MeeGo Touch Input Method Frameworks

Copyright: Linux Foundation, 2010
More info: http://conference2010.meego.com/session/meego-touch-input-method-frameworks

MeeGo Input Methods provides a pluggable framework for virtual keyboards and other means of input. It comes with core modules for text input. Throughout the session, we will demonstrate the flexibility of the MeeGo Input Methods by showcasing selected plugins on different devices, running with different toolkits. This session will mainly target developers interested in creating input methods for additional languages, writing systems and situations as well as application developers who want to use all features of MeeGo's on-screen keyboard or other input methods. The planned demos and the introductionary nature will make the session suitable for anyone with a general interest in input methods such as on-screen keyboards. We will give an overview of the MeeGo Input Methods architecture but focus primarily on the features that are relevant for input method developers.


Abstract:
User Experience

MeeGo Input Methods comes with a consistent look&feel for many types of MeeGo applications. Instead of switching to a dedicated fullscreen mode whenever input methods are activated, the application remains fully interactive. The MeeGo Input Methods also ensure that the focused widget remains visible, easing the typing of longer texts. We will explain how developers of input methods can support this.

Applications often want to customize specific keys dynamically. For instance, if the enter key triggers a default action, then this can be easily expressed by replacing the enter key's icon with a custom label such as "Send e-mail", "Close", etc. But by dynamically customizing other keys the application can do more. We will demonstrate some use cases and explain the API that input methods must support for this feature.

MeeGo Input Methods also supports smooth orientation change animations whilst input methods are actively displayed. For that, input method developers need to support a minimal synchronized orientation change API.
Feedback

Haptic and tactile feedback - in addition to visual feedback - can improve typing accuracy for virtual keyboard plugins. To reach the lowest possible latency in providing that feedback, we use colored reaction maps. Thus, the haptic and tactile feedback can often be faster than the actual screen update. We will explain how input method developers can make use of this technology.
Predictive On-Screen Keyboards

Word prediction and error correction has become an essential item for input methods on mobile devices. With the possibility to use custom prediction and correction engines it becomes feasible to implement predictive on-screen keyboards.
Open-Source

MeeGo Input Methods is open-source and part of MeeGo Core. We will explain our goals and give a status report regarding our ongoing efforts to work more openly. This will target developers interested in actively contributing to the MeeGo Input Methods.