T he emergence of the Internet has enabled interactions between people working in the same field across the globe. This has resulted in a number of efforts to solve the problems of supporting group communication, and providing access to information sources that are geographically and administratively scattered. Collaboratories are one way to provide virtual shared workspaces to users. We define a collaboratory as a wide area distributed system supporting a rich set of services and tools that facilitate effective synchronous and asynchronous communication between two or more people who are not co-located. The Upper Atmospheric Research Collaboratory is an Internet-based system that gives space scientists a virtual shared workspace for conducting real-time experiments as well as various asynchronous collaborations from geographically dispersed facilities. 1 During the past two years, UARC has integrated a significant fraction of the worldwide observational systems devoted to space physics and aeronomy. (Figure 1 shows the sources available to the last data-collection campaign in April 1998.) Moreover, the system now supports real-time outputs from supercomputer-class theoretical models. This has enabled simultaneous comparison of experimental and theoretical data on a global scale. Such comparisons would not be possible without collaboratory technology. UARC has evolved into a scalable Web-based collaboratory over a period of six years. Since its inception in 1993, UARC has supported more than 10 scientific campaigns, each of which represented a concerted effort to collect real-time data simultaneously from instruments around the world for several days at a time. The first few campaigns ran on a distributed set of Next systems on the NSFNET and could support about a dozen active par

Although Gigabit Ethernet has been wrongly perceived as a competitor to asynchronous transfer mode (ATM), each has its own place in the grand scheme. This perceived competition has caused users to hesitate in their migration to ATM. This study presents a reality check on the two technologies. It addresses the pros and cons of selecting each network based on some criteria.

No abstract