No abstract

Workstations and personal computers are increasingly being delivered with the ability to handle multimedia data; more and more of us are linked by high-speed digital networks. With multimedia communication environments becoming more commonplace, what have we learned from earlier experiences with prototype media environments? This paper reports on some of our experiences as developers, researchers and users of flexible, networked, multimedia computer environments, or "media spaces". It focusses on the lessons we can learn from extended, long-term use of media spaces, with connections that last not hours or days, but months or years. We take as our starting point a set of assumptions which differ from traditional analytical perspectives. In particular, we begin from the position that that real-world baseline is not always an appropriate point of comparison for new media technologies; that a set of complex and intricate communicative behaviours arise over time; and that media spaces connect not only individuals, but the wider social groups of which they form part. We outline a framework based on four perspectives -individual, interactional, communal and societal -from which to view the behaviour of individuals and groups linked by multimedia environments. On the basis of our long-term findings, we argue for a view of media spaces which, first, focuses on a wider interpretation of media space interaction than the traditional view of person-to-person connections, and, second, emphasises emergent communicative practices, rather than looking for the transfer of face-to-face behaviours.

In this organizational overview we cover some of the critical aspects of human interfaee research and application at Apple or, as we prefer to call it, the "User Experience." We cover what we do, where we are going (as much as we are permitted to say in public), and how we are organized. Some of our innovations in the product process and in the transfer of research from the laboratories to product should be of special interest to the HCI community.

In recent years system engineers, product designers, and human interface designers have become increasingly interested in developing ways of involving users in the design and evolution of computer-based systems. Some have turned for guidance and inspiration to an approach to systems design pioneered in Scandinavia and often referred to as Participatory Design. In this paper we examine the development of a computer-based design tool, Trillium, which on the surface looked like an example of Participatory Design in that users were directly involved in the development of the technology. Our analysis leads us to conclude, however, that Trillium's development departed in critical ways from our current model of Participatory Design and to suggest that the manner in which users are involved in the development effort plays an important role in the success of the endeavor.

An interface is presented that is designed to help users switch among tasks on which they are concurrently working. Nine desirable properties for such an interface are derived. It is argued that a key constraint to building interfaces that support task switching is that low user-overhead switching among tasks requires a large amount of display space, whereas actual display space is limited. A virtual workspace design is presented that greatly speeds the inevitable task-switching induced window faulting. The resulting interface is presented as a study in theory-based human-interface design. It is shown how in this case theory is important in inspiring a design, but design entailments outside the theory raise new issues that must be faced to make the design viable. These design experiences, in turn, help inspire new theory.