The ever expanding role of window based human-computer interfaces, while leading to dramatic improvements for users with normal vision, is in fact detrimental, and often highly so, to the millions of users who are visually impaired but not blind. Indeed, current trends threaten to widen rather than close the gap between normal and visually impaired users, to the extend that the latter group may soon literally find themselves locked out of the community of users almost entirely. The goal in this research is to prevent this from happening, by developing concepts and implementing a prototype for an alternative to the window-based interfaces of today which will allow the visually handicapped to become first-class computer users.