The focal point of many interactive systems is an information artifact being created and manipulated by one or more users through a user interface. The software components of such an interactive system perform their tasks relative to the data structures that represent the information artifact. System components interact with each other by changing these data and responding when relevant changes are made to them by other components. Perhaps the most difficult problem to be solved when building such data-centric systems is the monitoring problem. System components require the ability to watch for and respond to changes made to complex data structures. Previous monitoring approaches are geared toward monitoring single data items rather than entire data structures. This article describes an new monitoring approach called Nanites that is designed to simplify the task of monitoring complex data structures.

Conventional methods for maintaining telepointer consistency in shared windows do not work in the presence of per-user window customizations. This article presents the notion of a “smart telepointer,” which is a telepointer that works correctly in spite of such customizations. Methods for smart-telepointer implementation are discussed.

Abstract Multi‐user applications allows users in different locations to simultaneously interact with a common interface. Creating multi‐user applications is complicated by issues such as process synchronization, replicated window management, and floor control. As a result, several papers describing toolkits that simplify programming of multi‐user applications have appeared in the literature. This paper discusses several low‐level issues that must be addressed when implementing such a toolkit. Issues relating to replicated window management, supporting heterogeneous hardware environments, and multi‐user event handling are discussed.