Top Stories of 2001, #5: Linux Falters on the Desktop

For the past few years, you could confidently predict that every computer industry pundit in the world would write an article about the rise of Linux on the desktop. "1999 will be THE year of Linux on the desktop," they cried. "OK, seriously, 2000 is it! Well, maybe 2001." Maybe not. In 2001, the dot-com market officially dried up, the economy nosedived into a cyclical recession, and the market for Linux on the desktop remained nonexistent.

On a more positive note, Linux desktop advancements are quite impressive, and if you haven't tried this open-source solution recently, I recommend products such as Ximian Desktop and Evolution, two highlights of the Linux desktop movement. But as 2001 came to a close, any hope that Linux would supplant Windows desktop versions ebbed right along with it. Linux isn't a disaster, of course; the OS commands a healthy percentage of the small-server market. But the one-time Linux steamroller has stalled at the Gates of Windows, and the fact that the desktop is one market Linux will never conquer has become increasingly obvious. Besides, if you want a desktop UNIX, Mac OS X is so much more elegant.

Back in January 2001, however, Mac OS X hadn't shipped, and many Linux companies were planning heady desktop improvements. On January 11, Microsoft CEO Steve Ballmer even declared that Linux was the number-one threat to Windows. He described the OS as a "phenomenon" that the company must address.

The first chink in the desktop Linux armor appeared in late January when Corel announced that it would drop its popular Linux products and focus instead on its creative applications in an attempt to return to profitability. Predictably, the term profitability would be a recurring theme in 2001 as desktop Linux companies dropped off the map.

February opened on an ugly note with a trash-talking, anti-Linux article on Microsoft's Web site. The piece, titled "Linux to Fall by the Wayside," included some interesting Linux spin from Microsoft Senior Vice President Joachim Kempin. "Linux is too expensive," he said. "This is because each company has to invest an unreasonable amount of time and money in Linux verification. As a result, the total cost of system installation and operation (TCO) is too high, effectively making Linux a very expensive OS. Vendors who are building up their Linux businesses are making a serious mistake and need to wake up to that fact quickly." But Kempin saved his most vitriolic language for a question about Linux growth. "Linux is simply a fad that has been generated by the media and is destined to fall by the wayside in time," he continued. "Windows 2000 will gradually overtake the Linux share in the server market. In fact, the advent of Linux has spurred Microsoft's developers to move up a gear. The arrival of new competitors in applications or OS development provides us at Microsoft with the driving force to create even better software products."

In a Wired magazine article, Doug Miller, Microsoft's group product manager for competitive strategies, attacked the idea that free was better. "The recent security problems with Linux, coupled with the lack of key enterprise elements in the new kernel, really call into question whether Linux should be used at all," he said. Miller told Wired that the recent dot-com shutdowns also would be replayed in the Linux community. "Free does not sustain a business," Miller said. "Development costs money, \[quality assurance\] costs money, support costs money. We have yet to see a business model in the Linux world that has any chance of long-term success."

In March, we saw more indications that Miller was on to something. An International Data Corporation (IDC) study reported that Microsoft's desktop share had actually grown from 89 percent the previous year to 92 percent. But an even more amazing fact was that Microsoft's share of the server market had outgrown upstart Linux, leaping from 38 percent of the market last year to 41 percent by 2001. Linux also grew, but at a slower rate than in the past, and the growth of this open-source OS apparently came at the expense of Novell and various versions of UNIX, not of Windows. Linux grew from 25 percent of the market last year to 27 percent in 2001. On the desktop, Linux's share was barely measurable; IDC placed it at about 1 percent.

In a bizarre move, Microsoft announced in early March that it would expand the release of Windows source code to a wider range of companies. However, Microsoft heavily restricted the release, which it targeted more for its public-relations benefits than anything else. After IBM announced sweeping support for Linux that month, Microsoft CEO Steve Ballmer lashed out at the open-source OS, providing the kind of sound bites that just make people like me smile and shake their heads. "\[IBM supporting Linux\] will have an impact, but Linux is a toy," Ballmer said. "IBM has not participated in the modern software revolution. Who are you going to bet on--Microsoft or IBM? I'll take that bet any day."

In early May, Microsoft's anti-Linux crusade continued when Senior Vice President Craig Mundie used a New York University speech to trash the OS and its open-source roots. "As history has shown," Mundie said, "this type of model isn't successful in building a mass market and making powerful, easy-to-use software broadly accessible to consumers." Mundie said that open source software (OSS) might have made sense in the 1970s, when software was basically tied to hardware. Today, he says, hardware is wide open, so intellectual property is instead tied to software. Mundie's comments predictably garnered the collective wrath of the open-source community.

By mid-May, the most promising desktop Linux company, Eazel, called it quits. Founded by ex-Apple developers, Eazel had been working on easy-to-use desktop software. Eazel's stuff was good, and former employees promised to keep updating the software, but that effort mostly died off over time. Eazel's demise was probably the most damaging Linux desktop loss of the year.

In June, Gartner reported that Linux growth and market penetration were far less than previously thought. The company said that most of Linux's success had come from "white box" servers used at "mom-and-pop operations." And when you discounted these mostly meaningless boxes, Linux's market share on the server was closer to 6 percent, not the 27 percent previously reported. "Linux continues to be on a growth path," the report reads, "and \[Gartner\] believes the demand for Linux-based servers will grow to 10 percent of \[overall\] server shipments in 2001. The primary demand for these servers will be as Internet or infrastructure servers \[e.g., DNS, gateways\]." Gartner noted that its primary goal for the report, which Microsoft at least partly funded, was to establish realistic market-share data for Linux.

Microsoft couldn't stay out of the debate. Miller told eWEEK that Microsoft helped sponsor the study to see exactly who was using Linux. "There has been a lot of hype around Linux over the past year, and we wanted to ... find out the real story on its adoption," Miller said. "While I admit there has been interest in Linux, this by no means accounts for one out of every four new servers sold. That is simply ridiculous." According to Miller, the vast majority of Linux sales and downloads occurred so users could experiment with it, but the OS was rarely deployed.

On June 22, Microsoft Chairman and Chief Software Architect Bill Gates issued an infamous statement in which he compared Linux to Pac-Man. Mundie followed this statement with a late July appearance at an open-source conference. Amazingly, Mundie emerged unscathed, although reports differ on who won the debate in which he participated.

Citing an utter lack of sales, the world's largest PC maker, Dell, cancelled desktop Linux sales on its PCs in August. "Not that many customers are using Linux for their desktop systems," Dell's Jim Mazzola said. "We're a customer-demand-driven company, so as we see customer demand in certain areas that's where we opt to sell certain products or services." The bad news continued later that month when respected game developer John Carmack, who created DOOM, Quake, and other best-selling titles, railed against Linux. "It has been pretty clearly demonstrated that the Mac market is barely viable and \[that\] the Linux market is not viable for game developers to pursue," Carmack wrote. "Linux ports will be done out of good will, not profit motives. From an economic standpoint, \[developers aren't\] making a bad call if they ignore the existence of all platforms but Windows." In response to how well Linux games have sold, Carmack made some grim comments about the open-source solution. "\[The sales from\] all Linux games ever sold \[combined\] don't add up to \[the sales generated by\] one medium-selling Windows title," he said. That same week, staunch open-source supporter VA Linux announced that it would sell proprietary software to grow stagnant sales; the company later announced that it would remove the word Linux from its name.

Linux celebrated its 10th anniversary at the end of August, and although the OS had been quite successful, the mounting evidence that the revolution was stalling somewhat dimmed the celebration. That same week, Corel announced that it would sell off its Linux distribution rights for one-fifth the original asking price. But the company that bought the rights has yet to release its own distribution, effectively killing a once-proud OS release.

Netcraft chimed in on the Linux market-share issue in October, noting that Windows Servers actually dominate the Web with almost 50 percent of the market. Netcraft awarded Linux 30 percent of the market. By November, Gates attempted to take credit for inventing Open Source, but his argument is difficult to defend, so I mention it for historical (and humor) purposes only.

In early December, one of the few remaining viable desktop Linux companies, Ximian, released its watershed product, Evolution, a desktop-oriented personal information manager (PIM) package that apes virtually every feature in Microsoft's Outlook product. But Linux still lacks a credible Microsoft Office alternative; perhaps Ximian can work on an alternative in 2002.

Finally, 2001 ended on a controversial note with a Web Side Story survey that reported that only 0.24 percent of PC desktops use Linux, compared to about 98 percent for Windows. "For almost 3 years, Linux usage share has fluctuated between 0.2 and 0.3 percent, with no substantial growth," Web Side Story noted. Linux advocates, who had no particular data to back up their claims, refuted the assertion. So did others--one humorous example was a site called Low End Mac which purportedly contorted its server logs to show that Linux was in use by far more than 0.24 percent of the public. Note to Low End Mac: The few people who visit your site don't represent an accurate snapshot of overall Web use; sorry.

However you measure it--by absolute dollars or actual use--Linux has never achieved the desktop success that's so often predicted. That doesn't mean that it can't grow in this market; indeed, growth is much easier when there's no place to go but up. But for those people who had hoped to see Linux produce a credible desktop contender, one fact is now clear: It isn't happening any time soon. And it most certainly didn't happen in 2001.

Discuss this Article 26

richard (not verified)
on Jan 15, 2002
Why would a site such as yours spend anytime commenting about LINUX? Obviously Linux is not in your business plan. All your publications pertain to Microsoft products. Don’t be so transparent. Editor's note: I've explained this repeatedly, most recently in my End of Year review. Linux matters. --Paul
TSK (not verified)
on Jan 16, 2002
It is funny how people expect to see Linux Desktop replacing Windows desktops overnight. Of course, something like that is happening in Korea where 120,000 systems are going to use Linux. http://linuxpr.com/releases/4405.html Maybe I should ask some of my friends and replace the few hundreds of desktop computers that we have in the office this weekend. Seriously though, it is very uncommon if a system replaces another overnight. For instance, several years ago, MS wanted MS-Windows NT to replace UNIX and Novell. NT market shared grew, but not at the pace that they wanted. Also, MS expected Windows 2000 to be the new server OS but its adoption has not been as fast as expected. Some phrases in your article could apply to Windows 2000 as well. Allow me to replace some words of your article... However you measure it--by absolute dollars or actual use--Windows 2000 has never achieved the server adoption success that's so often predicted. That doesn't mean that it can't grow in this market; indeed, growth is much easier when you control the distribution channels. But for those people who had hoped to see Windows 2000 take over the server market with a credible server contender, one fact is now clear: It isn't happening any time soon. And it most certainly didn't happen in 2000. Posting via Netscape 6.1/Linux Mandrake 8.1
Tim Jansen (not verified)
on Jan 22, 2002
Few comments: - obviously expectations that Linux-based desktop system, in the state they are now or even were 2 years ago, could take a significant market share are ridiculous - as you suggested, its the progress that counts. The development of Linux desktops is currently going at a faster pace than the development of Windows. Assuming that the progress doesnt stop sooner or later Linux system will catch and then overtake Windows. Just don't expect this in the next 2 years. - there are working open source business models. They are just not the usual, high-profile "write something once and sell it a million times" business models: people pay developers for new features. If, for example, a company releases a new CPU it will hire a few developers to add support to the existing, free compilers (like gcc). These new feature must then be released as free software as well, because the original software's license requires this. This is a good deal for both sides: the CPU maker does not create a completely new compiler from the scratch and the community that has created gcc gets the new feature.
linuxuser (not verified)
on Jan 23, 2002
I keep hearing how Linux keeps missing these imaginary deadlines for desktop dominance. "Oh, if Linux can't be a player by such and such a date, it will never be a prominant platform in the general world of computing!" Well, guess what? Linux keeps progressing in spite of the imaginary deadlines! Windows keeps getting worse, and Linux keeps getting better. The trade press keeps asking " Is Linux ready for the desktop?", they should be asking "When is Linux going to be ready for the desktop?" The answer is "Pretty soon now."
SIG (not verified)
on Jan 22, 2002
Paul, you have to admit that if one is unable to do anything else than click and move the mouse, he needs badly Windows. For the rest of us there is Linux, OpenBSD, a.s.o Linux on the desktop is just a matter of time...that's all about it...wait and see
Bearcat M. Sandor (not verified)
on Jan 22, 2002
I've never been sure what all the "Linux is ready for the Desktop" hype is about. So what if Linux distros haven't taken over Microsoft's share in the 3 years they've been viable? I love my linux distro (Sorcerer). Everything on my machine has been compiled for my CPU. Not a 386, not a 586, my Athlon processor. I can grab the code and make any changes I like. Is linux more expensive to maintain? Not if you have a unix admin on hand. Security. I saw Microsoft state once that there were more reported security flaws in Linux distros than in Microsoft. Yea, REPORTED, and most were fixed with in a day of discovery. Cost of software? I am a professional developer and I admin my own web presence. I handle my own DNS, mail, ftp and web server. The total cost of the software on my system if I were using Microsoft products would be at least $200,000 (priced it once). Under the free linux tools all of my server and programming software is free. No I am not an average end-user. I like the command line. I like tweaking my config files and having my desktop look exactly the way i want it. I like being able to download software, add a feature that I want to it and then pass it back to the project admin for that software and then everyone in the world can have it. It's about community and power users. Let the Masses use Microsoft. I'll take my Linux, my Enlightemment my GTK and my souce code and be happy. And oh yes..I can run Microsoft, Mac, Commodore 64, Atari..and many other OS's programs under Linux too...
Joel (not verified)
on Jan 22, 2002
A very good article, all in all. Ximian is the shining star that could all make it happen. The desktop revolution will not happen soon, but it can only improve.
rob smith (not verified)
on Jan 24, 2002
I didn't find Paul's comments and thoughts unexpected on this site because it is, after all, "Windows" oriented. What I do find interesting is this constant chatter about a "viable desktop" which you can also find on many GNU/Linux sites; mostly from novice users. I believe what most GNU/Linux users experience, as they become more proficient with the system, is a gradual migration to the command line because of its speed and efficiency and the realization that the pretty pictures and colors are completely unnecessary. I can read text a hell of a lot faster than I can sort through graphics and there is no doubt in my mind that, using a GNU/Linux command line, I can do work faster, more efficient, more secure and stable than you can with any "Windows" OS.
John Madden (not verified)
on Jan 22, 2002
Keep in mind that not everyone wants or cares about "Linux on the Desktop." Since we're competing with an OS that's designed and marketed toward the masses (or rather, the morons), an arena few Linux distributions even attempt to encroach, the comparision isn't viable.
Gabriel Weisner (not verified)
on Jan 22, 2002
This is truly great! This is the funniest article I've read all day. Let's address the fallacies of Paul's World shall we? "The first chink in the desktop Linux armor appeared in late January when Corel announced that it would drop its popular Linux products and focus instead on its creative applications in an attempt to return to profitability. Predictably, the term profitability would be a recurring theme in 2001 as desktop Linux companies dropped off the map." Umm in actuality, Linux based companies (The real ones such as Red Hat, Mandrake, VajraMedia, Caldera, etc. not the overnighters) have no more or less less of a death ratio than the conventional dot-coms that utilized Windows based technologies.... Next! --- "Microsoft Senior Vice President Joachim Kempin. "Linux is too expensive," he said. (Let's address this by saying Microsoft is terrified that they have lost sales to Linux. Their new licensing schema has come back to haunt them and their trying to figure out how to counteract this. It's a damned if you do, damned if you don't scenario that Microsoft has boxed themselves into.) "This is because each company has to invest an unreasonable amount of time and money in Linux verification. As a result, the total cost of system installation and operation (TCO) is too high, effectively making Linux a very expensive OS. (This has been completely discounted by common sense alone. Any executive worth his salt would investigate all possible alternatives in this economic downturn we are now in. Again, Joachim Kempin is revealing his innate fear. He knows that Microsoft licensing schema's are costing the company money. Hence the real reason IMHO why Microsoft turned to a pro-forma accounting mechanism last summer. It's easier to mix the numbers to be brutally honest.) Vendors who are building up their Linux businesses are making a serious mistake and need to wake up to that fact quickly." (Again, there is story after story of successes with Linux implementations and the extra security and scalability that it provides over Microsoft's offerings.) But Kempin saved his most vitriolic language for a question about Linux growth. "Linux is simply a fad that has been generated by the media and is destined to fall by the wayside in time," he continued. "Windows 2000 will gradually overtake the Linux share in the server market. In fact, the advent of Linux has spurred Microsoft's developers to move up a gear. The arrival of new competitors in applications or OS development provides us at Microsoft with the driving force to create even better software products." (umm... How is this a Question? It is a statement not a question and he is simply unthinkingly repeating the Microsoft party line.) Next! "The recent security problems with Linux, " (umm... Are you joking? Microsoft Windows is THE most insecure OS out there! there are literally warnings from the FBI which state that using this OS is dangerous! Pot....Kettle....Black) coupled with the lack of key enterprise elements in the new kernel, (hmmm in the Win2k Beta that I tested had only support for 4 GB of memory. All of the sudden when Linux had 64+ GB capacity in the kernel 2.4.x series, Win2K had it. hmm.... I know.... I'm speculating... but it IS curious.) "really call into question whether Linux should be used at all," he said. Miller told Wired that the recent dot-com shutdowns also would be replayed in the Linux community. "Free does not sustain a business," Miller said. "Development costs money, [quality assurance] costs money, support costs money. We have yet to see a business model in the Linux world that has any chance of long-term success." (Hmmm... Gee... looks like Red Hat, Caldera, VajraMedia, SuSE are just non-existent eh? As I recall all four are posting profits.) Next! "An International Data Corporation (IDC) study reported that...." (I'm sorry but in the world of real business, where you must prove a business model in a due diligence process, IDC studies are laughed at. Just like the other so-called "analysts". These people are usually so clueless about what they write about it is the subject of email jokes.) Next! "As history has shown," Mundie said, (umm... you can't be serious? Mundie got his collective head handed to him during the debates. He was proven wrong so many times it's humorous. Keep in mind that Free Software has been around since before Microsoft's founders were even out of high school.) "this type of model isn't successful in building a mass market and making powerful, easy-to-use software broadly accessible to consumers." (umm... KDE2.2.x series desktops have just replaced over 730 in a well known brokerage house. It took us two days to teach these guys how to use it. They were so taken by the speed, the ease of use many are now using it at home. This will be the subject of a soon to be released study in Financial Services.) ... Next! "By mid-May, the most promising desktop Linux company, Eazel, called it quits. Founded by ex-Apple developers, Eazel had been working on easy-to-use desktop software. Eazel's stuff was good, and former employees promised to keep updating the software, but that effort mostly died off over time. Eazel's demise was probably the most damaging Linux desktop loss of the year. " (Paul now that is just made up from whole cloth. Eazel first of all had no business plan, IMHO. As a seasoned executive who has obtained venture capital, I would have never have funded their initial round. That goes with 80% of other dot-com businesses including many of those which utilized Windows technologies. There was no There... there. But to say that Eazel is not being updated is pure fabrication. It is part of the ximian desktop and coming along without a hitch. It's called product continuity son. It one of the strengths of Open Source. Next! Now... Let's skip down as I've a meeting to run to. "Netcraft chimed in on the Linux market-share issue in October, noting that Windows Servers actually dominate the Web with almost 50 percent of the market." (I'm not sure if they have yet, but I believe Netcraft has distanced itself from that survey as it was skewed by the discovery of a component in the new Internet explorer that made it appear as though there were a web server component.) Next! And last but not least.... "In early December, one of the few remaining viable desktop Linux companies, Ximian, released its watershed product, Evolution, a desktop-oriented personal information manager (PIM) package that apes virtually every feature in Microsoft's Outlook product. " (So first Open Source is criticized for not having applications. Then when a company builds them their criticized. Which is it?) "But Linux still lacks a credible Microsoft Office alternative; perhaps Ximian can work on an alternative in 2002." (No that is funny. Do you mean that the fact that out Office alternatives don't spread the multitude of security issues the MS-Office does makes it not a valid alternative? Please, try Open Office. Go there, download it. (they even have it for Win32 folks.) It uses XML for it's native file format. I've personally trained people on Open Office. Again, it's a two day affair as it looks so much like MS-Office, it's an easy switch. And Oh by the way it's (whisper....pssst....free) Uh Oh, there goes the TCO ratio tilting once again in the favour of Open Source. No vira included....sorry. That feature is one that Microsoft truly has the cornered the market on. Cheers, Gabriel
Jae Hawksworth (not verified)
on Jan 18, 2002
Paul, If you cut out the underlying editorial from you that Microsoft is bad and somehow Linux is good, the article isn't a bad chronicle of the Linux debacle. Do us a favor and cut out the anti-Microsoft rhetoric. This is a subscription based magazine site for Windows administrators. If we wanted the Linux hype, we can browse over to Linux Today's site and load up. Speaking of Linux Today, I understand they have a special call for journalist in love with Linux. Although we will really miss you here - not! Hurray and put your resume in anyway.
Debili Ipare (not verified)
on Jan 15, 2002
Paul, for someone who seems to crave data from the Linux supporters, you don't seem to provide much yourself. Rather, you provide other sites' arguments, add your own, and come to conclusions. Where's your data to back up your arguments. "Finally, 2001 ended on a controversial note with a Web Side Story survey reporting that only 0.24 percent of PC desktops use Linux, compared to about 98 percent for Windows" Exactly how would one go gathering such data? Looking at PC sales and noting what comes preinstalled? -- That would not work, because we all know how come Windows comes preinstalled on everything (for further data see Departement of Justice findings on Microsoft). Polling? -- That might work, but I still don't see any numbers. What data pool was used if that was the case? What about dual(mulit)-boot systems? Do they count as Windows only, do they count for both Windows and Linux? "The news was refuted by Linux advocates, who had no particular data to back up their claims, and by some others, including a humorous example in which a site called Low End Mac purported to contort its own server logs in such as way to show that Linux was in use by far more than 0.24 percent of the public." Yes, it makes sense to cite one idiotic example and disqualify all the others on that base. no data, no conclusion -- you're right about that
David Bakin (not verified)
on Jan 16, 2002
You forgot that 2001 ended on a positive note for Linux: After having started on the 2.4.x release in March 2000 (by having Linus open 2.4.x saying it was a "bug fix tree") and having achieved "code freeze" in June 2000, 2.4.0 was released Jan 2001 and finally "stabilized" at 2.4.15 in December 2001. Of course, 2.4.15 introduced a file-system killing bug. But never mind that, it was the start of the 2.4 stable release path! I'd say another top story of 2001 was that Linux DEVELOPMENT faltered. (Dates from Kernel Traffic mailing list archives.)
rm
on Jan 22, 2002
Look, I can debate all day long about which OS is better, but it's pointless because it it a really long road to travel. However, I will say this: say what you want, you cannot deny the course of evolution. Linux is evolving with our world and all aspects of mankind. Microsoft is not. Linux development is an evolving process. Microsoft development is an expensive, wretched process. That's why Microsoft claimed Open Source as being akin to a virus. Because it evolves. I won't argue with that. I believe in evolution, don't you? I saw the light several years ago. It just takes some people longer. I really don't mind seeing anti-linux articles. It really makes me laugh to see how scared the pro-Microsoft people really are.

Please or Register to post comments.

IT/Dev Connections

Las Vegas
September 30th - October 4th

Paul ThurottYou'll have the opportunity to experience:
• The Microsoft
Technology Roadmap
• Office 365 Implementation
• Hyper-V Optimizing
• Windows 8 Deployment
and much more!

Come See Paul Thurrott & Rod Trent in Person!

Early Registration Now Open

Upcoming Training

Mastering System Center 2012

During over 6 hours of training you can join John Savill from your computer as he will walk you through the key components and capabilities of System Center 2012, what’s involved in using the components, and the benefit they can bring to your environment.

Register Now

Current Issue

May 2013 - The NameTranslate object is useful when you need to translate Active Directory object names between different formats, but it's awkward to use from PowerShell. Here's a PowerShell script that eliminates the awkwardness.

CURRENT ISSUE / ARCHIVE / SUBSCRIBE

Windows Forums

Get answers to questions, share tips, and engage with the Windows Community in our Forums.