User:Kbcmdba/sandbox

Reference Data Set

A reference data set (RDS) is used in software development to indicate the minimum amount of data structure (DDL - data definition language) and supporting data (DML - data manipulation language) to allow an application to operate successfully. It does not contain any test data or test fixtures. A reference data set is typically a stand-alone piece of scripted or pre-compiled code that is "played" into a persistence layer (i.e. a database) allowing development, QA, and verification teams to test the functionality of the software when that reference data set is combined with one or more test fixtures.

A test fixture is a scripted or pre-compiled set of data (mostly DML) that is applied on top of a reference data set. A test fixture should be created to limit the scope of the fixture allowing the user of that data to properly verify a specific functionality or short list of functionalities against a very limited set of test cases.

How do Reference Data Sets and Test Fixtures work together in a software development environment? Ideally, prior to the start of a project, developers code up the data structures and supporting data into files that get "played" into the persistence layer (database for example). This helps developers because the persistence layer can be "reset" at any time to a known standard. It also helps testers (QA, verification) because they too can reset a data set they're using to test with.