Testing error feedback
Testing the user interface for errors is simultaneously far less frequent and far more critical than that of your primary use cases. It is less important because, hopefully, users encounter error conditions far more rarely. However, once they’ve hit an error, they will rely on your interface to correct them and help them achieve whatever goal they were aiming for. While these might be rarer cases, handling them well is vital for users who need assistance. This section considers how to present errors and internal failures, how to find spurious errors, and how to avoid displaying errors in the first place.
Preventing errors
As described in Chapter 7, Testing of Error Cases, your application should catch problems as early and as specifically as possible. If there are problems in the user input, then let the customer know exactly what is wrong – is the entered text too long, or does it contain invalid characters? Is a number field too high or...