Fable
Test pits and test river sections
A new engineer named Zheng arrived at the water bureau. His first task was to judge whether a river was navigable. Engineer Zheng dug a three-meter-square pit on the bank, put a model boat in it, and measured buoyancy and stability. The data was good, so he wrote a report: the river meets navigability conditions.
The bureau chief read the report and approved three cargo ships for trial voyages. As a result, the first ship ran aground less than two kilometers out. There were hidden reefs on the riverbed, whirlpools on the surface, and fishermen casting nets along both banks. None of this could be measured in Engineer Zheng's puddle.
The bureau chief was very angry and called Engineer Zheng in for questioning. Engineer Zheng explained: the water pit test verifies whether the boat can float and whether it is stable — this is called the test water pit. But whether the boat can run an entire stretch in a real river channel requires another kind of test — choose a real river section, have the boat run it fully loaded once, and measure the current, the hidden reefs, the docking, and the loading and unloading of cargo. This is called the test river section.
The director calmed down and understood the difference. A test pit is cheap and fast, suitable for quickly screening out unreliable designs early in a plan. But a design that passes the test pit does not mean it can go straight to navigation. There must be a trial river section in between—run under real conditions, at reduced scale, but the environment cannot be simplified.
From then on, the Water Resources Bureau set the rule: every navigation project would be accepted in two steps. The first step was the puddle test, verifying core technical metrics in the lab; pass it and you move to the second step. The second step was the river-section test: pick a representative stretch of river, with real loading, real scheduling and real weather, and run a full operating cycle. Only when the data passed would a formal navigation approval be issued.
The next year, Engineer Zheng led the evaluation of a new route. The puddle test took only two weeks and screened out two unreliable options out of three. The remaining option ran for three weeks in the river segment test and uncovered five problems that could never have surfaced in the lab. After all were fixed, the official route ran the full distance smoothly on its first day.
At the wrap-up meeting, Engineer Zheng said: the test pit helps us save money, the test river section helps us stay alive. Both matter, but neither can replace the other.
Later the county magistrate had the mechanism first run in a real warehouse for three weeks, recording how many times it broke, who repaired it, and whether the cargo owners were willing to keep using it. Only then did the excitement become a pilot.