Testbed Application Specification

Required:

A small application to test various aspects of interface design on a variety of users.



A simple disk-space navigating image viewer:

This application should provide a good range of interface principles. The program will take two forms: a realworld form and an abstract form.

The realworld form:

The user is immersed into an environment that models everyday surroundings. A shelf will hold various books which, when selected, can be paged to quickly browse the images, which can be selected for display in the picture frame (see Fig 1 below).


Fig 1
Sketch of the realworld application

  • Selection

    It is planned to keep things simple to avoid confusion about what part of the interface is effective. To select a book or image the hand is used to grab the object, and then it is moved to the relevant place.

    Paging through the book is done by grabbing a blank part of the page.



    The abstract form:

    This forms provides the user with a symbolic representation of disk space. This can either be an amalgamated view (a 3D grid landscape) or a network view (connected nodes with 3D grid landscapes alongside). Figure 2 shows a sketch of the network view. The amalgamated view would have a large single disk space block. Clicking on a block of the grid brings up a container which is used to browse the images or select an image to be displayed in the frame (which is similar to the frame in the realworld application).


    Fig 2
    Sketch of the abstract application



  • Selection

    As with the realworld version simple selection of images, blocks in grids, nodes and containers is done by grabbing with the hand.

  • Proportionality

    The picture frame is envisaged as a fixed structure, whereas the user can move around the collection of nodes to get to different disk spaces (by grabbing a node or possibly using a seperate movement widget or hand action).



    What we get out of all this:

    Hopefully, a better understanding of the 3D human computer link. The basic interactions - grabbing and placing an object, or moving around the network - should give an idea of the best ways to do the simple operations in a 3D environment.

    The two different interfaces will allow comparison of the effectiveness of representations which are drawn from the real world with those that are drawn from imagination and more abstract associations.

    Switching between these forms (which will be added afterwards) will provide a mechanism to test meta-level controls, which are an important part of the operating environment.

    Insight into control within an application may be gained from the mechanism for moving between different views in the abstract version.

    Possible extensions:

    The program could allow movement of images between containers.

    Various settings, such as grid or book colour, could provide more diverse interaction.



    Matthew Mundell
    Email with suggestions or comments welcome.

    Last updated 23 March 1999