Fable
The weaving workshop wove every thread change into a sample cloth
The Hexi weaving workshop wove sample cloth for trading firms. At first there was only one pattern, and the master changed threads, adjusted the shuttle and inspected the cloth all at the same table, managing quality by eye.
Later the trading company demanded more variety: some wanted speed, some wanted low cost, some wanted patterns modified from old ledgers. The weaving workshop changed threads, shuttles, and blueprints at the same time, yet the quality of the finished goods was hard to explain.
The manager's naive approach was to have every workstation paste the latest drawing on the wall. The walls became more and more covered, and no one could say clearly which drawing produced which bolt of cloth, or which thread change increased rework.
After a distinguished guest returned goods once, the weaving workshop discovered that the same pattern had been tested with seven kinds of thread, three kinds of shuttle, and two sets of cloth inspection methods. In the end only the finished product remained; the whole process was scattered in the masters' oral memory.
A pattern weaver kept a swatch book. Each trial weave recorded the input draft, yarn used, shuttle speed, rework, inspection score, and a photo of the finished cloth; different versions could be compared side by side.
He also turned common customer orders into fixed test questions. When changing blueprints, lines or fabric inspectors, they would first run an old customer order to see how speed, cost and pass rate changed. The weaving workshop later brought the sample fabric book to customer review meetings. When a customer complained that a batch of fabric was slow, the manager could point out whether the slowness was in fabric inspection or line changes, and what should be changed in the next round. Quality discussions finally returned from feelings to evidence.
The shopkeeper saw for the first time that a certain thread tended to break only in long patterns, not that quality was poor everywhere; that a certain design looked clever but made it harder for the fabric inspector to verify. The sample swatch book also helped newcomers avoid detours. In the past, when a master said "be careful with this kind of customer order," newcomers didn't know what to be careful about; now they could see which trial weave failed and which thread was changed before the failure.
Later the weaving workshop also put customer feedback back into the sample cloth book. Online issues were no longer just a complaint, but could be mapped to a trial, a version and a set of scores. Quality finally no longer depended only on luck. Later the manager wrote this experience into the daily rules: first look at the scenario, then look at the boundaries, and finally look at who can take over when it fails. Only when all of these are clear will a beautiful demo become deliverable work.