Make the condition visible before suggesting the answer

I ask which dimension the user needs the product to respect, not simply which dimension is easiest to collect. A nominal roll width can differ from the usable width for a particular job. The product should let the customer specify the relevant exclusions and understand the resulting layout boundary. I work with research, design and engineering to decide how those choices appear and what happens when a proposal exceeds them. The 152 cm example with 1 cm excluded on each side produced a 150 cm usable width; it was a defined customer condition, not a general rule for every fabric. My responsibility is to make the condition available to the product decision and visible to the user. Fabric inspection and manufacturing acceptance remain with the people qualified to make them in the actual setting.

An arrangement can look efficient because it has quietly relaxed a restriction. I want the product to preserve a user's specified orientation rules when it suggests where pattern pieces could be placed. Direction may matter to the fabric or design, and the software should not override it merely to produce a shorter result. When a supported suggestion cannot satisfy the stated constraint, the product should explain that limit instead of returning an attractive invalid answer. Our review led us to repair six suggestions and withhold three. I do not generalise that result into a claim about all fabrics or every construction method. The product role is to define the behaviour, review evidence with specialists and make the trade-off understandable, while engineering owns implementation and the customer retains the final judgement about use.

A printed page may be useful for discussion without preserving the dimensions required for a pattern task. I want those purposes separated in the product. If a user chooses output intended to retain scale, the preparation should make resizing visible and provide a reference that can be checked in the actual print path. A fitting option that makes the drawing occupy the page neatly can change the dimensions. Our task review showed how easily that change could be missed even when the outline looked complete. I work with users and engineering to improve the choice and explanation without claiming the software controls every printer or production step. A visible reference and clear mode support inspection; they do not substitute for the customer's own verification before physical work begins.

I organise product choices around a complete task and the judgement it asks of the user. Research can reveal a problem without establishing how common it is across the whole market, and engineering can demonstrate a capability without deciding whether the proposed behaviour is appropriate. I bring those perspectives together when allocating research and design effort and agreeing a delivery plan. The decision may be to expose a constraint, retain a manual route or narrow an automatic feature. I keep the cost and delay in the commercial account rather than treating a responsible correction as free progress. The aim is a product that makes a useful choice easier to understand. That requires explicit responsibility between product, engineering and customer specialists, instead of allowing a polished demonstration to stand in for evidence about the real task.