We present an overview and demonstration of open-source tools and technologies used to make large-scale computing connect readily to client visualization environments, bringing together multiple data sources to compose analyses in the McIDAS-V environment. McIDAS-V permits novel manipulations of atmospheric datasets distributed across the network using 3-D graphics and a highly literate data model implemented in Java. When coupled with plug-ins permitting it access to web services, cluster and grid computing can be made both easy-to-use and scriptable. Outputs can be sliced, subsetted and integrated into visualizations and further computations. This technology demonstration is intended to evolve into a toolkit and best practices for integrating heritage data processing applications with distributed computing and visualization.