Creating own gesture recognizer is vaguely described in the documentation. The documents talk about state machine and give hint that the gesture recognition should follow the QGesture state machine (Qt::GestureState), but neither of the state machines - Qt::GestureState or QGestureRecognizer::ResultFlag - or their state transitions are not defined in the documentation. Only their states are defined, but some of them leave open questions. For example description of QGestureRecognizer::MayBeGesture contais "Gesture recognizers in the MayBeGesture state may be reset automatically if they take too long to recognize gestures.", but how long is too long and what happens when the state is automatically reset.

Hence there is a need for improvement. The lack of proper documentation causes a lot of "trial and error" development and makes quick prototyping harder.