back to article Linux kernel R&D worth over 1bn euros

How much would it cost the European Union to cobble together the Linux kernel from scratch? The development costs would reach over a billion euros (or about £900m, or $1.4bn USD), according to researchers from the University of Oviedo, Spain. Jesús García-García and Mª Isabel Alonso de Magdaleno are set to present this open …


This topic is closed for new posts.
  1. James Green


    Thatis just the development cost, if this were a "real" marketted product it would need marketting cost too, it would also need to make a profit in "the real" world so 1 billion Euros is only the start of it.

  2. Trevor Pott o_O Gold badge

    Requires more $X

    The report concludes by saying that despite a lack of book value, commons-based innovation should receive a "higher level of official recognition that would set it as an alternative to decision-makers."

    Attention, adoption, legal protection and most importantly investment.

    $1B Euros isn't exactly a back-breaking amount for an entity the size of the EU, but think of how much value is derived from that $1B Euro product. Think how much more we could all get for $1B worth of new investment.

    All operating systems have their place, but given how dependant the entire world is on Linux...government support, uptake and legal protection would be very, very nice...

  3. M Gale
    Thumb Up

    Also notable..

    From the poster:

    "Legal and regulatory framework must allow companies participating on commons-based R&D to generate intangible assets for their contribution to successful projects. Otherwise, expenses must have an equitable tax treatment as a donation to social welfare."

    Tax breaks for OSS developers? Expect a certain operating system vendor to argue strongly against this recommendation.

  4. Charles Manning


    Average EURO31k/year programmers include Visual Basic developers and other low-skilled folk. Kernel developers are not found in the bargain bin. You probably want to double the programmer fees.

    Much of the value in Linux comes from all the testing which never really gets factored in. While programmers' contributions are protected by copyright there is no such recognition or protection for the people that test the code.

  5. Ken Hagan Gold badge


    "A base salary of 31,040 euros per year was punched in, based on the EC's Eurostat figures for 2006."

    I know it's all hypothetical, but if you want someone capable of writing an OS kernel and willing to work for that salary, you're wasting your time looking in the EU.

  6. jsg



  7. Anonymous Coward

    But why

    Why would somebody bother to work this out, its less useful than knowing how many grains of sand are on a beach.

    How much did it cost to find this percieved value, who paid for it and why!

    What next how many monkeys with typewriters would it take to come up with global cooling^H^H^H^H^H^Hwarming^H^H^H^H^H^H^Hchange. Useless.

    How many people died that could of been saved had the time/effort/money that was used to work this trivia out; That would be more interesting.

    What next - scientests have worked out why the chicken crossed the road - it was down to the city planners making a ring road right thru Mr Gregsons farm. Cost to find that out £2.7 million but nobody died in the process, though nobody was saved by it either.

    How much has Europeans lost in Euro's thru running windows; bet you that would be alot more.

    1. mike3

      It shows the POWER of the paradigm.


      Why would someone work it out? Well, it goes to show the true POWER of the Free Software paradigm, as opposed to the proprietary one.

      And the Linux kernel is not just a random chicken crossing a road, but a major part of computer infrastructure.

      And I'm not sure why it'd cost 2.7 mill GBP just to do this little calc. Finally, the cost is likely tiny compared to the much larger amounts spent on things like war, continuing ecologically destructive activities (as opposed to trying to get away from them), and more.

  8. Anonymous Coward
    Anonymous Coward

    Don't forget...

    that government software projects always overrun and are horribly over budget - and half the time don't do the job right. Are their calculations taking this into account? If the EU decided to make an OS it would be the biggest software engineering failure in history.

  9. Anonymous Coward
    Anonymous Coward

    Base Salary

    Is pretty meaningless in these things, my company charges my time out at about 2-2.5 times my actual salary, this covers supplies, electricity, premeses, heating, equipment, licences, training etc. and this is for internal customers, as it's all in-house, so there is no profit built into it.

    Hopefully all this has been taken account of too, it cost's a lot more to get someone sat in a chair coding, than just their salary.

  10. Jay Jaffa

    31,040 euros !!!

    Let me just say that 31,040 euros doesn't pay for an OS-level developer. I'd suggest you'd need to double that at a minimum to equate to the salary you'd need to pay to a decent kernel-level developer

  11. Anonymous Coward

    cheap EU

    Average EURO31k/year programmers?

    Go to cheap EU side: Romania, Bulgaria and with 500Ex12monts you'll find capable programmers to work on OS kernels.

    One employer from .ro once bragged: I can find good programmers to work for 80E per month.

  12. Sean Timarco Baggaley


    What a bloody ridiculous premise:

    "How much would it cost the European Union to cobble together the Linux kernel from scratch?"

    Why the hell would anyone bother creating another clone of a clone of a 1970s OS when the BSD variants already exist? Why duplicate so much effort for so much money over a period of 19 years to achieve so little, as GNU/Linux has done?

    The *only* reason Linux has proved popular is because (a) it has a lot of vocal fuckwits pimping it, and (b) it's "free". As in "beer". Nobody gives a slightly used condom about the "free as in speech" part.

    The correct answer to the premise is therefore closer to "£0.00". Politicians may not be the sharpest pins in the cushion, but they don't get to run entire countries by being *total* imbeciles.

    1. Keith Oldham

      Re : Huh? →

      Some might be impressed by your stupid ( and offensive ) reasons but they seem unlikely to have affected the choice of GNU/Linux for the bulk of supercomputer systems, the bulk of internet servers, all those the embedded systems and the more intelligent desktop user.

      I'm just a little puzzled that you seem so threatened by an OS that you both despise and yet clearly don't understand.

      Stick with Windows - you clearly deserve it

      1. Jamie Jones Silver badge


        You Linux zealots totally miss the point.

        When Sean Timarco Baggaley says:

        "Why the hell would anyone bother creating another clone of a clone of a 1970s OS when the BSD variants already exist? Why duplicate so much effort for so much money over a period of 19 years to achieve so little, as GNU/Linux has done?"

        He's totally right.

        The BSD's already existed, and would be the dominant Unix around now if it wasn't for Linux. They also perform better than Linux, and could have benefited from some of the Linux developers input, so putting it that way, the value of the Linux kernel is a negative figure.

        Go on fanbois, mod me down.

      2. Sean Timarco Baggaley

        I don't use Windows.

        I use UNIX.

        (BSD-flavoured with a commercial GUI, since you ask.) I hate all OSes. They're all shit. Some are marginally more tolerable than others, but it's a close-run thing.

        The article claims that Linux is "worth" a billion to somebody. It isn't. It's popular primarily because it's free and has had billions of dollars jizzed all over it in the form of *advertising*. It sure as hell hasn't added anything of any great note to the IT industry aside from this, unless you can explain exactly how a rip-off of a rip-off of an OS that's almost forty years old is of such a massive benefit in this day and age.

        Has it expanded our understanding of computers? Nope. Has it created a new way to develop software? Has it fuck—if anything, it's made programming harder! Has it, in short, done anything MINIX, BSD and System V weren't already doing? The answer is a resounding "no". And it's taken only 19 years to achieve this through the unique development process based on the "stupidity of crowds". (Seriously: how long does it *take* to come up with a single, unified install and uninstall system, or a single, consistent GUI, with proper guidelines? Microsoft cracked this in 1995 and even Apple have managed to pull it off. And on a BSD-flavoured UNIX at that.)

        That universities and research labs are more than happy to go with Linux for their supercomputers is hardly a great shock: it's a free OS based on UNIX, which was invented *by* researchers, for researchers and programmers! And QNX was running just dandy on embedded systems years before Linux came caught onto the idea.

        You don't hear BSD users going into a media-wank of self-congratulatory bullshit every blasted day on the internet, even though the free 386BSD project was already in development long before Linus released his first kernel. But anyone would think the GNU and FSF people *invented* the notion of giving stuff away for free, and that the long-standing concept of Public Domain is but a figment of my imagination.

        There's nothing remotely special about Linux other than the juvenile politics it comes bundled with. It's just a damned UNIX clone that took nigh-on twenty years to reach its present level of mediocrity and "good-enough-ness" (you know: that thing everyone used to accuse *Microsoft* of achieving.) It's not magic. It's just tragic. A technological dead-end.

        And yes, I doth protest too much. I'm angry. Over the last 30 years, I've seen far better operating systems flare briefly in the light of day, only to be snuffed out by far less deserving rivals thanks to humanity's love of wilful ignorance and short-termism.

        Bah! I'm off to go strangle a kitten.

        1. Ken Hagan Gold badge

          Re: I don't use Windows

          "(Seriously: how long does it *take* to come up with a single, unified install and uninstall system, or a single, consistent GUI, with proper guidelines? Microsoft cracked this in 1995....)"

          And came up with GUI alternatives in 1996, 1998, 2000, 2002, 2006 and 2007, to name but six. Prior to 1995, they had Win3 and OS/2 which were both different but both claimed to be built on IBM's CUA guidelines for interface design, dating from the 1980s. Just because something is a solved problem doesn't mean twats won't spend decades searching for a second solution.

          To be fair, they've stuck with much the same installer technology for most of the last decade. End-user sticks a CD in the drive, points and clicks, and then prays that the installer (written by a third party, not MS, and running with full admin privileges) doesn't ask the installer service (which *is* written by MS and would be trustworthy if it didn't just blindly execute whatever instructions were given to it) to do something nasty. Just because something is a solved problem doesn't mean twats won't spend no time at all searching for even the first solution.

          But I'm nit-picking. You're basically right. All popular OSes are shit. The UIs are designed by children obsessed with shiny things. The rest is designed by programmers obsessed with getting their pet projects top billing. Sadly the point of this cost estimate was not to argue that the EU should fund the development of one that isn't shit. (I'm not sure what the point *was*, mind.)

        2. Quirkafleeg

          Re: I don't use Windows

          “Seriously: how long does it *take* to come up with a single, unified install and uninstall system, or a single, consistent GUI, with proper guidelines?”

          What does that have to do with Linux?

          Linux, combined with userland stuff to make an OS (let's call it GNU/Linux, for convenience) – yes, there maybe you have a point. But that would be monoculture…

  13. Anonymous Coward
    Gates Halo

    What are they doing wrong?

    Ok lets indeed double the salary, so that makes a total of 2 billion dollars. A certain monopolist of the OS market spends 9 billion on R&D each year, yet still can't produce a secure OS.

    What are they doing wrong? Are they even trying?

  14. Giles Jones Gold badge

    Average salary

    The average is probably right, there would be a tier of different programmers on different wages.

    Some doing scripting and configuration (Linux kernel config screens), some doing simple bug fixes (stuff not compiling) on a lower salary, then those doing full blown development on a higher salary.

  15. Daniel 1

    Go, Sean

    Are the nasty people writing an operating system you don't understand? Are they starting to use it in games consoles? Never mind: there'll always be a place for Crayon-programmers like you.

    "His hobbies include writing about himself in the third person..."

    In the third person? Why? Because no one else will write about you?

    1. Sean Timarco Baggaley


      You found my WRITING website. Well done. And yes, I freely admit I used iWeb. (Until last week—check out Google's cache if you don't believe me—I was using Acquia's distro of Drupal. Overkill for my needs, but fun to play with for a while.)

      Perhaps I ought to post a full CV on it—I've written published games in assembly language, so I'm hardly a "Crayon-programmer". But please: do explain why you believe programming *should* be hard. I've never heard a logical explanation.

      For a community which so prizes the concept of providing source code with everything, this is a truly bizarre attitude. Why include that source code if you don't expect any of your users to know what to do with it? Anyone who isn't a "Crayon-programmer" would have built their own damned tool, assuming you have the same NIH problem as the many others of your ilk I've met over the years.

      The whole *point* of a computer is that it is a near-infinitely flexible device. Why should end users forever be kept from getting the most from it?

      1. Homard

        If you're as good as you say you are ......

        .......then contribute to Linux, and make it even better. We only have your word for it. Go on prove it, if you dare ;-).

        If not, please write your own, in which case I will be only to happy to offer a demonstrably literate critique of your O/S, equal in quality to your own, using all the expletives you have bundled in ! As a point of principle, my comments *WILL* be negative to make a point to *YOU*, regardless of whether I like your O/S or not ;-).

        I don't agree with you about the value of the Linux kernel. I think it is nothing short of a miracle that developers the world over have put this together and it is as wonderful as it is. Since the kernel is Open Source, and therefore free for all to use, with few restrictions, none of which are unreasonable, you could argue it has a 'value' of 0. **WRONG**. This just means it is given freely. What you fail to grasp is that this exercise has tried to demonstrate the value of the development effort that all the coders, testers, and coordinators have put in. I knew it was some huge sum, but had not really considered the magnitude. I think what it shows is just how much everyone who's contributed to the kernel and the software you can use on it has given to the world. THIS IS AN ENORMOUS GIFT ! And you criticise it ?

        There are other benefits. That old PC that is not fast enough to run the latest Redmond offering can still be used with Linux elsewhere. This means old kit can be used for education in Third World countries for example. There is no O/S license fee to use the machine. And when the kids that have grown up with Linux become developers ? Rise of the Penguins ?

        Your comment : "The whole *point* of a computer is that it is a near-infinitely flexible device. Why should end users forever be kept from getting the most from it?" - surely this does not add up ? In what way does Linux actively *STOP* you from doing what you want - other than user permissions, and hardware that has yet to be supported ? Neither of these are direct acts of user sabotage as the DRM brigade would have enforced. Some O/Ses include DRM features and the like to stop you from doing what you want. So again I disagree with you.

        Like you I have seen some fine O/Ses and software come and go, and like you I have been saddened by their demise, especially when the stuff that comes along and becomes popular is far from what you want ! Here is something **I** **LIKE** and enjoy using. Thank you to everyone who has made GNU/LInux possible : kernel and surrounding software coders/testers.

        Now be sensible and contribute to this effort. Help to make it better. Help to make it what you want. Everyone benefits. Far better than negativity towards such a wonderful free gift don't you think ?

      2. Daniel 1

        Oh :)

        Mate, I have lots of crayons: especially green ones (because Ilike to comment my code).

        My point was... stop trash-talking people who are cleverer than you, just because they're clever than you: it doesn't make you sound clever. It just makes you sound bitter and thrteatened. Especially if you have to brag about writing stuff in machine code.

        Live with it: we live in a world full of people much cleverer than you or I. Trash talking them does not make you (or I) look clever. We can compare crayons, if you like, however. I promise to snap all mine in half, to make sure all yours look bigger.

        1. mike3


          This statement is interesting: "we live in a world full of people much cleverer than you or I. " It makes it sound like they'd be the majority, and you must be in the dumbest minority then. How did you determine this? :)

          Finally, what the shouldn't have is big egos. Ego is evil. That's what ultimately people see and react to.

      3. mike3

        Why include source code...

        Why include source code? So other people who CAN program can make those changes. It frees the software from being tied to a single maintainer. If that maintainer dies, or quits, or whatever, someone else can pick it up and continue the software. And it's happened. It also means others can provide patches for bugs, or enhance the software program in various ways. And the Free Software paradigm is not just about including source code, but also about giving freedoms to the users that they would not have under restrictive (proprietary) models. With a piece of Free Software, I can legally take my disc and install it on as many computers as I please, or give a copy to my neighbor to help him, and other things. If I did that with a proprietary program, I'd be infringing the creator's copyright, i.e. violating the law. And I might not even be able to do some things that would otherwise be allowed under copyright law, because the license agreement said NO! to them.

        In other words, the FS paradigm does NOT expect that "none" of the users will know what to do with it.

  16. Dan 55 Silver badge

    31,040 euros/year

    Remember this is Spain, where one word (informático) covers everything from crawling round installing LAN cables to developing a new generation programming language. There's not much matching prospective employee's skillsets to projects as requirements will constantly change anyway, you might be hireed because you know a certain protocol, but the next week you might have to rip it out and start again. As a result the analysis is kept to a bare minimum and any problems are resolved by overtime and hammering the keyboard until it works. If you don't like it then someone else will be happy to take your place.

    So you'll get a kernel, sooner or later, it'll probably work, just don't expect it to be particularly optimised or anything like that.

    (Speaking from years of experience...)

  17. Anonymous Coward


    ...may functionally do the same as Linux. But it is significantly slower. And that includes MacOS X. It is not just a conspiracy of incompetence that made Linux successful.

    The BSD licence I like more, though.

    1. Jamie Jones Silver badge

      "Significantly slower"

      haha, total troll.

  18. Anonymous Coward
    Anonymous Coward

    @Jamie Jones

    Maybe *you* are the BSD/OSX fanboi, unable to objectively assess something ?

    I wrote a client/server application (a kind of Object Request Broker) and ran it on BSD, Linux and MacOS X. Linux was always faster in terms of TCP/IP requests processed/second. IIRC BSD and MacOS X was about 30% less than Linux. CPU-bound performance was not different, which is not surprising for any reasonable OS.

    So this seems to indicate that Linux is indeed an innovation in terms of performance.

    If your theory were right, Mr Thorvalds would have to be a marketing genius, but it seems to me the opposite is true, considering his regular insults towards a lot of people. I simply think his tech is better.

This topic is closed for new posts.

Other stories you might like