The rise of mobility and the cloud could elevate Linux to a level of
unprecedented dominance, reports Neil McAllister. Twenty years ago, when Linus Torvalds first announced his new operating system (OS) project to a Usenet discussion group, he had no way of knowing that his creation would become so popular. "It's just a hobby, it won't be big and professional," Torvalds wrote on 25 August 1991. Torvalds had begun the project as a fun way to teach himself about the Intel 80386 processor and nothing more. His greatest ambition was merely to see it work. Today, Linux - as Torvalds OS kernel came to be called - is available for just about every modern processor architecture and many archaic ones. Its OSes can power every kind of device, from PCs, netbooks and smartphones to mainframes, supercomputing clusters and beyond. Linux has won the support of industry heavyweights, including Dell, IBM, HP, Novell and Oracle. Red Hat, one of the first commercial Linux vendors, now has a market capitalisation of 7.3 billion dollars. The past 20 years haven't always been easy. Linux has made a few enemies, Microsoft foremost among them. It has faced its share of challenges, both technical and legal, and there are more hurdles ahead. Nonetheless, as Linux enters its third decade, its opportunities have never been greater. Computing is changing, and Linux is not only benefiting from this change but enabling it. Thanks to a shift beyond the PC, Linux is poised to become more than just an OS: it'll be one of the most transformative forces in computing history. The OS the internet built That Linux is helping to usher in a new computing age might come as a surprise to some. Linux has often been accused of following, rather than leading. Since 2003, the SCO Group has alleged that the open-source OS violates intellectual property relating to Unix, and it's true that much of Linux's early market share came at the expense of costly commercial Unix variants. Microsoft has also repeatedly claimed that Linux violates more than 200 patents. Yet Linux has been innovative from its inception. While commercial Unix variants ran on high-end systems based on reduced instruction set computer (Risc) processors, Torvalds designed his OS for Intel hardware, anticipating the trend toward low-cost x86 servers. And while Microsoft was famously slow to adapt to the internet, Linux has had the web at its core from the beginning. Torvalds has played a pivotal role throughout its existence and still co-ordinates each new kernel release - for example, it was he who dubbed a recent stable kernel Linux 3.0, despite it having "no landmark features or incompatibilities". As one of the world's most successful open-source projects, Linux represents the contributions of countless programmers worldwide. Anyone can download it, inspect it, learn from it, modify it or use it, free of charge, and with no obligation other than to allow others to do the same - all thanks to the internet. Some contributors have been individuals, and many have hailed from educational institutions. In recent years, however, the game has changed. Commercial interests are now the most prominent actors in the Linux development process. Of the top 20 contributors to Linux 3.0, more than half were acting on behalf of their employers. Linux is everywhere The relationship between Linux and the internet has been mutually beneficial. Just as Linux development has benefited from the rise of the internet, so too has the internet prospered from Linux's evolution. Today, the open- source OS is everywhere. It powers web servers, email servers, file servers, databases and more. In fact, according to Linux Foundation executive director Jim Zemlin, you use it "every time you surf the internet". For example, search engines have become the de facto home pages for most web users. Google leads the pack with 66 percent market share, and Google's servers run on Linux. So do most of Yahoo's, according to the Netcraft site survey. Even Microsoft Bing uses the Akamai content-delivery network, which runs on Linux. It's likely that none of these companies or their services would exist if it weren't for the massive scalability provided by low-cost Linux servers. Slow growth on the desktop Yet, as strong as Linux is in the data centre, it's notoriously weak on the desktop. Some PC and netbook vendors ship systems with Linux preinstalled, but sales of these are typically low. Many pundits have pointed to Linux's failure to displace Windows as evidence that it will never succeed as a client-side OS. But are they right? Where desktop Linux use can be measured, the OS seems to have found a niche among government agencies and large corporations, which typically enforce strict controls over employee desktops. Programmers also have a natural affinity for the OS, and in many cases they use Linux desktops that have been heavily customised for software development; for example, about half of all Google employees run a custom Linux variant. It would be a mistake, however, to assume that only geeks and specialists use Linux. It's as difficult to judge the number of desktop installations as it is servers. Because Linux is free and runs well even on older hardware, many hobbyists deploy it themselves. Some might erase another OS to install it, while others run it using virtualisation software, further confusing the numbers. Linux gained some ground as a consumer desktop OS with the debut of Ubuntu, a free distribution that excels in ease of use and user productivity. Ubuntu is currently the most popular desktop Linux, and founder Mark Shuttleworth has said his goal is to reach 200 million users by 2015. But that might be wishful thinking. According to analytics firm NetMarketShare, Linux desktops still represent less than 1 percent of the market. If we assume that users do most of their web browsing from their primary PCs, then Linux doesn't have much hope of gaining traction with consumers. The trouble is, that's an increasingly misguided assumption. Mobility and beyond Computing is changing in dramatic ways. More and more, users of all stripes are forgoing the traditional PC desktop in favour of new types of devices, services and usage modes. For example, in Japan, about half of all personal email is sent or received using a mobile phone. That trend seems likely to continue to the western world, where smartphone use is rising sharply - and that's a huge opportunity for Linux. Of all the smartphones sold in the UK market, Android now claims the largest share of any OS. Android is based on the Linux kernel, as are Nokia's Meego and HP's webOS (although the future of the latter platform remains in limbo, in light of HP's recent announcement to halt production of webOS-based devices). In Korea, Samsung's popular Bada OS also uses Linux, as does Aliyun in China. In other words, while desktop Linux use remains low, Linux gains where desktop operating systems lose as consumers forsake PCs for smartphones. Linux powers more than just phones, too. The market for Android tablets is growing and they, too, use Linux. Similarly, Google's Chromebooks, which offer a stripped-down experience that's little more than a web browser, rely on the Linux kernel. Even devices that barely resemble traditional computing platforms are often Linux powered. Leading e-book readers from Amazon and Sony are based on the open-source OS. You'll also find versions of the Linux kernel in networking equipment, GPS navigation systems, media players, TV set-top boxes and even TVs themselves. Many of these non-traditional devices are built using inexpensive, low-power processors based on the ARM architecture, which explains why Linux has been so successful in these markets. Robust, fully featured ports of the Linux kernel have been available for ARM since the late 90s. By comparison, Microsoft's only ARM offering to date has been Windows CE, and it won't have a fully featured OS ready for the architecture until Windows 8 ships in 2012. Microsoft may remain the biggest threat to Linux's continued growth, however. The rocky road ahead The rising value of the commercial Linux market may lead to an increased number of disputes among the Linux vendor community. For example, Oracle has frustrated Red Hat for several years by marketing what is in essence an exact copy of Red Hat Enterprise Linux. In response, Red Hat has become more guarded about how it releases kernel code patches. Equally important is the fact Linux's technical evolution isn't over. As successful as the open-source OS has been on mobile devices so far, it could do a lot better. Linux on the ARM architecture is a morass of redundant, device-specific kernel builds and distributions. Consolidation is sorely needed. Mobility is one frontier for Linux to conquer; parallel processing is another. Linux works well on today's multicore chips but, as tomorrow's chips grow to 48 cores or more, the Linux kernel won't be able to keep up. Between mobility and cloud computing, Linux has an unprecedented opportunity to become a dominant force the likes of which IT has never seen. But as it enters its third decade, Linux's greatest challenge may be to avoid becoming a victim of its own success. As the open-source OS has matured and stabilised and its code base has grown in complexity, Linux kernel hacking is losing its allure for new developers. Recruitment may soon become a top priority if Linux is to overcome the hurdles ahead. Linux's growing pains are over, but its grown-up problems have just begun. Search for old postings at: http://www.mail-archive.com/[email protected]/ To unsubscribe send a message to [email protected] with the subject unsubscribe. To change your subscription to digest mode or make any other changes, please visit the list home page at http://accessindia.org.in/mailman/listinfo/accessindia_accessindia.org.in
