摘要 |
A solution for distributing software products in a data processing system is disclosed. A set of models of each software product is to be distributed (for example, of the UML type) is provided; each model includes an indication of relationships among a plurality of available logical items (representing functions provided by the software product, such as use cases) and a plurality of available distribution artifacts (used to distribute the software product). A set of logical items is then selected among the available ones. The process continues by identifying a set of distribution artifacts required to implement the selected logical items (among the available distribution artifacts); the required distribution artifacts are identified according to the corresponding relationships in the models. A software package is built including the required distribution artifacts. The software package is then distributed to a set of data processing entities of the system (i.e., endpoints), so as to cause each entity to install the required distribution artifacts on the entity.
|