Visualization '91 workshop report: scientific visualization environments

• Authors: David M. Butler;Charles Hansen
• Venue: ACM SIGGRAPH Computer Graphics
• Year: 1992

The Workshop on Scientific Visualization Environments at Visualization '91 was organized with several purposes in mind. First and foremost was to bring together widely dispersed researchers so they could share information in a relatively informal and interactive environment. The organizers believed this would establish working relationships between the participants, which was a second purpose. It was hoped that a workshop could generate a focus and identify important research directions in the field. Finally, we believed the workshop report could provide a useful overview to a broader audience.With these purposes in mind, we established two specific objectives: (i) review the state of the art and (ii) identify key research issues and directions. Reaching these objectives in a short two day workshop would require an effective mechanism for organizing the discussions. The organizing principle we chose was a software design principle suggested by Dijkstra [ 1 ] some twenty years ago: the notion of an abstract machine. We considered a visualization environment as a special purpose, very high level computer ideally suited to the needs of scientific visualization. We then organized the workshop discussions around the questions of what the memory (data types), instruction set (operations) and control mechanisms (programming interfaces) of this machine might be.Using this organizing principle, we established the following agenda: Monday morning: position statements. Each participant presented a ten minute summary of their position paper Monday afternoon: data types and interfaces What are the (abstract) data types on which a visualization environment is based? What are the interfaces for importing and exporting data from the environment? Tuesday morning: control structures and architecture What are the mechanisms for sequencing and combining the primitive operations provided by the data types. What are the interaction and programming interfaces to the environment? Tuesday afternoon early: reference model Do we understand visualization environments well enough to construct a useful general model of their structure and function? Tuesday afternoon late: other issues and summary What issues are not addressed in the other sections? What have we accomplished? For each section we followed a simple procedure. The organizers presented introductory remarks consisting of a brief review of the state of the art and identification of research issues. The introductory remarks served as a starting point for discussions which followed. Participants were given copies of the organizers' remarks and asked to take notes for contribution to the report. The notes were collected at the end of the workshop and collated into this report. In addition, participants reviewed the first draft of the report and made additional contributions.