Visual Basic 2012 Programmer's Reference

SUMMARY

In practice, it’s extremely difficult to anticipate every condition that might occur within a large application. You should try to predict as many incorrect situations as possible, but you should also plan for unforeseen errors. You should write error-checking code that makes bugs obvious when they occur and recovers from them if possible. You may not be able to anticipate every possible bug, but with a little thought you can make the program detect and report obviously incorrect values.

You should also look for unplanned conditions (such as the user entering a phone number in a Social Security number field) and make the program react gracefully. Your program cannot control everything in its environment (such as the user’s actions, printer status, and network connectivity), but it should be prepared to work when things aren’t exactly the way they should be.

When you do encounter an error, you can use tools such as breakpoints, watches, and the development environment’s Locals, Auto, Immediate, and Call Stack windows to figure out where the problem begins and how to fix it. You may never be able to remove every last bug from a 100,000-line program, but you can make any remaining bugs relatively harmless and appear so rarely that the users can do their jobs in relative safety.

Chapters 7 through 12 focused on controls, forms, and other user interface objects. Chapters 13 through 17 focused on the code that lies behind the user interface. Chapter 19, “Database Controls and ...