I am currently building a test plan for the system I am working on. The plan is 5,000 lines long and about 10 years old. The structure is like this:
1. test title precondition: some W needs to be set up, X needs to be completed action: do some Y postcondition: message saying Z is displayed 2. ...
What is this type of testing called ? Is it useful ?
It isn't automated.. the tests would have to be handed to some unlucky person to run through and then the results would have to be given to development. It doesn't seem efficient. Is it worth modernising this method of testing (removing tests for removed features, updating tests where different postconditions happen, ...) or would a whole different approach be more appropriate ? We plan to start unit tests but the software requires so much work to actually get 'units' to test - there are no units at present ! Thank you.