The Haus

Hun-Speak

Click here for other Hun-Speak columns.

May 9, 2002 -- Lindows: Just a Bad Idea

For some reason, I've been thinking about Lindows lately. Lindows is an effort spearheaded by former MP3.com exec Michael Robertson to bring a version of Linux to the masses that can run every Windows application. You don't have to be a computer expert to understand what a mammoth undertaking that is. His stated goal is not to overturn Microsoft, but to provide people with choice. That is certainly a noble goal. Having said that, the more I look into Lindows, the more I am convinced that it is an effort doomed to fail. There are a couple of reasons I believe this.

Reason #1: Everything is run as root

Linux or other Unix variants are true multiuser operating systems. "Root" is the superuser account that gives you access to absolutely everything. Standard best practices call for users to use the root account only when absolutely necessary for system maintenance. Anything you run as root has access to the entire machine, making you vulnerable to worms, Trojans, viruses, etc.

Windows (at least Windows 3.x, 9x, and XP Home) is not a multiuser operating system. Whoever is using the computer can read, write, execute, or delete any file on the system. This is one thing that makes these versions of Windows more susceptible to viruses, etc. Once the virus is activated, it has access to everything on the computer and can wreak havoc.

If you would accidentally run an infected program as a regular user on a Linux system, the program could wreck your home directory, but the rest of the system would be safe. By having users run Lindows as root (and the preview version suggested that you not even have a password for the root account), they are removing a very important defense from Linux. One of the chief reasons for me and no doubt many others to run Linux is security. Lindows makes it just as vulnerable to attack as Windows 9x.

Reason #2: It relies on WINE

WINE is a recursive acronym (the Linux world loves those) meaning WINE Is Not an Emulator. Here's what it attempts to do: "Wine is an implementation of the Windows Win32 and Win16 APIs on top of X and Unix. Think of Wine as a Windows compatibility layer." WINE can run Windows applications with varying degrees of success. I use it to run Quicken 98, but it is quite flaky and prone to crash. I've used it with great success to run Jedi Knight II. I've used the CrossOver plugin which is based on WINE to run Quicktime, Windows Media Player, and the MS Word viewer. For those applications, it works reasonably well.

On their FAQ page, Lindows states its goal for running Windows apps:

Our goal is to run all Windows® software, however, that's an ambitious objective that will take time to achieve. Soon, LindowsOS users can expect to run a host of popular Windows® software titles. We'll make available a database of known useable applications in the near future. Watch as the list grows and we further enhance our Windows® compatibility features in the coming months.

To help them out, I can tell them exactly when they will be able to run all Windows software: when Hell freezes over. Trying to make a "compatibility layer" for Windows is shooting at a moving target. The specs are subject to change with very little, if any, notification at any time. While I applaud the WINE team for their efforts because they've done some amazing work, it is a dream to say that all Windows software will run under Lindows someday. Microsoft can't get all Win9x apps to run properly under WinXP! Why should Lindows have any greater success?

Reason #3: It's going to cost you

I'll grant you that this is a minor point, but one of the main advantages of Linux is that it is free. For the price of a CD-R, the whole OS and a slew of applications can be yours. If you want a book and some direct tech support, you can buy the boxed edition for a nominal fee.

Lindows plans on charging $99 per major release. While that isn't a terrifyingly huge amount and there are volume licensing plans, it still costs about $99 more than I paid for Red Hat 7.3. What is the advantage of paying $99 to get an OS that will pretend to be Windows when you can pay about the same amount for an OEM version of Windows XP Home that will run the apps natively?

Summary

I firmly believe that while Lindows' goals are noble, the whole project is just a bad idea doomed to go the way of the dot com. It tries to bring Windows to Linux, but ends up with the weaknesses of both operating systems and the strengths of neither. Linux's multiuser security is replaced with Windows' root access. Windows' native application support is replaced with WINE's troublesome emulation. It charges for an OS that is otherwise free. Somebody please tell me when this becomes a good idea.

Alternatives

If for some heretofore unknown reason you MUST run Windows applications--and I know that most businesses will give up on MS Office about as quickly and easily as a drug user will give up crack--check out Codeweavers' CrossOver Office which allows you to run Office 97 and 2000 and Lotus Notes with more apps to come. Pick a user-friendly Linux distribution such as Mandrake or the ubiquitous Red Hat. Granted it will take a bit to learn the differences between Windows and Linux, but in the end you will have the apps you need in a more stable and secure OS. Eventually, you might even be able to wean yourself off MS Office to a product like OpenOffice.org (not coincidentally, I'm using OpenOffice.org Writer to write this article) which, beside having great MS Office compatibility, is free of charge and free from draconian MS licenses.

Resources

If you would like to do some more research on this yourself, here's a couple of links to check out: