Bug: updating state to an existing value is (incorrectly?) triggering a re-render about react

[alexreardon]

Here is the full text of the set state optimization [as at 13th April]:

If the new value you provide is identical to the current state, as determined by an Object.is comparison, React will skip re-rendering the component and its children. This is an optimization. Although in some cases React may still need to call your component before skipping the children, it shouldn’t affect your code.

useState docs

I think that wording the caveat that the re-render optimization is skipped "in some cases" implies that the de-optimization should only occur in outlier circumstances (eg nested set states in an effect), and not for standard happy path cases.

I think that skipping the re-render optimization for a happy path case is a bug.

I appreciate that component renders should be pure, and so an additional re-render should not cause any functional problems. However, I still think that this current behaviour is incorrect.

The current behaviour hurt me when I was trying to write a test to assert that a component should not render after a particular event, but a render did occur. It would be fair to ask: "Why assert render behaviour in a test?". I think it is fair to assert re-rendering behaviour given that re-rendering can be expensive with react

[Andarist]

Might be a duplicate of #18098

[vkurchatkin]

On the linked documentation page it literally says that this is expected:

Although in some cases React may still need to call your component before skipping the children, it shouldn’t affect your code

[brayanjeshua]

It could be a joke but... Have you tried to remove the StictMode?

https://react.dev/reference/react/StrictMode

[alexreardon]

I can confirm that this bug appears without strict mode on