Biggest Story in IT for 2011

It depends what turns your crank but in the world of FLOSS, Android/Linux had to be one of the biggest stories. As well as the number of users who adopted the OS in 2011, >100 million, it seems to be a financial engine just getting warmed up. 2012 should be another great year.

While Android/Linux was huge, GNU/Linux was as well. The number of users is probably about the same as the number of users of Android/Linux. Even NetApplications reports nearly a 50% increase since July.

I still have no clue what changed last summer except for students being released into the wild after experiencing GNU/Linux at school. I am not surprised at the rate of growth of GNU/Linux but at the sudden reporting of that by NetApplications. NetApplications seems to have a particular clientele who pay to have their page hits counted. Probably they are businesses because schools and NGOs are not so generous with money. Has business seen the light and begun to adopt GNU/Linux more widely on the desktop? Is some smart thingie out there reporting as GNU/Linux rather than Android/Linux? Are there some more huge migrations happening? The magnitude of the change does not fit any explanation other than huge growth overall. The increase since July is about 0.4% of 1500 million PCs, 6 million PCs. That’s an order of magnitude larger than the Russian government. Perhaps it’s every school kid in Brazil. Perhaps NetApplications decided to count things more reasonably. Usually when they change methodology, GNU/Linux drops like a stone…

Perhaps people began to realize that the same benefits GNU/Linux gave on servers and Android/Linux gave on smart thingies could be had on other PCs.

Anyway, it will be very interesting to see how this change continues. At this rate, GNU/Linux will take over the world shortly: 2012 – 3%, 2013 – 6%, 2014 – 12%, 2015 – 24%, 2016 – 48% etc.

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.

30 Responses to Biggest Story in IT for 2011

  1. I am 61, more than two generations old. When I was 12 I was playing with radio, TV, high voltage PSUs and DTL (predecessor of TTL).

  2. Dr Loser says:

    Oh, and @Robert:

    Skills developed “over generations?” Are we turning Buddhist up there in the not-so frozen north?

    Neither one of us is quite old enough to have seen all that many generations go by, young feller!

  3. Dr Loser says:


    Rather late, I know, but I genuinely enjoyed your jeu d’esprit.

    Nice to see a bit of self-deprecating humour on a blog these days. I might even try it myself …

  4. Conzo says:

    If you put on a rotan skirt, you’ll look cute, when dancing around and away from the point.

    viz. the ‘uptick’, I don’t question the data, but it’s a glitch, and/or meaningless until backed up over more than just 6 months, and until the entire details of the metric (see the rest of the discussion) are clarified/proven to actually mean an increase in the way you argue.

    Until then, it’s just a not-too-well defined number increasing in too short a time to say anything meaningful about it.

  5. Conzo feels able and entitled to comment on my skills developed over generations but has nothing to offer in way of explanation for the uptick in NetApplications’ data.

  6. Clarence Moon says:

    “The absolute numbers are surely wrong but the trend must mean something, but what?”

    My statistics education memory is strained here, but I think the answer is that it is not materially significant. A sample, such as this, may or may not be representative of the general population, but let us say that it is. Else there are no conclusions to be drawn at all.

    So they have 160 million samples out of a total universe of how many? If you are correct in saying that there are maybe 1.5 billion PCs in the world in use today, that number of samples is a small number compared to the number of actual page hits in the time frame sampled. If people used the internet as much as I do, then there are trillions of hits from those billions of comuters.

    What it means is that there is a significant cause of error in the statistics, even for 160 million samples. They do not state their theoretical error margin, but if it were as low as 1%, it would still apply to all data. So the Microsoft percentage would be 83% +/- 1%, but the Linux percentage would be 1.4% +/- 1% as well and so a .4% increase would be insignificant, well within the margin for error.

  7. Conzo says:

    Wow, I thought I’d be nice and make the extrapolation a little bit less ridiculous by taking the whole year as a timebase for the 50%, but apparently you really, really, really want us to know you are such a hero of the toecheese brigade as to boldly extrapolate a sort of 50% increase in sort-of half a year to mean A now statistically backed up trend of ‘Linux’ usage doubling, did you hear me, doubling every year, from now on, for ever and ever

    Jeez, there’s just so much wrong with that.

    ‘Smart’ way to not ‘comment’ on your hilarious extrapolation skills.

    And to think, I offered you a beer ;(

  8. ch says:

    I wrote:
    “That’s what we all LOVE about you, Mr. Pogson”

    You choose to reply:
    “ch has nothing to talk about except HATING pogson”
    (Emphasis mine)

    Note the difference ?

    Oh, and I have some more numbers for you:

    Now please interpolate to your heart’s content. Really, we love you, or we wouldn’t be here.

  9. Clarence Moon wrote, “I don’t think that these statistics can be used as any sort of leading indicator of what might come to pass. “

    Agreed. The data is suspect but something has changed in the ecosystem of IT somewhere. It’s interesting to speculate on what that change is. I cannot correlate the change with anything on the web except school kids on summer break in Russia and Brazil, both places with huge roll-outs of GNU/Linux in government and education. The absolute numbers are surely wrong but the trend must mean something, but what?

    What happened in July 2011? Did NetApplications change, its clientele, or the world? The 160 million visitors come from somewhere, mostly USA/Europe, I expect. What are they doing differently? How long will the growth remain? It’s so huge that the end result can be quite dramatic. It certainly blows the “1%”ers out of the water.

  10. Clarence Moon says:

    I don’t think that these statistics can be used as any sort of leading indicator of what might come to pass. NetApplications gets its numbers from the huge number of website owners who subscribe to their tracking service and who end up contributing these counts. Unfortunately, the large, general population favorite websites are not subscribers to NetApplications services and so are never counted. If you only use a computer for your email and surfing on Yahoo, Google, CNN, etc., your visits are never recorded and so are lost for statistical evaluation purposes.

    Similarly, if you use your smart phone or tablet to connect to Netflix or Hulu or Amazon Prime or iTunes, you are not counted in NetApplication’s statistics. So that leaves a lot of stones unturned. NetApplictions says:

    “We use a unique methodology for collecting this data. We collect data from the browsers of site visitors to our exclusive on-demand network of live stats customers. The data is compiled from approximately 160 million visitors per month. The information published is an aggregate of the data from this network of hosted website statistics. The site unique visitor and referral information is summarized on a monthly, weekly, daily and hourly basis”

  11. ch wrote, “I don’t know what I’m talking about”.

    Of course ch has nothing to talk about except hating pogson. Sad, that waste of skin.

  12. ch says:

    “Since I don’t see their logs and know exactly what is doubling” I don’t know what I’m talking about – but I know that Linux will win in the end! Yes it will !

    That’s what we all love about you, Mr. Pogson – allthough for different reasons 😉

  13. OMG! I forget that so many people are numerically illiterate. The growth is ~50% in less than half a year… meaning it’s doubling per annum, or perhaps more. The actual % is 42% so the factor per half-year is something like 1.414… the square root of 2. Doubling per year it is.

    Since I don’t see their logs and know exactly what is doubling we cannot be sure it will continue but it is interesting and it leaves the “1%”ers up the creek without a paddle. I expect we shall know sooner or laters. Perhaps Dell and Ubuntu are gaining traction in China… They would tell us eventually. I just don’t see any event that matches the timing and continues for months.

  14. oiaohm says:

    Dr Loser Hypothetically and ineluctably I end up reducing to basically. I know I should not. But I can spell basically all the time. Hypothetically and ineluctably are words I have real fun spelling.

    Clarence Moon the web school one is normal for Linux to drop around Christmas. Also spike around middle of year. 5.3 at middle of year 5.1 near end of year is inside the normal flux. Dec and Jan is when you normally see growth. If you look back to 2008 2010 pattern look very much the same. Just higher.

    That w3schools statistic does have a bias. Note that site is mobile phone compatible. There is 1 percent turning up from mobile phones.

    Really the w3school statistics back my theory Clarence Moon of what is going on. So yes the numbers there is showing something.

    Question is how many sites NetApplications count system are mobile phone web browser incompatible. That might explain the difference between w3school and NetApplications completely. w3schools has seen a one percent in mobile. If that 1 percent is stuck on in Net-applications in a different place.

    As I say most of these web numbers are bogus. Too many factors can screw with them.

  15. Clarence Moon says:

    Wow, that is a real word, too! Thousands of NYT crossword puzzles and I had never seen it in print or heard it spoken. I’m going to put it on a 3×5 card and keep it handy.

    Another oddity that I have observed here is that the 40% growth in Linux heralded in this topic is not reflected in the usually more favored 3C Schools statistics. There, the Linux share is 5.1% today, but it was 5.3% a few months ago, showing that Linux is headed for oblivion long before 2016 rather than into the also-ran category thought to be a vast improvement over never-was.

  16. Dr Loser says:

    “Basically,” oiaohm?

    Is that the word du jour in your Markov chain nonsense?

    I look forward to the next one. “Hypothetically,” there’s a good one. Also “ineluctably.”

    I’m rather fond of ineluctability.

  17. oiaohm says:

    Really stone throwers you are incompetent. Do some homework what is going on is explainable.

    Population of devices that can run Desktop Linux has increased. Market share should increase accordingly.

    If you do homework you can install Desktop Linux in side Android in a chroot. So android taking off percentage of those uses may choose to add a Linux distribution to there phone. So yes we should see a market share growth for the simple fact there is more items that can run desktop Linux.

    How much is coming from the new platform space created by android devices for Linux Desktop Distrobutions. Who knows.

    How much might be faster conversions due to Munich success so admins have not be able to say conversion is impossible.

    Robert Pogson has made a mistake thinking that all the Desktop Linux that would be counted would be on a PC. You have to remember there is more mobile phones than PC devices in existence. So Linux desktop taking one percent from both market segments will appear as 2 percent.

    Once it crosses 2 percent I can start ruling out 1 percent installation in android devices. You would expect both markets of android and PC for hobbyists to convert at the same rate. So yes I am expecting at least an overall market share of 2 percent for Desktop Linux just from the sales of android.

    Also when you do the explain why you cannot see it comes common sense. If cannot see the 1 percent PC Desktop Linux that will mostly be stuck somewhere. What hope would you have of seeing 1 percent Desktop Linux on Android devices as well combining up to give 2 percent market share. Particularly thinking the android device using it contained desktop linux will just look like they are vnc to a remote server running Linux even that they are in fact vnc to the very device they are on. So yes the android Linux is highly invisible to the point of being unspotable.

    Basically desktop Linux on PC is highly more displayed than Desktop Linux on Android devices. If you cannot spot one I cannot expect you to spot the other.

    Any particular reason why someone might be installing desktop linux in android. Answer simple printer support. Desktop Linux printer support runs rings around Android printer support. For what you can do. Also browsers from desktop linux has less issues with particular sites than native android browsers.

    So the android devices showing up in the numbers I would expect.

    Basically if you could see Linux Desktop growth I would be shocked and this would suggest something more interesting is going on not just a simple case of runnable platform expand leading to expand in market share.

    This is why I have not focused on this. Its too early to know what is going on. Really revisit this at the end of 2012 maybe able to get a clue what is going on.

  18. Conzo says:

    I hear you (I swear to goddess, I do actually know about transfinite numbers, even though oddball Quantum Mechanics is more of a hobby) but since it’s way beyond my bed time anyway, I’ll have to get back to you on that sometime between my everyday musings

  19. Dr Loser says:

    Incidentally, Conzo, it depends upon what you mean by infinity.

    If it’s a countable infinity, like aleph-null, then I believe you are correct.

    If it’s (say) the infinity of real numbers, then you’re a bit buggered.

    But now back to our normal programming…

  20. Dr Loser says:

    Correct, but remember Cauchy.

    It’s quite possible for an asymptotic function to bounce about wildly (for illustrative purposes, we will consider the field of complex numbers between -1 and 1), and still be asymptotic.

    Somewhere within that range lives, or perhaps does not live, Schroedinger’s cat.

    Of course, Robert’s numbers are farcical, but we can leave such absurdities to the analysts at global banks. After all, they get paid for this rubbish.

    To his credit, Robert is stony-broke.

  21. Conzo says:

    Usually, an asymptotic would be decreasing ever more slowly until it approaches zero, like f(x) = 1/x, for x -> zero … An exponential function is increasing to an asymptote of infinity which is (if my rusty memory is correct) is technically indeed an asymptote, but usually classed as just ‘unbounded’. Another example of an unbounded function would be f(x) = log x for x -> minus infinity.

    Oh boy, this brings back memories 🙂

  22. Dr Loser says:

    Anyway, so: 11.39% (rounded up) in 2016. Yes, I can believe that.

    Nurse! Nurse! I need more powerful drugs!

  23. Dr Loser says:


    I’m (honestly) quite interested here. What would a non-monotonic asymptotic function be?

    I’m trying to imagine this. I suspect it involves Shroedinger and cats and stuff.

  24. Conzo says:

    Asymptotes are for sissies: real men know the power of extrapolating not even remotely close to the data 😉

    Don’t fear though – all the signs point to 2012 being the Year Of Linux On The Desktop!

  25. oe says:

    “Seriously, though, Robert. Didn’t they teach you about asymptotic functions during your Physics degree?”

    He beyond the Math-101,…while we’re looking at monotonic asymptotic functions, he’s familiar with sigmiod functions….

  26. Dr Loser says:

    Well, maybe it wasn’t that big after all.

    I’ll admit, it was a jolly good story, though.

    Now, Uncle Robert, can we return to nuclear armageddon in 2012? I seem to recall that the Iranians are going to be the good guys. My reading of Homer is at best shaky, but since we’re all sitting round the fire-side, gutting deer, let’s all have an Achilles moment, shall we?

  27. Dr Loser says:

    OK, I’ve gone off 2018.

    On your calculations, I make it more like 2022.

    Still not asymptotic, though, is it? But at least it gives eight billion people enough time to build little smart thingies that will accommodate the demand for Linux Desktops.

    Penguins be praised!

  28. Conzo says:

    Apart from that, even if you take the 50% growth on face value, it’s not the same as doubling, it’s more like … erm … one-and-a-half-ing. The base of the exponential function would be 1.5. Based on that, and – since I happen to ADORE penguins, dammit – starting with an optimistic number of 1.5% market share this here humble idiot arrives at:

    2012 – 2.25%, 2013 – 3.38%, 2014 – 5.06%, 2015 – 7.59%, 2016 – 11.39%

    Sorry dear Doctor – Penguin singularity will have to wait for some time longer.

  29. Dr Loser says:

    Seriously, though, Robert. Didn’t they teach you about asymptotic functions during your Physics degree?

  30. Dr Loser says:

    I’m looking forward to 2018. Based on your comprehensive mathematical wizardry, we’re gonna have to build a bigger world just to fit all those GNU/Linux desktops in.

Leave a Reply