7 Reason Why Linux Flies on the Desktop

GNU/Linux distros that work on desktop PCs have been around for 15 years or so. They are polished and effective. Still, some people think there’s no real advantage over the MacOS or that other OS, but I disagree:

  1. Cost: $free licences are just the tip of the iceberg in savings. Re-using code is efficient so the cost of production of the software is much less than non-free software. Having all your compatibility issues dealt with by the up-stream packagers is even larger. Instead of you having to hunt down drivers, updates for individual applications, security updates, upgrades to new versions, the package managers of many distros like Debian GNU/Linux allow you to update one machine by a simple command:apt-get update;apt-get upgrade. That’s all there is to it. You get the latest driver, kernel, applications, security updates/patches, bug-fixes for documentation, applications, services and the OS. Have a 1000 machines? No problem. There are many options:
    • insert that command in a script run daily, weekly, monthly, etc. on each machine
    • send that command to each machine, for ((f=1001;f=f-1;));do ssh machine$f “apt-get update;apt-get upgrade”&done
    • wake up all the machines you want at midnight and send them a whole new disc-image with CloneZilla
    • etc. Whatever you imagine, you can do with GNU/Linux.

    In addition to licensing and management, GNU/Linux just runs better, has fewer problems with malware, and will run on anything built in the last decade.

    Think I’m lying? Ask these folks, who have reported $millions in saving, not only with licensing but ease of management and keeping old hardware working longer:

    • French National Police:“the Gendarmerie estimates dramatic savings from using free open source software that amount to nearly 7 million Euros per year.”
    • City of Munich:“The current Linux desktop in Munich (Limux) costs the city 11.7 million euro, writes the Mayor in council bulletin number 54, published on 19 March. To offer a desktop similar to the LiMux desktops, on the basis of a proprietary system would costs at least 15.52 million euro, Ude concludes. And that would not yet include the costs for the renewal of the licences of the proprietary software, which would add another 2.8 million euro every three to four years.
      The Linux based PCs are also far less troublesome, Ude points out. Since 2006 the number of malfunctions reported to the helpdesk decreased from 70 to 46 per month in the same period that the number of Linux desktops increased from 1500 to 9500.”
    • German Foreign Ministry:“The Foreign Ministry is running desktops in many far away and some very difficult locations. Yet we invest only one thousand euro per desktop per year. That is far lower than other ministries, that on average invest more than 3000 euro per desktop per year.”
    • Ernie Ball Guitar Strings:“I know I saved $80,000 right away by going to open source, and each time something like (Windows) XP comes along, I save even more money because I don’t have to buy new equipment to run the software. One of the great things is that we’re able to run a poor man’s thin client by using old computers we weren’t using before because it couldn’t handle Windows 2000. They work fine with the software we have now.”
    • City of Largo, Florida:“The City has received international recognition for implementation of Linux open source technology as the platform of choice. Through the selection of open source solutions for word processing, electronic mail, spreadsheet, scheduling and office automation software, the IT department continues to provide extensive savings over more traditional alternatives in city-wide applications.”
  2. Compatibility: GNU/Linux depends strongly on open standards. Since all of it uses open standards all GNU/Linux systems work together with few problems. Compare that with Vista and “7” not speaking the same networking protocols so people are pressured to upgrade to “7” when even a few new machines with “7” enter the system. Who’s running your IT? You or M$?
  3. Making it work: There are still those who think it is easier to keep the system working with that other OS than with GNU/Linux. Ha! What about
    • malware?
    • slowing down?
    • re-re-reboots?
    • Patch Tuesdays?
    • Phoning home?
    • documenting the identities of machines and their “certificates of authenticity”?
    • needing a different driver for each device even though the same driver can run several devices using the same chips in GNU/Linux?
    • registry/file-system corruption?

    GNU/Linux has none of those problems and few of its own.

If facts don’t persuade you, consider the lies people tell when they spread the FUD about GNU/Linux:“A typical thing in a Windows setting is to establish some usage policies, and set up some limitations on the systems to keep them stable. Linux doesn’t have those types of standards out of the box.”
Yes. Maybe GNU/Linux doesn’t need its hand held like that other OS. Being more resilient means one can give end-users more freedom without breaking things. For example:

  • I can let users print to printers for which there is no driver on the user’s PC… cat file|ssh machine_with_driver “lp -D printer “
  • I can let a user with an 8 year old PC have the performance of a new PC by including a line in a file in their boot-scripts:X -query hot_new_machine
  • I can have 20 users running on a single PC each enjoying the experience where a single user running that other OS would be faced with annoying delays when they logged in or clicked on anything.
  • I can have users visit websites with images without fear the OS will execute an image and be taken over by malware…

Don’t see Why Linux is a desktop flop if you read and understood this post.

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.

41 Responses to 7 Reason Why Linux Flies on the Desktop

  1. Yonah says:

    A mixed bag indeed but Linux has more in the bag. Fecal matter, that is. “It’s not a problem” and “The user is stupid.” Ahhh…. Two classic gems that never go out of style.

    No, I don’t wonder why you use “serous-ally”, nor did I in any way imply that I was wondering. I’m aware of your disability. I’m asking why you don’t have just a little more self-respect and tackle a problem that would help people take you a little more… ahem… seriously.

    So, this is YOUR system that YOU have developed. Because your first posting on this thing used the plural word “systems” and made it sound like this was actually something other people could use. Can you give us the source code? Some pictures? Let us see it!

    I have an idea for an even better project for you. Dare I say it’s one that isn’t a complete waste of time. Automated GNU/Linux Defense Systems. Simply a program, running ONLY under GNU/Linux of course, that searches the Internet for anyone, anywhere, who writes anything stating GNU/Linux or any FOSS software has a problem or isn’t as good as the competition. It then scans those remarks and pulls up recomposed responses to counter those claims. Think about that! What you do now could be done with a software program, thus freeing up your time to work on your Automated Assault Systems, get more done at work, or have a girlfriend (assuming you didn’t build one yourself). I do caution, however, that you have all the pre-typed responses checked by at least 2 other people with at least High School level English. Ohh, and 2 other improvements. Make them short but sweet, and include lots of citations.

    Defend the honor of Linux, defeat those stupid people, and the princess could be yours. *wink*

  2. Phenom says:

    Pogson wrote: I don’t confuse the performance of applications with the performance of the OS doing video.

    Doing video? Where is the great achievement here? Even Xbox can do HDMI video since 2007.

  3. oiaohm says:

    Yonah without giving away exactly what I am doing in my “Automated Assault System” as this might give a few less respectful cheaters ideas how to be king.

    Take a demo video of a FPS game. Apply opencv Optical Flow tracker over the top and watch the video. Then remember the top of char in all current FPS games is the most killing point to the char to hit. Remember Optical Flow tracker processes very quickly on a real-time video stream on the right hardware. Faster than human reflexes by large margin.

    I remember because I was reminded my Automated Assault System does have a bug. Mech Warrior class games. Where the most fatal point changes with the char you are going head to head with. Does not cope because it has not really ID the other char just the other char motion and aims for the higher point what is normally fatal.

    Your targets you wish to shot are the items moving out of relative motion to everything else and the hud in most FPS games is fixed so you can configure in the fire point. Basically anti camping setting makes the Automated Assault System work.

    It does have a detectable feature. As person walks out past wall it shots up there body every single time. In effect it does more damage doing this. Since the person appears really short and grows.

    There are a few filters to remove smoke and other detraction to the Optical Flow tracker that would be given away by any video if it was compared to the game it was being used with. It does take a while to tune that bit.

    Over all its vastly superior to most humans playing the games. Behaviour is not the hacked games but it does have a unique behaviour.

    The vehicles in most games have a unique silhouette. Mech warrior there are a few Mech that look close to each other. in silhouette when they are bearing down on you. But these are avoidable in most FPS games by avoiding the maps with vehicles.

    My problem is not tactics about moving around map. I am great on that. My problem is the firing shots that will hit them and fancy foot work moves that I can script in. Basically I have proven that a generic game independent cheat is possible. Mostly started doing to to test out computer based vision systems. Its like the FBI working out that they can almost perfectly fake paintings without humans and mass print them. The FPS systems are not design to defeat computer based vision systems because if they are they would defeat the humans playing the game as well.

  4. oiaohm says:

    http://linux.slashdot.org/comments.pl?sid=2834521&cid=39925443 Other comments in that link as well.
    Reality its a mixed bag. Windows sux on some hardware Linux sux on others.

    Majority of the cause is items not built to spec. “Hardware is supported… well, kinda.. sorta… maybe….” That kinda applies to everything for all OS’s because of all this not to spec hardware.

    The link you provided is also funny. http://hostap.epitest.fi/wpa_supplicant/ Wpa is always supported maybe someone missed installing wpa_supplicant as is required under Linux and Windows so some cards support wpa. Because they don’t have a hardware wpa implementation.

    This is where it become stupid people complain about stuff that is not real problems. You have a card that only does wep install wpa_supplicant.

    Yes there are some issues. But there are a lot of issues that people complain about that are nothing more than stupidity. Linux there is no such thing as a card that cannot do wpa either in hardware or software. Of course if the software not installed and you card needs it no wpa. In fact there is no card that cannot do wpa2.

    Yes the fix is for windows and Linux. So that next time a person complains that there windows laptop cannot connection to your wpa2 network access point because there laptop does not do wpa2 kick them up ass for not having the right software installed.

    There are real issues. We need to talk about the real-issues we don’t need fake ones in the mix.

    Since its the same software for windows and Linux to fix a card that does not support wpa or wpa2 how could a windows user fail to know how to install it. Right they are idiots. This is where a my hardware is not broken comes from. Also I guess they go out and buy new wireless cards without need as well.

    This is the problem difference between broken and not broken hardware windows users don’t have a clue. They call good hardware broken and broken hardware good.

    You wonder why I use serous-ally so much how could people be this dumb. Yonah

  5. Yonah says:

    A great Slashdot thread, though biased as it is, describes a problem I’ve had with some Linux drivers. Hardware is supported… well, kinda.. sorta… maybe….


    Oiaohm, consider this bit of wisdom. If most people are breaking the law, perhaps the people are not wrong. The law is wrong. Always loved that one.

    Two requests, one for me and one for you. For me, can you post a link to a video showing an “Automated Assault System” in action? I need some sort of citation if I’m going to make my addition to Wikipedia.

    For yourself, just once a day, pass on posting a wall of text defense and repeated type “seriously, randomly,” over and over for 5 minutes. Briefly put down the FOSS torch and invest in something that despite our differences, all of us care about. It’s you, buddy. You’ll gain higher self-confidence based more on your ability and less on your ego. Don’t say it’s too hard. There’s a blind man here in Beijing who I think should inspire you.

  6. oiaohm says:

    Viktor even that it has improved. You still find machines with broken ACPI. http://biosbits.org yes is open source but is also the official software to reference check most of the bios. Made by Intel.

    You still find new machines that fail that don’t conform perfectly to ACPI. biosbits started out of the direct problem is when Linux is using the reference implementation of ACPI from the specification. So the reference implementation perfectly made to the spec but it don’t work Viktor.

    The reason windows works is that its broken in places and does not run fully to specification like disregarding quite a few power saving acpi options or is software patched over defective BIOS.

    Problem is your complete argument is not based on fact Viktor. The hard reality is that both freebsd and Linux truly do use the reference implementation of ACPI. The implementation that all other implementations should be identical to. So Linux having issues Windows not equals Windows not to spec somewhere. And if future version of windows comes into spec it will not work on that hardware either.

    So yes the Linux and FreeBSD guys have every right to be fuming over it. Its just like early EFI. It has some major bugs like writing into areas of memory its released where the OS could be now. Yet there is no firmware update coming for those motherboards. The makers have said live with it.

    You did not answer my question either what version of intel driver are you using and are they older than Linux 3.0.

    Viktor is simple to try to blame Linux but please do your homework first. The ACPI issue with Linux is a major problem that does not just stop at Linux it also effects people upgrading windows and at times updating windows drivers. You have the magical idea that you on Windows is safe from it. Reality you are not.

    This is why people getting up on there high horse over something works with Windows but not Linux find themselves in so much trouble. Because in a lot of places the Linux kernel implementation is the reference since its simpler to import the reference than code a new one from scratch and the license is compatible so why would you code from scratch.

    There are a lot of hacks in the Linux implementation to leave spec in places because of known defective bios stuff that only happens if you machine is detected on the defective list. Otherwise linux runs the reference.

    So a lot of machines are defective. Only middle of 2011 did this start to reduce as the more advance test-suites start coming on-line like http://biosbits.org .

    The real facts destroy most of these Linux Sux arguments. Instead you start wondering how such a poor grade of hardware was let out door.

  7. Viktor says:

    I have no idea what you’re saying, Pogson. At around the time of Windows 98, Windows’ ACPI implementation was equally bad. They have improved, Linux has not. The official ACPI specification is out there for everyone to read. It was once rather ambiguous in many places, it’s not today. The sad fact is not ACPI, the sad fact is that some Linux hackers can’t whip up a decent ACPI implementation.

  8. oiaohm says:

    “but it also worked in Linux before kernel 3.0.”
    Interesting question what are are you intel drivers for windows particular version. This will be interesting because they have to be pre power management acpi activation. 3.0 turns on the power management sections of the card that turns on a nasty trap. Cuts power usage in gpu idle by over half. If you are using older windows drivers than Linux kernel 3.0 that could explain it as well. Avoid the bug by not being upto date with drivers.

    From a secuirty point of view I am normally not wanting old drivers.

    “Yet somehow the closed source OSes manage to handle ACPI happily (no matter what), while the “omnipotent” (according to you) FLOSS fails miserably.”
    Phenom reality is windows with ACPI bugs creates hidden land mines waiting for the correct event to set them off.

    So neither OS is really handling ACPI that well because its a huge mess. EFI was ment to resolve it now that is turning into another huge mess with strange and bad behaviours.

  9. I don’t confuse the performance of applications with the performance of the OS doing video.

  10. Phenom says:

    Pogson wrote: Superior, if you ask me!

    Please show me the visually superior games for Linux with visual effect, superior to the games for Windows.

  11. Phenom wrote, “video in Linux is on par with video on Windows, right?”

    Superior, if you ask me! Same or better performance and lower cost.

    It doesn’t get any better than this.

  12. Phenom says:

    Yet somehow the closed source OSes manage to handle ACPI happily (no matter what), while the “omnipotent” (according to you) FLOSS fails miserably.

    But we are told that video in Linux is on par with video on Windows, right? 🙂

  13. “Linus: Yeah. I mean, suspend and resume really should be able to do all of that. A lot of people want to just do it with the same kernel. Because replacing the kernel is really hard. So, who knows? Let’s get the suspend and resume thing working first.

    The nastiest part of that is, it actually works on some machines, but it’s completely impossible to debug. Because when the machine doesn’t come up, you need an ice to know what’s the hell is going on, because you don’t have any devices that work. Even getting a serial line out isn’t very good, because you’re southbridge is hosed. And you can’t even get to the chip that does serial I/O, right?

    That’s why it’s really nasty. APM was a lot easier in that respect. When you had APM you had the BIOSes, but at least it made sure the serial line worked.

    Modern PCs are horrible. ACPI is a complete design disaster in every way. But we’re kind of stuck with it. If any Intel people are listening to this and you had anything to do with ACPI, shoot yourself now, before you reproduce.”

    see Doc Searls’ Linus and the Lunatics (2003)

  14. Viktor says:

    asus with intel 4500 is one of those. Knowing Viktor most likely does not really own the hardware but looked threw lkml logs has no clue that the fault is not fixable by anything other than a firmware replacement properly.

    Just shut up already.

    Not only does it work in Windows with Intel’s generic drivers — again: generic drivers! I don’t have Acer’s drivers installed — but it also worked in Linux before kernel 3.0.

    ACPI in Linux is fucked up, not the other way around.

  15. oiaohm says:

    dopeyjoe no there are just those odd bits of hardware that are truly defective.

    asus with intel 4500 is one of those. Knowing Viktor most likely does not really own the hardware but looked threw lkml logs has no clue that the fault is not fixable by anything other than a firmware replacement properly.

    This leaves the Linux kernel developers in a very tricky location. Masking over the acpi bug is not exactly valid since the right conditions still can cause it.

    Simple fact is windows users have this stupid idea that just because windows runs the hardware is fine. Its like one my boards here audio does not work under windows at all. Works under Linux. The audio chip is in fact damaged. Particular modes don’t work and cause it to lock up. One of those mode happens be the mode windows plays it start up sound on. As long as you feed it stereo and not particular freqs it loves you. Feed it mono or the wrong freq its stopped.

    So can I claim that windows is busted because the board is busted. No I cannot. Serous-ally this is what viktor is trying todo.

    This is the reality of hardware. Some of it out there is truly busted sometimes produced in mass volume. Ok might not appear to be busted to you yet. Like if I try to use that one mode my sound dies until power off and back on completely. Its a random land mine. Lot of hardware has them.

    People don’t want to accept the fact that some-things are broken. That is the reality. Linux world is normally fast to fix the one where the fix truly works. The one where its leaving a land mine not so much. Better to be a pain and not be run on that hardware.

    So this case they are on to Asus to provide firmware fixes to its fixed properly.

    Really people like you dopeyjoe are the retardedness. Who keep hardware that has key defects in them. That normally hurt you.

    Windows runs it fine attitude is so wrong its not funny. Worst example of this was a batch of foxcomm boards Linux refused to run on windows ran all right but every shutdown disc cache was not getting flushed to disc properly of course those machines had been in for repair a lot.

    Reality is if Linux does not run on something something is normally broken to spec somewhere in the machine. Sometimes this is OK can be software fixed by windows driver or Linux drivers. Sometimes this is not OK because it really cannot be software fixed.

    Viktor picked a case that its not OK. Reason Linux is not working is he has a timebomb that the Linux world cannot work out how to fully disarm.

    So really who is the retard here its Viktor. He has dead hardware yet he don’t want to admit it. Because its simpler to blame Linux.

  16. dopeyjoe says:

    oiaohm, you are the epitome of open sores retardedness.

  17. oiaohm says:

    I should have been more exact on the don’t install intel graphics drivers. Should have been don’t install those alone and install the ASUS ones first always on a reinstall otherwise you will suffer if the firmware is not fixed.

    You will find that its dependant on a ASUS driver to work correctly to a point that part fixes the acpi in windows in software. Until the intel has a overload reset and pulls table straight from bios so bypassing the ASUS fix. Yes this is two bugs. One that the intel 4500 does not take the new software ACPI tables as it should. Two that ASUS got it wrong in firmware. And the pull from bios is hard-wired into the intel chip.

    Basically you have a double lemon. Its only a question of time before you hit the bug Viktor in Windows. Really viktor you need to understand before you get on high horse. This case you did not understand the cause so you did not understand what asus did to hide there error and then did not understand that intel had made a different error as well.

    Welcome to the perfect storm. Most of the time software fix on windows works. You own one of the rare machines were it will not.

  18. oiaohm says:

    Viktor double check them you will find they are not the intel generic or at least not intel generic alone.

    If they are intel generic you will find asus driver installed overriding the acpi in software. This is not perfect. Basically install Windows 7 + just intel generic you are in trouble.

    The fact is the intel driver does work if the ASUS board tells the truth or if ASUS had release a firmware fix driver for Linux.

    This is the problem with Windows users they insist there hardware is perfect until you go hey lets install this machine clean with exactly min and I will show you that windows intel drivers are exactly as bad as the Linux ones on hardware detect.

    Methods to hide defects so they don’t have to scrap machines or release new firmwares. We will fix that in drivers and pray we get it right.

    Mask over a firmware bug is a nightmare waiting to happen.

  19. Viktor says:

    What ever you don’t install the generic intel drivers for your card on Windows you will regret it until you fix firmware.

    My God! Are you braindead!? I have the generic Intel drivers installed in Windows 7. They work! Stop blabbering about a failure in the firmware. The Linux drivers are made by Intel, they are open source, they don’t work.

  20. oiaohm says:

    iLia and Viktor. The difference is that if Linux works on a hardware everything is fine most cases.

    Windows works if the hardware is truly fine under or not.

    Viktor Acer has the power on screen backwards to Intel standard yet have acpi table as per a standard intel device. What ever you don’t install the generic intel drivers for your card on Windows you will regret it until you fix firmware. The option I gave before will turn Acer on can be placed in start-up scripts. Yes it is defective firmware. Yes by design setpci -s 00:02.0 F4.B=01 should be on. But acer its off and setpci -s 00:02.0 F4.B=00 is on.

    Yes windows does have the fault yes it will come back in a card reset where the hardware refers to the bios acpi tables. You are sitting on a timebomb if you are running windows newer than XP on it. Vista and Windows 7 added a reset gpu if detected non responsive this is what will cause you to black out as well.

    There are a few ways to fix this. Override the acpi tables this is what the windows drivers for acer do when they install this works perfectly for Linux because the Linux drivers do refer to the OS acpi tables on reset. Add setpci to Linux start up script works so so same risk as windows. Try the linux acpi table and hope acer has not stuffed that up what they have in a few models.

    You can force Linux to boot up standard vesa so allowing you to set the setpci or change the tables.

    Do check for a firmware update for you acer Viktor some already do have firmware updates for windows vista and 7 referring to interment black screens. Apply that fix Linux installs perfectly after that.

    Cause is exactly the same. acpi tables being wrong in bios.

    iLia this is not a case of more or less faults this is a case of exactly the same number faults. Viktor has a fault that cause 1 effect under Linux and a different effect under windows. But the fault is exactly the same thing. acpi in bios tables saying to turn screen on is off and to turn screen off is on. How is this not a hardware bug???

    Viktor I could if I was paid get you the command to push the acpi to switch screen on an off. You do it on any other laptops or desktop other than these defective ones it works perfectly. In fact if you had tried turning the screen off before installing the acer drivers it would have failed under windows then when you went to turn it back on blackout. Would have come back when screen suspended. Not fun with windows either.

    Yes this is a case of completely wrong. Linux is not the only one with the problem. Lose you acer driver discs you are in trouble as well.

    Pre 3.0 Linux did not try to use the acpi tables for this to turn screens on. So other with laptops that started with screens turned off by default were complaining and people were complaining that power saving was not working as well.

    Its a case you cannot win every time.

    Viktor that a fault is hidden does not equal no fault. The firmware update that fix the Linux problem addressing windows problems really prove this as not the case.

    iLia what is the worse out come. Linux fails on startup so you did not get to doing work. Or fails when you are half way threw work because the screen suspended under windows and you had missed installing the acer drivers.

    Random vs constant. Random is worse. I will give viktor that the process for fixing this for end users should be made simpler. But end user does need to take away information about the bug they have so they don’t get trapped by it on windows.

    Viktor and iLia really linux world never expected a problem like this that an option would be backwards to standard.

    Who writes the driver getting it wrong on Linux Viktor and iLia. Intel themselves. Yes its a true case acer as wired up a Intel chip wrong. Then not alter acpi tables to match. Intel 4500 reference works perfectly with the Linux default driver. The reference that was provided to all hardware makers.

    Production errors like this do happen. Linux does need a neater way of handling it.

  21. iLia says:

    Linux is constant with the fault. Windows is random with the same fault.

    When Windows has more problems than Linux — Windows is bad.

    When Windows has fewer problems than Linux — Windows is bad.

    Flawless logic!

  22. Viktor says:

    Linux is constant with the fault. Windows is random with the same fault. So windows user might think they don’t have it because they hit it less frequently.

    Ahem, no Ham. In Windows there is no fault. Despite it all being closed source. You can even read excuses on LKML somewhere for my specific problem, where someone accuses Acer (the maker of my notebook) of having bugged the hardware. I will search for the relevant thread.

  23. kozmcrae says:

    Ted wrote:

    “I didn’t even have to leave Germany.v”

    I would not be surprised if Microsoft had a list somewhere of organizations that switched back to Windows after trying Linux. That would be the place to look. I would think they would make a big stink about it. You shouldn’t have to go digging around searching for failed migrations.

  24. oiaohm says:

    “oiaohm:Duel boot machine method does not work in business environments

    Of cause, when people have choice they choose Windows.”

    Duel boot does not work windows to windows either.

    Its a universal way to fail. People will stick to the old version of windows on you if you provide duel boot as a option as well. All migrations with duel boot are on the path to failure question is how long before the person running the migration either aborts the migration or stops using duel boot in the migration process.

    iLia Maykop is small to all of Malaysia government and the failure there mirror Vienna.

    Yes Ted Vienna in Austria there are reports on that one why it failed as well. Its another one that included duel boot in the plan. This is fault 1.

    Biggest reason given for Vienna was that all there internal web services need IE. Now really this should be the first thing fixed before trying to migrate anywhere. Even firfox or chrome on windows would not work on there websites. So it was not possible even for them to migrate web browsers let alone OS. What are they going todo if IE stops supporting old IE designed websites most likely be up creek. Notice something this fault is worse is fairly much OS neutral. OS X or OSi or anything other than windows is basically screwed by it and they are at the wim of IE if there internal processes work. This is Fault 2.

    Next they had the insane idea of making there own upstream for a lot of applications. This is fault 3.

    Any one of those three doomed the project to absolute failure. Question we don’t know is what one or it was all of them.

    Libreoffice stayed so the project for Vienna still brought cost saving.

    “Today my Ubuntu-Linux box has rebooted twice, maybe it is because of my Huawai USB-modem”
    I can tell you absolutely it will not be that. All Huawai USB-modems are perfectly supported in Linux other than maybe a need to install usbmodswitch so they go into the right mode automatically this does not cause a crash or a reboot. Even a damaged Huawai or other USB device will not cause Linux to reboot. USB controller having bios that random-ally writes into sections of memory will do that it is also a time bomb for windows. Same with many other bios land mines.

    First thing iLia and Viktor get http://biosbits.org/ both your issues might in fact trace to a dead bios and that you should be after a firmware update.

    Yes some intel chip laptops to turn on the backlight need a firmware update. Same firmware update prevents direct x and opengl failure under windows. So I do question Viktor is your problem you device just you are not getting the message.

    Viktor if it is the firmware its normally this one that makes it turn on setpci -s 00:02.0 F4.B=00 with 00:02.0 being the pci address of your video card. Fun is if halfway or exit of game threw a opengl or direct x game under windows it calls a reset on the card this fault results in black out under windows. Quite a few people put this down to windows locking up what was not the fact. So that black out failure on the 4500 does not only effect Linux users. Linux users its lot more in the face since Linux tests a video card full reset on startup. Windows does not so if a full reset happens under windows because the card becomes non responsive for any reason on the 4500 with bad firmware same problem raises its ugly head.

    Viktor Linux is constant with the fault. Windows is random with the same fault. So windows user might think they don’t have it because they hit it less frequently.

    Big problem with a lot of these firmware fault once the device is shipped lot of hardware maker wash hands of it and never provide an update so leaving windows and linux suffering forever more.

  25. Ted says:

    You got me there. Geography is not my strong suit. Should read – I didn’t have to go far from Germany.

    It still counts as an organisation switching back to Windows, though. They do exist.

  26. Ted wrote, “I didn’t even have to leave Germany.”

    Uh… Vienna is in Austria, not Germany (Hitler did not get his way.).

    The reversal in Austria was a huge mess. The government with one hand acquired M$-only technology and tried to migrate to GNU/Linux. Not smart.

    That’s not really a commentary on GNU/Linux but an example of people locking their own cell. The correct response would be to chuck the M$-only technology, but I guess they did not want to think of all that money “thrown away”. The fact is the money was thrown away sticking with M$, not going to GNU/Linux.

  27. oldman says:

    “I wanted to try out the new Umbuntu 12.04 release, to see if they had finally fixed the problem that plagues the Linux kernel since version 3.0”

    I highly recommend virtualization if you wish to play. I have found that it is far safer to allow the hardware to be owned by the OS’s that can support it the best. Hence Linux has never been run on my systems as anything other than a virtual machine guest under windows.

  28. Ted says:


    “Keep looking, there must be another one out there some where.”

    I didn’t even have to leave Germany.


  29. Viktor says:

    I don’t have any problems with unnecessary reboots here.

    That’s great, Pogson. The last time my Windows 7 rebooted was when I wanted it to. That was about 11 days ago. The reason? I wanted to try out the new Umbuntu 12.04 release, to see if they had finally fixed the problem that plagues the Linux kernel since version 3.0, the problem that the backlight is not switched on when booting (this concerns Intels ubiquitous 4500 GPUs). Alas, I hoped in vain. Even esoteric workarounds didn’t help.

  30. Phenom says:

    Pogson, I don’t have any problems with unnecessary reboots on Windows, too, but this doesn’t stop you from bashing Windows. Why should we stop mock at loons?

  31. iLia wrote, “Today my Ubuntu-Linux box has rebooted twice”.

    You should fix it then. With FLOSS, you can.

    I don’t have any problems with unnecessary reboots here:
    “pogson@beast:~/US_DOJ_v_M$$ uptime;ssh ellaina uptime
    09:27:50 up 8 days, 13:33, 7 users, load average: 0.16, 0.14, 0.08
    09:27:51 up 113 days, 22:59, 1 user, load average: 0.15, 0.11, 0.07

  32. kozmcrae says:

    Viktor wrote:

    “Linux did not fly in the German Foreign Office. Never. Users complained about Linux from the start. On desktop computers, where dual-boot was offered, Windows was booted by users, not Linux.”

    Viktor makes this stuff up to make himself feel good. Do you feel better now Viktor? It’s tough when news of Microsoft gaining back office space is so scarce.

    Let us know when if there were any other switch backs will you Viktor because so far this looks like the only one. Keep looking, there must be another one out there some where.

  33. iLia says:

    Mr.Pogson: re-re-reboots?

    Today my Ubuntu-Linux box has rebooted twice, maybe it is because of my Huawai USB-modem, i don’t know, but it has rebooted without any interaction from my part. Thanks that I have not lost any data.

    oiaohm:Duel boot machine method does not work in business environments

    Of cause, when people have choice they choose Windows.

    So is that the best you can do? How many organizations that switch to GNU/Linux switch back to Microsoft (any version of their OS)?

    Russian city of Maykop, a city of 144K, switched back to Windows.

    Mr.Pogson:I suspect eventually Germany will have a government that dictates use of FLOSS as the default and there will be wholesale migrations.

    A government that dictates is called dictatorship.

    The goverment of Cuba does it.

    Rodrigez:Private software can have black holes and malicious codes that one doesn’t know about. That doesn’t happen with free software”

    But it has critical security vulnerabilities.

  34. Ted wrote, “With a plan to upgrade to Windows 7 and Office 2010. They only mentioned that in the same sentence.”

    Well, the migration to FLOSS had some good effects. The public and the workers know the elected officials at the top are the cause for the whiplash. That makes the future of that other OS doubtful, because it has a single point of failure… I expect we have not heard the last word on this. I suspect eventually Germany will have a government that dictates use of FLOSS as the default and there will be wholesale migrations. How many governments in Germany have switched on their own to good effect?

  35. oiaohm says:

    Ted no matter what you play the true remains the method of migration German Foreign Ministry did was a disaster almost 100 percent sure failure.

    “How many switch away from Microsoft in the first place? Very few and far between I’d think, considering the same usual suspects always get rolled out as the poster-children for Linux migration.”

    Most are sane enough to follow IBM Redbooks. So get to 80 20 migration states that are stable. 80 percent Linux 20 percent Windows.

    If you read the IBM redbooks you found out there new examples of migrations every year of course trolls don’t cover this. Instead jack hammer the same ones.


    Also notice the list of groups who go public about it. Not the ones that never go public about it redhat and ibm always have more nameless examples. But they are real migrations just names with held at client request.

    Like you hear nothing of Malaysia that 80-90 of there government departments run Linux only. And the remaining departments are a mixture of windows and Linux. Yes that includes their Foreign Ministry.

    Areas that work are making no noise at all.

    Yes deployments like Macedonia’s Ministry of Education and Science also make the German Foreign Office look small and not even of scale. 180,000 machines in 1 year. Serous to the Large Linux migrations the ones that failed look small to and should have been simple to pull off.

    This does directly suggest miss management and it does show when look look closer like examples of duel boot or insane migrate everything moves without any legacy application support plans. Even munich you can find miss management leading to its issues why it took massively longer than normal Linux migrations do.

    Really ted follow viktor and download and read the IBM redbooks and get a little informed on this topic. Wake up most of the failed migrations were predictable before they even started. Duel boot dead no plan for legacy application dead. Yes it all shows migration takes planning even on windows. There have be some massive failures of migrations done without a good enough plan. Windows and Linux are equal in this.

  36. Ted says:

    “Not only that, the Foreign Ministry switched back to XP!”

    With a plan to upgrade to Windows 7 and Office 2010. They only mentioned that in the same sentence.

    “That’s not what I would call a ringing endorsement of Microsoft’s latest OS.”

    That they switched back to Windows XP (two versions and over a decade old) from Linux is not what I would call a ringing endorsement of Linux. Quite the opposite, in fact.

    “How many organizations that switch to GNU/Linux switch back to Microsoft (any version of their OS)?”

    How many switch away from Microsoft in the first place? Very few and far between I’d think, considering the same usual suspects always get rolled out as the poster-children for Linux migration. The switches away from Windows are generally well-publicized because of their comparative rarity.

    And the fact that some of those that do migrate to Linux switch back to Windows later is again hardly a ringing endorsement of Linux on the desktop.

  37. oiaohm says:

    Viktor go read the IBM manual on Linux migrations.


    They also have manuals to perform windows to windows migrations. Ibm redbooks are bibles to system admins at times.

    One clear no no is duel boot machines. Why it results in arguments that rail road the process one way or the other.

    Duel boot machine method does not work in business environments. You have to choose one OS and run with it in that role.

    In fact when you read what the German Foreign Office did compared to best prac for a Windows to Linux or a Linux to Windows or windows to windows migration its no surprise at all that is failed.

    Virtual machine is valid but duel boot is not in all migrations cases(windows to Linux, Linux to windows and windows to windows). Its a universal constant of migration go against that the migration will be a cost over run or a failure.

    Now if German Foreign Office followed best practice for OS migration it would be a case. Yes you don’t duel boot XP with windows 7 if you are migrating a office to Windows 7 either. If you do users will glue to XP. Nothing strange happened there other than incompetence that is odd for Germans unless someone internationally wanted it to fail.

    Viktor please before putting your foot in it again get up to speed on what best prac for migration is.

    Wake up most of the reported failed Linux migrations if you deleted the Linux references stuck them in a pile with failed windows to windows migrations with windows deleted you would not be able to pick them out since the causes of failure are the same.

    Basic rules are broken leading to failure. Reason there are basic rules of business. 20) you cannot reboot threw day to change OS. So duel boot is always a non functional for migration. Its also confusing for users. Humans are spatial ware creatures. So a machine that is always windows or always linux at a location they get use to even change thinking when they go to that location. A machine changing from one to the other annoys the crap out of them so will get fort even if it was between 2 different versions of windows.

    Duel boot in business is how to annoy humans. Really I am surprised none of the BOFH stories worked this one out.

    Yes there are 19 other rules that mostly define how humans interact with tech in a business environment. Break any one of them migrations fail. Duel boot is 20 because most people are not dumb enough even to attempt that.

    Viktor how do you think you would cope if you desktop machine random-ally booted XP or 7 and you were stuck with that for 8 hours before you could change it. Not going to make you happy right. So you are going to linch the person who is always booting up with the different OS. Is this a Linux problem no its not. If a Linux or windows migration sets this up its doomed to failure. That is exactly what duel boot in the plan equals.

    Home user not sharing their machine can possibly get by duel boot. Any business IT office who thinks it is a good idea should be fired on the spot better rid of the problem.

  38. Viktor says:

    Yeah, ridicule yourself even more, Mr. Koz and Mr. Pogson.

    Linux did not fly in the German Foreign Office. Never. Users complained about Linux from the start. On desktop computers, where dual-boot was offered, Windows was booted by users, not Linux.

    Perhaps they should’ve asked Mr. Koz and Mr. Pogson to do the migration. Chuckle.

  39. kozmcrae says:

    Viktor wrote:

    “Hate to break it to you, Pogson, but the German Foreign Ministry switched back to Windows completely.”

    That’s news Viktor, big news. In fact, that is such a rare event that that could be the first and only time that has occurred this year. But wait, it didn’t happen this year did it Viktor? That news is over a year old. Not only that, the Foreign Ministry switched back to XP! That’s not what I would call a ringing endorsement of Microsoft’s latest OS.

    So is that the best you can do? How many organizations that switch to GNU/Linux switch back to Microsoft (any version of their OS)?

    Do us a favor and let us know the next time that happens will you Viktor. I know we can count on you.

  40. They telescoped. See cost…

    The German Foreign Ministry reverted to that other OS from a top-down burden. GNU/Linux still flew.“The German Foreign Office will migrate its desktop computers from Linux back to Windows. However, no truly compelling reasons for the decision appear to exist.”

  41. Viktor says:

    Hate to break it to you, Pogson, but the German Foreign Ministry switched back to Windows completely. It was widely publicized. The rest of this “think piece” is just as bad and inaccurate as always.

    And where are reasons 4, 5, 6, and 7?

Leave a Reply