Coding classes in Reading

The learner imagines the embroidery as 1,000 small details in a row and simulates two stages of copying, with rates we invented for the lesson. First the photographs: each detail has a half-per-cent chance of being altered, as the retouching did. Then the stitching: each detail has a one-per-cent chance of a slip. The program compares the finished copy two ways, against the photographs it was made from and against the original, and repeats the whole thing 5,000 times to average out the luck.

The table explains the pants story in numbers. Judged against its source, the replica looks like a fine copy with about ten slips. Judged against the original, it has about fifteen differences, a third of which were never the stitchers' fault. Blaming the last person in the chain for everything, as the often-repeated story did, is exactly what happens when you only compare the final copy with the original and skip the middle.

Then the learner extends the chain. If every copy keeps 99 per cent of details right, ten copies later about 90 per cent survive, thirty copies later about 74, and after around 69 copies only half are unchanged. That is why programmers keep one master source and copy from it, instead of copying the copies, and why version control records who changed each line. The rates here are ours; the lesson is how errors multiply and where to look for them.

Tracing a claim back to where it started is a strong reason for a Reading teenager to keep coding in 2026. The longer argument is in why teenagers should still learn to code in 2026.