[part 4 of the series on five traits of open source and its impact in the mobile industry]

Open source in mobile goes far beyond the confines of Linux-based operating systems for mobile phones. Examples are Sun’s Java, Motorola’s MIDP3 project, Microsoft’s Shared Source Initiative, Symbian’s use of open source, Adobe’s project Tamarin, Nokia’s S60 web browser, Funambol’s MDM server, the Eclipse Foundation’s open source development tools and the rising interest in open source hardware.

One of the biggest disruptions created by open source is in the case of mobile browsers. Since 2003, the mobile browser market had been dominated by three heavyweights, Openwave, Teleca (Obigo) and ACCESS (in addition to in-house browsers used by major OEMs). These companies had been responsible for the majority of mobile browsers shipped, while few manufacturers – most notably Nokia – had not only been sourcing browsers from third parties but also developing their own browser software in-house.

However, in the last few years, the browser market has been facing a number of challenges, namely:
- mobile browser per-unit royalties have been continually dropping, following the trend of software commoditisation. It is believed that browsers for mass-market phones today sell at a few pence per device.
- mobile browsers are inherently complex software, which have to cope with rendering malformed HTML (often called ‘street HTML’), the numerous evolving W3C standards around HTML, CSS and ECMAScript and the proprietary vendor extensions (e.g. rendering pages designed for Internet Explorer).
- as operator walled gardens are opening mobile devices are being exposed to the wilderness of billions web (HTML) pages, as opposed to thousands of simplified WAP pages that we previously the norm. The complexity and diversity of these web pages have called for advanced browsers, which typically take years to iteratively mature, as browser vendor Opera attests.
- the differentiating features of mobile browsers lie not in the HTML parsing and the rendering engine, but in the value-added features, such as intelligent zoom and navigation.

As these pressures were mounting, a critical point was reached in May 2007; within the space of one week, mobile browser vendor Teleca announced that it ‘halted investments into renewal of Obigo product’, while Openwave announced it was up for sale following a 50% tumble of its share price in 12 months. The industry impact has been significant, given that the Openwave and Obigo browser families have claimed the lion’s share of the mobile browser market.

Behind the scenes, this blow to the mobile browser business was struck primarily by Nokia’s S60 WebKit, Nokia’s newest browser based on an open source rendering and scripting engine for web pages. While business execution errors may have affected the demise of Obigo and Openwave’s business, it is the availability of WebKit, a reliable, open source, core browser engine that essentially drove browser pricing down. Nokia’s move towards WebKit also displaced some of its previous browser suppliers who lost a major customer. Furthermore, the open source WebKit has been developed into a first-class browser engine, under the auspices of Nokia, Apple and KDE. The corporate and community backing of WebKit implies that any further efforts to develop proprietary browsers is unlikely to be viable (although Opera and Access are still maintaining their proprietary browser products at the time of writing).

- Andreas Constantinou
VisionMobile

[Want to learn more about open source and its impact on the mobile industry? Register for the pre-workshop ‘A Crash Course in Open Source’ delivered by VisionMobile as part of Informa’s Open Source In Mobile conference taking place in Madrid on 17-20 September. Next on this blog series: Sun’s open source Java policy will mean very little for the mobile industry.] 

[part 3 of the series on five traits of open source and its impact in the mobile industry]

Linux is by far the software most commonly associated with (and often mis-identified with) open source and free software, where free refers to liberty, not costs. However the access to source code, ability to modify or redistribute, or the royalty-free nature of Linux are hardly the reasons why four out of five handset OEMs have adopted Linux. In other words, mobile Linux has not been adopted because of its ’free software’ qualities.

In 2007, handset OEMs have adopted Linux to varying degrees, from Motorola’s portfolio-wide Linux strategy to Nokia’s Internet Tablets segment-specific strategic experiment with Linux. The reasons behind the almost-unanimous OEM turn towards Linux are as follows:

- Reduced cost and time-to-market. The availability of a stable, high portable Linux kernel, hundreds of supporting royalty-free middleware components, thousands of Linux developer enthusiasts and a growing number of commercial mobile Linux software and service providers mean that mobile Linux is an effective operating system for mobile handsets, both in terms of time-to-market and cost of development. According to Nokia, one of the most successful corporate entities in working with open source, “Linux is the launching pad you need to stand on to be productive .. we have never managed to bring out a product in such a short time, with so few resources”.

- Wider choice: handset manufacturers have considerable freedom in selecting the middleware components of choice whether from open source communities, or in some cases from closed-source commercial components. A healthy exists in Linux-based software components such as graphics frameworks (e.g GTK+, Qt Core, FluffyPants), application environments (e.g. Qtopia, Hiker, Hildon, OpenMoko, SKY-MAP), multimedia frameworks, PIM middleware, file systems and telephony APIs.

- Strategic control: Linux-based operating systems afford manufacturers almost as much control of the platform roadmap as their in-house OSes. Manufacturers are much less dependent on a single software supplier, effectively lowering the cost of switching suppliers, an important strategic consideration. Furthermore, manufacturers are able to steer platform development of their own Linux OS variant in any direction they wish.

- Scalability: The Linux kernel has evolved over the years, to one of the most scalable and reliable operating systems, powering commercial mobile devices from low-end single-core feature phones to high-end smartphones. Manufacturers may easily trim unnecessary features or add high-end features such as USB support and VoIP protocols which are widely available for Linux distributions for PCs.

- Quality: Peer-review of popular Linux-based open source software provides for fewer software defects (‘bugs’). Both Nokia and Panasonic report that Linux-based software for mobile handsets has a high quality and very few bugs, compared to typical in-house software

- Innovation: The open, decentralised nature of Linux backed by strong developer communities, makes Linux-based operating systems a good choice for cultivating innovation. Chances are, a component will be already available somewhere in the Linux community ecosystem and can be adapted to a mobile Linux OS.

- Andreas Constantinou
VisionMobile

[Want to learn more about open source and its impact on the mobile industry? Register for the pre-workshop ‘A Crash Course in Open Source’ delivered by VisionMobile as part of Informa’s Open Source In Mobile conference taking place in Madrid on 17-20 September. Next on this blog series: How open source is shaking up the mobile browser market.]

[part 2 of the series on five traits of open source and its impact in the mobile industry]

Open source is in many ways the antithesis of corporate software development. The culture and dynamics of OSS development are defined by the nuances of a software community collaborating over the Internet. A community is typically formed by a combination of paid-for, pro-bono and hobbyist software developers with the same motivation towards solving a particular problem (‘scratching an itch’ in open source lingo). Community members are motivated by personal needs, peer recognition and last (and often least) financial reward. Communities are formed and organised ad-hoc around opinion leaders who are recognised based on the merit of their contributions to the community. This environment defies most rules of corporate software development:

- Processes and roadmaps: The mobile industry is accustomed to 100% specified and controlled development environments. However, thousands of open source software projects thrive despite a lack of project requirements and feature roadmaps. Open source development addresses features on an ad hoc basis; OSS projects are thereby evolved, not designed, driven by the needs and wants of individual developers or commercial participants into Linux development.

- Partner selection and management: Corporate software development projects rely on warrantees, indemnity clauses, non-disclosure agreements and service-level and marketing agreements. Each agreement is unique to the customer-supplier relationship and takes months to set up, adding up to an expensive relationship management. Moreover, software suppliers are chosen based on RFIs and RFPs which often consume extensive resources and time. On the contrary, open source software comes under oft-used licenses such as the GPL, LGPL and BSD, irrespective of the entities using or developing the software. Use of a few well-understood licenses in open source projects results in significantly reduced product time-to-development and time-to-market precluding customer-supplier negotiations. Moreover, the qualities of a software supplier are often evident through their OSS works, which are open to the community for inspection.

- Reversed customer-supplier relationship. In corporate software projects, the customer dictates conditions to the supplier and has control over project requirements, deliverables and roadmap. In open source projects, even if these are sponsored by a commercial entity, the community is the one who owns the project, not the sponsor. It is the norm for the sponsor’s corporate agenda to be in antithesis with the incentives of the community members; in these cases the community may take the project in a direction well beyond the control and the desire of the sponsor. As such, the customer-supplier relationship is reversed in open source projects. The community, which may be likened to the supplier, becomes the customer who must be appeased. Managing open source projects can be likened to walking on a tightrope, finely balancing the corporate agenda with community incentives. To win the community’s heart, sponsors must dedicate efforts, creativity and resources to the community.

- Innovation: Innovation in the software industry is almost always driven top-down; market segmentation and customer requirements filter all the way down to floor-level product decisions. Open source software is completely different. Innovation is entirely anarchic and ad hoc, often resulting in genuinely fresh concepts and product usage scenarios.

- Andreas Constantinou
VisionMobile

[Want to learn more about open source and its impact on the mobile industry? Register for the pre-workshop ‘A Crash Course in Open Source’ delivered by VisionMobile as part of Informa’s Open Source In Mobile conference taking place in Madrid on 17-20 September. Next on this blog series: ‘Mobile Linux is not about free software’]

[Over the next two weeks leading to the OSiM conference, I ‘ll be looking at five traits which characterise open source and its impact in the mobile industry, from community culture, to what open source means for mobile Linux, browsers and Java.]

Open source software is one of the most hyped, misunderstood, feared and high impact phenomena in the software industry today. The success of Linux, from a pet project to Microsoft’s arch-rival operating system has fostered hype for the success of the open source model. The unconventional business models where open source is employed, offer plenty of opportunity for misconceptions. Open source has also given rise to fear of IP contamination due to the ‘copyleft’ properties of the GPL license. At the same time, open source software has created tidal waves within many facets of the PC and mobile industries, from operating systems to browsers.

So what on earth is open source? A Google search produces more than 25 distinct definitions of open source, each one from a different perspective. In practice, there are three distinct contexts in which open source is used today:

- Software that comes with an OSI-certified license: The Open Source Initiative (OSI) is a non-profit organisation tasked with maintaining and promoting the definition of open source. The OSI defines 10 criteria for open source software, including that open source software must be freely distributable, access to source code and redistribution of modifications. Hundreds of open source-like licenses exist, of which the OSI has approved nearly 60. The vast majority of open source projects have been licensed under the GPL, the LGPL, the Mozilla Public License (MPL), the BSD License, the Apache Software License, and the MIT License.

- A social movement for making source code freely available
For many, open source represents a social movement among software development communities. This movement supports that software should be freely available to anyone interested in using it, modifying it or redistributing it. Community-based development and viral distribution are important characteristics of this movement. The very term ‘open source’ was coined to avoid misunderstandings arising from the earlier term ‘free software’.

- Open source as a collaborative development methodology
From a business perspective, open source is a collaborative software development methodology whereby a community of entities and individuals (commercial, non-profit or entirely voluntary) develop software through a transparent, distributed peer review process. Open source development can pool community efforts towards development of a software base that is of common interest to all participating parties, while allowing differentiation through derivatives built upon this software base. In this sense, open source as a business model is the polar opposite to commercial forums, which foster collaboration through exclusive or paid-for membership. Yet in some cases open source development may be equally or more effective at achieving the same goal. An example of a successful open-source based collaborative software development effort is the Eclipse non-profit foundation which is backed by over 150 industry players, including heayvyweights Google, HP, IBM, Intel, Motorola, Nokia and Wind River Systems.

Open source started in the early 90s as a social movement in favour of maintaining software freedom; the development of the Linux kernel as a free Unix alternative and the creation of the GPL license were the two defining milestones of that movement. Yet 15 years on, open source has evolved gradually and perhaps unexpectedly into one of of the most succesful methodologies for commercial, collaborative software development.

- Andreas Constantinou
VisionMobile.com

[Want to learn more about open source and its impact on the mobile industry? Register for the pre-workshop ‘A Crash Course in Open Source’ delivered by VisionMobile as part of Informa’s Open Source In Mobile conference taking place in Madrid on 17-20 September. Next on this blog series: how open source turns corporate software development on its head]

One day, sometime during the production of OSiM 2007, I looked up from my desk and found myself in the most extraordinary privileged position.

In researching the conference, I had spoken to almost one hundred individuals across the mobile and PC industries. I had first approached these individuals – strategists, evangelists, CTOs, developers and project leaders – for a research brief on their work in mobile software, and their priorities for an open source conference. Instead, our calls and meetings became something more…

Fifteen minute briefs became forty minute discussions on ideas and innovations. The voices on the other end of the line shared enthusiasm, inspiration, hopes and opinion. I heard experiences from beginnings of Linux in PC, the fragmentations, the forking, the hope that lessons can be learned. I heard excitement from the first mobile projects using open source, enthusiasm for what open source methodology could achieve, on how best to contribute back. I heard passionate debate on the philosophies behind Free Software and Open Source. I heard both scepticism and idealism on whether open source and Linux can really gain traction in mobile, and whether they should. I heard experience from the code-face and the board room.

In short, I had the privilege to speak with the most inspirational and forward thinking minds in mobile. Through this blog, I’d like to introduce you to them too.