The S1000D specification defines the Common Source Database (CSDB) as “an information store and management tool for all objects required to produce the technical publications within projects”[[4]]. In practice that means that it is the repository for all data modules and publication modules plus any graphics or other multimedia used. It must also contain supporting management files such as Data Dispatch Notes (DDN) and Data Management Lists (DML). The specification explicitly does not define how a CSDB should be implemented.

The most important concept behind the CSDB is that it is a single source for content and should not contain any duplicated data. Data modules should be sufficiently granular in their content to allow re-use in multiple outputs. The word ‘module’ in ‘data module’ and ‘publication module’ has been carefully chosen. Publications can be built up of many modules. There is nothing to stop different publications re-using the same data module, in fact this is positively encouraged. It means authors only need to update content in one location and multiple publications will contain the revised content.