Wintel Is French Toast

I love French toast. So tasty! So nutritious! So sinful!

Here we have another sort of French toast. In the past few weeks, France has made a huge move in page-views by GNU/Linux according to StatCounter. Curiously, similar moves have been seen in several other countries about the same time. It could be just a change in StatCounter’s universe, a big one. It’s in Canada, too, which is not tightly related to France except historically. Only about 25% of us speak French regularly. I studied the language for years and only know enough to be dangerous…

Other statistics: France has about double the population of Canada and Internet penetration over 80%. A move of 3% share of page-views amounts to 66million X 80% X 3% = 1.5 million users. That’s too large to be one government/education/business, the big players. Any concerted move like that would make the news of which we see none. There are moves about to secure networks and servers against the NSA and other evil-doers. Has that spread to the desktop? The government of France certainly is aware that FLOSS and GNU/Linux could save them a bundle of money. There are groups in France actively promoting FLOSS but they lack the resources to produce such a sharp move, so what’s up? Are GNU/Linux desktop PCs selling like hotcakes in France? The French seem to be having fun with Ubuntu GNU/Linux. Is this a popular revolution? Inquiring minds want to know.

See Top 7 Desktop OSs in France from 1 Apr to 27 May 2014.

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 and tagged , , , . Bookmark the permalink.

27 Responses to Wintel Is French Toast

  1. Thanks for the link, oiaohm. That demo-data is from an online seller of musical instruments. Their software seems to be .exe only… For what it’s worth, they are getting 2.9% GNU/Linux, 0.8% ChromeOS, 0.4% Android tablets and 9% mobile phones. So, we are close to ~10% */Linux, what I thought we had 8 years ago… They also show Ethiopia contributing 0.1% to their hits. If you search the whole universe for desktop+mobile phone, we find several countries where Android/Linux gets more hits than that other OS. Globally, StatCounter shows 14% Android/Linux. e.g. India 27.9% Android/Linux, almost double “7”, Kenya 28.5% Android/Linux and many where Android/Linux comes in second to “7”, e.g. South Africa 18%, Brazil 14%, all of Asia, 20%.

  2. oiaohm says:

    Basically all the MS troll quoted stats are good for is detecting distrubances. Other than that the error is too huge. +/- 60 to 70 percent. As I say too huge is joke. How can you find this out some countries only pc computers own to the governement and they have exact data on what is deployed and yes neither Netapplications or Statscounter comes close. So how far they are out based on those countries generate the possible error value.

    Net numbers have been taken by truth by Trolls when reality every time you test them they fail so massively its not funny. Robert is using the garbage numbers about the only way they are usable. To detect changes to current activies.

  3. oiaohm says:

    Of course Robert knows is blog.
    I did not say I was using top level data. Inside statscounter data you can drill down to state by state just like Netapplications. Netapplication displays state by state to public statcounter does not. Yes the same defect is in statcounter.

    Remember the population of Computers in California is larger than many small countries.

    Loser you have entered an arguement with a person who handles raw stat data. Google is not going to help you.

    This is the problem Loser you have never invested the money or the time to find out if any of the stats you love quoting are anything more than FUD. You say you want to prevent Linux people from spreading FUD its about time you get your own house in order and start facing the truth. We don’t know how much Linux or any OS is out there and most likely will never know. All collection methods are flawed to the point the allowance for error is that huge its laughable.

    Old saying Truth, Lies and Statstices. Without knowing how Statstices were collected you cannot know if they are telling the truth or telling lies or a mixture of both. If you cannot get how they were collected exactly then you search checkable events.

  4. DrLoser wrote, “There is no “Google migration disturbance to StatsCounter,””

    Well, that just shows that at the time StatCounter’s universe did not include stuff Googlers wanted to see on the web, perhaps because they were at work… NetApplications’ universe obviously detected the movement:
    Heck, maybe 2013 was the year StatCounter detected Googlers for the first time.

  5. dougman says:

    125 people, solely focused on one thing, can do the impossible.

  6. DrLoser says:

    BTW, a “spike” of 125 people (whether or not it slides back the razor-blade) is of no consequence whatsoever to anybody.

    Why would you waste valuable molecules of oxygen bringing it up?

  7. DrLoser says:

    Drivel, oioahm, sheer drivel. The “migration” you speak of was actually a “phasing out.” Even Google doesn’t do this over a weekend. (Also, they don’t fire people for not retraining over that weekend.)

    You’re on fire with these lunatic propositions this week, aren’t you? Well, here’s a mild damper for you, straight from the Champions of the Linux Desktop as quoted in your cite:

    High quality global journalism requires investment. Please share this article with others using the link below, do not cut & paste the article. See our Ts&Cs and Copyright Policy for more detail. Email to buy additional rights.

    The move created mild discontent among some Google employees, appreciative of the choice in operating systems granted to them – an unusual feature in large companies. But many employees were relieved they could still use Macs and Linux. “It would have made more people upset if they banned Macs rather than Windows,” he added.

    What, no love for Gubuntu? In the home of Gubuntu?

  8. oiaohm says:
    Loser sorry the Google migration disturbance is a real thing I was referring to a prior article of Roberts. The line was meant to remind Robert of prior event.

    Loser 10000 people can flip numbers on head on statscounter numbers. The Google disturbance was documented in one of the prior posts Robert has done.

    NetApplications and Statuscounter showed that one.

    One of the documented Migrations. Sorry Loser claiming there was no such event shows how little you know on the topic.

    Lot of this shows why these numbers are not trustworthy for market percentages.

    Loser when using collected data you compare against events and causes.

    There are also spikes of equal shape in the canada and france numbers that line up with Migrations.

    10000 people can generate a 50 to 80 percent swing in netapplication or statscounter, 2 percent could be as small as 125 people. Key thing to generate the spike is the fact they are synced.

    1.5 million users is in fact incorrect guess to generate the effect. 200 moving in unified way is all that is required to generate the spikes. But it must be unified. Political and civil problems will not generate the unified action required to create the spikes. If it did it would show up in the prior data aligned with political or civil events.

    If you really knew the topic you should be referencing these prior Migrations because they do explain the numbers. They also say what we are seeing in statscounter has no link to real world usage values.

  9. DrLoser says:

    Google migration disturbance to statscounter also tells your Robert…

    There is no “Google migration disturbance to StatsCounter,” oiaohm. That is an asinine observation.

    Consequently, it doesn’t tell you anything at all.

    But, given the vanishingly improbable case that it tells you anything at all …

    … the number required to generate these size spikes does not have to be huge.

    Well, hardly, little cardboard-trousered one. A “spike” is just that. (And I haven’t bothered to point out that one of the salient details of a “spike” is, er, the downslope: what Tom Lehrer calls “the razor-blade of life.”)

    How much more statistical ineptitude can you come up with, oiaohm? You’re so wonderfully stochastic, you haven’t even started to repeat yourself yet!

  10. oiaohm wrote, “the number required to generate these size spikes does not have to be huge.”

    That’s probably true. The analogies “share of page-views” == “share of PCs” == “share of users”, are not exact by any means and a motivated user can warp anything. As a teacher I saw students who were highly motivated. I still remember one young man who discovered that other OS would try to open everything on a cluttered desktop with “select all” and “open”… Strangely, XP did not die from this assault but would immediately crash when the unnecessary windows were closed… ewww! Another had such great hand-eye coordination and hyperactivity that he could race the pointer around a screen clicking everything clickable indefinitely. He really loaded up my terminal server… but, thanks to GNU/Linux, he just became the idle-loop and everyone else carried on.

  11. oiaohm says:

    It is interesting we are now seeing migration class activity in statscounter in more places.

    Statscounter may not be any good at working out exact populations but the one thing its historically proven to be good at is detecting when migrations happen. Google migration throwing numbers out completely is really the most public example.

    Google migration disturbance to statscounter also tells your Robert the number required to generate these size spikes does not have to be huge.

  12. wolfgang says:

    …ballmer run clippers…

    not likely. owner never run team. owner own team and get new coach when lose too many games. owner sit where he like and entertain mayor and other nabobs as he pleases.

  13. wolfgang wrote, “get to sit on bench with clippers”.

    I wonder how a basketball team run by a salesman will operate. I can see them killing off old box seats and adding new features to force renewal… Will they allow pickpockets in the stands to add excitement? How about re-re-rebootsforcing fans to leave their seats and re-enter the building in the middle of the fourth period of play? Worked for M$. Oh, wait, they had a monopoly. I guess that’s different.

  14. wolfgang says:

    …windows hanging on ropes…

    good thing for gates and ballmer that they have new hobbies now. gates get warren buffet money to play with along with his own and cure aids and build new wing on mit. ballmer get floor level seats and even get to sit on bench with clippers. all’s well that ends well.

  15. Mats Hagglund says:

    And another news from Finland too. Finnish Military Forces are adopting Android Linux-based combat system. Users will be all squad-, platoon- and company commanders and patrols too.

    More and more news showing how world is moving to post-Windows era.

  16. Mats Hagglund wrote of PCs, “- not more than 1 pc (or none at all)
    -3-4 tablets
    -4 smartphones”

    That would be my sense of things in my extended family. Folks love small cheap computers and have them in bunches while the dust collects on desktops. What M$ feared in the P3 era has come to pass. Small cheap computers make M$’s OS stick out like a sore thumb.

  17. Mats Hagglund says:

    Some pretty dramatic news from Finland too. Tablet-sales are growing fast. About 60% tablets sold more than pc (notebook+desktop) during the 1/3 2014.

    It’s really incredible. People – not only youngsters but older folks too – are abandoning their Windows notebooks and dekstops in families. There is a trend among young people not using notebooks at all. It’s likely we are marching to society where there is Windows-machines only in schools and companies (though Linux is getting more ground there too). In families (of 2 adults and 2 children) it migh look something like this:

    – not more than 1 pc (or none at all)
    -3-4 tablets
    -4 smartphones

    Windows is really hanging on ropes now. It has hardly more than 3-6% marketshare on those growing markets. Microsoft is indeed Neanderthal of our time. Too slow to adopt new environment and new conditions of life.

  18. rudregues says:

    France is a good example of how FLOSS can grow up. Remeber, french police likes Ubuntu since 2008.

    I think Brazil will have more Linux users in 2015. The reason is federals want more secure email service. For this, as you may know, “Expresso V3”, our FLOSS email service is being implemented. But now we have more numbers. Near 30.000 users already with V3, the plan is to reach 1,000.000 users at least.
    You can read more in portuguese in the following link:
    A good idea spreads with time, V3 idea will spread in Brazil, later another FLOSS apps and later GNNU/Linux.
    I’m warning you for when you see gs.stat don’t be so surprised. It’s happening since now, slowly and steadily.

  19. pogson says:

    Mats Hagglund wrote of the desktop segment.

    The 8-20 rule is about performance, not monopoly. M$ was at 95% a decade ago. There were effectively no competing technologies then because M$ had manipulated the market. Now there are several competing technologies all getting a share of the pie. Consumers, over half the market in established markets, now have choice and can make the choice. In my own city you can buy GNU/Linux online and on retail shelves desktops. You can also buy mobile. In effect, consumers are buying everything but M$. It still remains to be seen what happens to old XP machines but I would bet many are being replaced with GNU/Linux because folks want support. To compete on price with mobiles, OEMs and retailers have to sell GNU/Linux. Thus, M$’s manipulations have been replaced with choices leading to a 1/N behaviour, rather than 80-20. Consumers are diverse. That’s why M$ is advertising like mad.

  20. pogson says:

    Mats Hagglund wrote, “Linux has surely passed the 1-2%”.

    I think that was the case years ago but most machines were in schools/government with NATed IP addresses and web-caches. There are more machines mobile and more on retail shelves today.

  21. Mats Hagglund says:

    Firefox 47.59%
    Google 19.74%
    Safari 13.28%
    MSIE 8.7%
    Mozilla 5.52%
    Opera 2.7%
    Netscape 1.53
    Googlebot-Image 0.47%
    MSN 0.35%

    Those latest statistics of marketshare of my blogsite are showing huge growth of mobiles using Safari browser. Huge majority of these devices using Linux. Last year Safari got hardly more than 4%. Near half of Linux users visiting my page were using mobiles just like me now.

  22. dougman says:

    I like the 80/20 rule, as it permeates everything.

    Considering ALL markets., I would firmly say that Linux drives 80%, if you just look at the desktop segment then its 80% for Windows, with 20% left for others.

    That being the case Linux has surely surpassed the 1-2% manta the trolls spew forth consistently.


  23. Mats Hagglund says:

    Most of Scandinavian blogsites are giving much bigget marketshare for Linux than 2%. And IT sites commonly 10-18%.

    There are lots of reasons to questionize both StatCounter &NetApplication figures. E.g in Finland some 1% of all pc are those in schools using just one Linux system called OpinSys (=Ubuntu)…perhaps that “hunreds of pcs behind one ip-address effect” there?

  24. Aldiio wrote, “@pogson: Which software do you use for the nice charts with R^2, and regression line and formula? Thanks!”

    I use FireFox web browser to download CSV files, and Gnumeric spreadsheet to generate the charts. I save the charts as SVG images. I can create templates easily and copy/paste data too but it is really easy to configure the charts. I examine the graphs locally with ImageMagick and scoot them over to my server with openSSH. I use AutoKey to generate paths and HTML and CSS strings for the blog and openSSH. All of this software was obtained from the Debian GNU/Linux repository.

  25. Aldiio says:

    @pogson: Which software do you use for the nice charts with R^2, and regression line and formula? Thanks!

  26. pogson says:

    DrLoser wrote, “if you select the same graph for Germany, Robert, you see something that looks a lot like a flat-line”.

    It’s actually rising but very slowly, reflecting the small slice of government and schools in a huge installed base. It rises fro 2.24 to 2.94 in mid-term then wanders back down. That probably means the government expanded seats first faster then slower than the rest of the economy. It’s still a far cry from the days of ~1%. I think that shows that business is still seriously locked in even as consumers shop elsewhere.

  27. DrLoser says:

    And if you select the same graph for Germany, Robert, you see something that looks a lot like a flat-line, with a high of 3.02% and a current low (I expect a regression to the mean) of 2.34%.

    Odd, this, since Germany is a hot-bed of active FLOSS supporters.

    It’s just confirmation bias either way, isn’t it?

Leave a Reply