Blog Post

The Mobile Linux War

A recent report from ABI Research highlights the rise of mobile Linux, estimating that 23 percent of the world’s smartphones will have a Linux operating system by 2013. It appears that much of that growth will come at the expense of Nokia’s Symbian, and that LiMo and Android will be the main beneficiaries. What the report doesn’t note is that last year ABI predicted that 31 percent of smartphones will have Linux by 2012.

Either there’s something to explain the change in numbers, or we should perhaps take our analyst reports with a grain of salt. However, Linux is undoubtedly moving fast: 15 handsets were launched earlier this year with LiMo, and after several demos and prototypes, anticipation for the Android is running high. But the jury is still out on which framework will win out with carriers and application developers.

LiMo has the backing of NEC, Motorola and Samsung as well as SK Telecom and Verizon. Android, through the Open Handset Alliance, has T-Mobile, NTT DoCoMo, China Telecom, Telefonica, Google and several others. The stated goal behind both efforts is to eliminate some of the costs associated with developing mobile applications for multiple operating systems by using open source. It’s a laudable goal, but the fight between the two for market share demonstrates how hard it will be to lower costs, as developers will still have to build for multiple platforms.

photo courtesy of the LiMo Foundation and NTT DoCoMo

24 Responses to “The Mobile Linux War”

  1. Thanks for all the feedback on the forecast. I was the author. To be clear we are forecasting a market that now includes high mid tier, smartphones and MIDs. Our forecast is actually more positive in that we are appropriating more share to Linux across a larger market. I hope that explains the confusion.



  2. We should all take analysts’ research reports with several grains (or a bucket) of salt, particularly in growing markets where technology, market and consumer trends are hard to figure. It is hard enough to forecast numbers in a 12-18 month period (e.g. who’d have thought about Google’s Android a year and half ago), let along over a 4-5 year time period.
    Especially in these technology markets, it is hard to trust any numbers beyond a 3 year period.

    The most that one could trust the reports is on their being “directionally correct”, which is a cute euphemism for being 50% correct. Last I checked, any techie with an ounce of grey matter could make a “directionally correct” forecast.

    Coming to the more substantial part of post, there is tremendous fragmentation in the mobile OS/platform market. Symbian, Windows Mobile, Palm, Blackberry, LiMO, Android, Apple’s mobile, etc. all vie for attention from the same set of consumers. And while the market is growing, other players also jump in. For a more cynical or pragmatic (take your pic) view of Android, see:
    However, sooner or later, Operators will start reigning in their development and maintenance costs. Vodafone has already declared its intention to limit their supported OS. More carriers will likely do that. At the same time, new OS versions will likely spring up that offer consumers more variety and freedom (such as the Open Handset alliance supposes to do so).

    In a few year’s time, though, there’s bound to be consolidation to a fewer number of mobile OSes. And it will be consumers and developers who decide it when they encounter incompatibilities in their services. Local and fringe offshoots will continue to exist, but they are unlikely to be affect the market in a big way.

    That’s my 2c take.

  3. Let’s think about a couple of things for a second. The open source community *thrives* on a variety of options, selections, and the ability to customize. This is one of the very things that has made it so popular. In many recent Linux migrations, one of the reasons stated for the migration was to avoid vendor lock-in. People like freedom to choose. Diversity creates this. You think people are going to be bothered by two main mobile phone distributions? That pales greatly in comparison to the number of different processor architectures currently supported by the Linux kernel, some of which include: x86, alpha, amd64, arm, hppa, ia64, mips, ppc, ppc64, sparc. Not only that, how many Linux distributions exist? At *least* 348, likely much more. You think that developers will be phased in the slightest? It is my experience that given the current variety of architectures and distributions, having 2 major mobile Linux platforms will be but a start. If history is an indicator of the future, having two mobile Linux distributions will only last as long as enough people want and are willing to support additional ones.

  4. There’s some misapprehension, apparently, about ACCESS’ involvement here, and the LiMo Foundation’s general direction. ACCESS is not simply “developing our own version of Linux”, we’re a member of the LiMo Foundation and are working closely in that context with the other members to ensure that the ACCESS Linux Platform is, indeed, LiMo-compliant moving forward. That’s something our customers want, and something we’re going to deliver. So we are, in fact, “Flying with LiMo” (Android is flying squarely _against_ the world of mainstream open source, but that’s another story entirely…)

    ACCESS is committed to using mainstream open source code as much as possible, and for all the usual reasons: no sense in reinventing wheels, and we get a great deal back by working as a member of the open source community than we would trying to work against it, or without regard to it.

    Linux doesn’t need reinvention (a la Android). It does require that some gaps be filled in the basic platform to enable full cell phone functionality (e.g. telephony, SMS and MMS messaging, device management according to OMA specifications, etc.) By and large, however, the APIs that you’d use to program an ACCESS Linux Platform phone (or, indeed, any other “LiMo platform” phone) are going to be the same APIs that you’re used to on desktop Linux. GTK+, Gstreamer, BlueZ and other open source projects are core components.

    LiMo will enable the smartphone market by providing a common set of APIs for third-party developers to use. “Smart phones” are all about the applications you can add, and LiMo is very much committed to the creation and support of a thriving developer ecosystem.

    And, most certainly, ACCESS and other LiMo members are fully committed to observing the license obligations of working with open source code. Those are the rules, and we play by them.

    ACCESS works _with_ the open source community, not outside it: we’re a member, to give just one example, of the GNOME Foundation Advisory Board and are (for the second year in a row) a Silver sponsor of GUADEC, the GNOME Users and Developers Conference.

    David “Lefty” Schlesinger
    Director, Open Source Technologies
    ACCESS Co., Ltd.

  5. Vipin

    “how hard it will be to lower costs, as developers will still have to build for multiple platforms.”

    Very well said. I don’t understand when will these guys realize how hard is it for mobile application developer to port it on different platforms esp when all platform do not provide same capabilities.

  6. charlie

    I think the point of Linux phone is that the “smart-phone” market is slowly degenerating into a “featurephone+” market. You only need one or two developer applications outside the core set the phone provides out of the box. Developer mindshare becomes less important, and so multiple versions of Linux can flourish.

    UNIX on phones. What a waste. Ten years ago we would have laughed at it, but between Apple and the Linux Phones it may be 30% of the market soon.

  7. Thanks for the post!

    I think that it is detrimental that developers have to build their applications for two systems. From what I’ve seen though, these Linux systems are great and will be taking enough market share to make themselves long time platforms. Hopefully that will mean they create some development standards so the application creation gets easier.

    My vote goes for Anroid though. With Google backing them and with the super sweet gyroscopic mechanism they are developing for it, I don’t think it could loose.

    Check out this fun video Google released about their Street View implemented with Android:

  8. Brian

    Can someone, such as the author, place some intelligent commentary, in the body of this article, explaining how much greener or better the mobile world with LiMO because frankly, I’m not seeing it – just a bunch of hype. I understand how LiMO will improve feature phones but I don’t see it with smartphones.

    Give me some context here and why I should care about a war!

  9. Ollie

    I got a new work phone a few months ago, all I want is one that can run linux, I had to choose between an iPhone, n95, Sony p1i and blackberry, all of which did not cut it but I had to go with the Sony in the end. A good ‘open’ linux phone would be great to have access to all the tools I use and have bash with me at all times, would save me from having to carry a laptop with me most places I go.

  10. Herman Manfred

    So why are Access and Palm developing their own versions of Linux rather than taking Limo or Android and flying with THEM, are they making their own versions sufficiently unique that they WON’T really be “Linux” (or compatible) anymore, and will the source for those theoretically-GPLed pieces of software be freely available?