One action leading to another

These texts were created for this question and concern imaginary workshop projects.

Text One: A participant's account of building a chain-reaction machine.

Our section began with a wooden ball rolling down a channel. It struck a tile, which pulled a thread and released a paper flag. We tested each action separately until it seemed dependable. However, when our section was joined to the next group's, the final tile did not push their lever far enough.

We raised one end of the channel and tried the combined sections again. This time the machine completed its whole sequence, and everyone cheered. The workshop leader immediately asked us to reset it and repeat the test. On the next attempt, the ball missed because the channel had shifted while we were putting the pieces back. We marked its position on the table before trying again.

At the exhibition, visitors saw a sequence that lasted less than half a minute. Preparing for each run took much longer. Every ball, tile and thread had to return to the correct place, so we followed a checklist rather than relying on memory.

One public run stopped at our flag. I moved the thread by hand, and the remaining sections continued. We recorded the intervention instead of claiming that the machine had worked without help. The stoppage showed us exactly which connection needed further adjustment.

Text Two: A report on a different chain-reaction project.

Several teams built separate sections of a chain-reaction machine across a row of tables. Early tests showed that most sections worked on their own. Once connected, however, movement from one table shook a neighbouring ramp, and a ball was released before the receiving section was ready.

The teams agreed common positions for every connection and added supports beneath two ramps. A numbered reset list was also introduced. One student prepared the machine while another checked each stage, reducing the chance that a piece would be overlooked.

The complete machine was tested five times. Four sequences reached the final bell; in the other, a ball left its cup and a student placed it back. Observers noted where delays, early releases and human assistance occurred. The report did not treat the loud final bell as proof that every earlier stage had performed reliably.

For the exhibition, the teams kept the individual sections simple enough to reset quickly. They also stationed a student beside each table so that a stopped sequence could continue. The report concluded that repeated testing and accurate records were more useful than presenting one successful run as though it represented every attempt.

Synthesise what you learn from both texts about building and testing chain-reaction machines.