The Ajax application development environment has gained considerable attention in the enterprise IT world, particularly in light of the support of IBM for the Open Ajax movement, but its potential as a mobile platform, and a possible challenge to the seemingly unstoppable Java, has been less discussed. One significant boost to mobile Ajax appeared earlier this year ‘ and was widely ignored. This was the incorporation of Ajax support in the Opera Platform, which aims to extend the influence of the Opera browser in the mobile world by putting it in the forefront of a possible shift to a browser-only client.
As in the PC world, the practicality of using the browser as the universal mobile client is increasing ‘ though problems persist that are peculiar to the mobile device, such as screen format. However, the advantages of the browser client ‘ universal access from different devices and locations, support for web services and so on ‘ are ensuring that considerable work is going into improving the user experience, especially as mobile systems that are geared to open internet access rather than carrier-controlled ‘walled gardens’ evolve. On cellular networks, Java’s mobile incarnation, J2ME, is emerging as the dominant application development and content delivery vehicle, but a browser-only model could support devices with smaller footprints and price tags.
Opera, whose browser has a tiny share of the PC market, has always done better in the mobile world against Internet Explorer and others, and now believes it could ride the wave of interest in browser-only mobile clients to gain a significant market position. Its Opera Platform comprises its browser running in full screen mode; access to the phone’s native functionality through an abstraction layer; and an Ajax framework for running multiple applications and ‘widgets’ (a concept common in the Mac world, and adopted by Ajax, which revolves around downloadable, interactive software objects that provide a single service such as a map). Ajax provides a mechanism to implement widgets on browsers.
Ajax is supported in the 17m Opera mobile browsers shipped last year, enabling users to access desktop services such as GMail and Google Maps. Not only is the Opera Platform the first mobile Ajax framework, but it also has a browser framework that, unlike others such as Backbase, is fully designed for mobile units, using the same code base on the browser and device. Desktop widgets can also run on the mobile device with minimal recoding, a major attraction for developers. Other advantages are that widgets can call other widgets, allowing for the creation of complex applications from simple browser-based components, and the Opera Platform allows access to device application programming interfaces, for better integration with legacy functions. This last is essential to make mobile Ajax really useful, since it enables access to data elements that are tightly coupled to the device itself, such as text messages, phone book and the SIM card. In the past, few browser-based platforms have been able to access these functions.
Ajax is not a pure browser environment, since the client does need some form of local software, although this is installed by the manufacturer rather than downloaded or plugged in. And for now at least ‘ though perhaps Open Ajax and other nascent standards moves will help ‘ Ajax widgets are not transferable across multivendor platforms. Also, in its current form Ajax is nowhere close to Java for rich content applications such as games. But it does offer a consistent and simple set of tools for developers and, once supported by the device manufacturer, a cheap and flexible way to create and update applications ‘ to the extent that very small user groups, with specialized requirements, can be economically catered for. This will be of particular interest to the MVNO virtual operators, many of which are very reliant on targeting ever more specific user bases.
Certain key factors are needed before mobile Ajax will be a real challenger to Java downloads. The primary ones are:
‘
Standards support, notably from the OMA (Open Mobile Alliance). This will come with increased vendor and operator interest, and we would expect Nokia, for one, to take a keen interest in mobile Ajax and in Opera Platform in particular. Nokia is increasingly pushing the internet tablet device, such as its 770 Linux-based, non-cellular product, as a universal enterprise client, as part of its challenge to the monopoly of Windows in this arena. Mobile Ajax would fit with the strategy very easily, and Nokia already supports the Opera browser. IBM will be another important influence as it seeks to dominate the mobile enterprise environment through its key software technologies, with Open Ajax likely to join WebSphere and Java on its list.
‘ Greater bandwidth and local storage will also be essential on the devices to support serious browser-based activity.
‘ The user experience remains far better with Java downloads for now, but as the browsing architecture changes, the gap is narrowing. The main problem has been fragmentation, with many different approaches to critical issues such as user interface and navigation. The potential of Ajax is to provide a unifying force as Java has in downloads, which in turn expands the appeal to developers and the base of applications available ‘ especially as, because the level of abstraction shifts to the browser, applications are easier to update and target to small groups. Ajax has the chance to foster the creation of a widget authoring market such as exists in the Apple world.
Mobile Ajax is also helping to solve some of the user experience disadvantages that plague mobile browsers, with significant progress in data management that will be aided by large vendor input, and superior widget authoring capabilities.
The Ajax platform-led approach closes the gap between browser-based and downloaded applications by improving the user interface and offering integration with device capabilities, and driving down development costs. Ajax is the most open source environment choice and has a wider range of deployment models than Java, and far lower porting costs between operator platforms, even though Java is currently richer and better specified. Other alternatives include Symbian/WinMobile, which offers more phone integration, but is not portable and is entirely proprietary. Adobe Flash is another option, but can actually work with and enhance, rather than compete with, Ajax.
With the operators’ focus shifting to low cost, flexible deployment of frequently updated applications, rather than costly development of a few ‘killer apps’, the browser approach will gain in appeal, especially since it offers the promise of creating one app for all platforms (though as we have seen, Ajax does not yet deliver this entirely). Java will remain the platform of choice for rich content applications, notably games, but for more ‘casual’ apps, Ajax will be a cheaper alternative. In the end, much of the relative success of these two platforms will depend on support of large vendors ‘ and how Microsoft responds to either system ‘ and on how far Ajax is taken up on the enterprise desktop. In the mobile enterprise, there will be strong reasons to support the same environment on the mobile device as is dominant on the desktop. For inherently IP/web-oriented products like the internet tablets and Wi-Fi or WiMAX phones, mobile Ajax will have many advantages, and may be the kickstart Opera needs to take a key position in that browser world.