Escaping the Horror

Today was a great day. There was an all-day family-affair with feasting, Easter egg hunting and a unique competitive quest where the younger generation travelled on foot around the community doing really strange things like eating “balut” and buying coffee for strangers…

After it was all over we got the instant replay on a wide-screen TV from all the gadgets young people carry around: smartphones, handicams and digital cameras. To my horror, one of the presentations involved using the TV as a computer monitor. The computer ran that other OS, the one I haven’t used in years. Why did they use it? That’s the OS that shipped on their PC. What applications did they use:

Nothing those folks did with a PC required M$’s OS. It was all basic utilities found on GNU/Linux and FLOSS applications.

This incident strengthened my belief that M$’s OS is a tax on IT, over-priced and unecessarily bundled with most legacy PCs sold today. The sooner retailers start selling GNU/Linux on PCs the better. The world is wasting tens of $billions today for no benefit. I recommend Debian GNU/Linux but there are hundreds of useful distributions of Free Software developed by the world and freely shared for $0.

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.

73 Responses to Escaping the Horror

  1. oiaohm says:

    DrLoser I guess you did not go to HP and use there matched Linux images.

    Funny enough Windows 7 on those HP Mini-Note was a pain in but as well.

    http://en.wikipedia.org/wiki/Ubuntu_Netbook_Edition

    Linux still required you to know your hardware. Something like the first Acer or HP Mini-Note required specialised spins of Linux distributions and Windows to work.

    Ubuntu 11.04+ and Windows 8+ contains detection for horible atom processors and shoves them a new Microcode to make them sane. SLED commonly failed from HP on Mini-Note because binaries not built to be intel Atom compatible would be downloaded and installed. And when you asked a horible intel atom todo particular instructions it did not support from 2010 and before it locked up. Microcode update from intel in 2010. So it returns the general error when it receives and instruction it does not know. Programs cope with this or kernel can cope with this( detect the error and drop back to software emulation of that bit of code ).

    DrLoser basically your acquired something with a bored cpu. Windows 8 and 11.04 ubuntu/other latter Linux distributions know how to fix it. Kinda a few years before Windows 8 was release did Linux Distribution appear knowing how to fix it.

    Yes atom might be x86 but early on it had to be treated as a completely different architecture since general x86 binaries would cause it to fail.

    What would have been installed would have been SLED 10.2 Netbook edition Dr Loser. With its highly unstable package management.

    George Wilson Linpus is universal known as one of the worst Linux Distributions in existence. Only thing worse is the lite addition.

    http://en.wikipedia.org/wiki/Linpus_Linux

    Yes you picked another horible early atom machine.

    Reality we have had some really horrible hardware. Main reason why Linpus was chosen is kick back to Acer for every time someone paid to upgrade to Linpus full.

    Linpus is a good example of Linux crapware. Luckily this is not that common.

    Linpus falls into the same camp as those Windows machines you get from Acer full of crap software that cannot run because they are full of crap software. Yes just because a maker supports Linux does not mean they will do a good job of it.

    Lenovo provides pure Linux laptops. I have quite a few of them. Yes Windows, Linux/Windows duel boot and Linux options. Normally Redhat Linux or customers Linux image for the Linux only image.

  2. oiaohm says:

    bw “That is an interesting thought although you miss the larger lesson. To wit, if you need a lot of decimal places to distinguish your product’s success from zero, then you have a pretty poorly selling product. ”

    You are the nit wit. Now here is a interesting point but. Are the categories valid. Mint and Ubuntu are binary compatible. Ubuntu version number with 2 decimal points is marking service packs. Last number is service pack number.

    You could make Microsoft Windows appear less if you broken it down by service pack. Heck even less if you broke it down to if machine was running classic or current mode.

    OS X is possibly over broken down as well in steam numbers for number of decimal points. As its 12 displayed for really 3 true different versions.

    bw there are some issues with steam numbers we might as well be truthful about them.

  3. DrLoser says:

    My HP Mini-Note with, I think, a specially spun SLED 10.2 borked itself inside six months on an upgrade of Pascal.

    Ironically, the only OS that got it working again (and believe me, I ran through a lot of Linuxen) was Windows 8 RC.

    And your point, again?

  4. George Wilson wrote, “you can’t “buy” it, but only subscribe to it. It’s not intended to be put on Desktop PCs sold at Walmart & Co.”

    The HP Mini-Notes were sold retail to consumers with SLED although they were intended for education/business. HP has retail stores and connections with retail chains.

  5. George Wilson wrote, “It’s easily imaginable that it’s so defective that the user couldn’t get to a graphical login.”

    Linpus is widely used in Asia and it does get to the GUI pretty sharply. Ask yourself why an OEM would ship units that wouldn’t get to the GUI? What manufacturing process would produce an occasional unit that would not get to the GUI? It’s more likely a defect in manufacturing than any flaw of Linpus. Linpus Technologies regularly visits OEMs to help them promptly with any problems of implementation.

    “Linpus is the only Linux vendor with research and development facilities in both Taipei and Shanghai, strategically positioned next to the main hardware manufacturers. As part of our service we promise to be on site to solve problems within two hours, 24/7, 365 days a year”

    see About Linpus

  6. George Wilson says:

    Every consumer-product can probably find pro and con responses. Defective units do ship despite quality control. Units are damaged in shipment too.

    Yes, but doesn’t it seem odd that you don’t need to look long to find the expected results? Indian people (at least those shopping at FlipKart) have no love for Linux and like to use Windows.

    Contrary to your often repeated claims that the Linux revolution will begin in India, China, and so on.

    Linpus Lite is not a command-only system. It’s a GUI. So, the comments make no sense.

    Yes, they do. Linpus is a truly bad Linux distribution. It’s easily imaginable that it’s so defective that the user couldn’t get to a graphical login.

    SLED is a desktop edition of GNU/Linux. It does ship on PCs bound for consumers.

    No, it’s not. It’s an Enterprise Linux distribution. Hence you can’t “buy” it, but only subscribe to it. It’s not intended to be put on Desktop PCs sold at Walmart & Co.

  7. Every consumer-product can probably find pro and con responses. Defective units do ship despite quality control. Units are damaged in shipment too.

    Linpus Lite is not a command-only system. It’s a GUI. So, the comments make no sense.

    SLED is a desktop edition of GNU/Linux. It does ship on PCs bound for consumers.

    See also, SJVN with With HP in all OEMs now ship desktop Linux

  8. George Wilson says:

    Come on, Pogson, please free my post from the evil clutches of your spam filter.

  9. George Wilson says:

    1. SLED has nothing to do with the consumer market.

    2. The notebooks in question from Lenovo are dual-boot systems and come with Windows 7 Professional.

    3. Like China: sell computers without OS, users put illegal copies of Windows on them.

    Don’t believe me? Let’s look at some of the customer reviews for the first notebook listed at the link to FlipKart you gave (I’m simply going from top to bottom):

    My system’s software is conked. The preinstalled Linux isnt loading.

    What a bummer!

    The pre installed linpus linux is a lite version of linpus linux(only text commands ll wortk.no GUI).so i downloaded linux 12.04 from net and when i try to install through pen drive,it doesnt work(pen drive installation is not supported).so i found a dvd writer and burnt the installation iso. Installation is also very bad(since this machine is built to support windows 7&8)it took 4 hours to figure out wrong things in installation and sort out.after 2 days of working with linux i got frustrated(most app wont work with linux).then finally i downlaoded win 8 trial version and installed.win 8 is bit slow but no errors and al.

    This customer is not happy with Linux, it seems.

    Win7 with all aero effects enabled runs very smooth, no glitch whatsoever. Tried running a 1080p full hd vid, runs smoothly.

    This customer is very happy with Windows 7.

    +comes with a driver disk which relieves you from downloading all drivers from the net.

    A driver disk … wouldn’t one of these be necessary for, yes, Windows? What a bummer! Again, not a Linux user.

    No issues if you Load Win 8 . Suggest 64 Bit DVD to be used and it will take care of all drivers and install the product. Require formating of existing partitions.
    Suggest to go for Win 8 Professional offer at Rs 3499 as its an excellent introductory offer.

    A happy Windows 8 user. Is he a paid Microsoft shill? It can’t be!

  10. bw wrote, “One relevant question would be in regard to what these people would select in the event that they someday become prosperous?”

    India is a very diverse country with a wide range of rich and poor and middle-class. The government is trying to lift every boat with moves like this. It’s not about being inexpensive but bridging gaps. Lower-priced IT allows them to do more faster. That’s the same in most public educational systems anywhere.

    India has the 4th largest economy in the world and GDP is growing rapidly well past $1trillion. GDP per capita is sad however. Education is key to increasing wealth for more of the population. I am hopeful the government continues on its path to modernize just about everything including IT with GNU/Linux.

  11. bw says:

    “Lenovo sold 380,000 units in India.”

    A tough business to be in, in my opinion. To operate in these areas requires an extreme tolerance to discomfort. I would rather have the North American zone business than this.

    http://www.youtube.com/watch?v=DQzs4izA27Y

    One relevant question would be in regard to what these people would select in the event that they someday become prosperous? Will they stay with the cheap stuff or will they want a Mac?

  12. bw says:

    “You know you don’t have enough decimal points when you start seeing all zero numbers in stat collections”

    That is an interesting thought although you miss the larger lesson. To wit, if you need a lot of decimal places to distinguish your product’s success from zero, then you have a pretty poorly selling product. Whether the unsatisfactory sales volume is up 50% or down 50% from month to month doesn’t change your overall certainty of failure which looks like 100% to anyone without bias.

  13. oiaohm says:

    Ted rounding or truncating both produce an error. Both the error is made large the the more categories you break your numbers into.

    Now my problem is it might be insignificant at this stage. What happens if we end up with 500 distributions listed.

    Ted this is the problem we have to be aware of going forwards. The defect may not remain insignificant thinking all of Apple in this survey is about 3 percent. Yes at 500 Current .20 and .42 swing possibility only minor-ally changes location. If we end up of a swing of a few percent this is very problematic.

  14. Ted says:

    “Unless you can come up with some rational argument why shares should be two-digit integers/10000 exactly,”

    A rational argument? How about I propose that Valve used the arcane and little known practice of rounding the numbers up to the nearest two decimal places?

    “it is you who are hand-waving.”

    Not in the slightest. I simply say Valve’s numbers are what Valve says they are. They’re there on the page. I am not making things up from whole cloth, as you are. You still have not produced a lick of actual evidence for you bumping the numbers up to suit your agenda. *Evidence* would be the unvarnished numbers from Valve, or a statement from them that says that they truncated the numbers without first rounding them up.

    Why don’t you just follow your usual MO and deride this new batch of numbers as hopelessly wrong or in some way influenced by Microsoft? Is it because you’ve hung your hat on the Steam Hardware Survey as an example of a success for Linux, and another nail in Microsoft’s coffin?

    And you have still not addressed one of my first points – you’re quibbling over statistical noise. Even if you’re right – and it’s highly likely you are not – any gains you create with your dishonesty over this supposed truncation are insignificant.

  15. George Wilson says:

    Reality tells us otherwise. You will see what I mean when Mr. Pogson frees my post which apparently has become stuck in the spam filter.

  16. oiaohm says:

    George Wilson the problem here MS trolls want to believe OEM don’t give a crap about Linux.

    The reality OEM do give a crap about Linux. Reality is also that Linux world has made it hard to market products by lacking information on what distributions are dominate in what area. So hard for OEM to release products so the OEM have been defaulting to the Linux Distribution neutral option FreeDos.

    George Wilson Dell is the first OEM to release something in response to the Steam data. Others will follow.

    So gamer machines expect to see them with Ubuntu not Freedos as they have had.

    George Wilson this is your problem you are using Myths. Research will kick your ass over this point. OEM do care about Linux.

    Linux issue is its fragmentation and the confusion that has caused for OEMs. Not OEM lack of support for Linux.

    Confusion/Fragmentation has made Linux hard to market.

    Better survey data removes the confusion.

    Reduction in confusion also results in developers having a better time choosing what platform to support.

    Steam support of Linux is really a domino effect.

  17. George Wilson says:

    1. SLED has nothing to do with the consumer market.

    2. The notebooks in question from Lenovo are dual-boot systems and come with Windows 7 Professional.

    3. Like China: sell computers without OS, users put illegal copies of Windows on them.

    Don’t believe me? Let’s look at some of the customer reviews for the first notebook listed at the link to FlipKart you gave (I’m simply going from top to bottom):

    My system’s software is conked. The preinstalled Linux isnt loading.

    What a bummer!

    The pre installed linpus linux is a lite version of linpus linux(only text commands ll wortk.no GUI).so i downloaded linux 12.04 from net and when i try to install through pen drive,it doesnt work(pen drive installation is not supported).so i found a dvd writer and burnt the installation iso. Installation is also very bad(since this machine is built to support windows 7&8)it took 4 hours to figure out wrong things in installation and sort out.after 2 days of working with linux i got frustrated(most app wont work with linux).then finally i downlaoded win 8 trial version and installed.win 8 is bit slow but no errors and al.

    This customer is not happy with Linux, it seems.

    Win7 with all aero effects enabled runs very smooth, no glitch whatsoever. Tried running a 1080p full hd vid, runs smoothly.

    This customer is very happy with Windows 7.

    +comes with a driver disk which relieves you from downloading all drivers from the net.

    A driver disk … wouldn’t one of these be necessary for, yes, Windows? What a bummer! Again, not a Linux user.

    No issues if you Load Win 8 . Suggest 64 Bit DVD to be used and it will take care of all drivers and install the product. Require formating of existing partitions.
    Suggest to go for Win 8 Professional offer at Rs 3499 as its an excellent introductory offer.

    Come on now! Not a single Linux user? India isn’t what it used to be.

  18. George Wilson dropped by to drop this steaming turd, “No OEM gives a shit about Linux.”

    Hmmm…

    So, Bunk!

  19. George Wilson says:

    No OEM gives a shit about Linux.

    Except Dell. What do they offer? Their overpriced “Sputnik” notebook for “developers”. And a horrendous Alienware machine which is supposed to be put into the living room (unfortunately there’s already a real console there).

    I can sense real mass appeal here!

  20. oiaohm says:

    bw know how you wanted more from factory Linux machines.

    http://alienware.com/ubuntu/

    Yes direct result of steam survey. Companies can direct anyone who complains that Ubuntu is the wrong distribution to the steam survey.

    Yes the offers on that are still insane. Windows RT tablet or a Windows laptop.

    The reality is once Linux world starts getting more good surveys of what Linux users are using and doing hardware makers will ship more hardware setup out box that way.

    Yes Linux fragmentation is why you have not seeing more Linux installed stuff. Why you were seeing the Dell N class stuff. Dell has no clue what OS buyers of N class are using.

    Load the wrong OS bad enough Linux guys might just pave over with the one they want. Have the wrong hardware configuration for what the Linux users want they will pave over something else.

    Steam survey gives OEM’s a nice idea what the Linux guys want so can provide to them.

    Web surveys are too generic to be any use to Hardware makers. Like 90 percent Windows. Ok what are they doing so what machines do I need to build.

    Large Wastes in PC production come from the fact of OEM’s not having enough data on Users.

  21. oiaohm says:

    Ted get over it Nats do hide users.

    Its very darn in your face with china that there is a collection problem. Since we know there is more Internet users in China than any other country. China keeps exact counts of there Internet users. USA carries have exact counts of there Internet Users. China is bigger.

    So if Nat don’t explain it. Exactly why does not Statscoutner or anything else get it right.

    There is another option. Called proxy servers.

    Ted a proxy server can filter retrievals based on user-agent. So only 1 request per different computer type in the network.

    Even so something that is logged like steam or other websites punches straight through this.

    Ted between nats and Proxies the page counting methods used don’t work.

    It also gets worst that a lot of sites that depend on advertising for revenue are reporting over 50 percent of users are using ad-blocking software of some form. Yes that blocks user agent being reported to the likes of NetApplications and StatCounter. To avoid adblocking or adreplacement stats have to be run on the site hosting server.

    Some evil ISP replace the advertisement on site and redirect it to themselves.(legal in china)

    bw the more cats you have worst rounding error is.

    There is from my quick count 42 Linux cats on the steam site. The general worse case would be that ever two would total up to 0.01. So approx 0.20 under truncated.

    Rounding not a good idea in fact.

    Worst cases rounding. To 0.01 +-0.20 Because you may have incorrectly rounded up or down.

    Even worse case with truncation that you were have like 0.xx9999999 on everything you are down by 0.42.

    Truncated tells you that you are under quoting. Rounding you may be under or over quoting. Now if Linux does get to 100. Two decimal points is too small.

    If there were only 10000 people responding there would be no 0.00 entries. Since that is a 1 to 1 resolution with 100.00%. So the number of people in survey have to be in excess of 10000. How much in excess of is critical for divide rounding errors as well.

    You know you don’t have enough decimal points when you start seeing all zero numbers in stat collections

    –As stated, the survey is optional and is not apparently a statistically controlled sample in the first place.–
    Most ground surveys are optional. Steam could publish what percentages don’t take part. But they don’t.

    Just like web surveys don’t publish how much they have been ad blocked or proxy blocked.

    I am not saying steam is the best. Its about the best of a very bad bunch.

  22. bw says:

    “Because truncation is easier…The mystery would have been solved with just another digit”

    ??? How is truncation any easier or rounding harder than the other?

    As to the number of decimal places available, it would be more useful to know the sample size. Are they even getting 10000 people to respond? As stated, the survey is optional and is not apparently a statistically controlled sample in the first place.

    It would also seem to me that a resolution of 0.01% is a pretty fine resolution already for any rational business purpose. Anyone worried about that digit’s variance in their sales plan is in a lot of trouble. One part in 10,000 seems like enough to me.

  23. bw wrote, “Why do you think they are truncated and not simply rounded to the nearest .01? Then the totals would tally a lot better.”

    Because truncation is easier. Rounding is just a nicer form of truncation and it works on average but in such a small sample, they roundings could all be in one direction about one in a million times. Without detailed information I assume the worst. Why did they stop at two digits? The mystery would have been solved with just another digit.

  24. bw says:

    “See the truncated values come to a much different total than the true values.”

    Why do you think they are truncated and not simply rounded to the nearest .01? Then the totals would tally a lot better.

  25. Ted wrote, “You still have NO *evidence* that Valve’s numbers are anything except what Valve have published them as.”

    Unless you can come up with some rational argument why shares should be two-digit integers/10000 exactly, it is you who are hand-waving. Here are some random numbers. How many of them are XX0000?
    0.0821557133
    0.2907592179
    0.680791242
    0.9092756344
    0.1024252402
    0.2763211513
    0.8131378966
    0.1717418588
    0.1107164442
    0.5791130711
    0.0826627475
    0.0128928307
    0.8175259945
    0.633252074
    0.6259751585
    0.065819188
    0.8731977751
    0.0343410801
    0.3159773927
    0.6485938907
    Hint: None of them. The numbers you imagine are prevalent are actually quite rare.
    Here’s another set of the magnitude in doubt:
    0.0202672231 0.02
    0.030015545 0.03
    0.0105139743 0.01
    0.0292412787 0.02
    0.0163776928 0.01
    0.008487154 0
    0.0300067802 0.03
    0.0266145321 0.02
    0.014601892 0.01
    0.0024690102 0
    0.0191610909 0.01
    0.0001531467 0
    0.0118353404 0.01
    0.0047770545 0
    0.005265476 0
    0.000900178 0
    0.0083183573 0
    0.0132620354 0.01
    0.0022886853 0
    0.0061583205 0
    0.2607147673 Totals 0.18
    See the truncated values come to a much different total than the true values.

  26. Ted wrote, “NAT does not hide User-Agent.”

    Sure it does. If I have 10K PCs running GNU/Linux coming out of a single IP address with NAT then NetApplications sees a single visitor for a site when there could be 1K because they only count one visit per unique visitor per day. Do any of them track MAC addresses? Nope. Too much work. They don’t have any clue whether or not I have 10K GNU/Linux and 1 that other OS or 1 GNU/Linux and 1 that other OS as long as one or more of each gets through to the counted site.

    Statcounter does a little better by counting page-views but they are then drowned by the unknown activity of users. “Every month, we record billions of page views to these sites. For each page view, we analyse the browser/operating system/screen resolution used and we establish if the page view is from a mobile device.” Of course one page view by GNU/Linux from one PC should count the same as any number of page-views by that other OS from one PC but they are not even trying to count that. I spend 8h per day browsing from GNU/Linux. I should be able to shift the count of GNU/Linux for all of Canada but they list Canada lower than USA…

    Neither NetApplications nor StatCounter can or do count number of PCs running various operating systems. They count something only tangentially related to the real issue, share of units out there.

    Further, StatCounter gives these breakdowns for sample-size by country. They show USA as #1 even though China has more users of the Internet. China is not even counted in the top 10. Think NAT has something to do with that? How about language? Why is a Chinese pageview counted less than a USAian PC?
    “4217059839,”United States”
    1308190515,Turkey
    1128880659,India
    732874454,”United Kingdom”
    701039219,Brazil
    494757188,Canada
    475967494,Germany
    430455704,Thailand
    370607997,Mexico
    331858106,Spain
    314091473,Indonesia
    308106207,Argentina
    284459373,France
    275096518,Italy
    262043649,Australia
    251139355,China”

  27. Ted says:

    “Not in the least. People can read my work on the web. e.g. here

    I have done my share of number-crunching and measuring things.”

    You appeal to authority again, and it’s still yourself. You still have NO *evidence* that Valve’s numbers are anything except what Valve have published them as.

    “NAT hides a lot of other bright spots.”

    NAT does not hide User-Agent.

  28. oiaohm wrote, “User login based and IP based web stats don’t come close to agreeing.”

    I often thought of using accounts to rid myself of spam but on this site the vast majority of commentators would use GNU/Linux, I suspect. On a general-interest site, using accounts would be a great way to judge usage. It’s too bad Google and some of the other popular players on the web don’t publish that information. Thanks to Google Sunnyvale, CA, USA shows up as a bright spot on the web-stats. NAT hides a lot of other bright spots.

  29. oiaohm says:

    George Wilson Wikimedia filters by IP. We know different countries have different levels of users to each nat. China is over 50 to 1.

    So what Wikimedia counts as one is not 1 user or 1 computer.

    I will tell you people using sites using User logins that see threw the IP issue. Are telling me different to what the Wikimedia shows.

    George Wilson wikimedia numbers are mostly shot from a collection point of view.

    Mats Hagglund never told us how that is collected. IP or User login based??

    George Wilson that question I just asked is the question you need ask. User login based and IP based web stats don’t come close to agreeing.

    Problem is majority of numbers are IP based or IP based with some voodoo maths applied so rendering them garbage for anything other that possibly spotting trends.

  30. Ted wrote, “This is irrelevant, anecdotal…”

    Not in the least. People can read my work on the web. e.g. here

    I have done my share of number-crunching and measuring things. That’s what web-stats are all about, counting experiments. If you have some random number of a few digits taken by counting unique IP addresses and divide by some other random number, the total number of hits, the probability of getting zero truncation in a two-digit result is tiny. Since it is pretty random stuff, the next digit is a zero only 10% of the time. The chances of all the numbers not being truncated when displayed as two digits is ~10-20, not enough to worry about, but you go right ahead.

  31. Ted says:

    “I do.”

    No, you do not.

    To quote you;
    “The problem is those displayed numbers have been truncated and could well be”

    You assert they have been truncated with *zero* evidence of such, and you all but admit making up your numbers with three decimal places – “could well be”. This is not evidence, it’s supposition.

    “After doing years of nuclear physics,”

    This is not evidence either. This is irrelevant, anecdotal, and an Appeal To Authority where the Authority is yourself. This is on a par with “It is, because I say it is!“.

  32. George Wilson says:

    Anybody who runs a few webservers can tell you that Linux is now over half the hits and visits. For anything, cheap consumer products through expensive industrial products. Just look at the server stats.

    Yes, just look at the server stats of, say, Wikimedia:

    http://stats.wikimedia.org/wikimedia/squids/SquidReportOperatingSystems.htm

    Astonishing. Even with Linux Android included you only get 7.45 percent. Yes, I’ll be the first to admit it:

    That’s over half by looney bin standards!

    But perhaps you want to show us the statistics of “anybody who runs a few webservers” that prove your point.

    (Enter Pogson, who will tell us how the Wikimedia stats have no meaning at all.)

  33. Mats Hagglund wrote, “Not all Linux were smartphones most of them were.”

    That’s cute. Visitors do their thing and they all see that */Linux works for them. I can see “positive feedback” which M$ exploited to create its empire.

  34. ram says:

    Anybody who runs a few webservers can tell you that Linux is now over half the hits and visits. For anything, cheap consumer products through expensive industrial products. Just look at the server stats.

  35. Mats Hagglund says:

    CasioIndia Club website gave these statistics (1st Feb to 3rd of April):

    Windows 136475 78.42%
    Unknown 18833 10.82%
    Linux 14344 8.24%
    Macintosh 3762 2.16%
    OS/2 603 0.35%
    UNIX 11 0.01%
    OpenBSD 2 0%
    FreeBSD 1 0%

    Not all Linux were smartphones most of them were. Here’re some examples:

    -Linux (720x1280x32)
    Safari 5.3430 (JavaScript: 1.7, Cookie: Yes, Java: No)

    -Linux (1001x1254x32)
    Safari 5.331 (JavaScript: 1.7, Cookie: Yes, Java: No)

    -Linux (600x1024x32)
    Safari 5.34 (JavaScript: 1.7, Cookie: Yes, Java: No)

    -Linux (720x1280x32)
    Safari 5.3430 (JavaScript: 1.7, Cookie: Yes, Java: No) (visitor from Australia)

  36. Mats Hagglund says:

    Isn’t it curious how non-commercial non-pay-per-click sites just like this one:

    http://www.elsua.net/

    is getting remarkable great deal of non-Windows machines visiting it? Both Linux and Mac having 8-14% daily market share.

  37. Ted wrote, “You have no evidence for this assertion – you do not know the numbers have been truncated. You surmise it is so, because it suits your agenda.”

    I do. After doing years of nuclear physics, I know that the result of counting experiments for totally random events are random numbers with distributions approaching “normal” distribution. Such numbers are unlikely to have only one or two significant digits considering they are talking about many thousands of users. e.g. Take some integer and divide it by some much larger integer. It is extremely rare that any two such integers divide evenly. To do so they need a bunch of common factors, sort of like opening a combination lock at a random spin. So, maths and the natural world support my argument. For the numbers not to be truncated the numerators would have to be something like 800 and the denominator something like 800000, a highly unlikely combination. Things like 831/815354X100 are more likely. The result? 0.1019 instead of 0.1000. This knowledge is ancient and was very well known in the 1960s when the ability of computers to multiply was used to produce pseudo-random numbers.

  38. Ted says:

    “–Vanishinly few–
    This what numbers do you have that its true.”

    You asking for evidence is a bit rich, but how about, just as an example…

    THE STEAM HARDWARE SURVEY

  39. oiaohm says:

    Ted –I don’t recall anyone ever seriously claiming NO gamers on Linux. —
    Head of Blizzard Entertainment did at one point. That arguement is absolutely dead.

    –Vanishinly few–
    This what numbers do you have that its true.

    Answer none. That is might be the next myth to fall.

    Ted basically we need the steam number to run and stabilise and tell us what is correct for the gamer market.

  40. Ted says:

    “If you go from the displayed numbers like
    [numbers possibly from Valve]
    and add them up, you get 0.06, right?. The problem is those displayed numbers have been truncated and could well be
    [numbers that certainly did not come from Valve]”

    You have no evidence for this assertion – you do not know the numbers have been truncated. You surmise it is so, because it suits your agenda.

    For all you know, the figures could have been rounded up and then only formatted to two decimal places. Doesn’t look good for the distros on 0.00% if that’s the case.

    “so the total should be 0.084, quite a bit larger than 0.06.”

    You’re quibbling over thousandths that add up to a few hundredths of a single percent.

  41. Ted says:

    “Ted it does bust the MS Troll mythical arguement there are no gamers on Linux.”

    I don’t recall anyone ever seriously claiming NO gamers on Linux. Vanishinly few, yes, but not none. Even I’ve played Super Tux Kart. It’s crap, but I’ve played it.

  42. oiaohm says:

    bw pardon Redhat does fly places to seal deals at times this include desktop deals.

    Sorry bw your complete idea is wrong. Closed source on Linux and Support companies on Linux do send their sales people.

    –So there are no full time sales people evangelizing Linux on the desktop–

    There are full time sales people pushing the Linux desktop. Just they have no interest in the small end businesses or entry level consumers.

    As normal your information is based off your limited and very limited point of view bw.

  43. oiaohm says:

    George Wilson of course you miss since Jan even with the correction it still up by 0.50 in march compared to jan. Windows is down 0.11 in march compared to Jan.

    Yes rounding errors could be there. Linux is still progressing in the direction of 2 percent and we still don’t know how big it really is.

    Ted, George Wilson and Robert Pogson. Steam numbers are not old enough to stable yet to draw any absolute idea where it going to stop.

    Ted it does bust the MS Troll mythical arguement there are no gamers on Linux.

  44. George Wilson says:

    You used the same data to “prove” things in favor of Linux. And you were told that the data doesn’t allow to draw such conclusions.

    Robert Pogson, Linux evangelist, caught once again in his own web of hypocrisy.

  45. Ted wrote, “Even adding the various Linuxes together (and I have) Linux is still less than 2%.”

    I guess I need to give a lesson in floating point/integer arithmetic.

    If you go from the displayed numbers like
    0.01
    0.03
    0.00
    0.02
    and add them up, you get 0.06, right?. The problem is those displayed numbers have been truncated and could well be
    0.013
    0.036
    0.008
    0.027
    so the total should be 0.084, quite a bit larger than 0.06. Splitting the count for GNU/Linux into more bins just increases the accumulation of truncation error which is of the order of 0.005N. The average error could be 0.005 so with N being more than 20, the error could accumulate to 0.1 or more. It could be that the average truncation error is more like 0.008N. We just don’t know.

    So, the presentation of the data hides a lot of information. They added bins for the second report but did not give us even one more digit. We don’t even know whether they rounded or just chopped the numbers off.

    You can also consider that 15% of Linuxers are gamers so 2% could be 15% of 12% or so. It depends on whether a young male GNU/Linux user is more or less likely to be a gamer than a young male user of that other OS. I would think taking risks is a bit more like a GNU/Linux user. I do know that students are much more ready to take up GNU/Linux than teachers. Perhaps gamers are as well. So, you can’t make much of a conclusion about these data except that there are more GNU/Linux gamers now than before.

  46. Ted says:

    “There is a bias here. Windows OS are not split on service packs. But Linux and OS X are.”

    So what? Even adding the various Linuxes together (and I have) Linux is still less than 2%.

    “So all of Desktop Linux being 1 percent is fairly much busted.”

    If all PC users had Steam installed. And they do not. Neither do all desktop Linux users, for that matter. To use Mr Pogson’s own number from an old post I can’t be bothered to dig up, gamers are a tiny “niche” on about 15% of total PCs. 2% of the 15% is less than 1% of the whole.

  47. bw says:

    Given enough time, just about everything is doomed, even the sun rise in the east. It is interesting to speculate how computers will go, even so.

    Back in the 60’s IBM was the king and was being besieged by mainframe competitors like Univac or RCA and numerous others. The competitors with Unix and DEC came along and took a lot of new business away from IBM although it kept a lot of its old business. The Unix workstations and VAXStation business gradually died out when PCs increase their horsepower. Now there are tablets and phones to do much of the work previously done on a PC or Mac. Is that the end of the conventional PC?

    I think you have to look at it from a different angle. The dweebs want to look at semiconductor physics and bits and bytes of software, but I think you have to look at the situation from a more traditional commercial point of view. Technology change has been with us since the wheel became rounder and rounder.

    New models and new technologies take over from the old when someone sees a profitable way to bring about change. The IBM mainframes were replaceable on the low end of performance certainly by minicomputers which became more and more capable. Minis started losing out to PCs ten to fifteen years ago as faster and faster x86 processors became common and businesses like Autodesk arose to threaten minicomputer workstations and Computer Vision or Intergraph style software on high end PCs. There was a lot of money in doing that.

    Today, though, it is not likely that phones and tablets will really threaten the workstation world for PCs and the rest of the business is just too cheap to support any really focused effort.

    An Autodesk salesman can still afford to fly to Peoria to seal a deal with Caterpillar for graphics software, but the majority of PC users are not readily addressable with anything to replace their PCs in a commercial setting. The technology messages and sales efforts needed for switching, say, from Windows PCs to a Linux environment are just as complex as they were for switching from Unix workstations, terminals, and servers, but no one is really incented to do that.

    Maybe all the yak around here about economies of Linux are true, but they are economies for the end user and are not profits for the advocate and won’t pay for airfare to the customer’s location. So there are no full time sales people evangelizing Linux on the desktop, just the sort of bombastic techo-dweebs found in computer hobby forums such as this one. They don’t do a very effective job overall and it is a very big world that needs to be addressed.

  48. George Wilson says:

    Microsoft is doomed.

    If every time a “Microsoft is doomed” evangelist uttered those very words I had received 0.01 Euro, I would be a rich man by now.

  49. Mats Hagglund says:

    I’d blame older generation – those who are now teachers at schools and boards making decision of their IT-systems. These now middle age people are generation of Microsoft Windows ecosystem.

    In my country the situation isn’t much different though Linus was born here. However there is some progress – 45 000 students and teachers are now using e.g OpinSys Ubuntu-based ecosystems in our schools. Worst are city councils and libraries. I know only two library using some other machines than Windows there. Only few small towns are using Linux though several domestic scientific studies have revealed how much they could save money by adopting Linux and FLOSS.

    Basically i see the whole picture: this is nothing more or less than modern humans (=Linux/FLOSS) living side by side with Neanderthals (=Microsoft ecosystem). Of course evolution of IT will be another “survival of fittest”. Microsoft is doomed.

    http://gizmodo.com/5993071/dna-is-the-linux-of-the-natural-world

    “DNA Is The Linux Of The Natural World ”

    ” We probably all vaguely assume that computers will overthrow us someday, which may be why it’s so unsettling to learn that computer code is evolving much like genetic code. By comparing bacterial genomes to Linux, researchers have found “survival of the fittest” acting in computer programming.”

  50. oiaohm says:

    Ted that bug lines up with the Ubuntu LTS update.

    –Specifically, only systems that had run the survey prior to the introduction of the bug would be asked to run the survey again.–
    So you can fairly much say this is Ubuntu double counting.

    This is why we need 12 months of data. We can bet surveying Linux is going to bring some glitches.

    Ted I would say the Linux share has dropped because the double counting.

    If you total up all of windows even with the 9 percent gain over all it has dropped compared to Jan numbers. There was a fairly much sharp spike in Ubuntu in feb. I could possibly say too sharp.

    Ted so we most likely have two errors at once. New machines not being counted as they should have been. Old Linux machines being counted twice or more.

    Hopefully that is the end of the bug.

  51. oiaohm says:

    George Wilson and Robert Pogson Linux Feb miss reading is explained by this. http://fridge.ubuntu.com/2013/02/14/ubuntu-12-04-2-lts-released/

    You can fairly much bet a few Ubuntu users got double counted. There is still growth compare to Jan numbers.

    I saved the all the numbers 24 hours ago. I can tell you the changed numbers on the steam hardware survey have changed since then. So those are totally unstable.

    George Wilson problem here we need 12 months worth of data. We don’t have that yet. Why 12 months with distribution releases and other things messing with the numbers. Like particular games could cause bigger moves than others.

    Currently we are still in guess what is going on.

    “That was a short hurrah.” Most likely no thinking the error is explainable. This more more you having egg on face.

    There is a bias here. Windows OS are not split on service packs. But Linux and OS X are.

    Remember how I have been saying we will have to wait for the numbers to stabilise. We still don’t have a clue yet where that will be.

    OS X had also had a update in March. So it will be interesting to see if OS X dips in April.

    Maybe steam cannot count OS upgrades correctly.

    Still Ubuntu alone is over 1.17 percent. Mint is also grown fast. So all of Desktop Linux being 1 percent is fairly much busted.

    George Wilson lot of Linux surveys suggest Mint is the dominate not Ubuntu. So I would not be surprised if Mint just keep on growing. Mint vs Ubuntu is what I am truly interested in.

    There is something interesting from the numbers so far. Even that Ubuntu releases lots of versions.

    In the LTS versions the version that is 12 months old you are fairly much seeing disappear completely.

    So I would expect 12.04 no .x to most likely disappear completely in April. I am also expecting 12.10 to pull the same thing.

    13.04 coming up will be another LTS.

    George Wilson how fast Linux users migrate is data we have never had. Apparently its a lot faster than what we thought.

    Basically steam numbers is telling us lots of things.

    Steam is now reporting on how many Windows server additions are in use.

    George Wilson the 12 month data will bring us more information that the Linux world we have never had.

    Its really bad thinking that the Linux world has never really known what is the leading distribution 100 percent.

    Its very hard to market without this data.

  52. Ted says:

    “Without knowing how they were obtained they don’t show much.”

    How they were obtained? How about exactly the same way as the February results? You didn’t question those, did you?

    http://mrpogson.com/2013/03/03/gnulinux-gains-steam/

    And on Valve’s page;

    “There was a bug introduced into Steam’s survey code several months ago that caused a bias toward older systems. Specifically, only systems that had run the survey prior to the introduction of the bug would be asked to run the survey again. This caused brand new systems to never run the survey. In March 2012, we caught the bug, causing the survey to be run on a large number of new computers, thus giving us a more accurate survey and causing some of the numbers to vary more than they normally would month-to-month.” [emphasis mine]

    So it looks like a new raft of Windows 7 and Windows 8 (9%) machines have finally been surveyed, and thus Linux’s share has dropped.

  53. Those numbers seem questionable. For instance some are marked as decreases even though they obviously are increases. Without knowing how they were obtained they don’t show much. However, a sudden growth in even poor data shows GNU/Linux works for gamers.

    e.g.
    Other in February showed 0.82% with an increase of 0.06% while in March they showed 0.03% with 0.0% change. What’s with that? Clearly something is unstable about the mechanism of the survey. They increased the categories of GNU/Linux from 4 in January to 7 in February to 38 in March. Could the drop be just truncation errors?

  54. George Wilson says:

    Brilliant! Linux numbers are down in March 2013 Steam Hardware Survey. That was a short hurrah.

  55. bw says:

    “14 legacy PCs”

    Why? I have my desktop, where I do mostly serious stuff, including work at home. My wife has her own desktop/all-in-one that she uses for her social tasks. I have this laptop that is more convenient to use than going upstairs to the office where I check mail and things like this forum. She has an iPad, too, that is mostly for show with her friends.

    I think the laptop gets the most work and I take it with me on trips. I cannot fathom what you would do with 14 PCs other than count them periodically. Maybe you could open up one of those gambling site internet cafes if they are still legal in Canada.

  56. bw wrote, “Your goal seems to be to never buy a computer from anyone, preferring to resurrect junk box items on Debian, if I get the sense of most of your posts.”

    I have bought a lot of computers for my employers and for myself. In my home I have 14 legacy PCs and two smartphones all of which were purchased by me or a member of the family. Ten shipped with GNU/Linux, one with Android/Linux and others were built by me from parts or shipped with that other OS now over-written with GNU/Linux. So, “never” is not even close to correct. My computers range in age from 0 to 8 years old. Average age is about 5. One of the computers is in a case about 10 years old. It travelled thousands of miles with me in the North and the side panel can no longer fit but it has a 2 year old motherboard and a quad-core 64bit CPU and serves as a PC, server, database, backup…

  57. oiaohm says:

    –Bad news for you: it’s still wrong. The Japanese clearly talks about a “classical music program”. And since we’re talking about FM radio, where the radio is always tuned to one station, it makes much less sense to listen to an “FM broadcast” when you’re told by the Japanese that someone’s listening to a certain program by means of FM radio. Because that invariably means that the radio is tuned to a certain station. Simple logic therefore dictates what the translation should be.–

    You need to go listen to many sections of the USA south. They don’t use the word station for FM AM or anything else radio. So FM broadcast is valid there. Other sections of the USA is FM Channel not Station. Yes station in some of those places only mean train station not radio anything. Google American translation to Broadcast instead of Station is an Americanism. Heck they could have gone really odd ball and gone Channel.

    Yes Channel and Station when it comes to radio are 100 percent interchange right. The answer is wrong. A few english places get nightmarish. Channel being cb and other public bands and Station being commercial radio stations and other areas are also reversed. Yes sections of Africa. Broadcast is kinda generic there as well.

    George Wilson English is a ass of a language. inconstant item. Apply simple logic you are screwed George Wilson as you have miss translate for English speakers in particular places. And goggle translation is wrong for where you are. In effect there is no correct translation for that bit. It will always be incorrect somewhere in the english world.

    Simple logic my foot. English is not that simple. Across the English speaking world we have different acceptable translations. And what is acceptable translation to one english country is completely wrong for another.

    Dyslexia when I was diagnosed they had the idea teach you more and more English will fix it.

    So I learnt all the horrible of the English world.

    –I doubt it. Japanese grammar is vast. Chinese is not.–
    That is the point. Japanese Grammar is more vast than Chinese so Chinese can map in non destructively to Japanese. Maybe gruff but it maps in non destructively all the Chinese grammar points remain.

    Chinese Grammar is more vast than English(particularly with english inconstancy). So translation straight to english can be problematic since stuff is lost.

    More of the basic Grammar and words that cannot 1 for 1 translate. The higher the miss translation rate is. This does not matter human or machine doing it.

    I did not say todo Japanese to Chinese. This is going from a more complex grammar to a less complex grammar translation errors from hell.

    George Wilson tracking and finding translation errors is part about knowing what language should be translated to what language.

    There are very few languages that translate to english safely.

    George Wilson knowing Japanese puts in in the nice location to be able to use automated tools to read Chinese almost 100 percent. Now accepting and using those tools can be a hard bit with the errors you see going from Japanese to other languages.

    –You’re probably more thrown off by the fact that a kanji can look very different form-wise when written as a traditional or simplified Chinese character.–
    No I am not.

    I badly typo at symbol. I should have typed symbols. Yes typo. I was aiming to refer to the compounds some of those differences really screw you over.

  58. George Wilson says:

    Remember I said Google translate Chinese to Japanese is the best. All those fine genitive bits map from Chinese to Japanese where they don’t to english.

    I doubt it. Japanese grammar is vast. Chinese is not. For example, let’s see a machine translation of Japanese to Chinese or vice versa when the Chinese or the Japanese contains polite/reverent/humble language.

    A machine translation can operate on rules. But we all know that learning a language by rules is not the best way to do it.

    Bad news here that is not wrong. broadcast FM or broadcasting FM instead of FM station. Depends what section of the English world you are translating for.

    Bad news for you: it’s still wrong. The Japanese clearly talks about a “classical music program”. And since we’re talking about FM radio, where the radio is always tuned to one station, it makes much less sense to listen to an “FM broadcast” when you’re told by the Japanese that someone’s listening to a certain program by means of FM radio. Because that invariably means that the radio is tuned to a certain station. Simple logic therefore dictates what the translation should be.

    放送 in Japanese can be used for Station, Broadcasting…. In Chinese this means RUN. …

    No. In Chinese it means exactly the same.

    No matter how much you focus on your knowledge of Sino-Japanese there are words in Chinese that have the exact same symbol as Japanese with completely different meaning.

    That’s kind of inevitable as the Japanese have only borrowed Chinese characters. So you will find that a certain character can have another meaning in Japanese. But most characters have the same meaning in Chinese and in Japanese (it’s a different story for compounds, though). You’re probably more thrown off by the fact that a kanji can look very different form-wise when written as a traditional or simplified Chinese character.

  59. bw says:

    “Lots of brands sell GNU/Linux on DangDang”

    Presumably that is what all that Asian text says, but of what good is that to you or me? I would never buy a computer from DangDang and I really doubt that you would either. Your goal seems to be to never buy a computer from anyone, preferring to resurrect junk box items on Debian, if I get the sense of most of your posts.

    I think there is enough business in the good old USA and even Canada to keep Microsoft humping for as long as computers are used. No matter what happens in China. Or Japan, or Bolivia, or Venezuela, or any other far corner of the earth. All that international money mostly stays international, due to the tax laws, and it has no material effect on the North American economy.

  60. Google Translate:” According to the agreement, Lenovo will ensure that Microsoft software is installed in its production of computer authorized and encourage dealers to sell pre-installed genuine Microsoft software computer. Gome will be dedicated to computer preinstalled Genuine Windows system sales in all stores, and strengthen efforts to publicize the existence of the risk of the popularization of the use of legitimate software to consumers the benefits of pirated software.”

    We shall see whether Lenovo and Gome see a drop in sales because their machines are not price-competitive with GNU/Linux. While Lenovo and Gome are big in China there are many other outlets M$ does not control. It is also possible that the government of China may respond to such exclusive dealing on the grounds that it is anti-competitive. I would love to know whether the agreement is an exclusive deal or just an agreement to ensure legitimate copies of M$’s OS are installed when that OS is installed on PCs in that channel. It may well be that no OS and GNU/Linux are still sold . Last year M$ sued Gome so I doubt there is a love-in happening. Gome is just one of many e-tailers there. The Alibaba group is into GNU/Linux. It’s no problem finding GNU/Linux installed on Lenovo PCs on DangDang which has a Gome store…

    Lots of brands sell GNU/Linux on DangDang.

  61. George Wilson wrote, ” Dietrich T. Schmitz has written a nice criticism of Debian”

    I wrote a reply to that which has not yet made it through moderation… It’s along the lines of “Debian GNU/Linux is for people who actually want to get things done with their PC instead of fiddling around.”

  62. oe wrote, “I really appreciate the ability to move the drive to a new system and still have it boot and mount up. With windows that was a very iffy proposition at best….”

    Amen. The Linux kernel is key to that and a huge slice of IT on everything from smart gadgets to mainframes. In schools with a few different PC-types having a single image work on all hard drives is a great help.

  63. oe says:

    Pogson you forgot to mention all the time spent on registry rot and malware driven re-install. On over 50 odd linux installs to date NEVER had to reninstall; once up its good to go for the life of the hardware. And on that note with the 2-3 hardware failures I really appreciate the ability to move the drive to a new system and still have it boot and mount up. With windows that was a very iffy proposition at best….

  64. oiaohm says:

    George Wilson I stated that problem about coming to english.

    –Chinese to English there are some major syntax points that cannot be translated.–

    This is the same with Japanese to English.

    The syntax error you can result by getting a word wrong using Sino-Japanese words will ruin you.

    Basic Chinese grammar is required either way.

    Remember I said Google translate Chinese to Japanese is the best. All those fine genitive bits map from Chinese to Japanese where they don’t to english.

    –The translation of FM放送 is wrong.–

    Bad news here that is not wrong. broadcast FM or broadcasting FM instead of FM station. Depends what section of the English world you are translating for.

    When translating the problem becomes very important to be aware stuff has multi-able possible translations.

    By pure bad luck you have picked on of the examples of why Japanese reading Chinese script is a really bad thing.

    放送 in Japanese can be used for Station, Broadcasting…. In Chinese this means RUN. If you have used goggle translate to go from Chinese to Japanese then read. What is basically a dictionary swap. You do not make complete disaster translations. No matter how much you focus on your knowledge of Sino-Japanese there are words in Chinese that have the exact same symbol as Japanese with completely different meaning.

    George Wilson there is a correct method for you to read Chinese since you know Japanses. This is using a dictionary conversion to remove at min the overlapping word with different meanings and replace them with equal in Japanese. Google translate and other translators on line provide you this.

    Not doing it the correct way is playing with fire.

  65. George Wilson says:

    A correct translation, by the way, would be along the lines of:

    In the taxi’s radio a classical music program of an FM station was playing.

    If you want to keep the transitive nature of the verb:

    The taxi’s radio was playing a classical music program of an FM station.

    Note that the verb’s translation is anything but literal. In Japanese there are many, many verbs where a literal translation taken from a dictionary will fail.

  66. George Wilson says:

    Google translate is more exact than someone attempt to use Japanese to read Chinese.

    No, it’s not. Basic Chinese grammar and good knowledge of Sino-Japanese words will get you much further than any machine translation.

    And as I wrote, the exact translation I spoke of was done by my Chinese colleague, you dimwit.

    Here’s an easy enough example (in Japanese, as that’s my forte):

    タクシーのラジオは、FM放送のクラシック音楽番組を流していた。

    That’s not a complex sentence. After Google Translate has mangled it, out comes this:

    Radio taxi, had shed a classical music program broadcast FM.

    Google Translate fails:

    – タクシのラジオ is not the subject.
    – The first の has the function of a genitive here.
    – The translation of FM放送 is wrong.
    – The second の has the function of a genitive here.
    – The tense is wrong because Google does not realize that the ている form describes a state here.
    – The verb’s translation is wrong because Google does not take into account the context.

    Six mistakes. In one simple sentence.

    Sorry, I’m not impressed.

  67. oiaohm says:

    Gome Electrical Appliances Ltd George Wilson is also the one of the ones that partner with Dell providing Ubuntu machines into China.

    So You have Windows machines, Ubuntu machines and no OS machines in their stores.

    So number of those machines leave the store with Linux installed. George Wilson

  68. oiaohm says:

    George Wilson
    –The other 30% will undoubtedly run a pirated Windows before long. Not Linux.–
    This is a presume not based on any numbers. This is normal Microsoft Presume anything non Windows is Pirate.

    Look at ChromeOS sales from Amazon 10 percent of all notebook/netbook devices are ChromeOS devices.

    Remember Linux people do pave over. So no All Windows does not change the number of machines that Linux can be installed on.

    I know both languages. I know that you can possibly read with good knowledge of kanji. But I know doing that you screw up meaning massively at times.

    –I checked an exact translation afterwards–
    Google translate is more exact than someone attempt to use Japanese to read Chinese. Even so google translate not 100 percent. Still should be checked if you are 100 percent depending on it to make a point.

    George Wilson anyone who really knows translation knows what you attempted is only a last resort if you don’t have access to goggle translate or equal.

    So please learn from you mistake this time and not attempt to use your Japanese to read Chinese when you have Google translate and other services at offer. You Japanese can allow you to see errors in google translate at times. But your Japanese alone the error rate you have is way to high when reading Chinese than using Google to go from Chinese to Japanese. Chinese to English there are some major syntax points that cannot be translated. Even so Google Chinese to English is more exact than using Japanese to read Chinese.

    This is the problem I deal with finding translation errors.

    Still you should have been able to confirm in this case that the Google translate to english was fairly close to exact and been correct and used that on a english speaking blog.

    Its common to get a Japanese idiot thinking there good skill in that is better than google translate.

  69. George Wilson says:

    No, you dimwit. I stated that my Chinese is rather *bad*. But my Japanese is rather *good*. And if you knew anything about Chinese and Japanese you’d know that it is possible to actually read Chinese rather well with a good knowledge of kanji. And yes, I checked an exact translation afterwards — a translation done by one of my colleagues who happens to be a Chinese. And the text says exactly what I thought it did.

    You simply are unable to grasp the significance:

    – Lenovo has a vast network of retail stores in China. And Lenovo has now agreed to sell in those stores computers with licensed Windows pre-installed, meaning the number of computers without OS where one could in theory install Linux will diminish.

    – GOME, a large electronics chain, has agreed to the same.

    So where Microsoft previously was ruling China with pirated Windows copies, they’ll now rule China with licensed Windows copies.

    http://microsoft-news.tmcnet.com/news/2013/03/29/7026142.htm

    70% of those 3 million PCs will come with genuine Windows pre-installed, a big number in Piracy-loving China. The other 30% will undoubtedly run a pirated Windows before long. Not Linux.

  70. oiaohm says:

    George Wilson the reality Dietrich T. Schmitz criticism of Debian is the normal outsider of Debian.

    Debian has changed. Ok Debian rate of change appears slow this is why its liked for particular things.

    Debian compared to most distributions insists on a massive amount of quality testing.

    2005 http://piuparts.debian.org/ starts on Debian. Now is run by default against every new package. Less cases of malformed package that don’t uninstall correctly.

    http://jenkins.debian.net/userContent/about.html

    Now jenkins 2012.

    Other tools are under development to automate more of the Quality control process.

    So the only way debian can release faster is reduce it quality control cost. Please note Ubuntu and most other distributions don’t release on as many platforms as debian or run the same quality controls.

    Some of the debian monitoring is funny http://edos.debian.net/weather/ Yes this is 2008.

    George Wilson debian is one of the leading distributions in automated quality control. Every 2 to 3 years a new automated quality control feature gets added to debian.

    Debian renown stability is not without it causes.

    This is the problem George Wilson how to have fast release cycles and high quality controls.

    Debian release speed is in fact the about the same as Microsofts. Yes slow compare to Linux Distributions about comparable to most closed source operating systems.

    George Wilson make a choice High Quality Control or Rapid release cycle. You can only have one.

  71. oiaohm says:

    George Wilson you posted a link that was not in english he don’t speak Japanese or Chinese.

    I have some bad news for you the artificial does not say what you think it does. George Wilson

    I really don’t bother normally answering morons who try to make out they have something special when it exact same thing just in another language.

    In future don’t link to crap you cannot read.

    http://translate.google.com/translate?sl=auto&tl=en&js=n&prev=_t&hl=en&ie=UTF-8&eotf=1&u=http%3A%2F%2Fwww.stdaily.com%2Fkjrb%2Fcontent%2F2013-03%2F28%2Fcontent_587404.htm

    Fairly much a stock standard Microsoft Genuine Advantage Campaign. I do like the new title for it in China “really safe”. Result Microsoft is paying advertising.

    George Wilson really you had stated in your post that you did not understand what the artical said. So why should we have bothered answering someone who did not have a full clue about what they were posting about.

  72. George Wilson says:

    I also see that your good friend Dietrich T. Schmitz has written a nice criticism of Debian, the laughable OS you so heartily recommend. It seems there are some people in the Linux community who are not yet deluded. Good for them.

  73. George Wilson says:

    Desktop Linux is garbage. Users and OEMs both know it. It’s telling that you never answered here, where I showed you that the Microsoft-Lenovo partnership is for real and was all over the Chinese news. So even in your holy grail country of Linux growth people want Windows — they way it always was, the way it always will be. In China and everywhere else.

Leave a Reply