The Precipitous Fall From Grace of That Other OS

A few numbers say it all. M$’s share of hits on Wikimedia:

Month Share (%)
2012-11 71.03
2012-12 68.21
2013-01 58.40
2013-02 57.06

I guess we know what many good girls and boys found under their Christmas-tree, eh? Tablets and smartphones not running that other OS. iThingies apparently more than doubled their use of Wikipedia and Android/Linux increased 50%. I suppose that’s biased to English-speaking regions mostly because Android/Linux is kicking iOS’ butt lately but I doubt the story about M$’s share was much different in other parts where I see folks walking around with a smartphone up to their ears. Acres of smartphones on retail shelves do make a difference in the marketplace:

How low will M$’s share be a year from now? On another note, “8” has just passed GNU/Linux share months after release. M$ used to do that with a new release in the first month. I’ll bet Ballmer is thinking his best days are long gone.

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.

61 Responses to The Precipitous Fall From Grace of That Other OS

  1. oiaohm says:

    bw basically the smart thing right now its sit back and lets see where the steam numbers level off. Or be looking for other groups collecting numbers the steam style with logins/credentials.

  2. oiaohm says:

    bw that is the problem. If you look closer they are not the minimum color pixel possible. Microsoft did manage to sneak an other bar in there with a gradient fill.

    But there is a problem we don’t know source so we cannot confirm method as valid.

    –I think you have to get real numbers from real places if you want to make such a case.–

    Yes this is what I want Real numbers. Real numbers taken by a Real valid method. Anything basing on IPv4 addresses we know is shot.

    Carriers sid is a very good one it shows that web stats are shot so badly we cannot trust them to tell us anything. Steam is another fairly safe reference. But we do need better and more solid.

    Google gmail status would be good. Problem is goggle stopped releasing stats long ago.

    bw the requirement is quite simple. The site collecting the site must have user logins. Then the values base on per user information. Where able users must be unique as able. Of course this is still more buggy than items like Carrier sid. But its vastly ahead of web stats.

    bw problem is too many take stats at face value without checking method. Steam method of collecting is valid the question there is if it user base is broad enough to be fully representative.

    Web stats bad collection method. No way to check if the sites are truly fully representative or stacked for the two most common quoted.

    I have said for a long time I want good numbers so we know what in heck is going on. Steam has Ubuntu already in its own right running 1.8 percent. So next question is how long will the .4-6 percent a month growth in steam stats last with Ubuntu. So there is a darn good chance that Ubuntu will over take OS X.

    Other distributions are only starting to have steam in there default packages.

    bw there is a high risk that everything you think you have known about the Linux numbers is bogus. All because you have been trusting sites with methods that when inspected and compared are broken.

  3. bw says:

    No numbers in any of it, though. It looks to me like the width of the color bands for Apple and Linux are just the minimum number of pixels that can handle the color and the gradient fill between them. I think you have to get real numbers from real places if you want to make such a case.

  4. bw wrote, “All there is a slide that is inconclusive about any numerical information and no representation that it is to scale or even where the lines might be drawn.”

    Ballmer made that presentation to analysts so it was market-manipulation if untrue. SEC cares about that.

  5. bw says:

    “To fudge the data would have gotten them in trouble with SEC if ever found out.”

    What data? Search high and low and there is nothing that Ballmer or anyone else from Microsoft ever said in terms of numbers or about Linux or Apple OS at all. All there is a slide that is inconclusive about any numerical information and no representation that it is to scale or even where the lines might be drawn.

    The SEC could care less about sales pitches, too, unless they are about selling stock to investors.

  6. oiaohm says:

    bw the graphic matches one true ground survey company style. This is why I wish he had. The fact he did not is a problem. Maybe he did not want to admit to the world web stats are complete crap.

  7. bw, endlessly apologizing for M$ wrote, “I think that the bands on the doughnut chart to show Apple and Linux were just big enough to show up rather than being an exactly scaled value. It was not a scientific chart, just an illustration of an impression. Ballmer wasn’t “lying” and he was talking about an entirely different topic.”

    Have you no shame? It was a 3D pie-chart no doubt produced from the best information M$ has available. To fudge the data would have gotten them in trouble with SEC if ever found out.

  8. oiaohm says:

    bw
    –There are lots of ways to measure things and the most reasonable is to measure the ratios of OS usage for popular activities such as searching Wikipedia or general access for internet sites. There are various counters of these activities and they all seem to show Linux as a very minor player.–

    This has not been reasonable for a long time. Because the means to identify users is gone.

    These use the IPv4 address as the Unique identifier. So can you explain why China is at least 50 users to 1 IP address.

    Also you don’t know with rolling IPv4 address that the two IPv4 address you have just seen at 2 different people or the same person. So home users in some countries are multi counted. Users in other countries are under counted. All because of using IPv4 base unique ID.

    Use of IPv6 is under a few percent. There is only 1 ISP in Australia that you can have a IPv6 address with. Everyone else mandates IPv4. This is not unique globally. Yes we run out of IPv4 address quite a few years back. Since then carriers globally have been stacking users into less and less IP addresses.

    bw steam numbers that are true per user numbers based on steam credentials already show Linux exceeding 2 percent and matching OS X.

    The reality bw is web numbers are method fail. Advertising is worse since many sites are reporting over 50 percent of users these days have add blocking installed. Since the Linux users are the more skilled users due to complexity they are also more likely to have advertising blocking.

    Also bw to work out how trustable a number is you need to check it method.

    Net Applications and other web stats you can check its method against telephone carrier numbers of connected devices.

    Problem here none of the web numbers even slightly line up to what carriers report. They know the mobile devices they have unique sids(System Identification Code) so even a phone connecting by a different sim card to a different carrier still counts as one. Carriers count by the sids connected in the last 30 days. Even changing the firmware does not change the sid(System Identification Code).

    You cannot see sid numbers Internet side. bw Carriers report in all numbers Android is ahead of IOS(apple devices). That is not devices sold but active in use devices.

    Checking against those numbers the only one that gets kinda ball park is StatCounter Global Stats. But this could be good luck.

    Its not only android and IOS numbers off the ball. With IOS over reported. Windows mobile is under reported as well. Yes Windows mobile by carriers is 3 percent. Yet web studies report around 1 percent like Linux.

    bw so Linux could truly be 3 percent due to the error factor web numbers show on Windows Mobile.

    This is the problem we know the web numbers are badly wrong. Out by factors of +-6% in places if not more.

    Symbian/Series_40 by carriers is basically non existent. StatCounter Global Stats and Net Market Share are magically able to find over 10 percent of them and wikipedia is over reporting there number by a factor of 6.

    This is the problem bw. Real world carrier survey numbers completely just rip web stats to shreds. There is just no relationship from web states back to the physical world.

    Particularly when you have a case of a basically non existent OS being reported in numbers. Something is badly wrong.

    bw I am using steam stats because I know I cannot trust web numbers and steam stats are way more stable. I do wish I had a more broader survey other than steam users. But at least steam stats are not reporting non existent.

  9. bw says:

    “So why would Ballmer lie and claim that MS seems Linux and Apple as tied on the desktop?”

    If you look all over the internet, all you find is a copy of a PowerPoint slide that Ballmer used in his talk. His point was that Microsoft had much more competition from piracy than any actual competitor. “We have met the enemy and it is us!” like Pogo once said.

    I think that the bands on the doughnut chart to show Apple and Linux were just big enough to show up rather than being an exactly scaled value. It was not a scientific chart, just an illustration of an impression. Ballmer wasn’t “lying” and he was talking about an entirely different topic.

    There are lots of ways to measure things and the most reasonable is to measure the ratios of OS usage for popular activities such as searching Wikipedia or general access for internet sites. There are various counters of these activities and they all seem to show Linux as a very minor player.

  10. oiaohm says:

    Hans Gruber Problem was he never told us we he got those stats from. Wish he had.

  11. Hans Gruber says:

    you can debate failed techniques to determine which OS is use all you want, although I still remember that Linux was claimed to be around 5% quite a few years back before the whole 1% number started coming out…

    but there is only one man I trust on this because he has no love for Linux and that man is…
    Steve Ballmer.

    Ballmer was talking to investors in February about 2-3years ago and there was a chart used that Im sure you can find online which shows OS uses (i cant remember the breakdown but I remember there was somethign about R&D, browser an OS use) and it had Linux and Apple tied on the desktop and maybe even a shade more for Linux.

    So why would Ballmer lie and claim that MS seems Linux and Apple as tied on the desktop?

    is there anyone else that well placed (and without a horse in the race) who has claimed anything else about what their multi-billion company perceives to be the OS landscape?

    Why should Ballmer not be trusted with these claims?

  12. oiaohm says:

    bw also in case of a closed source provide if its broken you have to put up with it until the closed source provider decides to fix it.

    You can treat FOSS the same way and use it and forbid your coders from working on it. So you don’t lose any of your developer time.(yes right). Writing hacks over bugs is some how magically productive.

    The reality is a FOSS supply is superior to a closed source supply. You can treat a FOSS supply the same as a closed from your coders point of view if you are so worried about losing code your company produced. Of course you still have to provide the non altered form.

  13. oiaohm says:

    bw
    –I am happy to see that, within all the demented ravings you produce, you do agree with my point, namely that it is a bad idea to use FOSS code in your product since you have to provide the source code that you created for your product.–

    No I don’t agree with your idea at all.

    Its a bad idea not to pay for the software you use from third parties. Companies get into trouble for providing Windows machines into the market using copyright infringing copies of windows. Because they did not keep track of what they were meant to be paying for.

    bw using source code that is not your own you have to pay for. If it closed source it comes in form of binary that you have to pay that vendor for how many copies you used in most case. Keeping exact counts and exact tracking.

    FOSS is simpler. All you have todo have is provide the source code. If happen to produce more or less no fowl.

    Using closed source supply is in fact way more risky than using FOSS. bw.

    You might preorder millions you don’t sell with closed source. Or if you happen to produce a few too many units and lose count that you never acquired enough licenses. Closed Source comes back and bites.

    The payment requirement of releasing source code to open source parts you have used is quite minor and very low risk of having issues.

    License fee side of closed source is not without its troubles bw. If you don’t want to pay bills code everything in house. That normally end you up with a poor grade product and not a very compatible product.

    Depending on what you are doing lot of cases FOSS is better choice. Particularly if you don’t know how many units you will be producing FOSS as a hardware maker.
    1) No risk of over ordering resulting being stuck with a stack of closed source product you cannot move because the closed source making release a new version.
    2) No risk of screwing up and being under ordered on software and being sued by the closed source maker for providing copyright infringing software this can result in product destruction.

    But in there is a price to pay for these two advantage with restrictive Open Source licensed projects. Releasing the source code of what you alter that owns to the project.

    The idea that is bad to use FOSS due to having to release source code is wrong.

    Its bad to use third party parts if you are not willing to accept the terms and work to what the terms define.

    Closed source massive tracking auditing and counting then pray you don’t suffer from human error anywhere.

    Open Source terms its simpler to be sure that you are in 100 percent compliance with the licensing terms.

    Like it does not matter with FOSS if a factory somewhere did not run of like a 1 million for testing that allowed to be sold of your product without checking the number of licenses company holds. This is a big problem with closed source provide and patents. Results if this you find yourself without enough licenses and in legal hot water.

    bw so there is good and bad to closed source provide and open source provide. On balance the least risky path is the FOSS path. Because you know when you paid your bills and it cannot be a human error like someone typing 1 000 000 in instead of 100 000 into an automated production run. or 100 000 instead of 10 000.

    bw you want to make out releasing the source code is bad. What is worse. Hardware running copyright infringing firmware can be taken at customs in most countries and crushed at your expense.

    All the lose of possible IP to your competitors is nothing to producing a product and losing the lot to a crusher.

    In every FOSS case against a company they have never asked for the offending products to be destroyed. All they have asked for is that the company that shipped it come into compliance.

    Microsoft with copyright infringing copies of windows in retail has asked for the machines to be crushed and destroyed.

    FOSS will not be out to destroy your inventory. They will want a sorry and maybe some legal expenses and that you do put policies in place that you don’t infringe closed source or open source licenses in future but this is quite minor and the last bit you should have been doing anyhow.

    bw
    –If you use a FOSS alternative, you are faced with having to educate those same buyers as to how they are not giving up anything in exchange for any possible price difference.–

    There is a possibility the existing market is the same size as OS X for particular Linux distributions. Do you have to educate those same buyers about OS X. Answer is no there is an existing customer base that wants it.

    bw you are presuming there is not already an educated market of Linux users.

    There is also a problem not providing to customer expectations. From the numbers showing up with FOSS solutions this could be the case. The presume of customer expectations could be completely wrong.

  14. bw says:

    “This is why I say don’t mess with stuff you did not make. Stuff you did not make you have to pay for in some form. Closed Source pay a license fee. Open Source provide the source code. Both are payment.”

    I am happy to see that, within all the demented ravings you produce, you do agree with my point, namely that it is a bad idea to use FOSS code in your product since you have to provide the source code that you created for your product.

    When you sell a computer with Windows pre-installed, you get the benefit of immediate recognition by hundreds of millions of potential buyers. If you use a FOSS alternative, you are faced with having to educate those same buyers as to how they are not giving up anything in exchange for any possible price difference.

    When you consider the difficulty in having to supply such a detailed technical message in the kind of distribution channel available for such a business, you can see where it is unwise to deviate from customer expectations.

  15. oiaohm says:

    bw you are doing a troll arguement that has no base in law or reality basically.

  16. oiaohm says:

    –As I said, the licensing for Linux doesn’t allow a manufacturer to keep things to itself. It has to give its secrets away even if there are none.–

    Wrong. I can explain the busybox law suit.

    1) makers did not include the fact busy-box and other gpl parts were inside.
    2) there was no offer to end users to provide the source code of the GPL parts.

    The manfactor is taken advantage of GPL parts not to write those parts themselves. There is a price to pay.

    Like the manufactures custom configuration Web UI code. That was not forced to be released by the Busybox court case.

    bw the router cases they have messed with source code they have not made. They have not obeyed the license.

    –All they were doing was selling Wi-Fi boxes with Linux inside. I don’t think they even had anything proprietary at all, yet they get sued.–

    There was and is proprietary inside Wi-Fi and other routers Linux guys never asked for that.

    GPL and LGPL licenses only require you to provide the source code that is directly effected by those licenses.

    Web UI on a router that a company has written themselves the Linux guys have never asked for.

    This is why I say don’t mess with stuff you did not make. Stuff you did not make you have to pay for in some form. Closed Source pay a license fee. Open Source provide the source code. Both are payment.

    What would you say it if was a case of vxworks inside that the router maker had failed to pay for and intel sues them over it?? Would intel be the bad guy??

    bw the Busybox case is simply using software without payment. Just in Busybox case payment is source code of the used Busybox to your end users not cash.

    Basically if you disagree with the Busybox case you basically are saying a company should not have to pay for software they did not make.

    Busybox case was an example of what Microsoft people like to call software piracy. Or correctly called copyright infringement.

    bw playing by the licensing rules you are allowed to keep secrets. Like a device might contain a Linux kernel and everything above that be made by a company or be BSD. Result the only source code that has to be release to pay licensing bill would the the Linux kernel since its GPL.

    Now if you use Linux Kernel and Busybox. Correct payment is release source code to those and everything else you have created is yours.

    bw its all fairly straight forwards. Most companies found to have issues not meeting GPL payment requirements have also been found to be using closed source software without payment as well. Its all about having the correct processes in place to document your license payment bills. Yes source code release is a bill to pay.

    Really were you stupid enough to thing the only source code a wifi router contains is Linux kernel and busybox. Neither provide a web interface. Neither configure firewall and bridging completely out box. Yes there is some proprietary code in there that the Linux and busybox people are not asking for.

    bw also read the complaint busybox never asked to be given the source code directly. As per GPL source code had to be provided to who ever you gave device to. Then if it made it back to busybox from there so be it.

    Remember device release cycles are basically once every 12 months. So you do some in house modification to busybox you ship device with written offer to end users. It could take 12 months before your competitors are using it. Yes using GPL there are advantages to modification.

    In fact it becomes really legally messy if you don’t release the source code. Since the payment method for FOSS is source code release it valid in law for payment of damages to be in source code. That is how you can lose your proprietary code parts. Judge decides all source code in the device be release to cover damages basically.

    bw I will not say that FOSS does not have teeth in court it has huge ones. Now if you use it legally you are allowed proprietary code holding secrets from your competitors. There is a reason why in embedded Linux user-space drivers are popular. Linux user-space drivers are not limited by GPL. Yes release Linux kernel of the device not release the user-space driver anyone trying to fire up on that device is bricked. Fully legal.

    Your Source code is Your Source code. Your Source in a restrictive license FOSS project is payment for right to ship.

    Release restrictive license FOSS project source code to end users is also payment for right to ship.

    Your Source code not linked to restrictive license FOSS you are free to keep secret. This is where you code your companies special sauce.

    bw there are a lot of places where you can code your companies special sauce and not fall fowl of the FOSS license requirements.

    FOSS yes gives companies a lot of freedom. But it comes with responsibilities. This is why FOSS is zero cost is wrong. There is a cost in maintaining your responsibilities.

    bw the busybox case does not prove router makers cannot keep secret parts while using FOSS. Just proves the FOSS parts under restrictive license cannot be kept secret.

  17. bw says:

    As I said, the licensing for Linux doesn’t allow a manufacturer to keep things to itself. It has to give its secrets away even if there are none.

  18. They were distributing the software without a licence because they did not provide the source-code. That’s not good.

  19. bw says:

    ” If you do come up with something different with Linux that is more appealing you can keep it to yourself as long as the item that make it more appealing is not mess with libraries and applications source code you did not make.”

    That would be great, but how do you explain things like this:

    http://www.softwarefreedom.org/news/2009/dec/14/busybox-gpl-lawsuit/

    All they were doing was selling Wi-Fi boxes with Linux inside. I don’t think they even had anything proprietary at all, yet they get sued. Not good.

  20. oiaohm says:

    bw
    –If you come up with some wrinkle that makes your product more appealing than someone else’s, you have to give it to you competitor who can easily copy your work. That takes all the motivation for in-house development out of the picture.–
    BOGUS. You have not read Linux Licensing. You are listening to stupid moron MS Trolls again.

    Higher grade hardware is one way to make your product better than competitor that your competitor cannot copy without spending the same amount.

    Also you need to read GPL. GPL does not require giving source code to everyone. Only to your customers at first. LGPL stuff as long as you only link to it your source code is still yours.

    bw
    –Manufacturers are apparently free to design unique presentation layers for their devices and keep them proprietary.–
    This was true on Linux netbooks by the way bw. Pre android in fact.

    True most don’t because there is a price of maintainer-ship.

    Also look at Manufacturers making Windows devices they don’t alter the OS UI at all. They install extra programs that are closed to attempt to split themselves. Dell does ship some open source programs that only work with there hardware in there Ubuntu images. Just because you have the source code does not mean it will work on the hardware you have bw.

    Hardware and Software is linked to a point.

    bw next you get Linux licensing wrong.

    The two biggest Linux X11 GUI KDE and Gnome. These are both LGPL licensed. So any extension not done to a core library or application is yours. In fact both support plugins to the interface that can be closed source.

    bw Linux does in fact allow companies to differ if they choose to. There does come a question if you should.

    Kernel mode drivers most likely being not shared is a bad thing. Your own developers are only so skilled and a kernel failure will cause users a lot of pain. Some tweak to the user interface that the user can disable if it not to there taste is still a feature difference.

    Linux allows vertically integrated if makers choose.

    Reality you said a huge stack of bogus MS troll stuff about Linux. The key feature of Linux is that it allows your product more appealing to others by extending it. Of course you are free also to choose not to extend it.

    Compare to Windows where you cannot see source code inside libraries provided. If one of those libraries in windows is causing you major issues you cannot fix it so now you have to place to hack in your application and hope MS does not fix that fault causing the hack to break.

    Yes the API/ABI of Linux is better behaved when it comes to defects. Find a defect have it fixed or at least find out if it is a defect not just you using the API/ABI the wrong way.

    –Linux is essentially the same way in that, if you manage to come up with something more appealing, you must give it over to everyone else so there is little reason to even try to make things different from one another.–

    Completely BOGUS basically. If you do come up with something different with Linux that is more appealing you can keep it to yourself as long as the item that make it more appealing is not mess with libraries and applications source code you did not make. Reality under Windows you cannot mess with source code you did not make because Microsoft does not show it to you.

    What went wrong with Linux Netbooks is that manufacture and lot of Linux Distribution are nuts and distinguished themselves a step too far. Don’t distinguish yourself that far that you break application compatibility with each other.

    Android manufactures have been a little smarter this time around. Of course manufactures could have been a little smarter and choosen debian/ubuntu/… as common base as well.

    Basically bw Linux gives manufactures that much freedom they can blow there own feet off. Since this is the fact it taking manufactures a while to get use to it.

    bw this is the problem most of the arguments against Linux are a house of cards. Fall over with a little investigation.

    Its not like Linux world is going to force valve to release the source code to all there applications and steam provide system itself. Ok they would like to see the source code. Like and demand/force are two different things.

    People take about Linux Desktop being fragmented. There is a reason. Linux users like there machine to feel unique to them.

    Valve with steam solve a problem Linux has had. How to distribute an application to many different distributions without have to remake it every single time.

    Of course every major software house could do this on Linux.

  21. bw wrote, “they must distinguish themselves from their competitors in some way. For a few, that is having the lowest price around and you seem to favor that approach yourself, based on what you post about.”

    Nope. Location, location, location… If your customers can find you and you are competitive your business will thrive. There was a time on the web when it was easy to find anything for offer. Now there are millions of hits in the search-engines. You need bricks-and-mortar for mobile consumers, a great website that’s easy to use (think different from Dell’s) and a great product. You could be identical to 47 other businesses in all respects and still survive if your customers can find you. Any single factor, like being a “partner” of M$, is no guarantee of survival. Clearly, many consumers have no loyalty to M$. I hear M$ is dropping Phoney “8”…

  22. bw says:

    “Nonsense. A business sells products and services.”

    You are not a businessman. A business sells these things, true, but they must distinguish themselves from their competitors in some way. For a few, that is having the lowest price around and you seem to favor that approach yourself, based on what you post about.

    In the case of personal computers, it is a fairly well established business in which the first question asked is whether to get a PC or a Mac. For a long period of time that was not the question, but Apple has done a masterful job in making people aware of the distinction.

    If one chooses “Mac” then there is little more to do other than pick a size and option set. If one chooses “PC”, there is a little choice on the level of the OS, that is Home, Pro, or Ultimate, but that tends to suit a price strata for the base model chosen.

    The choice for a PC, other than laptop or desktop, is mostly a perceived quality level for the hardware involved. Some units have better finish or appear more robust or have enhanced capacity, but they all work almost exactly the same way. Microsoft will not allow them to differ. Linux is essentially the same way in that, if you manage to come up with something more appealing, you must give it over to everyone else so there is little reason to even try to make things different from one another.

    Parenthetically I might note that this does not seem to be true with Android. Manufacturers are apparently free to design unique presentation layers for their devices and keep them proprietary. So you can’t buy a cheap $79 Android tablet and load it with the same software as a Kindle Fire and have it look and feel the same. That part is secret to Amazon. Or Barnes and Nobel or Galaxy or whomever is making the device. Of course the low price producers use the default sort of interfaces and do all look alike. They have to distinguish themselves on price vs how cheesy they look.

    I think that real progress is only made when vendors can separate themselves in some way from the crowd and are rewarded for doing so. That requires proprietary development and protection for ownership of the results.

  23. bw wrote, of GNU/Linux, “the licensing, which seems to require that anything done differently has to be disclosed to everyone else, makes that a questionable undertaking. If you come up with some wrinkle that makes your product more appealing than someone else’s, you have to give it to you competitor who can easily copy your work. That takes all the motivation for in-house development out of the picture.”

    Nonsense. A business sells products and services. They can still sell a service because being able to read, run, modify and distribute the software takes nothing away from that. It still takes some knowledge to deploy GNU/Linux whether as an individual or an organization. There is a huge market for anyone supplying services related to GNU/Linux. They can still sell products containing GNU/Linux because it’s less work for a consumer or business to buy products with GNU/Linux than to install them. I know a business who sells PCs either ready to run, or as components. They charge $50 to assemble a PC from components you can buy from them. They can do that business regardless of the distro, the configuration or the customization of the unit. There may be some businesses who don’t want to show the code but that’s a tiny part of the whole ecosystem of goods and services that can be sold with GNU/Linux. You can see that with Android/Linux. Hundreds of millions of units are sold annually and it’s a growing business despite the fact that the Linux kernel including drivers are made public. In fact, the OEMs making Android/Linux devices share the cost of all those devices/drivers in Linux and it reduces the per-unit cost of production, a very good thing for their business.

    So, the idea that a business will be reluctant to show the code for units that ship in the millions and be replaced by a new model next year is silly. There might have been some hesitation early on but there’s certainly none now. That’s why we see a proliferation of Android/Linux and GNU/Linux devices on the market. Lots of startup love being able to ship a decent OS with little or no development costs. That profitability dwarfs and issue of showing the code and they can easily afford to contribute and share */Linux.

  24. bw says:

    “He should be required to show hard evidence that GNU/Linux is not much larger than ~1%”

    A more significant question is “percent of what?” If you are in the business of selling the OS software, you care about how big the business can be and how much of it you have in your pocket.

    I personally have nothing to do with selling or buying computer OS software, but I have been associated with other commercial products that are sold in similar circumstances. What matters here is the notion of who customers might be and whether or not they are actually customers for my product or not.

    Many companies are vertically integrated and produce their own components. An example of this is Apple who produce their own OS and use it to create differences between their computers and those from other manufacturers who buy OS software.

    This “in-house/out-sourced” distinction is very important when evaluating a business to be in. Linux, I think, basically gives a manufacturer of computers the ability to be an in-house kind of supplier, but the licensing, which seems to require that anything done differently has to be disclosed to everyone else, makes that a questionable undertaking. If you come up with some wrinkle that makes your product more appealing than someone else’s, you have to give it to you competitor who can easily copy your work. That takes all the motivation for in-house development out of the picture.

    So if everything is going to be the same-old, same-old form, you might as well buy it from Microsoft and take advantage of their ability to force consistency between suppliers.

  25. oiaohm says:

    Its just scary what we don’t know. If Mint is as large as ubuntu and ubuntu is bigger than OS X. The result could be that OS X is position 4 on market share.

    This is the really big problem. No one has really known how big Linux is really in particular markets. Yes a lot of game developers have said Linux people can just use Wine. Result is Linux numbers uncounted.

    Its a high risk that companies have been justifying making OS X versions of games when there is no valid reason to be. If it turns out Linux is bigger in single distributions there will be a lot of questions why MS Office is not ported to Linux.

    Microsoft will be hoping single Linux Distributions don’t have numbers higher than OS X.

    Thinking that Linux will have achieved this without major retail shelf space.

    That real method to count and compare Linux users has appeared.

    Also remember these Linux user counts are before Valve releases the steam box. Valve has not hardware pushed Linux yet. Right now even a 1 percent success of Valve idea sees OS X behind Linux.

    Linux moving to second position now moves making applications for Linux up the list.

    Really its all nice timelines lining up.

    cgroups/user namespaces. 3.8 kernel. This is a new level of virtualization. This should appear in the next releases of distributions.

    Wayland due to Ubuntu over Mir has been accelerated by Gnome and KDE.

    Better distribution independent bundling. http://people.gnome.org/~alexl/glick2/

    Most of the major problems all are ending around the same time. And if Linux already has more than OS X in market share there is only one competitor left to go after. Microsoft.

    Now if Linux can prove market share bigger than OS X it has good grounds for retail shelf space.

  26. oiaohm wrote, “This is the problem is how far past 3.24. Heck it could turn out that Ubuntu has as much market share as apple. We just don’t know yet.”

    Yep. GNU/Linux is growing rapidly from a tiny installed-base on Steam. I expect those high rates of growth will continue for a few months at least giving GNU/Linux close to 10% of Steam. Further, it will be interesting to see whether GNU/Linux takes share from M$ or Apple or both.

  27. ram says:

    I know for sure my companies’ servers are hit by far more Linux users than any other operating system. BSD and its derivatives are in second place. Mainframe Unixes (e.g. AIX) in third place. Only then does anything from Microsoft show up.

    The industries we sell to are just that, industry, not consumers. Still, Microsoft was never as big as they claimed to be. The whole ‘Monopoly’ thing was a consumer product sales tactic. Now even English speaking consumers are moving to Linux in a big way.

  28. oiaohm says:

    Der Balrog the Steam’s survey has not stabilised yet. I am also very suspect that the dropping Windows XP value is not windows users. Wine projects default reported OS is still Windows XP.

    There is something surprising. from steam.
    MacOS 10.8.2 64 bit 1.71%
    MacOS 10.7.5 64 bit 0.74%
    MacOS 10.6.8 64 bit 0.62%
    MacOS 10.7.4 64 bit 0.08%
    MacOS 10.5.8 64 bit 0.04%
    MacOS 10.8.0 64 bit 0.02%
    MacOS 10.7.3 64 bit 0.02%
    MacOS 10.7.2 64 bit 0.02%
    Apple is only 3.24% It has more games released for it than Linux currently. So with Linux at 2.48%

    We are at a very big possibility that the number 2 PC OS in gaming is Linux. And possibly has been the complete time just hidden by the fact they are using Wine cursing and swearing at it and showing up as XP in the numbers.

    Der Balrog steam survey is 100 percent sure shows the market of PC hardware gamers.

    Out of all PC buyers. Who spends the most on individual machines. 3d cad and 3d animation comes in at number 1. Number 2 is PC gamers.

    Most businesses buy cheep and crappy.

    If the possibility turns out to be true Retail has to ask themselves some serous questions. Most retail is giving shelf space to OS X. But if its number 3 in gaming. The question is what else is it number 3 in.

    Der Balrog we don’t have a good business or home survey. We can write the current web servey numbers off. Carrier grade nats fairly much ensure they are going to come more and more crap.

    Der Balrog to a place like Wikimedia due to China super big carrier grade nat. The complete population of china appears coming from a limited number of IP blocks. So the max you could ever see from china is about 10 million. The 2012 9.25 million spike at wikimedia is it about seeing every IP address that china has got for general popluation. China carrier grade nat is based on cisco system tech. Also in 2012 cisco upgrade there carrier grade nat tech to fit more users per IP address. Magically chinas connect to Wikimedia start dropping off.

    We know how many real Internet users are in china. China keeps count.

    http://en.wikipedia.org/wiki/Internet_in_China
    538 million Internet users repesented by 10 million unique IP identifiers. At least 50 to 1. Suspect this is quickly heading to 100 to 1 and higher.

    China is not the only place with huge career grade nats. Telstra in Australia is scarily at 2 million to 1. Yes 1 IP address representing 2 million users. Makes the china warping look minor. Use the same tech testra is using china to reduce the ip address from china to 269 IP address. At this point you fairly much cannot identify who is behind there.

    China Internet population has grown but wikimedia and others are seeing less of them.

    Main reason why china carrier grade nat is going to pack users tighter. So free up IPv4 address for Websites. So one day 269 IP address representing all of china could happen.

    Der Balrog something like steam or some site needing user logins can cut threw a carrier grade nat.

    I would love to see facebook using its user logins do a OS survey. It would be one of the best surveys we would get. This will not happen while people don’t make it profitable. Ie supporting the bogus surveys don’t make it profitable.

    Now if I cannot get facebook or some other site requiring userlogins to provide good numbers I will trust stream. Its simply the best I know of not to be completely distorted to hell.

    Der Balrog really I would like to have more survey that does not fall apart under collection method investigation. Steam does not. Shows possible bias to gamers yes. Does not show collection method is going to miss count users. If you miss count users you have balled it up.

    Its a bit like getting the carrier numbers on mobile phones and comparing to web stats the two don’t line up at all. It comes very clear web stats are crap.

    Yet 1 percent arguement has been based on webstats.

    Steam stats are validish. Steam has not been on Linux long enough or with enough games yet to reach stable state of population count. Ie Linux growth above what ever number it is now is expected. The only argument is how far. If it over 1 percent big questions about OS X as the number 2 OS.

    In fact Steam will give us better numbers than http://distrowatch.com/ It will also let us see how correct distrowatch is as it become stable. By distrowatch mint should be number 1. It also has not had steam as long as Ubuntu. Same with most of the others in the top 10 of distrowatch. It will be very interesting to see if the top 10 on distrowatch end up the same on steam. 6 percent is ball park possible. Higher is possible.

    This is the problem is how far past 3.24. Heck it could turn out that Ubuntu has as much market share as apple. We just don’t know yet.

    Der Balrog steam is more likely to be a correct representation once it stabilises than webstats. Including Wikimedia stats.

  29. oiaohm says:

    bw Canonical does not really compare to redhat.

    Redhat has Scientific Linux that they share development with. Some packages shipped in redhat enterprise are not built by Redhat but built by Scientific due to the build servers and test results being open to both sides to inspect.

    So yes Redhat is making it profit without having todo all the work. Scientific Linux does not provide phone support at all.

    bw Canonical is known for being the least cooperative firm going in the open source world. SUSE also does fairly well from the fact they do share resource with OpenSUSE. Yes OpenSUSE and Scientific Linux are free to make selections for there users independent of upstream.

  30. Der Balrog wrote, “Unless you have hard evidence (which you never have) “

    I can point to long lists of individuals and organizations who use GNU/Linux. That would not be the case for a ~1% solution. What organization would entrust it’s IT to software that was not mainstream? Organizations of all kinds use GNU/Linux from families to governments and global corporations.

    Further, Der Balrog does not get to claim the answer by default. He should be required to show hard evidence that GNU/Linux is not much larger than ~1%. He cannot. Even M$ years ago recognized the significant share that GNU/Linux holds and GNU/Linux has grown a lot since then.

  31. Der Balrog says:

    Can we claim that Linux has indeed some 6% market share of pc’s?

    No, we can not. Unless you have hard evidence (which you never have) that Steam’s survey is representative (most surveys are not) for all PCs out there.

    But do go on. Look for statistics that confirm your bias. Hint: you’ll find some in the sidebar of this site.

  32. Mats Hagglund says:

    Just take a look at latest Steam statistics:

    2,48% of those players are using Linux. And as we all know they are still mostly Ubuntu-users. Can we claim that Linux has indeed some 6% market share of pc’s? Probably yes – there are many studies showing that some 30-35% Linux PC’s are indeed Ubuntu computers. Near in future there will be rapid increase of Mint and Mageia-users playing in Steam.

  33. Der Balrog says:

    Hmmm… The world does not owe those heads a living.

    Typical Pogson drivel. The famous null argument. The world does not “owe” anyone a living. Not Red Hat, not the Linux Foundation, not the FSF, not IBM, not Google, not Canonical, and so on.

    What’s your point?

  34. bw says:

    “The world does not owe those heads a living”

    Certainly not. Microsoft has to earn their money each day the same as any other hard working company. They have been successful for decades now by keeping OEMs and other distributors convinced that they can mutually benefit from their strong partnerships with Microsoft. Many billions have been made by companies such as Dell or Compaq/HP who have hitched their own wagons to this rising star.

    Perhaps this mother lode is playing out or maybe there is still a lot of ore to be refined in the next decade or two. I don’t really know and I do not think you know either, but we are both entitled to our favorite guesses.

    OS, in the form of Android, is just a sideline for Google and it is not particularly useful to compare them to Microsoft if you want to learn anything useful from the comparison. I wouldn’t bother with it. I’m suspicious of Google in terms of investment opportunity because I have little faith in their business model that sells “click-throughs” to small merchants and so generates most of their revenue and all of their profits. I am surprised that such a grizzly as yourself would have such faith in such fluff.

  35. bw says:

    “Reality the numbers don’t show MS as a company doing ideal”

    Google for “stages of denial” and try to figure out where you are at. You statement lacks a definition for “ideal”. If you are just interested in relative, though, Microsoft is maybe a dozen times better than Canonical in terms of sales per employee and likely to be hundreds of times better in terms of profitability, if and when Canonical ever turns a profit, though. At $60K per head revenues, that doesn’t leave much for ordinary costs, like salaries, before they go into the loss column.

    It is easier to contrast them to Red Hat with about 10 times the profit per employee. Your mileage may vary, eh?

  36. bw wrote, “$720K or more per head”.

    Hmmm… The world does not owe those heads a living. That disgusting number is reason enough to spend elsewhere. Google, which has servers doing most of the work, gets $29 billion revenue with 54K employees, about $600K per employee. I would rather pay for servers at work than M$’s bloat.

  37. oiaohm says:

    bw
    –But Microsoft is selling twice as much product today as they were back then and they are still growing year by year.–
    Sorry no Windows 7 sales only came up to be in-line with XP sales rates. 4 years of fairly much flat.

    So stalled from XP to now. Also you have to take into account lost product lines as well.

    –As of November 30, 2012 Red Hat had approximately 5,300 employees worldwide.–
    This is in the redhat investor faq.
    Redhat is on about 200K per staff. Also Redhat does not spend billions on advertising or have any debt to support.

    You can fairly much say the 1.5 billion spent advertising Windows RT has fairly much gone down the drain.

    http://www.wikinvest.com/stock/Microsoft_%28MSFT%29/Data/Net_Income/2012
    http://www.wikinvest.com/stock/Microsoft_%28MSFT%29/Data/Revenue/2012
    Reality not moving twice as much product as 2008.

    Cost of production is way up there. Notice 13 billion more reported revenue but Net_Income is lower than 2008 in 2012.

    Reality the numbers don’t show MS as a company doing ideal.

  38. bw says:

    “vista was a flop”

    You all say that a lot and it has become some kind of legend. After all, wasn’t it replaced in a couple of years by Windows 7? But Microsoft is selling twice as much product today as they were back then and they are still growing year by year.

    Vista was just a name, like Ubuntu calls their new versions names like Daffy Duck and Fabulous Frog and Ice Cream Sandwich. Microsoft is still in the business and making scads of money. A lot more than Ubuntu, I bet.

    It’s hard to get trustworthy numbers, as you say, but Canonical was boasting of $30M in sales with 500+ employees, which is about $60K per head. Microsoft has $72B in sales (documented) with slightly less than 100,000 employees worldwide. That’s $720K or more per head, and that is much more better than Ubuntu. They had profits of almost a quarter million per head as well. You can run the numbers for Red Hat and SUSE, but I bet they come out on the short side compared to Microsoft, too.

  39. oiaohm says:

    bw that growth rate is slower than Vista in the same timeframe after release.

    –Windows 8 has increased some 200% in the 4 months of data you have linked.–

    It should be looking to be 400% to be inline with Vista. And vista was a flop. Now either the numbers are wrong. Yes I know the numbers are most likely wrong or Microsoft is in trouble.

    Remember Microsoft has a huge company it need large income to support.

    bw 500 times debian is that a problem. Is debian depending on the income. Answer no its not.

    bw we need better numbers. Steam numbers in a lot of ways are better than current web studies due to the fact not being effected by carrier grade nats.

    Facebook if they wanted to do a really good OS survey.

    Notice the common thing here to do a study now you need a login so you can tell who is who. Carrier grade nats remove IP addresses from being unique identifiers any more.

  40. bw says:

    “There are certainly more people familiar with GNU/Linux than with “8″”

    Don’t take this as any endorsement of how you extrapolate numbers to define trends, but it looks to me like usage of Windows 8 has increased some 200% in the 4 months of data you have linked. In absolute terms of page accesses, it is already about 500 times as popular as your “recommended” Debian distribution and more than twice the aggregate usage of all forms of Linux on PCs.

    That doesn’t look like bad news for Microsoft at all.

  41. oiaohm says:

    Der Balrog
    –But all the distributions out there utterly lack coherence.–
    Strength and Weakness.

    As Ubuntu found out recently there is coherence. They want todo mir. They got a universal smackdown by Gnome and KDE. Because both projects had already decided on Wayland.

    “greatest common denominators” is why I hated the old head go gnome. He did not take part in the standard process.

    http://www.freedesktop.org/wiki/Specifications

    Der Balrog I remember when every X11 Windows manager had its own way of storing list of application. Also remember when gnome applications started with a g and kde applications started with a k and they would not run in each others environment.

    Utterly lack coherence in today distributions is not true. Yes I will give coherence can be improved.

  42. dougman says:

    2% market share?

    What you been smokin??

  43. Der Balrog says:

    Think how far we have come in the past 10-years, then look ahead 10-years.

    You have come to at the most 2% market share. Yes, clearly something to brag about. Perhaps in ten years you’ll be at 4%.

  44. dougman says:

    Everyone I ever demo’d Linux to had no qualms in using it.

    Think how far we have come in the past 10-years, then look ahead 10-years. Where is M$??

    Still struggling to ship something that no one wants and pissing off users with their petty rules, regulations and reboots.

  45. Der Balrog wrote, “Wrong question! “[L]inux people” are not comparable with the average consumer.”

    Nonsense. Folks like me who have never seen GNU/Linux before can use it successfully. I have introduced GNU/Linux to thousands and only a tiny handful, who already have problems with that other OS, had any difficulty with it. For most people the login screen is the main stumbling block. They are not used to seeing no feedback on the typing. There are certainly more people familiar with GNU/Linux than with “8”. They are far more than just geeks or FLOSS-lovers.

  46. Der Balrog says:

    Do linux people like unified look and feel. Answer is fairly much no.

    Wrong question! “[L]inux people” are not comparable with the average consumer. “[L]inux people” have specifically sought out Linux to get that kind of “choice” (which really is a pseudo-choice and ultimately bogs you down). The average consumer will not.

    And in La-La-Linux Land not even the greatest common denominators (compositors, desktop environments, distributions, window managers, you name it) can agree on pretty much anything.

    Linux has the dubious distinction that it “kind of works”. But all the distributions out there utterly lack coherence. The end result is then a “customized” OS like that of our distinguished host Mr. Pogson — it looks like crap. Yeah, it may be good enough for him. But it’s not good enough to persuade anyone to put this abomination on his or her computer.

  47. oiaohm says:

    bw just look at how many Linux distrobutions there are. Do linux people like unified look and feel. Answer is fairly much no.

  48. oiaohm says:

    bw kernel stuff decided if OEM will even consider your OS in the phone Market.

    Victory for Linux is the fact we have more hardware developers working on the Linux kernel.

    In fact it was the OEM who demand Google mainlines the kernel. For better quality control.

    So without kernel.org and Google agreeing to mainline the biggest vendor selling android that is samsung would not be. bw the reality is without kernel.org android would not be in as many places as it is today.

    bw if the OEM don’t touch the product end consumers will never see it.

    Other vicory is now those vendors are talking about supporting Ubuntu, Firefox, Tizen and other Linux based OS’s into the mobile market.

    Result is going to be a sea of different Looking devices for users to choose between.

  49. bw says:

    “Android has been a victory for FOSS/Linux”

    I would read it that way. Instead, I would say that Google grabbed a bunch of Linux code that did things that Google needed done but did not see as having any effect on their ability to sell Android to the OEMs. Whatever the kernel stuff does, it does it unnoticed and out of sight of the people who may or may not decide to buy a tablet or phone with Android.

    It is a “Who cares?” kind of thing, just like whatever is in Windows 8 to do the same sort of functionality is a don’t care for Windows. People only see the face of these products. Apple did the same thing, but they chose a version of the freeware that didn’t require them to disclose all their other secrets. I would not be surprised to find that Microsoft did a lot of the same things. They have been accused of stealing the “network stack” or something like that from Unix.

    How many ways are there to do these things anyway? What difference does it make which way you pick?

  50. oiaohm says:

    Der Balrog Android has been a victory for FOSS/Linux.

    Just the problem is you don’t want to see it.

    Now that the hardware makers are accepting the Linux kernel we are going to see more FOSS operating systems on phones.

    Ubuntu wants to make themselves a foot hold there. What is Ubuntu goal merge Phone and Desktop.

    Welcome to the assault. If one pulls it off others will follow.

  51. oiaohm says:

    Der Balrog —

    “This is not a characteristic of Linux. There have been retail versions in the past.”

    And yet, even back in the day you already could download Linux freely. Linux has never needed retail.–
    What was the retail versions. Support subscriptions.

    Yes Linux does have retail forms. Just not the software.

    There is no particular reason why Support subscriptions could not be sold retail. Retail does not particularly like the fact those don’t have that much on going.

  52. Der Balrog says:

    This is not a characteristic of Linux. There have been retail versions in the past.

    And yet, even back in the day you already could download Linux freely. Linux has never needed retail.

  53. Der Balrog says:

    Thank you for summarizing your attitude to any kind of numbers so succinctly.

    I wanted to write the same thing. But I thought it’d be like casting pearls before swines. Logic was never the strong point of FLOSS fundamentalists. But hypocrisy is. Hence Pogson stumbling along, desperately trying to redefine corporate giant Google’s success with Android as the “victory” of FLOSS/Linux/whatever.

  54. ch says:

    “We can’t trust much these figures …”

    “However there is no doubt that these statistics are showing …”

    Thank you for summarizing your attitude to any kind of numbers so succinctly.

  55. oiaohm says:

    Der Balrog
    –And even the other (Windows-centric) magazines offer Linux articles in nearly every issue. Every month you can buy dozens of magazines with Linux Live CDs and detailed installation instructions.–

    Now that is a big question why. Turns out a lot of those magazines have noticed a drop in sales when they do a book without a Linux. Some as high as 80 percent drop.

    Wikimedia numbers are bias by schools and other places requirements on students.

    Der Balrog
    –Isn’t it a defining characteristic of Linux that you don’t need a distribution channel via retail?–

    This is not a characteristic of Linux. There have been retail versions in the past.

    –For instance there are surely twice as many Android-devices than iPhones and iPads. But Wikimedia is showing 2:1 or even 3:1 for iOS? Why?–

    Mats Hagglund
    This is a good question particularly thinking it possible to go to the carriers and get exact number of those devices connected to the telephone networks by 2g 3g and 4g world wide.

    http://en.wikipedia.org/wiki/Usage_share_of_operating_systems#Mobile_devices Comscore uses USA carriers.

    So yes its almost 1.5 to 1 in android favour in the usa. Yet for some reason ios devices are being counted more often.

    Is this carrier grade nats? Or maybe Android people are not that big of web surfers.

    Der Balrog that is the problem I have with web numbers. Mobile phones where we get exact details from carriers web states don’t even come close to lining up.

  56. Der Balrog says:

    As is well known, M$’s OS is widely found on retail shelves so is widely used by consumers who know nothing else.

    As is well known, Linux distributions can be downloaded from the internet for free. Either blame people or don’t. If they can operate a browser then they are perfectly capable of downloading a Linux distribution.

    Isn’t it a defining characteristic of Linux that you don’t need a distribution channel via retail? In Germany there are as many as six dedicated Linux magazines on the shelves. And even the other (Windows-centric) magazines offer Linux articles in nearly every issue. Every month you can buy dozens of magazines with Linux Live CDs and detailed installation instructions.

    You’re contradicting yourself again and again and again. If it’s about retail and pre-installation then Firefox shouldn’t have stood a chance. But how did it then happen that Firefox took away major market share from Microsoft’s Internet Explorer? Something doesn’t add up here.

  57. Mats Hagglund says:

    During the last 4 years Wikimedia shows slow increase for Linux with non-mobiles. From 1,5% to 1,85-1,9%. That’s not much. However, as we should understand these statistics are based too much of American, and other english-speaking countries.

    For instance there are surely twice as many Android-devices than iPhones and iPads. But Wikimedia is showing 2:1 or even 3:1 for iOS? Why?

  58. Mats Hagglund says:

    We can’t trust much these figures like 7,4% for Linux, 57% for Windows and over 25% for iOS/MacOS. Everybody should understand that less than 10% of all devices are using Apple systems. That’s a fact.

    However there is no doubt that these statistics are showing the clear trend of Windows fall down. Not only in mobiles but non-mobiles too. It was something like 89,5% (April 2009) to less than 75% (Feb 2013) in 4 years if mobiles are excluded.

  59. Der Balrog wrote, “Desktop Linux stays flat at about 1.4 percent.

    Isn’t that interesting? Those computers you claim are sold with Linux seem to make such a big impact.”

    As is well known, M$’s OS is widely found on retail shelves so is widely used by consumers who know nothing else. Schools, governments and businesses which have rolled out huge quantities of GNU/Linux desktops are counted as units NATed by firewalls.

    e.g. There are school districts in USA using many thousands of GNU/Linux desktops/notebooks and yet they don’t show anywhere in the stats. Google, on the other hand with just a few thousand desktops in California gets to shift the entire USA and the state of California. I guess they have one IP address per PC unlike most others. QED.

  60. Der Balrog says:

    The more important thing for you loons seems to be:

    Desktop Linux stays flat at about 1.4 percent.

    Isn’t that interesting? Those computers you claim are sold with Linux seem to make such a big impact.

    Didn’t some people claim that Desktop Linux has a Mac-like market share by now? Well, it sure doesn’t show in the Wikipedia stats. “Expensive” Apple computers that actually work: almost 7 percent. Desktop Linux crap: 1.4 percent.

    Windows may “fall from grace” (whatever that means), but Desktop Linux just does as bad as ever. If it wouldn’t suck so much, it would’ve usurped all those former Windows people. But those people have a brain (like Miguel de Icaza) and rather go Mac than put up with the abomination that is Desktop Linux.

    Nice try, though, Pogson.

    One man still believes.

Leave a Reply