In the previous chapter, you learned about troubleshooting and how it can help us identify problems in our code. However, the biggest question is--How would you define a problem? If you don't have a performance measurement checklist that shows you what to expect from the program and doesn't collect supporting data that can be used to validate the checklist, then you will not be able to define the performance problem properly. And if you don't define the problem properly, then what will you fix? It will be like the following experience published in Reader's Digest:
My mother was rushed to the hospital following a serious tumble. There the staff placed a band around her wrist with large letters warning: Fall Risk.
Unimpressed, Mom said to me, "I'll have them know I'm a winter, spring, and summer risk too."...