DATA CREEP

From Test Automation Patterns
Jump to navigation Jump to search
.................................................................................................................Main Page / Back to Process Issues / Back to Test Automation Issues

Issue Summary

There are countless data files with different names but identical or almost identical content

Category

Process

Examples

  1. Nobody knows what is being used and where, so nobody wants to be responsible for deleting eventually needed data
  2. To edit or remove the the data files is too much work: one has to look up all the places where they are used and change the referrals. If the files are similar rather than identical, a unified file has to be created.

Questions

Is the data documented?
Are there standards regarding naming and documentation?
Who creates the data? How? Who uses it?

Resolving Patterns

Most recommended:

You should already be applying these patterns. If not, do it!

Other useful Patterns:

  • GOOD DEVELOPMENT PROCESS: apply this pattern if you don't have a process for developing test automation. Apply it also if your process lives only on paper (nobody cares)
  • LEARN FROM MISTAKES: apply this pattern to turn mistakes into useful experiences
  • KILL THE ZOMBIES: Apply this pattern for a start
  • DEFAULT DATA: use this pattern if your tests use a lot of common data that is not relevant to the specific test case
  • DOCUMENT THE TESTWARE: you should be already applying this pattern. Retrofixing documentation is quite an effort. Do it in the future for all new projects and everytime you have to update something old
  • KEEP IT SIMPLE: Always apply this pattern!

.................................................................................................................Main Page / Back to Process Issues / Back to Test Automation Issues