Abstract: The subject disclosure is directed towards a graphical or printed keyboard having keys removed, in which the removed keys are those made redundant by gesture input. For example, a graphical or printed keyboard may be the same overall size and have the same key sizes as other graphical or printed keyboards with no numeric keys, yet via the removed keys may fit numeric and alphabetic keys into the same footprint. Also described is having three or more characters per key, with a tap corresponding to one character, and different gestures on the key differentiating among the other characters.

Abstract: A dedicated text prediction key for inserting a text prediction candidate is described. The text prediction key may be configured and/or output for display in conjunction with an on-screen/adaptable keyboard when a prediction for entered characters is detected. The text prediction key provides a mechanism for explicit insertion of a text prediction candidate that can be employed to avoid passive/non-explicit insertions. In at least some embodiments, the text prediction key may be selectively displayed by replacing/modifying one or more keys of the on-screen keyboard. In another approach, the text prediction key may appear at a designated location within or proximate to the on-screen keyboard. The text prediction key may also be configured in various ways to notify a user that a prediction candidate is available.

Abstract: Font/script association techniques are described. In an implementation, a client includes one or more modules to determine which of a plurality of scripts are to be used to display characters and one or more fonts are located that correspond to the determined script. Representations are output of the located one or more fonts in a user interface that are selectable to cause the characters to be displayed using a selected one of the fonts.