When determining a solution to a problem, I have found it invaluable to ask "Are we solving the root problem?". The first problem that is observed is typically the symptom, the surface problem, it is the observed affects of the root problem. The root problem is usually deeper down and requires further analysis.
When a problem arises, a motivated team is quick to jump to solutions. For example, I was recently on a project where two test environments operated on vastly different dates - often years apart. We had operations that spanned both test environments and in order to complete our integration testing we needed a way to convert dates between the two environments so our web service calls and batch jobs could operate seamlessly.
To solve this problem, our team built a small utility to intercept data between the two systems and modify the date. Within a few days, our utility could handle date conversion for all of our interfaces. We had solved our problem. Or so we had thought.
A few weeks into testing it became clear that our little utility wouldn't be able to run without manual intervention, the two systems moved at unpredictable rates. That is, one system would move one day at a time, another would jump months. To solve this next problem, it would require a much more intelligent and complicated tool to calculate the conversions. So instead, we pushed on and used the tool for part of the operation and manual calculations for the other parts. Manual intervention of this form was error prone, difficult, slow and tedious.
The problem we had solved (the dates are different) was only a symptom, a surface problem the root problem still existed. The two test environments are still operating on different dates and our solution did not address that, it worked around it. By solving the symptom and not the illness we left ourselves vulnerable to other symptoms, in this case the other symptom took the form of the variability in the rate that time moved.
A solution that does not solve the root problem is called a band aid solution it covers the wound today, however you can expect it to bust open tomorrow and cause more pain later.
In a perfect world, we would always solve our Root Problems, however as we know the world we live in is far from perfect and sometimes band aid solutions are good enough - the hardest part is knowing where to draw the line. In the example shared above - it turned out that fixing the root problem was too difficult and beyond the expertise of our team, it became a cost to our development process that we had to manage.
