Categories

Open always wins, unless it's Symbian or LiMo or Openmoko or Qt

In June 2008, Nokia made its first big move to turn around the platform, and announced that it was acquiring Symbian, with the intent of turning the OS into an open source project.

Two years later, the move to open source has proved to be a miscalculation that is slowing down Symbian’s development. It would be better for Nokia to take full control of the OS, according to Wood. A lack of support from other vendors means Nokia has to do most of the work itself, while the open nature of the platform allows competitors to keep a close eye on its progress.

via Nokia on long comeback trail after smartphone misses – Digital Lifestyle – Macworld UK.

Then there’s LiMo foundation open source mobile Linux. Maemo is/was open source, Openmoko and Qt Extended and PalmSource/Access moving to open source and there was the Motorola Linux OS that launched years ago. If Open always wins, whatever it wins, it’s not market share.

  • Ben

    In android's case, open is nothing more than a gimmick.

    How many android users actually downloaded the source code?
    How many of those who downloaded actually understand what's going on?
    How many of those who understand actually made any contribution?

    • J Osborne

      Disclaimer: I have not touched android.

      Not every user of an OS needs to look at or use the source to benefit. On other OSes I have managed to debug some tricky stuff with just a stack dump and libc source to figure out what my app did wrong by unraveling what libc did when I abused it.

      Other times I have fixed OS bugs or worked around them rather then blindly being for d to figure it all out from "my layer".

      The end user benifits from a better product without having to understand to OS source at all.

      Open source is valuable, but it isn't the only thing that is needed to make something great. You can put a world class clutch in a '81 VW bus and you still won't get a sports car from the mix. (or you can be Lotus and make great sports cars despite the crapy clutch). Open Source is a good ingredient, not magic pixie dust that makes it all better.

      (most of these benefits could come from religavly restricted license … But if you try that Donahue will be wary of allowing their cords to look at the source, or will take forever to process a "modest fee")

    • Uri

      Open source is about having acess to the code, to promote innovation, cooperation etc. When we say so the "user" can download the code and make modifications it's not that non developers can do it, rather developers who want to get involved. To begin with, android is built, and would not be possible, with out open source. This is the true power of open source..

      P.s.

      The fact you still hear this level of arguments in 2010 is truly disappointing. I mean just read about open source on wikipedia if you are not sure with the basics..

      • Ben

        Really. How many of those 3rd party developers actually made contribution to it.

        As far as I am concerned, with all the "openness" of Google, they are the one who used the kill switch, not the "closed and evil" Apple.

        The fact that you are still making arguments about open source promotes "innovation" in 2010 is simply hilarious. At least be honest to yourself, where would the Android be without the iPhone.

    • Sam

      According to these stats, at least 4000 people "understand" and have made ~2 contributions each (Android developers blog). A little more than a gimmick imo

  • Rob

    Ah, but you're leaving out the fact that Symbian sucks. Have you ever looked at the SDK? Have you ever tried to use a symbian phone for more than basic phone calls and text messaging? If anything, your conclusion should be that open source doesn't magically make your platform viable, not that opening your platform somehow makes it more difficult to improve. Nokia has had over 10 years to make the Symbian platform viable, and they simply blew it.

    • http://www.asymco.com asymco

      Ah, so Open always wins except when it sucks. And how are we to know if something sucks? Symbian's long life while sucking and its current HUGE share relative to the presumably non-sucking Android does not give us any hints about how much it sucks now or was sucking back in 2003 when it launched.

      • Rob

        Well, the OS was born in the late 80's, but you might say the modern OS was from about 2001.

        The platform is difficult to develop for. I feel confident you haven't tried it, spoken with a developer, or even read the executive summary on Wikipedia. In the era of "app phones" the lack of a good quality high-level API is a massive disadvantage. The fact that the platform was opened does nothing to solve the fundamental problems with the OS. Opening the platform could not have been expected to solve this at all, let alone in the 18 months since it happened. As I said before, being open won't magically save your crappy platform.

        Nokia has been betting the company on this platform for 8+ years, so it shouldn't come as a shock that the market share is large. However, Nokia has been losing market share since 2008, which is when (you guessed it) the "app phone" platforms started to gain traction. When they saw this happening, they bought out the joint venture they had with Sony/Ericsson and opened the OS hoping it would get some developers on board. So far it hasn't been the magic bullet they were hoping for, because writing apps for Google or Apple is a lot easier to sell than fixing Symbian.

        This isn't a problem with openness, it is that openness as a last-ditch effort (or even more accurately, openness without community) doesn't often have the desired effect. Your article seems to suggest that opening a platform isn't a good thing to do, but in fact Google's platform which has been open and well documented from the start (and has a thriving developer community) is far superior to Nokia's last-minute effort to save what is, from a development standpoint, essentially a niche product. You can't blame a lack of interest in Symbian on their open-sourcing the platform.

      • Rob

        TL;DR: your wish to somehow blame openness on Symbian's demise doesn't change the fact that it was actually just a lousy platform.

      • Tom Ross

        Actually asymco is quoting Macworld UK who are quoting some analyst.

      • http://www.asymco.com asymco

        My point is not that opening Symbian caused it to lose, but rather than "open always wins" is obviously bogus.

      • http://www.asymco.com asymco

        Rob,

        I worked at Nokia for 8 years in Mobile Software and Strategy. I know about Symbian.

      • Bill

        Time scale required can be huge. How many decades was Unix open and with a truly impressive community?

      • zunguri

        I, too, found S60 to be difficult to develop for in the early days. Later on, however, the tools got better and the problem became one of target creep. In terms of ease of development, however, I think you need to go back and try out Qt. I actually found the experience to be easier than doing my first Android app.

      • Jason Lotito

        "Open always wins" doesn't mean "Anything open always wins."

        But "open always wins" is pretty accurate. Look at how the iPhone succeeded (built on open standards and numerous open source/open standard technologies). Look at the innovation over the past 2 decades, and open standards and open source is at the root of it all.

        You can discount it all you want, but "Open" is the great innovator. Nothing any single company has done can compete.

  • WmTel

    What you've correctly identified is that OSS survives. Thriving, ie most of the money that's being made on mobile devices (though it's just a snapshot) is happening via iOS. iOS is also blessed–or cursed–with a vision. "Where there is no vision, the people perish"

  • Space Gorilla

    I think the point is more that the 'open vs closed' debate is kind of dumb. We have many closed systems in our lives that we're all perfectly happy with, such as the cars we drive, the water we drink, the houses we build, the list goes on. Where's all the bitching and moaning about how I can't plumb my house any way I want, how I have to follow these darn rules called building code, how I'm told how I have to 'develop' my house? Everything should be open so I can do whatever I want, right?

    • Michael Houghton

      This comment is hilariously misplaced.

      However strict your building code, plumbing is an open system. A set of standards (think GSM, HTTP) establish rules and protocols within which any plumber, any manufacturer of pipes, and (local natural monopolies notwithstanding) more or less any water supplier, etc., can operate.

      If the building code prescribed that you must buy your pipes and fittings from one supplier, and kept the specifications of those pieces secret from plumbers unless they paid a licence fee, then it would be a closed system.

      The strictness of a building code has little to do with its openness unless it is being used to introduce private monopolies via the back door.

      • Space Gorilla

        My point stands. There are many closed systems all around us, and none of the open zealots complain about those systems. I expected someone to nitpick about my examples. I can make an argument about cars not being a closed system as well, all you have to do is shift your definition of closed, which is what you've done re: plumbing. While you are allowed to buy fittings, fixtures, pipes, etc from many different suppliers, all of those materials must be approved before they can be sold (sound familiar?) and used in your building. And, an inspector must pass how you've installed those materials and how you have built your plumbing system (again, sound familiar?). You are not allowed to simply buy the materials, build a plumbing system, and then turn it on. You are also not allowed to build a plumbing system out of non-approved materials (think Flash in the app store). We can nitpick forever about degrees of open and closed, but I think it's obvious that many systems that work very well are essentially closed. And none of the open zealots complain about those systems.

      • Space Gorilla

        Further to my point, here's Google's CEO on their model: "The Google model is completely open. You can basically take the software – it's free – you can modify whatever you want, you can add any kind of app, you can build any kind of business model on top of it and you can add any kind of hardware."

        Now tell me again how plumbing is open by Google's definition of open. You cannot modify whatever you want re: building codes and materials. You cannot configure the plumbing system any way you like. You cannot add any kind of hardware. All of this is regulated and must be approved at various stages of development. Nuff said.

      • kevin

        To further your point even more, Schmidt is wrong; you actually can't build whatever app you want and sell it,. because Google has already shown that they can kill it.

        And you can build whatever business you want on it, but you shouldn't count on Google for any support (or even the next release). Witness Google and its support for Android-based tablets or the China OPhone (though here I presume China doesn't actually want any further support).

    • Mitchell

      "We have many closed systems in our lives that we’re all perfectly happy with, such as the cars we drive"

      Where did you get that idea? I am not at all happy with how closed my car is!

      When I was a kid, my dad could change the oil in the garage with an oil pan and a few paper towels. Now I need to buy special expensive machinery, or take it to the manufacturer. The service manual used to describe how the car fit together, which was enough to make most any minor repair. Today the service manual shows how the pieces fit together, but you need special proprietary tools and computers to actually do anything more complex than change a lightbulb.

      Cars have gotten a little simpler to use, and so hard/expensive to maintain that you have no choice to do it yourself, even if you're not afraid of using a wrench (gasp!). I would have gladly paid a few thousand dollars more for a more hackable car.

      I find it depressing to think that what happened to cars is now happening to computers. :-(

      • Tom

        When the Google CEO said anyone can do anything they want with Android, the telcos grinned. They took over the OS. Android is only as "open" as the telcos want you to believe. The manufacturers, HTC, LG, SE, etc., do what the telcos want. This is not truly open to the consumer. We have seen that the customer for android is the telcos, while the customer for the iPhone is — the consumer.

  • Ted T.

    Both comments are on the mark. Making something open source doesn't necessarily make it better or worse, but right now "open" is meaningless marketing nonsense being spewed by Google, which at it's heart (search and adwords) is about as closed source as closed source gets.

    Many companies this side of Microsoft have open source projects — Apple has Webkit (the Safari rendering engine) and Darwin (the Mac OS X kernel).
    It neither guarantees success nor failure.

  • Eric

    Open for so many people is a religious expression. An object of devotion. A choice that they wish to impose on all people. A choice they make with their hearts, and not necessarily with their heads.

    Open is cool. Apple believe in open standards. (Ever hear of Webkit?) Proprietary system, like RIM, iOS4 and others can do just fine when they interact with open standards. And though Apple is often accused of being a cult (it’s getting to be more of a sect rather than a cult – it’s too big to be the latter) they are pushing for removal of DRM, use of open standards for interoperability. They don’t always live up to their ideals, but then who does?

    Open, in the end is good, as long as it doesn’t become even more oppressive than the closed systems.

    • kevin

      Those who say "Open always wins" really have no clue what they are talking about, because obviously open doesn't always win – which means there are other variables that count as much or more in determining whether it wins.

      And those who associate the Open philosophy with only particular companies wear blinders. No commercial company is truly open in every part of their business, and most commercial companies have open parts and closed parts. Companies monetize and profit from the closed parts, while using open to commoditize components of the system or ecosystem in which competitors have an advantage or lead over them. Apple does it. Google does it. Nokia does it.

  • JJ

    Open works if it complements your core proprietary business model. In googles case – it allows them to collect detailed info to deliver ads to. They will never "open" the search engine, gmail etc.
    Closed works if it supports your business model.

    For sun Microsystems open did not work.

    • http://www.asymco.com asymco

      My article was about mobile OS's and in particular a closed system that was open sourced, but your point is well taken.

  • http://www.itmustbef8.net/ Hank

    There are two versions of open: the pragmatic, practical version and the religious one. Apple likes the practical side, both as a user and a contributor; it gives them engineering benefits and makes it easier to interact and integrate with the outside world. The religious side doesn't like Apple, because they reap the benefits but don't truly believe. Me, I'm not a true believer; I've seen success in open, and failure, and everything in between. Open's good for a lot of stuff, but not for everything. And it relies on a community, one that sees benefits from their efforts.

    Being somewhat open isn't like being somewhat pregnant, but it's a balancing act. Sun's approach met with limited success, although it wasn't enough to keep them a viable business. Apple's is doing very well for them so far. It might be that they've found the right balance, or it may just be that they have better taste both in how much they open up and how they build the closed part. Or both.

    • Marcos El Malo

      Well put, Hank.

      Sometimes the "religious" bigotry and fervent evangelism behind platforms is off-putting, even (especially) when it's its source seems to be social idealism.

      Another point: The theoretical benefits of the OSS model aren't always practical or real benefits. Here's an example: I was recently asking a dev about how to do something with his application. His answer was that it was possible, that I should look at the source code and have a go! :) I appreciate in theory both the fact that he had neither the time nor inclination to implement the feature I was asking about and that I had the freedom to implement it if I had the skill, time, and inclination. But practically, it's a non-starter. I haven't the skill nor the time. I still appreciate the theoretical advantage, at least in theory.

      (On the practical front, I am able to use this app in the first place because it is open source. So there is that.)

  • Fredrik Olsson

    'Open' is that extra spice that can make an already good software dish taste as if sent from heaven. But if the dish is made out of a turd then no amount of 'open' can ever make it edible.

    'Open' is not a silver bullet that turns any software project good. It is the extra spice that can turn a good software project into great. But even then it can kill your company unless 'open' is also compatible with your business model.

    In many ways 'open' and communism do have allot in common (Ballmer is right for once). It is a glorious ideal, but very seldom works in practice due to basic human nature.

    • FalKirk

      Communism is NOT a "glorious ideal". It is a soul-crushing, life-destroying mistake. And I am not saying this because I'm some 1950's ideologue who hated communist Russia. I studied communism, fascism and democracy in college. It's not the practice of communism that is flawed, it's the theory. Saying that communism doesn't work "due to human nature" is like saying that arsenic doesn't provide us with sustenance because human nature is weak and selfish.

      You really don't want to equate open source with communism. They aren't at all alike. The mantra of communism was "from each according to their ability, to each according to their need". That's forced labor where the weakest are rewarded and the best and brightest are oppressed. Open source is volunteerism at it's best. It's voluntary and collaborative. Communism is mandatory and collective.

      (Sorry about the lecture and going off-topic. As you can tell, I feel strongly about this.)

      • Tom Ross

        The GPL is not voluntary.

      • Roberto

        Yes it is voluntary. If you do not want to use GPL software, you do not need to use it – nobody forces you, really. If you do not want to be subject to the GPL as a developer, just do not include GPL code in your project. There is no law forcing you to include GPL code in your projects.

        You voluntary decide to use GPL. If you see some library you like, but it is under GPL, and you do not want to use that license, write your own replacement.

        With closed source you do not have access to you would have to reimplement the stuff anyway.

        So, the GPL given you more choice than closed source solutions.

        Of course the BSD license gives you even more freedom, but it is the choice of the original developer to set the conditions under which HIS code has to be used. It's HIS code, his private property. GPL works within the capitalist system, in fact.

        If you dream of a world where every line of code must be licensed under the BSD or even public domain, well, then THAT's communism ;-)

      • almonte23

        The same can be said about a world where every line of code must be licensed as GPL. The communist part is not the license, it's the authority that's forcing you to license your code with any OSS or closed license, the creator choose which license for his work, now where the GPL is a little bit like communism is where they say if you use GPL code you must license as GPL code, the GPL takes from you your right as a creator to license your own code, and it's debatable you are using GPL code to build your library, but is that reason enough to take your liberty away from you??? As I say people have different opinions about the GPL.

        I'm only sure of one thing, the GPL takes more freedom from you than the BSD or MIT license does, everyone can and should license their code according to their needs, this people who follows the FSF and the GPL as if it was a religious sect have given open source a bad name for far too long, I don't use the GPL on my code and I try to stay away from the GPL so I can avoid embarrassments like "hey you used GPL code here and there you must change your license from MIT to GPL or dual license it.", for these people it's not about using open source licenses it's about using the GPL license.

      • Fredrik Olsson

        The mantra of communism was “from each according to their ability, to each according to their need”. That’s forced labor where the weakest are rewarded and the best and brightest are oppressed.

        Your own words prove my point best. The mantra you quoted is fairness embodied: Each and everyone should out in their share according to their ability, in return everyone should bet everything they need, regardless of their disabilities. But as you continue; human weakness made sure the idea turned into opression and an upperclass feeding on the weaker.

        I am not saying that open source and comunism are equal in any details. What they have in common is that they only work if you assert qualities in humanity that does not exist.

      • FalKirk

        The mantra I quoted is the opposite of "fairness embodied." It's soull-sucking tyranny and a race to to the bottom where sloth and inability are rewarded and deified and ability is deemed shameful and martyred. Since it's irrelevant to the subject at hand, I'll just leave it at that.

    • http://www.asymco.com asymco

      Open, to me, is something that allows a company whose software is more than good enough to compete on a new basis; namely conformability to a new value chain. The cases where open source works well is are those when the software is "commoditized"; where improvements along existing dimensions of performance are neither valued nor consumed. Opening the source permits it to evolve in new dimensions. This is what allowed Unix to reach new customers: well after improving the kernel was not the basis for commercial competition.

      • http://www.asymco.com asymco

        I should also point out that a corollary is that if there is commercial value in improvements to the software and if it's not yet "good enough", keeping it closed makes it much more responsive to whatever signals the market is sending. This is why opening something like a mobile OS which is clearly still not mature as a user experience makes it instantly uncompetitive.

  • http://www.can-con.ca Dave Girard

    While I appreciate Qt's role in bringing a lot of the 3D apps I use to OS X, as an iPhone/iPad end user, I know all too well what Jobs means when he talks about the lowest common denominator of cross-platform development. Recently Maya 2011 just got the miraculous ability to not leave windows over other apps while rendering in the background in OS X and it's still a problem with Maxwell Render (both Qt apps from different devs). Nuke and Maya use an open/save dialog that looks like something from 1982 – no image/font/video/Quicklook plug-in based previews or Spotlight searches like the OS X native ones. The end result is that it feels like a Linux VM running a single app that doesn't care what my OS does – because the devs have used a quick and dirty solution that works on all platforms. Would I like to wait 12 months for Qt to get iOS 4 multitasking updates and then for no one to use them? And Qt is the best of the lot. I am indebted to Qt for my 3D work but if I was Jobs, I'd also be using what leverage I had to keep the lazy developers from crapping stuff like this on end users and basically forcing Apple to deal with another CodeWarrior – an albatross to drag the platform down.

    • Danny77

      "I’d also be using what leverage I had to keep the lazy developers from crapping stuff like this on end users and basically forcing Apple to deal with another CodeWarrior – an albatross to drag the platform down."

      You forget that CW was successful because Apple's own development tools back in the classic days were terrible. Apple were the lazy ones in supporting 3rd party developers, unlike MS. As a former CW developer, I remember how crazy-difficult it was just to code a basic window with buttons on Mac OS9. I'm surprised as much software existed as it did.

      Qt is successful for a similar reason – It's a great alternative for those of us who want to write apps using *standards* like C++, instead of strange niche technologies list Obj-c and Interface Builder. Apple wants you to develop in XCode/ObjC/IB and nothing else just like it used to expect all developers to develop in ObjectPascal/Toolbox back in the old days. They're just as anti-developer as they've always been.

      • Tom CF

        Apple isn't anti-developer, they're pro-customer. Linux is pro-developer and what's-a-customer?

      • http://www.can-con.ca Dave Girard

        This is really what the argument is about: developers want to do as little work to target multiple platforms, which is understandable. What Jobs is saying is "you come to my restaurant and apply for a job, you're going to cook our food. There are plenty of talented cooks willing to do it our way and our customers expect our food."

      • Ted T.

        Objective-C stopped being niche a while ago — I recently saw a list of most popular languages. Obj C has entered the top 10 Some that no one will call niche are way behind it.

      • Steves

        @Danny77

        you do realize that it is no longer 1997? XCode and IB are amazingly powerful, fast and efficient means of creating apps targeting multiple Apple platforms. As for being Niche? Not any more. That ended a couple years ago when the iPhone started excepting Apps to the App Store.

  • SockRolid

    Smartphones are moving hard and fast into the consumer space. Most consumers aren't technical, and don't care whether their OS is open source or not. The 'user experience' is everything.

    And having said that, it's important to minimize that experience. A mobile OS must be excellent at getting out of its users' way because users want to go straight to their apps and their data. I think most consumers (and many tech savvy commenters here) are blurring the line between the mobile OS and the apps that it runs.

  • Mark Hernandez

    We always get into arguments when we make the mistake of trying to decide which is better – black or white. The world is nothing but shades of gray, and the trick is to optimize the tradeoffs and find a better balance, while it's all in motion, shooting for where you've determined the puck is going to be, getting out from under as many dependencies as possible to remain agile when things suddenly change direction, love what you do, and never letting your eye off the main goal – pleasing the customer.

    When one sums it up like this into the basic essentials, it's interesting how many of these points we've heard come out of Apple's mouth in recent years.

    Many companies think there's another path, and I wonder if there is.

  • John

    I didnt choose an iphone because it was an open or closed platform. I chose it because it had wifi and 3G together. Everythin except some top of the line rim phones at the time were 3G only.
    For me this meant I could use my phone at home, for free. Where iPhone os won was by supporting open standards. Unfortuneately the license every app developer must sign is Completely incompatible with the common open source licenced, so you can't reuse other great open software when making an app for sale. Doesn't seem to have slowed the platform down though.
    Also as far as dymbian goes, it was fine on the two nokiS I have had but there were no apps for what I wanted and nokia kept trying to make cheap under powerd phones with it. Fine phones, but crap PDAs.

  • Mark Hernandez

    Here is a new post by Marco Arment, author of Instapaper.

    His post directly relates to this conversation here. Forget about the products mentioned and follow the point from the title to the last sentence.

    Great Since Day One
    http://www.marco.org/769340032

    • FalKirk

      Thanks for the link. A great read.

  • Carniphage

    Don't care about open or closed.
    Care about good or bad.

    C.

    • Roberto

      Well, people that care only about closed or open and want only open, also care about good or bad. They think that closed=bad and open=good, which is a perfectly valid point of view. I am not saying I necessarily agree with them, but you need to specify the context of your quality assessment ;-)

  • http://www.androiddevelopmenttalk.com Android developer

    Well its all the end user , how the platform is served to users , how does it benifit them , how does it make life easy for them , how does it do everything by not ruining their expectations .
    Thats Android and iOS all about .

  • Vint

    Based on this evidence, I'm not convinced that you've found anything wrong with "open always wins".

    First of all, the timescale here is tiny. The iPhone has been around for less than 3 years. I can point to dozens of closed platforms that have been around for far longer than that before being displaced by an open system. If you want to say a closed product being huge for 3 years is evidence that open doesn't always win, then wouldn't Microsoft Windows be a much better example? Even where open solutions have completely displaced proprietary solutions, the proprietary solutions were around for a lot longer, e.g., TCP/IP is universal today but in 1974 + 3 years it was not at all obvious that DECnet and X.25 were doomed.

    Second, looking at iOS, how much of it would there by if they hadn't used open-source components with which to assemble it? To me, running a GNU compiler and an open-source Unix kernel and an open-source browser engine is a lot closer to being "open" than most proprietary platforms out there. You compare it to fully-open-source platforms and remark how they got beat by a closed-source one, but you could just as easily compare it to a fully-proprietary platform and remark how a half-open one beat the pants off them.

    I would be willing to bet that in 2015, the median smartphone will be more open than the median smartphone of 2010. There's hiccups along the way, on any sufficiently short timescale, but the curve is still clear.

    • http://www.asymco.com asymco

      As I've said before, opening source is natural for components which have entered the "good enough" phase of their existence. Commercially speaking, good enough is end of life. One cannot create value past the point of over-service. This is basic disruptive innovation theory. Commoditized (i.e. at a point where improvements are no longer valued) software can live on to compete on the basis of conformability or on becoming deeply modular.

      I would agree that in 2015 or soon thereafter the median smartphone will be open, but that means there will be no commercial interest in developing it further. Capital will be deployed in a new market by then. Open always wins in the same sense that death always wins or the second law of thermodynamics always wins. When I write that open does not win, I mean open does not win when value (and its corollary, wealth) is being created. In the end we are all dead but that does not mean we should all be acting as if we already are.

      • Vint

        Well, I'm not sure how your conclusion isn't trivially true, then. On a small enough timescale, the open-win curve isn't monotonic? I don't think anyone ever disputed that, even RMS himself.

        Aside from that, even if you assume that people only make creative things for money (itself not a very strong claim), the innovation before a system reaches the "good enough phase" isn't the only way to create value. Is Webkit in "the 'good enough' phase of [its] existence"? It's arguably one of the most innovative browser engines in the world, and it's completely open-source.

      • http://www.asymco.com asymco

        In the case of webkit the innovation is to make the webkit module conform and enable value generating layers above it. The theory is simple: integration is necessary when systems are not good enough because iteration is more rapid and the knowledge of what needs doing comes from market signals. When systems are good enough they become modular as the basis of competition shifts to customization and conformability to layers above.

      • Sander van der Wal

        By this argument, Windows for PC's should have been Open Source years ago (with WinXP). Windows 7 is not Open Source at this time, though, and it doesn't like it is going to be Open Source anytime soon.

        Further, the second most successful PC OS, Mac OS X, is not Open Source either.

        The Linux distributions are Open Source, but they have a minute market share on desktop PC's.

        This means that the argument cannot be correct, as it is failing to predict the Open Source-ness of the most popular desktop OS in human history.

  • creativereason

    Open has huge wins on Linux on the desktop. There really has been a ton of innovation to the GUI / xwindows end user experience. And as predicted several times over the last decade Linux is finally dominating market share for desktop machines. And all of those awesome apps like Photoshop, etc. /sarcasm

    Look, I love Linux as a web server, but Open doesn't always win