Windows Won't Eat Where Linux Has Walked?

If it's the 21st century, why is the tech industry acting like it's stuck in a "Deadwood" rerun when it comes to Linux support?

By Carla Schroder | Posted Mar 6, 2006
Page of   |  Back to Page 1
Print ArticleEmail Article
  • Share on Facebook
  • Share on Twitter
  • Share on LinkedIn

The more things change, the more they stay the same. Just a few years ago I was looking forward to a happier Linux future; one where as Linux uptake increased, so did supported hardware. Hardware support has improved a lot, but events have taken some interesting twists and turns, and finding certain types of supported hardware, such as wireless networking devices, printers, and video adapters is still ridiculously difficult. Yes, I know that NVidia and ATI supply Linux drivers, which is better than nothing, but what about some open, unencumbered video adapters that do not bork system files like ATI and NVdia do?

It is silly for such basic necessities to still be such a pain. Are the people who run Linux-unfriendly businesses stupid? Is it pressure from our favorite Evil Empire that keeps them in line? Is it true that the managers who make these decisions can be bought with a couple of trips to the local nekked nightclub? Or are there sound, well-reasoned business decisions at work?

A Cow Won't Eat Where a Sheep Has Walked
How did this bizarre philosophy come to dominate computing? I don't see it in other industries. It's like the olden days of the wild wild West, when cattle ranchers were dead certain that cows and sheep could not graze on the same ground, so sheep had to be driven off, or even better, killed. I look out my front window and see sheep and cows grazing in the same pasture. And I can take a short walk and find folks who still insist that a cow won't eat where a sheep has walked.

In any other industry we have innumerable choices, whether it's tools, vehicles, houses, clothing, building materials, furniture, knick knacks, you name it. When you purchase a pen it will write on any paper. When you buy a Bosch cordless drill, you know it will take bits made by Ryobi, Black and Decker, DeWalt–anyone you want. When you buy a car, you know you will have abundance choices in third-party replacement parts and customizations. You can drive on any road, and fill up at any fuel station. Have you looked in the junk-food aisles of the food store lately? How many different brands of tortilla chips and salsa do we need? How many different brands of fizzy brown soda water? The people making these products are all doing fine. Their kids are not going hungry, and you don't see their senior executives in line at the food bank. This tells us that having a large number of competing brands and companies works fine.

And no matter how you mix and match, they all interoperate. Just imagine driving a car that accepted only a specific brand of gasoline, or salsa that declared it was illegal to use on any but approved chips. Of course it's absurd, but the computing industry does this every day.

We Must Protect Our Eye Pee
For years, Linux-unfriendly vendors fobbed us off with the excuse that "there is not enough demand to support Linux. We target the 93 percent Windows market, not the 3 percent Linux market or the 4 percent Apple market," or whatever fictional numbers they have handy. Hogwash, and they know it. Linux/Unix users have been buying their products all along and making them work, often in the face of vendor opposition. There is a huge pool of Linux coders who already write drivers for free, and ask only for specs and test units. The response these fine folks often get is "we have to protect our intellectual property." Piffle. What are they afraid of, unless that is codespeak for "misappropriated code and patent violations" ?

Suppose those numbers are true, which I do not believe they are. I think the total Linux market share is much larger, over 10 percent, and Apple is close behind. But just suppose. You don't have to be a business school PhD to understand that no one owns that 93 percent; rather, it is chopped up among a sizable number of companies with very little to differentiate them. I won't bore you with listing all the dozens of hardware manufacturers that play shuffle-the-same-three-networking-or-video-chipsets amongst hundred of different brands and models. Suffice it to say that all they need to support are a small set of drivers that work across hundreds of devices.

Printers work in much the same fashion. Every printer manufacturer has lines of printers that operate from common drivers, so even if they have fifty printers, a few basic drivers support the lot. This is no secret– you see it on their download sites. So that takes care of the "it's too hard to support Linux" argument. Actually it's easier than supporting Windows–just let the devs and distribution packagers do the work.

So let's say Vendor A, a giant in its field, holds a 10 percent Windows market share. Vendor A has Windows 9x, ME, NT4, 2000, 2003, and XP to support. But Vendor A claims Linux is not worth it, even with a free and willing labor pool to write drivers, and even with a common driver code base with Mac OS X and every Unix on the planet. Again my head hurts to even try to follow this line of "reasoning." How can it be smart to ignore this huge, under-served market? Especially when Linux is growing far faster than Windows, and OS X is healthy and prospering?

Phenomenal Linux Growth
While we're discussing numbers, keep in mind that the numbers trotted out by various analysts, which show 22-percent-plus annual growth and sales in the billions for Linux, don't account for the probably larger number of re-purposed and new boxes using free-of-cost Linuxes. The software is free, but the hardware still costs money, and any hardware vendor who is not aware of this market segment is too stupid to stay in business.

The Real World Is Mixed
It's short-sighted to not serve users' real needs, which is mixed environments. Hardly anyone is a 100 percent Linux or Unix or Apple or Windows shop. Nobody likes being locked in, and why restrict your own potential market share anyway? Why make products that only work, or only work well, with your own stuff, or on a single platform? When your widgets work with everybody's, you've just increased your potential market share manyfold.

Interestingly, in my conversations with companies that do not want to support Linux, I have yet to hear any of them mention customer's needs, just their own. It's nice to be so valued.

Invitation To Vendors
I am bored to the gills with wading through jargon-laden, uninformative Websites trying to find basic product information, like, you know, will it work on Linux? I love how they can lard up product specs with junk like operating temperatures and physical dimensions, and never quite reveal the secret information about which operating systems it runs on- "Turns any PC into a networking powerhouse!" Well no, it doesn't, only Windows PCs, so that would technically be a lie.

So here is your big chance: I formally invite you to send me information on your Linux-supported products, and I will share it with my legions of readers. Send it to carla@bratgrrl.com. Anything and everything is welcome, but especially printers, wireless networking hardware, and video cards. It's time to join the 21st century, friends!

Comment and Contribute
(Maximum characters: 1200). You have
characters left.
Get the Latest Scoop with Enterprise Networking Planet Newsletter