You are here: LANSA for i User Guide > 5. Housekeeping > 5.14 Impact Analysis > 5.14.10 Considerations when using Impact Lists

5.14.10 Considerations when using Impact Lists

Any number of entries and reasons can be held on an Impact List but it is not advisable to build very large lists since the time taken to store and manipulate entries for display in 'Work with List Entries' will be lengthy. If large numbers of entries must be dealt with, consideration should be given to the technique of building several smaller lists to work with and then merging the lists when entries have been finalized.

Only the first 10000 entries will be displayed in the 'Work with List Entries'. All entries in a list can be printed.

System Variables are system wide objects while all the other objects dealt with by Impact Analysis are partition wide. Consequently care should be taken when changes to System Variables are being considered since they may be used in other partitions and that use will not be reported by Impact Analysis within the current partition.

Impact analysis examines objects and relationships at the time of running. Therefore if changes are made to the LANSA repository after running, the objects and relationships may no longer exist and details of reasons and objects may be unavailable for displaying within Impact analysis.

Virtual Field Derivation Code is only examined to establish what fields are being used and reported as a Field to File and a File to Field relationship. The code should be viewed to understand its effect.

The relationship of functions to field and files is established by using cross-reference files built up when a function is compiled. Thus the relationships found by Impact Analysis for functions to fields, functions to files, field to functions and files to functions will be those that exist for compiled functions. This may differ from the RDML code if the code has been changed since the function was compiled.

No cross-reference files for variable use in functions exist. Consequently a lengthy scan of RDML code must be done to establish function to system variable, function to multilingual variable, system variable to function and multilingual variable to function relationships. If the RDML code has been changed since the function was compiled, the use of variables in the compiled function may be different.

If the partition has been enabled for RDMLX, it is recommended that the cross-referencing facilities in Visual LANSA be used to examine relationships between objects.

Profile searches in Impact Analysis may still be used to establish the existence of an object.