overbearing. I defy Bill Gates himself to remove Internet Explorer (not the icons, the program). Ditto for about a dozen other Microsoft programs: windows messenger (always running in the background!), media player, the start menu, and on and on. Don’t like the desktop layout? You can’t really change it. Lets say you’ve thought up a neat way to interface with your computer. With Windows you are stuck.

So is Linux the solution? It will be. It is quickly catching up to Windows as a usable product for an end user, especially distributions like Ubuntu. I think it may run into problems once it has caught up and everyone says, “now what?” But that is beside the point.

But can Microsoft’s stranglehold in the operating system department be broken? Why does everyone use Windows? I’ve come up with a basic list of the four most common explanations.
· Games only work on Windows
· It is what I am used to (I’ve always done it)
· Good hardware support, it just works [which is only true if you have the drivers, many linux distributions come with all the drivers you need]
· Widest variety of software available
These are worth taking a look at, because the reasons TO use Windows are also the reasons NOT to use Linux.

First, the gamer’s dilemma. It is common for gamers to be slightly techy (or pretend to be) and want to show your technological superiority by running the ubiquitous Linux OS. Ok fine, but then you find out about MD5/SHA-1 hashes you have to worry about with iso’s, partitioning, and then you find all your games won’t work. If you do it anyway, you’ll probably still spend most of your time in Windows anyhow.

This has an easy solution that will probably be forced on most PC gamers whether they like it or not. The death knell for PC games can be heard anywhere. Video games are a huge industry, for consoles. PC games are often pirated, low volume sales (you need a high end computer to run games now, and only a minority of people can afford to). Only the biggest games turn a noticable profit (like Half-Life 2 or Doom 3). A PS2 game is not as easily pirated and generally higher-volume. Which would you make if you made games?

Second the tradition problem. This is really the hardest to overcome. People with Windows experience expect other OSs to be like Windows. If everyone grew up on Linux; Ballmer and Co. would have some significant problems too, showing us why we want a less reliable proprietary OS. But that isn’t the case.

The solution to this seems to be imitation. Currently the Linux GUI closest to Windows is KDE. KDE seems to have imitated Windows quick-launch icons, system tray, start menu, control panel, and on. I’m not saying that is good or bad, just easy for Windows users. Personally I like windowmaker. But you give KDE to my grandma she can check email and surf the web. You give her Gnome and she’s up a creek without a paddle (unless Grandma is a former OSX user in which case Gnome is a lot closer to what she is used to than KDE would be).

Hardware support shouldn’t be much of a problem for long. Hardware companies are starting to realize that a critical mass of friendly geeks is powerful. If geeks bash something long enough they will begin to discredit it (like IE). If someone knows anything about SCO it is bad. But then how about IBM? Mac OSX? They have pretty good vibes, because geeks like them.

So hardware companies are starting to support the OSS movement more and more. Drivers have moved towards the ‘plug in and work’ phase (though they still have a way to go). This will continue. The problem with hardware (other than the proprietary issue, which ought to go away with driver), is that new hardware support makes its way to Linux late. iPod support? Still dodgy. As I said, wireless support took a long time in coming. So you have to be choosy in selecting your wireless card, or your portable mp3 player if you want it to work in Linux. The same will probably be true in the future, unless Linux gains more public credibility as a viable OS.

Lastly the software issue. It is true that most programs do not have Linux ports of any kind. Fortunately there are often open source equivalents. Photoshop? We have the GIMP. Microsoft Office? We have OpenOffice or KOffice, and IBM is working on Symphony. Internet Explorer? We have dozens of browsers, Konquerer, Firefox, Epiphany, Galeon, etc. This substitute solution covers about any major application type out there - except video editing. People may still be deterred via the tradition problem, switching from Photoshop to GIMP is hard. But for those willing to try, they’ll find good alternatives.

Some specialty software has no equivalent yet. This can be a real problem for some people. But for most is ok. As more people switch to a Unix-based system, more people/companies will write code for it. The more code writing, the more programs. Linux (by virtue of being open source) sees the snowball effect to a spectacular degree and is very easy to write code for.

But Linux is not the new Windows. Neither are any of the BSD flavors or Solaris. There will be no new Windows. I do not expect or hope anyone will again gain such a huge market share; OS diversity will improve a lot of things. How hard is it to write malware to affect everyone when there are 5 different OSs with 20% or more market share? It will be very difficult to get something to the same degree as Sasser or Netsky. OS variety will improve security, drive innovation (competition) and force interoperability as well as other strengths.

In sum, Linux will grow. Windows will fade. Unless Longhorn is really tight, widespread adoption (over XP) will be very slow. This is because XP is a durable good. In fact if Linux gets moving on some major user-friendly improvements, now might be the perfect time to get some market share. When Windows users face the end of XP support and updates, they may look elsewhere rather than shell out $300 for Longhorn. They should anyways.

I think the biggest barrier to people adopting Linux is that users don’t care. They’ve learned the Windows idiosyncrasies and it works just fine. Only when XP doesn’t support the newest technology (assuming there won’t be widespread upgrades to Longhorn) and Linux does; people may be driven to alternatives. I only hope it comes soon and we are ready.