Reviewing Your Own Work

Read it as if someone else wrote it. Leave it a day, then open the code and see what you have to work out again. Whatever you have to reconstruct is what needs a better name or a comment.

Then ask the questions a reviewer would ask.

The questions.

Can I tell what each component does from its name and its props?

Is there state here that could be derived instead?

Is there an effect here that is really just a calculation?

Does every list key belong to the data rather than the position?

If the network fails at the worst possible moment, what does someone see?

Is there anything here I added because it seemed advanced rather than because it was needed?

✦ What to write in the README

What it does, in one sentence a non-programmer would understand. What data it uses and where that comes from. How to run it. One paragraph on a decision you made and why, and one on something you would do differently. That last paragraph is the one that reads as an engineer rather than a student.

When it is finished. When another person can use it on your deployed URL, from a phone, on a bad connection, with a keyboard, and never sees a blank screen or a spinner that does not resolve.

That is a real application. Put it in your portfolio.