Head-mounted displays for virtual environments facilitate an immersive experience that seems more real than an experience provided by a desk-top monitor [18]; however, the cost of head-mounted displays can prohibit their use. An empirical study was conducted investigating differences in spatial knowledge learned for a virtual environment presented in three viewing conditions: head-mounted display, large projection screen, and desk-top monitor. Participants in each condition were asked to reproduce their cognitive map of a virtual environment, which had been developed during individual exploration of the environment along a predetermined course. Error scores were calculated, indicating the degree to which each participant's map differed from the actual layout of the virtual environment. No statistically significant difference was found between the head-mounted display and large projection screen conditions. An implication of this result is that a large projection screen may be an effective, inexpensive substitute for a head-mounted display.

This paper describes an architecture for supporting interface attuchments -small interactive programs which are designed to augment the functionality of other applications. This architecture is designed to work with a diverse set of conventional applications, but require only a minimal set of "hooks" into those applications. In order to achieve this, the work described here concentrates on what we will call observational attachments, a subclass of attachments that operate primarily by observing and manipulating the surface representations of applicationsthat is the visual information that applications would normally display on the screen or print. These attachments can be thought of as "looking over the shoulder of the user" to assist with various tasks. By requiring very little modification to, or help from, the applications they augment, this approach supports the creation of a set of uniform services that can be applied across a more diverse set of applications than traditional approaches.

Essentird prerqnisites to asynchronous work with shared artifacts include things such as an abfity to effectively communicate information, an abtity to understand the actions of co~aborators, and an abtity to integrate work from others. Systems designd to support ubiquitous co~aboration -co~aboration that can scale to communities the size of the ktemet -face a number of important chtienges in providing these prerequisites. For example, when the set of potential co~aborators becomes large, and co~abomtive media becomes richer, stiple interopembfity of application pro=~qnicMy becomes a diffictit issue. Further, various market pressures, along with the rapid wowti of a diverse kteme~~, for the most pm make these problem worse rather than better. This paper considers a new strategy for asynchronous coMaboration that is designed to support large-sale, u~lquitous, co~aboration through shared artifacts. This strategy is based on hmn comdle su$ace represe-m -representations of information that are designed to be direcdy understood by people. Printing on paper is an important example of such a representation. h the work presented here, we concentrate on 2D images or pictures (what wa be ctied pictorial su#aces, or simply su~aes) as one such representation. A prototype system that supports surface-based co~aboration Id be described. This system supports a co~aborahve process inspird by paper-based co~aborations, and is designed to demonstrate concepts of surface representations as a first step towards utilquitous collaboration. Ke~ordsAsynchronous co~aboratio% ubiquitous co~abomtion, shaed artifacts, pictorird surfaces.

We describe an architecture which allows any external agent (human or software) to point into the visual space of an interactive application. We describe the visual design of a scheme for highlighting any information in any application. This architecture requires the application to provide information about its semantic structure as part of its redraw algorithms. Based on this semantic map generalized pointer descriptions are defined and used to reference objects to be highlighted. The architecture is demonstrated using a multibookmark agent framework and several example applications. THE MULTIAGENTPOINTING PROBLEM