Look beyond the symptoms to find the real problems. When customers are calling the support line nonstop and complaining about a problem in code that you wrote, it is tempting to fix the symptom and not the problem. That is frequently the quick solution. Of course the intent is always to return to the real problem later and find a resolution, but you never do. That is human nature.
When you fix the symptom, you can actually make the problem worse—sometimes much worse. Fixing the symptom hides the problem. If you subscribe to "out of sight, out of mind" as a motto, this is a perfect solution. However, the hidden problem may harm the application in more insidious ways, which highlights the fallacy of this approach.
A man goes to a doctor’s office and says, "Doctor, it hurts when I raise my arm over my head."
The doctor replies, "Then don’t raise your arm over your head."
The doctor resolved the problem based on the symptom. That was quick and easy. However, tomorrow the man could drop dead. He had an undiagnosed broken arm with internal injuries and was bleeding to death. But the doctor never knew. This is the classic example of treating the symptom and not resolving the real problem.
Don’t try to fix a software problem. I have seen developers resolve a problem simply by trying a few things until the problem disappeared. However, this does not mean that the problem is fixed. Any observations made in this manner are anecdotal at best. A more scientific approach is probably advisable. If you are fortunate enough to fix the real problem with the "try" approach, consider yourself lucky. The stability of your application should not be based on luck.
Never make assumptions. Assumptions are almost always made from symptoms and not facts. By definition, if you had the facts, you wouldn’t be making an assumption. Invest the time to investigate software problems beyond the obvious and obtain the facts. Because of the Web, you have almost an infinite repository of information at your fingertips. You will most likely find someone on the Web who had the identical or similar problem previously.
