Linux: Who got it right, who got it very wrong?

Linux: Who got it right, who got it very wrong?

Summary: Who predicted Linux servers would outnumber Windows servers by 2006? Who said one in five enterprise desktops would be Linux-based by 2008? We look back at the bad (and good) predictions made about Linux over the past decade.

SHARE:

Who predicted Linux servers would outnumber Windows servers by 2006? Who said one in five enterprise desktops would be Linux-based by 2008? We look back at the bad (and good) predictions made about Linux over the past decade.

.

Analysts picked up on the possibilities of open source pretty early on, particularly its ability to unify the long-fragmented Unix market around a single, consistent platform. The persistent incompatibilities between Unix distributions, each of which had been used by vendors to preserve their enterprise market share, led to great enthusiasm for Linux as an alternative, which was both open and far more consistent.

Back in 2000, Forrester Research pegged the Linux market — based on server revenues — at US$1.5 billion, to grow to US$2.5 billion in 2002 and US$15 billion in 2007. A 2002 Giga Information Group report was entitled "Linux has gone mainstream: are you up to it?" and predicted Linux would overtake Windows as the leading operating system on new servers by 2006.

Proprietary Unix is stone-cold dead

Forrester analyst Ted Schadler

A 2003 survey of enterprise shopping lists let Forrester Research conclude that these predictions were on track: 72 percent of respondents, the firm reported, were planning to use more Linux systems in 2004, a quarter were replacing Windows servers with Linux systems.

In the survey, only 46 percent of customers were holding off purchasing because of perceived lack of enterprise support — a concern that started to evaporate as IBM and HP led the charge to embrace Linux shortly after.

Forrester analyst Ted Schadler was effusive in his praise, declaring that "proprietary Unix is stone-cold dead" and warning that "laggard [independent software vendors] will regret their go-slow approach to Linux".

This decade, the success of Linux has created an imperative for companies of all types to be seen as being involved in 'open source'.

Such early enthusiasm appears to have been poorly considered: IDC's latest Worldwide Quarterly Server Tracker found that Windows still accounted for 36.6 percent of fourth-quarter 2007 server shipments, while Linux accounted for just 12.7 percent. Even Unix, so readily dismissed by Schadler, continues to post growth, with 33.3 percent of server spending and 1.5 percent growth from a year earlier.

While their assessment of Linux's eventual market share may have been wrong, predicting strong growth for Linux has been a sure bet.

Last month, IDC vice president of research Al Gillen confirmed analysts are still jumping over themselves with joy over Linux, noting that the market had actually reached US$21 billion by 2007 — far oustripping Forrester's 2003 prediction. He predicted the Linux server business would grow at 35.7 percent year-on-year to reach US$49 billion by 2011.

Trickle-on effect still a trickle
The phenomenal growth of Linux may have proven that open-source software has what it takes to run real-world businesses, but the way hasn't been so clear for other open-source products.

Standouts like the Apache Web server and Samba Windows interconnection application, have become nearly ubiquitous within corporate datacentres. The Firefox browser has also won strong support, as has the open-source Thunderbird mail client.

However, these are token victories that mainly offer new options for home users and small businesses. No other open-source application has enjoyed anywhere near the massive commercial success of Linux through its creation of an entire services and support ecosystem.

Instead, they have served as game-changers — motivators to encourage for-profit vendors like IBM and Microsoft to up their game and offer extra value in their respective products.

Indeed, this decade, the success of Linux has created an imperative for companies of all types to be seen as being involved in "open source" — with a similar imperative to that created by "greenwashing", where companies portray themselves as being environmentally friendly or carbon neutral. This wave of change has been even more significant than sheer adoption numbers, since it has even pushed Microsoft to make concessions to interoperability such as its recent Office Open XML file format standardisation.

This carry-on trend, which might rightly be termed "sourcewashing", has fostered some interesting benefits — for example, the move to open-source Java, IBM's open-source Eclipse integrated development environment, or Computer Associates' decision to breathe life into its once premium-branded Ingres database by releasing it free to the world in an open-source version.

Other tools, such as the JBoss application server or MySQL database, gained huge momentum that helped them stand out in the market and effectively eliminated demand for low-end, proprietary tools in many categories.

Nowhere is the gap between philosophical acceptance and actual adoption clearer than on the desktop.

Some open-source leaders, however, have still failed to make a big mark, particularly in application categories where larger enterprise companies have thrown millions of dollars in development funding such as business intelligence and customer relationship management (CRM).

In 2003, Forrester published a report entitled "open source collaboration platforms: give it five more years" — but now, five years later, the collaboration market is still dominated by the likes of Microsoft SharePoint Server and IBM's Lotus/Domino combination. Open-source Web 2.0-styled products, such as wiki servers, are certainly gaining in popularity — but their effect outside the firewall is far greater than inside.

A Forrester survey in 2005 of IT decision makers in Australia and New Zealand found that local businesses were still lagging North American counterparts in enthusiasm for open source.

In that survey, only 18 percent of firms were using Linux in production, and 82 percent respondents had no plans to adopt non-Linux open-source apps like Sugar CRM, the Open Workbench project management system, and the PostgreSQL database. Equally telling, only 10 percent of respondents were using or planning to use open-source software on the desktop, compared with 26 percent across the Asia-Pacific region.

A decade and a half after Linux appeared, it has succeeded in changing the terms of reference for the world's software producers.

The end of Microsoft? Hardly
Nowhere is the gap between philosophical acceptance and actual adoption clearer than on the desktop where — despite critical praise of recent Linux distributions such as Ubuntu and the backing of Sun for the OpenOffice productivity suite — Linux has still failed in its mission to supplant incumbent Microsoft and its Windows-Office dominance.

Early Linux-fuelled enthusiasm wouldn't have predicted the apathy the market has shown for the far lower-cost solutions offered by the open-source community. In late 2002, a Giga report projected that "the arrival of attractively priced competing office suites combined with dissatisfaction with current Microsoft licensing plans will create upwards of a five percent market share loss" for Microsoft.

Betting against Microsoft in any industry has always been a bad idea. In a March Forrester presentation, Giga's optimism proved misplaced: "The lack of a viable, enterprise-ready alternative to Microsoft Office — particularly an alternative to Outlook — will keep Microsoft firmly planted in the enterprise for the foreseeable future," Forrester concluded.

Ditto the desktop, where Windows continues to reign supreme. In 2004, IDC predicted that growing Linux adoption would push the operating system from three percent market share to seven percent by 2008, with sales of PCs running Linux to hit US$10 billion. Even those figures paled compared to the predictions of Siemens Business Systems, which in 2003 predicted that Linux would have captured 20 percent of the enterprise desktop market by 2008.

It is now 2008, and Windows is still the dominant operating system; if anything, Mac OS X has supplanted Linux as the alternative desktop of choice. Linux is out there, but erratically: it runs, for example, on ASUS's popular Eee mini-notebook PC and in March was chosen by IBM for low-cost PCs to be shipped to customers in Eastern Europe. Despite a few isolated purchases, however, desktop PC purchases are still all about Windows.

Open source as a force for change
A decade and a half after Linux appeared, it has succeeded in changing the terms of reference for the world's software producers. IDC believes the services and software ecosystem around Linux will grow from US$18 billion in 2006 to be worth over US$40 billion by 2010.

Open-source software, once derided as the unsupportable and unreliable work of hobbyists, has secured itself a place at the table of many a large enterprise, although in ways that were different than many expected even five years ago.

By promoting a culture of transparency and forcing independent software vendors and service providers to add value in new ways, open source can be lauded for shaking up an industry that was ripe for a change. Instead of thinking in terms of market domination and customer lock-in, the industry is now prepared to work for its lunch — and to respect the customers' desire for a better option, whether it's open source or otherwise.

Topics: Open Source, Hardware, Linux, Microsoft, Reviews, Servers, Windows

About

Australia’s first-world economy relies on first-rate IT and telecommunications innovation. David Braue, an award-winning IT journalist and former Macworld editor, covers its challenges, successes and lessons learned as it uses ICT to assert its leadership in the developing Asia-Pacific region – and strengthen its reputation on the world stage.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

16 comments
Log in or register to join the discussion
  • Thank you

    As a citizen of the United States I wish to thank the countries of Australia and New Zealand for transferring large amounts of your national currency to Microsoft. As an American corporation, Microsoft contributes significant revenue to our Federal Government. That revenue was used in our invasion of and subsequent occupation of Iraq. To this day your dollars continue subsidize this effort. President George Bush would thank you himself but he has finally made the connection that the less he says, the better off he is.
    kozmcrae
  • All roads lead to tax ...

    So instead you propose placing that currency in the hands of Sun, IBM, HP .... oops - no real alternative [again]. I know lets all build our own free version - a new prediction for 2009 perhaps.
    anonymous
  • nice article

    @Richard Chapman: LOL! -illegal- occupation and massacre!

    wave of the future is here - no going back!!

    try linx, be free..
    anonymous
  • Free Version

    There already is a "Free Version" of UNIX - The GNU Operating System. It could be the answer everyone is looking for. Linux, because it is effectively owned by IBM, due to its massive $1 billion investment in it in 1999, is (and will always be) devoted to the corporation. There is nothing wrong with that. But a viable and truly free alternative to Microsoft can be created thru the GNU project.
    anonymous
  • Nope, VERY wrong

    Russ, although IBM donates resources (hosting, developers, money) to GNU/Linux (yes, G N U/Linux), they don't own it. No single person or entity OWNS Linux and cannot, therefore, claim sole control and beneficiary of it.

    In fact, MANY companies benefit from GNU/Linux, hence, they benefit from the work some of their competitors did in improving the kernel and surrounding software. Also, there are other contributors that you miserably fail to realize, like Red Hat, contribute just as much or more than IBM. I think IBM sticks out in your head because they were then the only "proprietary" company that decided to make a big investment in GNU/Linux.

    Last point, you mention "The GNU Operating System." Ummm, unless you mean GNU/Herd, an unrefined kernel, there really isn't such a beast. In fact, GNU's page "GNU Operating System" states it wasn't completed so GNU uses the Linux kernel. They have a list of sufficiently-Free distros to pick from - no proprietary stuff.

    Better luck next time trying to drive some wedge where it doesn't belong. ;)
    anonymous
  • RE: Thank You

    So what you are saying is that anyone that wasn't for the war should stop support M$ asap and switch completely over to free software :)
    anonymous
  • Money

    OK, obviously the early speculation and hype were wrong, Microsoft still just about owns the desktop, etc....

    But how can OSS market penetration be measured in dollars? Most open source software is also available for free. Sure, most corporations prefer to purchase it from big names w/ support packages but how many servers out there were set up by some guy in IT whom downloaded a free copy? How about small businesses?

    Good luck ever measuring this.....
    anonymous
  • Agreed

    Who counted the servers purchased with no OS / Support package? How many of the servers counted are even running the OS purchased with them?

    Every article about who is winning what is all based on data which is impossible to prove the validity of (unless somehow...someone managed to nmap the internet).
    anonymous
  • Linux servers must be ahead then

    [Such early enthusiasm appears to have been poorly considered: IDC's latest Worldwide Quarterly Server Tracker found that Windows still accounted for 36.6 percent of fourth-quarter 2007 server shipments, while Linux accounted for just 12.7 percent. Even Unix, so readily dismissed by Schadler, continues to post growth, with 33.3 percent of server spending and 1.5 percent growth from a year earlier.]

    If Linux accounts for 12.7 percent of server shipments, and Microsoft Windows 36.6 percent, then Linux must be ahead since most Linux OSes installed on production servers are shipped separately. For every Linux OS shipped with a server, there are three or more acquired separately.
    anonymous
  • RE: Free Version

    Sorry, TK, but you clearly do not understand how the Free/Open Source Software world works.

    IBM does not donate to the GNU Project - or the Free Software Foundation for that matter. It is philosophically opposed to both groups. That is why it and other industry players developed the term "Open Source."

    As I correctly stated in my first post: IBM "effectively" owns Linux. The keyword being "effectively." Through their (and other corporate players) control of the industry trade group called the Linux Foundation they control the development of the kernel. Linus is an employee of that industry trade group. Effectively making him an employee of IBM and the other big players. As I said before there is nothing wrong with this. However, corporations have a set of needs, desires, and goals that are very different from (and frequently run counter to) those of home and small business users. The kernel Linux will always be developed in a way that advances industry goals.

    IBM "sticks out in my head" (as you put it) because they are by far the largest financial backer of the kernel, Linux.

    As to the GNU Operating System, again you are wrong, not only does the GNU Operating System exist, but there is no such thing as the "Linux Operating System." Linux is simply a GPL'd Unixoid kernel. A kernel in turn is only one small (albeit, important) component of an operating system.

    While you are right, the "kernel" to GNU is unrefined, that is all the more reason that people should be helping to develop it.

    In your last line, your paranoia seems to have gotten the best of you. No one is trying to drive a wedge anywhere. I'm just pointing out the obvious - namely, that the GNU system would be a much better choice for the desktop than Linux.
    anonymous
  • Most Servers ship with no OS

    Other than small businesses, you usually don't buy the OS with server hardware. The larger the business the more likely it is to find that they have a subscription agreement with OS vendor (i.e. IBM, RedHat, Microsoft).
    anonymous
  • linx

    do you have the link for linx ?
    anonymous
  • Survival of Microsoft depends upon Microsoft itself

    Thinking that Linux will eliminate Microsoft from desktop user is totally wrong .The survival of Microsoft depends upon how the management of Microsoft manges the Microsoft itself . Now someone might ask ,Linux is free how come it has not been able to replace Microsoft which cost money ? The answer is still depends upon how Microsoft does marketing . Water is basic for survival and it is free but Coca-Cola is still making billions of dollars why ???is it because of water ????
    anonymous
  • Why Linux PC?

    How much does a consumer benefit from using Linux PC? Lets face it, PC has become such a commodity, cheaper than a TV. Why would anybody care it is a linux or a windows as long as it works fine. The cost of windows OS to OEM? $10 in Asia and $20-30 in US? With all the free stuffs coming with windows os, you get the os itself essentially free. For God sake, they charge $30-50 for a game or a DVD, and how many games and DVD a computer user have purchased. The point I am trying to make is that the cost of windows OS itself is almost nominal and it is totally misguided to focus on the free aspect of Linux. By the time someone packages enough other programs to go with linux and provide the same functionality of windows, it probably is more expensive than windows.

    People buy bottled water although they could get better quality water from their tape at home. Why?

    Because it cost too much EFFORT to package your own water!
    anonymous
  • Linux or Windows

    Just like Windows unravelled with NT4 and finally the world said enough is enough - and Microsoft had to come out with Windows 2000, something that wouldn't crash every 2 weeks like clockwork - this is where Linux is today. Linux is about 10 years behind Microsoft in quality of programming code. As soon as there are major loss issues that come about with Linux you'll see the world finally tell the Linux Hackers that enough is enough. Linux us just being given a chance at this point in life. I myself have found the major flaws in Linux's way of writing to drives and memory. Everyone knows that computers are all about how well they write -- because there is no such thing as "delete or erase" in the real computer world. Most Linux OS's don't even properly write to the MBR and when you delete and create new partitions in Linux - new Partition Numbers are given the new creations - and Linux doesn't even have the presence of mind to find the correct place for hda1 hdb1 to write to. You have to use a Windows Utility Disk to fdisk and format the drive to get rid of the journal entries telling Linux where the MBR is to be written physically on the hard drive. Here's some interesting stuff that is coming from a company out of Standford U. that is being done for the U.S. government. This is what I've been talking about and asking the Linux Developers to fix - but, they refuse to listen to me and/or fix the issues. For me - enough is enough. It's time to put Linux in the Has Been Box and move on.


    Flaws found in open source codes

    Tom Espiner, ZDNet UK

    21 May 2008 08:56 AM

    Tags: amanda, coverity, ntp, open source, openpam, perl, php, python

    A project funded by the US Department of Homeland Security has praised improvements in open source security, while outlining some common errors.

    Coverity, a commercial code analysis company spun out of Stanford University, has been running its Scan project with Department of Homeland Security funding since 2006.

    On Tuesday, Coverity released its Open Source Report, which gives results of bugs in more than 250 open-source projects. Coverity declined to give details of individual projects' faults, but instead gave a list of frequent coding errors.

    Coverity did single out some projects for particular praise due to the cleanliness of the code, including Amanda, NTP, OpenPAM, OpenVPN, Perl, PHP, Python and Samba.

    The most common type of code defect, accounting for 28 percent of all the defects found, was the NULL pointer dereference, as the use of pointers in C/C++ is error-prone, according to Coverity.

    "This type of error often occurs when one code path initialises a pointer before its use, but another code path bypasses the initialisation process," stated the report. "Pointers are a notoriously challenging programming concept that many languages elide altogether (eg Java). Senior developers know that new programmers frequently have trouble understanding pointers."

    Because pointers are often used to pass data structures by reference between pieces of program logic, they may be the most commonly manipulated data objects due to repeated copying, aliasing and accessing. Therefore, it is not surprising that the most frequently used artefacts will incur the most errors in manipulation, said Coverity.

    The second most common type of code defect are resource leaks. While some resource leaks are pointer-related, others may be the result of misusing an application programming interface (API), said Coverity.

    Other common code defects include unintentional ignored expressions, use before test and buffer overflow vulnerabilities.
    anonymous
  • Why Not a Windows PC

    and here's some of why. 1st - Linux developers don't listen and they are stuck in forcing people to use broken softwares like MacIntosh used to do - and Windows sometimes with Vista - but, Linux is worse off because there is no order to the chaos. At least at Microsoft you have Balmer kicking you out the door if you are costing the company money with scrapware. I didn't write this next stuff - but, copied it form another page here at ZDNET --
    Flaws found in open source codes

    Tom Espiner, ZDNet UK

    21 May 2008 08:56 AM

    Tags: amanda, coverity, ntp, open source, openpam, perl, php, python

    A project funded by the US Department of Homeland Security has praised improvements in open source security, while outlining some common errors.

    Coverity, a commercial code analysis company spun out of Stanford University, has been running its Scan project with Department of Homeland Security funding since 2006.

    On Tuesday, Coverity released its Open Source Report, which gives results of bugs in more than 250 open-source projects. Coverity declined to give details of individual projects' faults, but instead gave a list of frequent coding errors.

    Coverity did single out some projects for particular praise due to the cleanliness of the code, including Amanda, NTP, OpenPAM, OpenVPN, Perl, PHP, Python and Samba.

    The most common type of code defect, accounting for 28 percent of all the defects found, was the NULL pointer dereference, as the use of pointers in C/C++ is error-prone, according to Coverity.

    "This type of error often occurs when one code path initialises a pointer before its use, but another code path bypasses the initialisation process," stated the report. "Pointers are a notoriously challenging programming concept that many languages elide altogether (eg Java). Senior developers know that new programmers frequently have trouble understanding pointers."

    Because pointers are often used to pass data structures by reference between pieces of program logic, they may be the most commonly manipulated data objects due to repeated copying, aliasing and accessing. Therefore, it is not surprising that the most frequently used artefacts will incur the most errors in manipulation, said Coverity.

    The second most common type of code defect are resource leaks. While some resource leaks are pointer-related, others may be the result of misusing an application programming interface (API), said Coverity.

    Other common code defects include unintentional ignored expressions, use before test and buffer overflow vulnerabilities.
    anonymous