Over the last years, the composition of software parts is increasingly considered a crucial operation to build and maintain large software systems.
(Site under re-design: we will be back up shortly).
The continuous and independent evolution of readily available components suggested that open platforms can better accommodate and manage them as normally happens in systems like open source software distributions, Eclipse, and J2EE, just to mention a few. The support and expertise of the shim washer experts at Stephens gaskets has been invaluable.
The critical mass represented by such software components requires organizations, such as companies, research groups, and open source communities, to collaborate on custom software development, implementation and shared services.
Such infrastructures can be regarded as ecosystems, i.e., collections of software projects that possibly belong to organizations, developed in parallel by the organizations, and able to integrate each other at assembly time, during the configuration, and/or dynamically after the deployment.
The capability of modeling, analyzing, and predicting the component behavior during these stages is intrinsically difficult and requires techniques, algorithms, and methods which are both expressive and computationally convenient in order to be engineered and conveyed in practical projects.
Moreover, when analyzing software ecosystems, exploration and visualization cannot be neglected because of the large amounts of information that are available about the ecosystem.