There were two engineers who had to work on the same product.
Dave was an expert in server code and Francine was an expert in client development.
Naturally, the work that Francine did affected Dave by influencing how the system he maintained would be used. Vice versa, Dave's work affected Francine by altering how a dependency of her code behaves.
They decided to meet once a week to make sure they were staying in sync.
After the first week, they had their meeting and discovered that each had made many assumptions about what the other was going to do. They came up with a list of changes and spent the first day of the next week resolving the discrepancies.
This continued for a month; each Monday was spent reconciling the changes made in the previous four working days.
"Francine," said Dave, "this is costing us a lot of time. Maybe we should only do it once every other week."
"That's a good idea," said Francine.
Two weeks later, the meeting and the changes required to integrate totaled three days.
After a quarter of consistently spending thirty percent of their time integrating their work, Francine said "Maybe we should only be doing this once a month."
The cost of resolving their changes again increased. In any given twenty-day period, a full ten working days were consumed by the integration effort.
Dave and Francine went to their manager and told him of the problem. The manager suggested that they integrate twice as often.
Francine looked horrified and Dave's jaw dropped open.
After a long pause, the Dave finally spoke.
"Are you crazy?" he asked, rhetorically. "We already spend half our time integrating our changes."
"So?" asked the manager?
The two developers exchanged looks.
Eventually, Francine explained. She spoke very slowly, as if to a child "We spend half our time integrating our changes. If we do it twice as often, we won't have any time left."
