Inference of shape type of demonstrations could be modified based on usage seen while at schools

[dhood]

Currently, the inference about which letter a demonstration was made for is done by combining information about the closest shape to the demonstration's location, and the 'active shape for demonstration.' The active shape for demonstration is used to break ties in the event that they arise, i.e. because a demonstration has been made equally-close (in terms of the ShapeDisplayManager's grid) to multiple shape types. Priority is given to the active shape for demonstration in this case, which by default is the most recently drawn shape. The active shape for demonstration may be manaully assigned, however, with a gesture (swipe with the finger on the tablet) at the position of another shape type.

At schools, it was seen that this inference scheme is not sufficient. On numerous occassions a demonstration was allowed to be sent to the controller which did not fit this protocol. It may be better to have the active shape for demonstration be used as the shape for demonstration all of the time, rather than just in the event of a tie between types. However, children were observed using the gesture capabilities for other purposes/accidentally, and this should be considered also.