Comment: Breaking the cycle of OS fragmentation
The mobile software world is evolving at a rapid rate buoyed by the infusion of open source principles, methodologies, philosophies and licensing. While this is exciting and invigorating, it is also giving rise to a number of issues – some new, while others give us a sense of “déjà-vu”.
Déjà-vu all over again
One example of the latter case is how fragmentation is playing out in the mobile industry today – largely a consequence of varied approaches to platform licensing and governance. Fortunately, there are forward-thinking organizations in the mobile industry driving initiatives that hold the potential to break the cycle of fragmentation within the mobile ecosystem.
Fragmentation occurs when there are multiple inconsistent versions of a single platform causing an inability to develop applications that can directly, or with limited reconfiguration to address variances in form factor, run across multiple devices that are ostensibly based upon the same software platform.
Within the desktop computer industry, it is widely believed that the biggest hurdle to desktop Linux adoption was fragmentation – more precisely, the fragmented communities, who were each using a different flavour of Linux (Red Hat, SUSE, Gentoo, Ubuntu etc).
Given that these flavours are not uniform in nature, PC vendors are unable to predict their behaviour and have been reluctant to adopt them.
For a long time, operating system fragmentation has also been the Achilles’ heel of the mobile industry – hindering innovation and preventing the democratisation, widespread adoption and use of the mobile Internet.
Applications must be ported from one device to another, translated from one language to another, and even customized for specific mobile operator needs. Once a developer builds an application, it is often a further challenge to get it to run on enough phones to create a commercially viable addressable market.
A handful of mobile platforms
While it is not realistic to envisage a single operating system for all mobile devices -n or would one wish to create an OS stranglehold similar to Microsoft’s position in the PC marketplace – it is essential for the mobile industry to coalesce on a handful of mobile platforms and this process has already started.
The industry has finally come to realize that while proprietary and vertically integrated systems are very good at addressing niches, the only way to unleash innovation is through open, independent and collaboratively developed industry platforms.
See also: How openness and Linux are unlocking innovation [Part 1]
This has resulted in the formation of a number of consortia offering a variety of “open” operating systems and development environments, each promising to be the solution that can deliver exceptional ROI for developers. While on the surface, these organizations appear to be operating in the same way and delivering similar outputs, there are subtle but critical nuances in their governance, collaboration and licensing model that highly impacts their ability to solve fragmentation. In other words, the devil lies in the details!
It is now fully acknowledged that focusing on common platform technologies alone will not guarantee successful management of fragmentation; instead, a well defined infrastructure incorporating a strong and sophisticated IPR policy and platform licensing is required to ensure broad adoption by the various constituents of the mobile value chain while preventing forking and inconsistency in future releases.
For fragmentation to be controlled, a strong copyleft policy is a prerequisite, as it obliges users of the platform to contribute fixes and optimizations before the platform is commercialised. Another key element in combating fragmentation is compliance testing which ensures that the right to distribute a platform is limited to compliant implementations. However, to be effective, it is also essential that compliance regimes are built from day one and not as an afterthought.
While platform coalescence is a great way of addressing fragmentation, it is unfortunately not a short term strategy. Rather, it will take the industry some time before it settles on a handful of platforms. In the meantime, standard runtimes and web technologies can serve as the bridge that developers use to reach the billions of mobile devices in the market.
Even the Web runtime space was highly fragmented until the advent of OMTP’s BONDI initiative, whose aim is to consistently and securely open up access from web applications to device and network resident capabilities. BONDI is defining the key mobile interfaces which are currently not standardised, but present the best opportunities for developers to provide great mobile web services. By enabling access in a consistent manner to these interfaces, BONDI is providing the developer with the ability to produce new web applications which will be capable of running across different devices and platforms.
In the mobile space, ensuring compatibility and interoperability across a broad diversity of platforms is a real nightmare for developers and as a result, is impeding true innovation. Convergence around a few common platforms will no doubt reduce the burden of development, maintenance and suppo rt for developers.
However, well defined licensing and IPR policies as well as support for multiple application development frameworks will be the key elements that determine an industry platform’s ability to address fragmentation.
The broad adoption of mobile standards such as BONDI will enable mobile operators and OEMs to manage their device portfolios more efficiently, ensuring that key applications can be supported consistently across all of them.
Consumers will benefit from a more diverse range of applications and more choice of devices while application developers can ensure that their applications work across different device platforms and operating systems, helping them get the maximum value from their investments.
Morgan Gillis is executive director of LiMo Foundation, an industry consortium dedicated to creating the first truly open, hardware-independent, Linux-based operating system for mobile devices.Tags: comment, LiMo, LiMo Foundation, open-source, os, software