Application of data-centered approach to Year 2000 problem

The data-centered approach uses variable classifications, dependence analysis, generalised program slicing and ripple effect analysis. In the Year 2000 problem, this approach can be useful. Variable classification is used to identify all input and output variables, and once the initial set of variables that are Year 2000 related is identified, dependence analysis can be used to identify all the variables that are potentially affected by the initial set of variables. The second set of variables is then examined to see if they are Year 2000 related. This process is repeated until all the Year 2000 related variables are identified. This process is essentially ripple effect analysis and uses generalized program slicing and dependence analysis. Once Year 2000 related variables are identified, generalized program slicing is performed to identify all the statements that potentially need to be changed. Once a statement is changed, it may induce additional changes. Ripple effect analysis can be used to ensure that all the parts that need to modified are examined. Finally, the changed software should be validated and regression testing can be used in this stage. Ripple effect analysis can be used in this stage by identifying the relevant test cases that needed to evaluated. This is done by maintaining traceability links between the software and its test cases