Potential repro for possible bug w #26763 #27465
Conversation
facebook-github-bot
added
CLA Signed
React Core Team
| <div | ||
| onClick={[Function]} | ||
| > | ||
| Expand | ||
| </div>, | ||
| <div> | ||
| Item 2 | ||
| </div>, |
There was a problem hiding this comment.
both items should be "expanded", so this should be showing <div>Item 1</div><div>Item 2</div>. But somehow only the useEffect from the second item takes effect.
There was a problem hiding this comment.
@rickhanlonii see comment here
|
Comparing: 0fba3ec...9fe5904 Critical size changesIncludes critical production bundles, as well as any change greater than 2%:
Significant size changesIncludes any change greater than 0.2%: (No significant changes) |
| const isExpanded = selectionContext.expanded.has(props.id); | ||
| const didInitiallyExpand = React.useRef(false); | ||
| const shouldInitiallyExpand = | ||
| props.initiallyExpanded && didInitiallyExpand.current === false; |
There was a problem hiding this comment.
technically reading the ref here is wrong but it's easy to move it into the effect (and likely unrelated to the issue here)
|
When I run Looking through the CI failures, I see two reasons:
|
Possible bug w #26763. See comments for more. This is a simplified version of an internal test that failed after #26763, where we observed a different set of useEffects fired prior to taking a snapshot with ReactTestRenderer. This PR isn't a perfect repro, since manually reverting the changes from the above PR doesn't produce the expected result on this snapshot. But this is potentially in the ballpark of a repro, since the original code internally did fail and the expected result is different than expected in a similar way to the internal failure: a different set of useEffects runs as part of the act() call.
Related to #27463