Scoping Honestly

Three questions that cut scope fast.

What happens if this feature does not exist? If the answer is 'the application still solves the problem', it is not version one.

Can a person do this another way for now, even clumsily? Then let them, and build it later.

Does this feature exist because it is useful, or because it is interesting to build? Both are valid reasons and only one belongs in version one.

Write acceptance criteria before you write code. Not a feature list, which is a description of what you will build, but a list of statements that are either true or false when you are finished.

'Users can save items' is a feature. 'A signed-in person can save an item, see it after a refresh, and delete it, and a different account cannot see it' is an acceptance criterion. The second one tells you when you are done, and it also tells you what to test.

✦ This is the same standard as every capstone so far

React, Django, APIs & Datasets, SQL and TypeScript each ended with twelve criteria checkable by another person. That was practice for this. The difference is that here you write them yourself, and choosing good ones is most of the skill.