Many IT professionals ask me when Linux will finally "make it" on the desktop. How will they know when Linux has made it? What's holding it back? In what ways is Microsoft working behind the scenes to inhibit the adoption of Linux desktops?
|John H. Terpstra|
For some time, I have pondered and researched these questions. Then, a recent experience lifted the clouds of uncertainty. In part one of this column, I'll relate a true story of two Linux desktop purchasers who ran into multiple roadblocks. Then, in part two, I'll analyze those problems and discuss how Microsoft and electronics manufacturers and retailers created them. Finally, in part three, I'll predict a future that could happen, one in which a monopoly leads the U.S. IT industry to second-class citizenship, and the opportunity that could change that scenario.
Most people think that application availability drives adoption of a computing platform. While applications are very important, there are more factors at work, as the following story of two Linux newbies demonstrates.
A recent adventure
Recently, Joe purchased a new laptop computer. Because he had an older laptop that still functioned more or less adequately, he decided to purchase a machine that had Linux installed from the factory. He had heard that Linux does not suffer many of the problems with viruses, worms and malware, and his Norton AntiVirus subscription had expired, so there was nothing to lose in trying Linux.
Joe figured that since Linux is free, the cost of a laptop computer pre-loaded with Linux would cost less than one that shipped with Microsoft Windows. Wrong! The cost estimates he came up with were between $300 and $500 more for a system with Linux than for one with Windows.
Joe did a Google search to find Linux on laptop suppliers and obtained five price offers. Many Windows laptop specials offered a free bundled LCD monitor or a free bundled printer. No such offer was found for a Linux pre-loaded laptop.
Although Joe could not understand why it should cost more to purchase a laptop that has no bundled licensed Windows operating system than one supplied with it, he decided that it made sense to purchase the lower-priced system and just junk the bundled Microsoft Windows XP Home Edition.
After shopping around, Joe stumbled across a sweet deal and purchased an HP Pavilion dv1000 laptop. Joe's friend Dennis was with him that day and was suckered into ordering a customized HP zv6000 series laptop direct from HP.
Both Joe and Dennis were determined to try Linux and agreed that if Linux is ready for the desktop, they were ready to climb onboard. Dennis' laptop arrived 10 days after Joe had gone home with his purchase.
Joe installed SuSE Linux 10.0, while Dennis purchased SuSE Linux Professional 9.3. Dennis chose the 9.3 because he wanted the 64-bit support that was available with the AMD Athlon 64 CPU. At the time, a 64-bit version of SuSE Linux 10.0 was not yet available.
Joe's installation of SuSE Linux 10.0 was an immediate success. The built-in PRO/Wireless 2200BG network card worked perfectly. Well, it had one little problem: If he pressed the built-in function key to turn off the wireless card, it would not restart without a reboot. Even so, the video card and LCD display operated at the full 1200x768 resolution. Another small problem occurred when Joe tried to use the built-in digital media slots. He found they did not work, but he figured that was a small sacrifice.
Generally, Joe was impressed. His HP DeskJet printer worked the first time he plugged it in. SuSE Linux 10.0 instantly recognized the printer, asked him if he wanted to configure it, and in seconds he was able to print a test page.
Dennis was not so fortunate. He has been unable to get X-Windows working. It seems that the video chip set is not supported in SuSE 9.3, and he has not been able to get the built-in Broadcom wireless card to function (not even with the ndiswrapper drivers).
Needless to say, Dennis is not a happy camper. He will most likely reinstall Microsoft Windows XP Home from the recovery disk that came with the system. He feels forced to use Windows and believes Linux is simply not ready for prime-time use. He has seen a sad outcome to a project that started with great promise and expectation.
The plot thickens
Joe was so happy with his new Linux laptop that he decided it was time to install Linux on his old Windows laptop. This laptop has no built-in wireless card, so he purchased a Netgear RangeMax Wireless Router with a Netgear RangeMax wireless PC card.
Joe's installation of SuSE Linux 10.0 on the old 15-inch Sony Vaio RPG600, 1.8 GHz P4 laptop was another flawless installation. The screen worked perfectly at 1600x1200 video resolution. Joe was delighted with his second Linux laptop.
Unfortunately, Joe's story now takes a bad turn. The Netgear RangeMax wireless card could not be recognized by SuSE Linux 10.0. Joe called Netgear, which explained that the company does not support Linux. Joe was told that Netgear had no plan to provide Linux drivers for Netgear RangeMax wireless cards.
After returning the Netgear wireless network card to the store, Joe purchased a Belkin Pre-N F5D8010 Notebook Network card. The new card was able to recognize that an Ethernet controller had been inserted into the computer, but it could not find a suitable driver. Joe then found out that Belkin does not support Linux and that no suitable driver is available.
Joe did an Internet search, which revealed that Belkin's wireless card chip set is manufactured by Airgo Networks Inc. He found a link on the Airgo Networks Web site that offered hope. (You can see why by visiting AirgoNetworks.com.) Alas, this was a blind alley, because the General Public License Linux package only contains open source software (OSS) that Airgo Networks has modified and for which it has made the source code available.
Joe felt that Airgo Networks should be commended for its honorable handling of OSS, but he still had no luck there. When he called the company, he found out that no Linux drivers will be available for the Airgo chip set until late 2005 or early 2006.
Joe went back to the store to return another useless wireless card. Not one wireless card that was on the shelves at CompUSA or Best Buy listed Linux driver support, so Joe gave up. That's right: Not one wireless card currently sold at CompUSA and at Best Buy mentions that it is suitable for use with Linux.
The good news is that there is one wireless card that does work with Linux. A friend gave Joe a Linksys Wireless-G Notebook adaptor V3.0 card that works perfectly in his Sony Vaio laptop with SuSE Linux 10.0.
So Linux desktop computers cost more than Microsoft Windows PCs do, and it's hard to find devices and drivers for Linux. Is that such a big deal? Well, in this story of just two Linux PC buyers, such difficulties stopped one from using Linux and the other only succeeded by being very persistent. Multiply that by millions of PC users, and you have a big deal.
Joe had to pay for Microsoft Windows when he had no desire to use it, because he would have paid more for a machine without it. Why should consumers suffer cost increases to use a free operating system? Why are governments around the world so silent on this matter? Isn't it time for the consumer to be better informed of the graft and corruption in the IT retail industry?
There are layers upon layers of roadblocks being placed in the path of Linux. In the next installment, I name some of those participating in the blockade and how they're hurting consumers and businesses.
The open source software (OSS) movement started as a result of dissatisfaction with the proprietary software world. It is a global initiative that is correcting a seriously broken system in which vendors are taking undue advantage of consumers and depriving the consumer of choice.
Clearly, many players in the IT world have roles in blockading Linux. In part one of this column, I described the barriers placed before Linux adopters Joe and Dennis. Let's drill deeper and find the roots of this anti-Linux conspiracy.
PCs for the rich only, thanks to IP laws
The IT consumer market caters to a mere 10% of the global population. Unix and Linux are the only platforms that provide desktop support for many countries and languages that would otherwise not be able to use modern computing tools in the consumer's native language character set. Most consumer software available in the world today is suitable only for use in English-speaking parts of the world.
So-called intellectual property (IP) protection keeps software in the English-only category. Proprietary licenses built on IP laws make it unprofitable to create software for minority language areas. In other words, there are some customers that the incumbent solution providers do not want.
Non-proprietary, low-cost OSS could bridge the commercial chasm between profitable and unprofitable markets. To make this happen, however, the companies selling IT products to the masses worldwide need to support Linux and OSS. Also, the interoperability problems presented by proprietary devices, drivers and software need to be erased. Otherwise, every OSS user could face the problems that plagued Joe and Dennis.
Who isn't onboard for widespread access to PCs & IT?
Despite the illusionary problem of commercial viability for commercial software vendors, there is a deeper problem in the IT industry. It's apparent that the commercial IT retail market has no desire to provide real consumer choice. Let's look at the situation:
- CompUSA, Best Buy, Circuit City, Fry's Electronics and other major consumer electronics retailers do not offer Linux pre-loaded PCs for sale. These stores do sell some PCs that will work with Linux, if consumers download Linux themselves; they only sell PCs bundled with Microsoft and Apple operating systems.
Stores could sell a lower cost desktop, or laptop, computer at a lower price and with higher margins, thus making it possible to attract a larger consumer base into the active market. Why are these retailers not interested in doing this?
A desktop computer can be purchased for as little as $400. A laptop computer can be purchased at a price point below $550. Linux is free. Microsoft Windows, coupled with its bundled software, must cost at least $40 per machine. So, if Linux were to be pre-loaded, the retailer could offer the device at the same price and make an additional 5% to 10% gross margin.
Obviously, there are forces at work in the IT industry that cause retailers to choose not to participate in being more profitable. These stores don't offer the consumer the choice of a desktop platform other than Microsoft Windows and Apple. Why?
- The aforementioned stores don't carry peripheral hardware suitable for use with Linux. This forces the consumer who wishes to use Linux to shop elsewhere. Clearly, these stores have made a decision that they are not interested in having Linux users as customers. Why?
- A few smart vendors offer limited support for Linux. Dell, Hewlett-Packard and IBM offer a very minor selection of laptop PCs, desktop systems and servers that are compatible with Linux.
If companies really seek to attract the largest number of potential consumers, why are their practices so restrictive? What commercial arrangements have been made behind closed doors so as to keep Linux out of the public eye?
- Server, PC and peripheral hardware vendors today introduce products that lack any form of Linux support, thereby delaying the availability of Linux drivers for these products. Linux developers have to rush to build drivers after major vendors' products are first shipped. This is a major deterrent to Linux adoption by users, as demonstrated by Dennis having to revert to using Microsoft Windows.
How is it possible then for the consumer even to try Linux without significant added costs, and with radically limited choice of supported hardware?
- A store manager of one of the major consumer electronics retailers told me that his store had received complaints from customers because it had sold a network card for which the Microsoft Windows driver had not been certified by Microsoft. When he contacted the peripheral hardware vendor/manufacturer in question, he was told that Microsoft certification for the driver would require a royalty payment to Microsoft. The royalty would add as much as to $10 to the cost of each unit sold.
There are no certification or license fees for Linux drivers. Assuming that Microsoft does charge a royalty or any type of certification fee, why do vendors choose to pay for the privilege of providing a driver for Windows, when there are no such costs for a Linux driver?
Linspire, take a bow
Michael Robertson, CEO of Linspire, deserves public credit for his initiative over the past few years to make Linux pre-loaded laptops and desktop PCs available in the retail channel. Wal-Mart was among the first to sell Linspire-equipped systems to consumers. It would be unjust of me to criticize the IT retailers and OEMs, without giving credit where it is due.
However, I am astounded that consumers know little about Linux, largely because they don't see Linux when they go shopping. It's not like Linux was born yesterday. Linux has a stable track record over 14 years, already accounts for up to 35% of the installed server operating system market and is ready for the desktop with many free desktop applications that outperform the Microsoft Windows platform equivalents. Linux makes it financially feasible for more people in the world to use modern computing tools.
The fact that Linspire's Robertson has had to fight, and fight hard, to convince a few retailers to carry Linux pre-loaded PCs is clear evidence of the stranglehold Microsoft has on that channel. Major retailers are not interested in giving customers a less expensive, more reliable PC platform. They more interested in not damaging the relationship with Microsoft. This layer of the anti-Linux movement has wide repercussions, as I'll discuss in the concluding installment of this column.
The open source software (OSS) initiative has already created a highly productive, globally supported alternative software platform. The choice of free applications for Linux is growing daily. Smart application software vendors are already reaping a financial reward from the sale of their applications. Companies that have not yet embraced OSS, by porting their application to Linux, are likely to go the way of the dinosaur.
Software vendors are getting onboard. Hardware vendors are playing on the sidelines, and electronics retailers are out in the peanut gallery, still hogtied to Microsoft. Considering this situation, I think it is time for Open Hardware Manufacturing (OHM) and for the creation of a new global IT solutions retail infrastructure. Yet, I fear that U.S. players are still making enough money from their Microsoft connections that they feel comfortable with the status quo.
It's a good possibility, however, that a new-styled OHM will emerge out of China, where labor costs are low, and the desire to forge ahead is stronger than it is today in America. I've heard rumblings that Chinese manufacturers might establish new retail operations throughout North America, Europe and Asia, resulting in even more of the U.S. domestic technology market being sacrificed to the scrap-heap of history. Consider this: China has already stolen the march on the U.S. textile industry.
Will China use Linux to kick some U.S. butt?
The question we must answer is this, "Will China be the hub of future IT innovation and consumerism?" If the answer is yes, Linux is the tool that is waiting for the right Chinese entrepreneur who has a vision for creating the future.
I do see some inkling that U.S. players will play the Linux/OSS card and do it well. I'll wager a bet that Novell understands the dynamics of the market. Novell purchased SuSE, a major Linux development house. Recently, Novell released OpenSuSE, which means that the professional Linux desktop is now free. Novell sells a fully supported version, but the core product is freely available from OpenSuSE.org. This is a smart and timely move.
If I understand Novell's strategy correctly and my anticipation of the market is correct, Novell is positioned for a meteoric rise. That success could come through local OHM regenerative growth. Or, Novell could even take advantage of the entry of Asian competitors, whose OSS/OHM- oriented consumer retail channel would make the advantages of OSS and Linux desktops famous.
Widespread adoption of Linux desktops and OSS is going to happen. It's going to happen with or without the help of U.S. IT vendors and electronics retailers. Surely, there must be someone in North America with the entrepreneurship, the vision and the determination to take advantage of this opportunity before it is too late. I cannot believe that the current consumer IT retail industry is willing to fall on its sword without reconnecting with a profitable and loyal customer base. I wonder who will step forward.
Users, IT pros: Speak out now
Consumers and IT professionals, this is your call to action. If you want freedom of choice, please write to the managers of your local electronics stores, advising them that you want the choice of Linux on your next desktop or laptop computer purchase. Tell the store manager that if they refuse to be more consumer choice oriented, you will no longer purchase from them. Retail stores value consumer feedback. The few store managers I polled told me that no consumer had ever requested a Linux system, and until at least a dozen do, it does not make sense to offer such systems.
Why should IT pros join this battle? Your users buy their PCs from retailers. If they're only familiar with Microsoft products, then you'll encounter great resistance to any effort bring the lower costs and greater reliability of Linux desktops to your company. Many of you have already fought and lost that battle … for the time being.
You and I are consumers. We have the right and the responsibility to make our wishes known. If enough consumers speak up, retailers will jump to attention faster than you can imagine.
Windows problems with viruses, worms and malware exist because consumers failed to tell Microsoft that they had had enough. Instead of making our dissatisfaction known to the company that could do most to solve the problem, we delighted in complaining to each other. Make your wishes known to your IT retailers now. Remember that your actions, when properly directed, can change the world.
About the author: John H. Terpstra is chief technology officer of PrimaStasys Inc., an IT consulting firm, and a member of SearchOpenSource.com's Editorial Advisory Board. He is author of Samba-3 by Example: Practical Exercises to Successful Deployment, 2nd Edition and The Official Samba-3 HOWTO and Reference Guide, 2nd Edition.