Determining what to check

So far in this chapter, we have considered many different tests to run, but we haven’t covered the vital next step in the test process—what should you check to ensure the test passed? There are many levels to consider. Most superficially, you can tell whether the application continues to work as described. In our example of a signup web page, does the user receive an email after entering their address? For everything written and saved, you can check where it is subsequently displayed to the user. Is it correct in all those places?

For changes that transition your program into a new state, does it make that transition correctly in all cases? For inputs that trigger other changes in the application, do you have a complete list of those changes so you can check them all?

The feature specification should include the user experience for each output, so you’ll need to refer back to that extensively for these checks. If anything is missing...