As the mobile industry catches up with the iPhone, device fragmentation will increasingly become an issue and the costs and effects of potentially needing to rewrite (and maintain) every application for every device and platform combination will become increasingly noticeable, says Jim Morrish, Principal Analyst at Analysys Mason. This has recently been highlighted by Adobe’s desire to use Flash CS5 to write software once for deployment across multiple mobile platforms, and Apple’s recent actions that effectively prevent Flash-compiled software from being admitted to the App Store.
In the medium term, defragmentation may be driven by developers of more-sophisticated applications, mobile operators, handset manufacturers and the use of browsers as runtime environments, and combating device fragmentation is currently high on many agendas in the mobile industry.
This is in the context of a boom in smartphones worldwide, forecast to grow at an annual rate of 32% between 2010 and 2014, according to a Analysys Mason report, Smartphones forecast to grow to 26% of handsets by 2014.
More than 50% of this growth will be generated in developing Asia–Pacific and Latin American markets. Western European, North American and developed Asia–Pacific markets combined will account for less than 30% of total growth, so it is clear that the market for smartphones will transform over coming years.
This means initiatives to combat the effects of device fragmentation are now moving centre-stage. The GSMA recently announced the Wholesale Application Community, which will take the JIL (browser widget) and BONDI (handset API standardisation) initiatives and develop them into a common standard for widget-based applications. It will also take advantage of the work that the GSMA has already undertaken in developing open network APIs (OneAPI).
Others are taking an alternative approach to the same problem, promoting a ‘write-once, deploy-many-times’ scheme. For instance, Tech Mahindra is working to develop software that will ‘translate’ applications between incompatible OS environments, and Nokia is seeking to exploit the Qt application development environment, which is capable of compiling applications into code suitable for a variety of devices and platforms. Writing a mobile application once and then compiling it into many OS environments (cross-platform compilation) is also something that Adobe is pursuing with Flash CS5, as are Appcelerator with Titanium, Metismo with Bedrock and Nitobi Software with PhoneGap.
Another theme in the battle against device fragmentation is the extension of a ‘smart-like’ environment to non-smart phones. Opera Software is mining this vein, with its server-side widgets that are accessible via a handset web browser client, and shift application processing requirements off the handset and into the network. Another example is Silent Communication, which has written a proprietary software client that it claims will work on the majority of handsets shipped in developed markets during the past two years. The client acts as an abstraction layer, enabling the effective roll-out of standardised applications across a large range of handsets. An example is visual voicemail, which the company soon expects to roll out across the whole range of handsets from a leading manufacturer.
From a developer’s perspective, many of the defragmentation initiatives can simply seem to result in more fragmentation. Apple’s recent stipulation that all App Store applications must be originally written in Objective-C, C, C++ or JavaScript as executed by the iPhone OS WebKit engine – not, for example, in Flash and then compiled – adds further complexity (although cross-platform compilers that use Apple’s SDK may sidestep this restriction).
Moving forwards in a fragmented world It is unclear how this situation will evolve, but fragmentation is certainly here to stay for the foreseeable future and will continue to affect different applications in different ways. For example, it might be easier to port sophisticated games that depend on specific handset capabilities (such as a touchscreen interface) between OS environments for handsets with similar capabilities than it would be to port them within OS environments for handsets with different capabilities. Such an effect may drive some standardisation among high-end smartphones – for example, manufacturers and developers might adopt the latest iPhone capabilities as a lowest common denominator. The reach of sophisticated applications beyond this set of phones would then be determined by a simple business case: the incremental cost of tailoring the application compared with the addressable market per additional handset type, combined with an expected sales-conversion rate.
Many simple applications are relatively untroubled by fragmentation within OSs. For example, Nokia’s Ovi app wizard enables users to construct an RSS newsfeed client in a few minutes, and the resulting application runs on nearly all Nokia devices that support Ovi Store. Other relatively simple applications may simply be developed as browser widgets.
Operators and handset manufacturers are likely to be the driving force behind the development of more-sophisticated applications that are suitable for a large cross-section of devices. Such effects suggest that in the medium-term new dynamics may emerge in the market for mobile content and applications, including:
- a range of high-end handsets (running various OSs, but similar in terms of functionality) that benefit from an extensive and sophisticated developer community, using cross-platform compilation and application translation tools
- a significantly larger group of diverse handsets (not necessarily smartphones) that will have access to a more restricted set of moderately sophisticated applications for which investment by operators or handset manufacturers to overcome device fragmentation is financially viable
- an extensive range of less-sophisticated applications, available via a browser interface across an extensive range of handsets.
![Reblog this post [with Zemanta]](http://img.zemanta.com/reblog_e.png?x-id=167662cb-eb40-41c2-b7d0-5ff69298a674)
No comments:
Post a Comment