Web Stats for June, 2011

W3Schools has the stats up for June. That other OS is down to 85% share. Android/Linux has had 100% growth in the last six months, 300% per annum, while iPhone has been static, iPad has had 50% growth. GNU/Linux’s share rose 8% in the last year while that other OS’s share dropped 3%.

It looks like “7” might finally catch XP after two years of retail sales. Since the world is getting off the Wintel treadmill, this may be the last time any release of that other OS reaches 50% share. Between switching to thin clients, ARM and longer refresh cycles, the world will soon find that other OS is not “the standard” desktop.

About Robert Pogson

I am a retired teacher in Canada. I taught in the subject areas where I have worked for almost forty years: maths, physics, chemistry and computers. I love hunting, fishing, picking berries and mushrooms, too.
This entry was posted in technology. Bookmark the permalink.

26 Responses to Web Stats for June, 2011

  1. Contrarian says:

    #someone had asserted that Linux was “dominant” in software areas outside the desktop and that is plainly not true. If you want to say that Linux is dominant except for servers and clients, I guess that is true. I think that what is also true is that Windows is dominant in any environment where the device supplier is paying someone else for the OS platform software.

    Some day that market opportunity will start to decline, but not so soon, I think. Meanwhile, it just continues to grow more slowly than in the past due to the side attractions of smart phones and tablets.

  2. Servers: GNU/Linux approached UNIX revenue in 2011. They were close in 2010 but GNU/Linux was growing rapidly while UNIX was shrinking rapidly.

    see http://www.idc.com/getdoc.jsp?containerId=prUS22360110

    While GNU/Linux server revenue may not catch M$ soon, GNU/Linux was at $1.7billion in 2010 and growing well.

    GNU/Linux server revenue was $2billion in 1Q2011 as counted by IDC, while UNIX revenue was $2.6billion. That’s only counting factory shipments. GNU/Linux grew 16% compared to that other OS growing 10%. UNIX grew 12.5%.

    GNU/Linux definitely is in the race and positioned to gain second place soon. There are all kinds of places where GNU/Linux has great share not counted by IDC like HPC and web serving. Where that other OS has a lock is holding the hands of the client PCs rendered so pathetic by that other OS. As that other OS declines on the client, it will on the server as well.

    Other accounting of server share show GNU/Linux far ahead of UNIX already.

    On client PCs GNU/Linux is also growing steadily as Canonical gets more OEMs to ship Ubuntu and other distros also are growing. The netbook is not dead and several new versions shipping with ARM processors run GNU/Linux or Android/Linux in 2011. That will only increase as ARM is still ahead of Intel in power consumption. Tablets and smart phones are taking some “PC” share and that will increase as the new quad-core CPUs and tablets rub off on consumers.

  3. Contrarian says:

    “As Linux and the distributions built on it are open source, they can’t simply be killed by low market share. It’s always going to be there, steadily improving and growing”

    That is the popular thought amongst Linux fans, I agree. But realistically speaking, desktop Linux has been a sub-single digit player and its more organized proponents such as Linspire (nee Lindows) and some of the European distributions that I do not recall the names of have vanished from the effort. Ubuntu is carrying the torch today, but it is hard to see where it is making any move toward “steadily improving and growing” in terms of acceptance.

    Unit volume-wise Linux is running a distant third to Windows and branded Unix in the server arena as well. Dollar wise, it is in even worse shape.

    If you want to count telephones, go ahead, but it is not the same thing and I think you all know that.

  4. Contrarian wrote, ” Linux was never even in the game, starting far too late to ever be considered.”

    GNU/Linux has been in the game since 1991 and will not stop no matter how much such prattle is repeated. GNU/Linux on servers and desktops has a major share and it’s increasing. That other OS has a major share and it’s decreasing. M$ cannot win without making exclusive deals and that’s happening a lot less these days. Even Intel is supporting GNU/Linux. Every industry/business category is adopting GNU/Linux and consumers around the world are beginning to see it on retail shelves.

  5. Someone says:

    There is, though, one significant difference here. The products that you mentioned were all proprietary. So when they entered the marketplace, market share and the money it brought were their lifeblood, and when that dried up, they died, probably never to be heard from again.

    As Linux and the distributions built on it are open source, they can’t simply be killed by low market share. It’s always going to be there, steadily improving and growing. It’s already dominant nearly everywhere but the desktop, so I doubt very seriously that it’s going to fade away. Even Microsoft is unlikely to maintain its monopoly forever, and since Apple is too content to maintain its low volume/high margin business model, when Microsoft eventually does fade, the thing that replaces it will likely be some form of Linux (or maybe BSD, but even that would be an improvement). What else could replace Microsoft? It’s already killed any and all of its competition that could be killed, and building another full OS from scratch seems too much like reinventing the wheel unnecessarily when Linux and BSD are already out there and mature.

  6. Contrarian says:

    “but how is this in any way a shortcoming of Linux”

    There is no “fault” involved for Linux. The missing functionality of Linux in terms of popular product support is due simply to its overall lack of popularity. There can only be on winner and second or third place in software businesses is always extremely far behind. MS Money could not come anywhere near Quicken and so is now history. 123 lost to Excel, WordPerfect lost to Word, dBase lost to Access, and so on. The losers disappear from the commercial scene once the public has voted with the near universal adoption that follows their product selection.

    Similarly Google search is not threatened by Bing, Facebook and Twitter have no real competition either. That is just the nature of the PC experience. IBM’s OS/2 lost to Windows in the early 1990’s. Linux was never even in the game, starting far too late to ever be considered.

  7. Someone says:

    And is it Linux’s fault that Adobe chooses to only write its software for Windows and OSX (a close cousin to Linux)? Is it Linux’s fault that Microsoft, its most venomous competitor, chooses not to release Microsoft Office for Linux? Not that these programs are necessary to a great many people (including Windows users), but how is this in any way a shortcoming of Linux?

  8. GNU/Linux should not be judged on the merits of other peoples’ programming errors. e.g. coding for that other OS. I can write programmes that will run on any OS. So can others but they drink the KoolAid.

  9. Contrarian says:

    “They would have been kicked aside 15 years ago if not for their illegal exclusive dealing”

    That is said a lot, it seems, and the law is supposed to demand corrective action for any market imbalance caused by violations of the antitrust laws. That is apparently why Microsoft had to offer a version of Windows in the EU that did not have the MP3 player software pre-installed so that others would have a chance to get their product installed.

    When you look int the exclusive dealing, though, the facts of the matter are that Microsoft was not found to have violated any law regarding exclusive dealing. They were found to be under the limit in terms of restricting market access by the original district court ruling. The legal limit seems to be set at 40% of the available market and Microsoft had not exceeded that limit.

    I think that the real answer is that Linux is not competitive since it requires so many changes to popular programs in order to run them under Linux. Plus the procedures are kind of complicated, for example using the WINE interfaces, and the programs do not run exactly the way they should in a lot of cases.

    Saying that there is FOSS software that is “just as good” is empty words for most people. They want the real thing and that is why they do not consider buying a Linux desktop computer.

  10. oldman says:

    “The PC OS is a commodity and M$ is not competitive.”

    This is nonsense. People run applications not operating systems, and microsoft not only provides the premiere Platform for desktop applications, but it even runs FOSS as well. Linux in comparison, is severe limited in desktop applications.

    I lived through the same period that you did Pog. Whatever shenanigans microsoft might have been engaging in, their position rested on the applications that they and their ISV’s delivered to their user community. To say otherwise borders on denial of reality.

  11. If the relevant market is the “PC OS” then M$ is doomed. Their product costs $50-$100 and does less than GNU/Linux or Android/Linux. “Less” means not doing 100% of what needs to be done. e.g. phoning home, inviting malware with needless APIs, etc. The PC OS is a commodity and M$ is not competitive. They would have been kicked aside 15 years ago if not for their illegal exclusive dealing.

  12. Contrarian says:

    “the whole issue of market share is to me bogus”

    I think that #pogson is misidentifying what is the relevant market. He wants to lump anything with a processor together and then count how many units are produced. In that way a cell phone is counted on the same basis as a mainframe or server. The reality is that there are dozens of relevant markets that manufacturers focus on.

    Microsoft triumphs in the PC OS market and almost anything sold for money in that market results in their income rising. People may avoid buying a new copy of Windows if they install Linux, but I don’t think many people ever actually do that. I have tried it as an experiment, but I am a techie sort. No regular person that I know has ever done it.

    Apple sells its computers with its own OS and maybe sells some copies to existing Mac users as an upgrade, but I don’t think that is very much money when compared to Windows revenues.

    In other markets, Microsoft doesn’t do so well. They do not make a phone or a tablet and the money in those markets seems to stem from selling the hardware. No one is selling much software in those arenas and even the apps are mostly free or very low cost. I don’t think that anyone is making very much money there with the large number of vendors and the hundreds of thousands of apps available. You cannot stand out very well from promotions when the prices are so low and you get lost in the crowd.

    Some day in the next year, the Nokia deal with Windows Phone will get off the drawing board and it will be interesting to see how the market for phones reacts to that. I have some confidence in the power of the brand and the two together are a force to be considered.

    Microsoft does make a lot of money from office automation software and also from game consoles and software, too. Gaming is bringing in a couple of billion dollars a year now and they have recouped all of their previous losses on Xbox.

  13. Thanks for the feedback. I can easily produce graphs from the w3schools data and others. Love that stuff. I have a script I can run on Wikipedia data. What I really want is Google’s data and they are not selling it at least at a price I can afford. That should be much less biased than the usual web stats.

  14. oldman wrote, “trefis is an example of a group of idealogues producing suspect statistics pro Linux”.

    “Trefis.com was founded by MIT Engineers and former Wall Street analysts who realized that most people do not understand the seemingly familiar companies around them. For example: Apple, Google, Coca Cola, Walmart, GE, Ford, Gap, and others.

    We understand that multiple factors, some quantifiable, and some not so easily tackled influence a company’s stock price. The Trefis view is a very detailed fundamental model of the business, one of the most rigorous one can find in the industry, yet simple and easy-to-use. However, it is still just a model – an artificial representation of the real company, which is much more complex. By providing a transparent, interactive, and intuitive platform we hope to empower you to better make sense of companies and the businesses around you.”

    see https://www.trefis.com/about

    In what way are Trefis’ analysts idealogues? They are number-crunchers experienced in evaluating stocks.

    They don’t even follow RedHat…

  15. oldman says:

    “Web stats are biased and easily manipulated by the likes of M$. ”

    That cuts both ways Pog. And trefis is an example of a group of idealogues producing suspect statistics pro Linux.

    “Also, any way you look at it, the share is dropping.”

    Only if you look at linux or tech biased sites. In the end Pog, the whole issue of market share is to me bogus. The desktop market is for all intents and purpose based on commercial software (Apple and microsoft) as is the mobile market (Blackberry OS, Android, MS).

    Pure free software is basically nowhere in either of these spaces, and that IMHO is where it will remain.

  16. radu says:

    While a less than ideal sample can distort share, trends from one measurement to the other tend to be relatively accurate unless there is a clear difference between the sampled population and the general population.

    Therefore the change over time of the OS/browser shares would be meaningful even if snapshot percentages are wrong.

    Mr. Pogson, I would appreciate it if you focused on the trends in future posts.

  17. Web stats are biased and easily manipulated by the likes of M$. The stats show shares like 86% but Trefis, having examined better data, uses 76%. Also, any way you look at it, the share is dropping. Only a couple of years ago, W3Schools was showing 90%+. Now its 80%+, down a few % in a couple of years. NetApplications says it’s Apple taking up the slack but Apple published unit shipment numbers nowhere near the share necessary. There is 5-10% share missing from the numbers. As the share of M$ declines, NetApplications has to come up with bizarre fictions to make it work. It’s getting harder all the time.

    M$’s share of the desktop is no longer a given.

  18. oldman says:

    “Yes. It’s part of the mythology of IT that GNU/Linux is not making it on the desktop.”

    It depends on where you look Pog. I’m not surprised about the higher stats in europe, though I think Mr. Hagglund’s statistics are probably skewed by the clientelle. IN contrast, by several statistics gathering sites in North America, Desktop Linux has dropped below the 1% mark and is now bundled in with other.

    One of the things that never gets discussed is the reality that the older systems that are “resuscitated” with desktop linux simply don’t count anymore – the money has been made on them, and the people who own them are unlikely to contribute to anyones bottom line. In that sense, desktop linux doesnt exist.

    This may sound unfair, but it is nonetheless a reality, as is the fact that the third world would still rather pirate western commercial software than use FOSS.

  19. Dann says:

    “but W3Schools shows IE: 23.2 % and Firefox: 42.2 % … WTF?”

    Firefox is more standards compliant than IE.
    Friends don’t let friends use Internet Explorer.

  20. Yes. It’s part of the mythology of IT that GNU/Linux is not making it on the desktop.

  21. Mats Hagglund says:

    Statistics of my blogsite are showing that there are indeed more and more Linux-users (now steadily over 10% from Sweden and Finland mostly). It’s interesting coz that blogsite consists hardly any geek or even computer related writings. It’s mostly society, politics, culture etc…

  22. Ray says:

    8ut when wikimedia is at 1.56%, that’s quite low.

  23. W3Schools has been around a while and are consistent. They also reveal their sample, their own site. Those others keep that secret, so we cannot know much about their sites. M$ spends so much on advertising they could easily fund a whole bunch of counter accounts pointing to themselves or their partners. There are literally millions of sites associated with M$ and its partners.

    The ~1% numbers are just not believable. Proper surveys showed GNU/Linux at several % years ago, some of them by M$. Android, alone, counts for about 1% and it got that share within a couple of years. GNU/Linux has been growing geometrically with higher growth than that other OS for much longer.

  24. W3Schools is heavily into technology that runs only on that other OS, like .asp and .net and VBscript. Further, it’s English. That’s a huge bias just like NetApplications sample of the universe or Wikipedia.

    I like the data because it is not from NetApplications, a supplier to M$.

  25. JairJy says:

    Lets compare W3Schools with other stats sites.

    Net Applications shows IE: 53.7% and Firefox: 21.7%
    StatCounter shows IE: 43.6% and Firefox: 28.3%
    W3Counter shows IE: 36.3% and Firefox: 28.2%
    StatOwl shows IE: 61.28% and Firefox: 19.85%
    Wikimedia(May): IE: 39.5%; Firefox: 25.9%

    but W3Schools shows IE: 23.2 % and Firefox: 42.2 % … WTF?

    Now if we look after Linux market Share:
    AT Internet (Apr): 0.9%
    Clicky Web Analytics: 1.28%
    Net Market Share: 0.95%
    Global Stats: 0.74%
    StatOwl (May): 0.70%
    W3Counter: 1.53%
    Webmasterpro: 1.3%
    Wikimedia (May): 1.56%

    But again, W3Schools shows a strange 5.2% to Linux.

    Why do you follow W3Schools?

  26. Contrarian says:

    “W3Schools has the stats up for June.”

    This site seems to me to be very oriented towards technical web developers and is not a very likely destination for ordinary users. Since so much of the web is based on Linux servers and open source programs such as PHP and Apache and MySql, you would naturally expect that there would be more users accessing it with Linux. Also, such users would categorically be inclined to have the latest Windows version, too. So the decline in XP would naturally be accelerated for their own workstations.

    What puzzles me is that, with some 70% of the web sites running on Apache and Linux today, only 5% or so of the web developers actually use Linux. Doesn’t this seem odd to you, #pogson?

Leave a Reply