You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[Due for payment 2026-04-07] [$250] Expense - Tooltip "Try it out" remains in incorrect position when recipients list is filtered #85799
Select the "Track" intent during the onboarding flow
Tap the FAB button
Select "Manual"
Enter any amount and tap "Continue"
Type "manager" in the search input field
Expected Result:
Tooltip "Try it out" is displayed next to the Manager item and correctly follows it when the recipients list is updated or filtered, maintaining proper alignment without any visual gaps.
Actual Result:
Tooltip "Try it out" remains in its original position when the recipients list is filtered, while the Manager item moves in the list, causing misalignment and a visible gap between the tooltip and the item.
Please re-state the problem that we are trying to solve in this issue.
The "Try it out" tooltip on the Manager McTest item in the recipients list stays in its original position when the list is filtered. When the user types "manager" in the search field, the Manager item moves in the list, but the tooltip remains at the old coordinates, causing a visible gap and misalignment.
What is the root cause of that problem?
The EducationalTooltip measures its position once via measureInWindow on the initial onLayout and renders as a fixed-position portal detached from the list. When the list is filtered, no re-measurement is triggered because:
uniqueID never changes:InviteMemberListItem computes uniqueID={${sectionIndex}-${index}} (InviteMemberListItem.tsx:116), but ListItemRenderer never passes index or sectionIndex props (ListItemRenderer.tsx:64-113). Both default to 0, so uniqueID is always "0-0". The useEffect in BaseEducationalTooltip (line 85) includes uniqueID in its dependency array and would call renderTooltip() to re-measure — but since uniqueID never changes, it never fires.
Filtering is not a scroll event: The shouldHideOnScroll mechanism (BaseEducationalTooltip.tsx:87-114) only responds to CONST.EVENTS.SCROLLING events. Filtering changes the list data, not the scroll position, so no scroll event is emitted and the hide/re-show/re-measure cycle never triggers.
onLayout doesn't call show.current(): Even if onLayout fires at the new position, the handler (BaseEducationalTooltip.tsx:164-172) only updatesshow.current — it does not invoke it. The useEffect that calls show.current() (line 122-137) depends on [shouldMeasure, shouldRender, shouldSuppressTooltip], none of which change on filtering.
What changes do you think we should make in order to solve the problem?
Pass the item's actual index from ListItemRenderer to InviteMemberListItem so that uniqueID reflects the item's real position in the list. When filtering causes the Manager item to move to a different index, uniqueID will change, triggering the existing useEffect at BaseEducationalTooltip.tsx:85 which calls renderTooltip() to re-measure and reposition the tooltip.
Specifically, in ListItemRenderer.tsx, pass the index prop (which is already available as a parameter) through to the <ListItem> component. The InviteMemberListItem already accepts and uses index — it just never receives it from its parent.
What alternative solutions did you explore? (Optional)
Call show.current() directly from onLayout: This would re-measure on every layout event, which is more robust but could cause unnecessary re-measurements and potential flickering.
Add BoundsObserver to EducationalTooltip: The regular BaseTooltip uses BoundsObserver (backed by BoundingClientRectObserver) to continuously track position changes. Adding this to EducationalTooltip would be the most robust solution but is web-only and adds complexity for a case that the uniqueID fix handles cleanly.
Next Steps for Contributor+ team: Reply with @MelvinBot implement this to create a draft PR, @MelvinBot <your feedback> to refine this analysis, or explain why you are rejecting Melvin's proposal.
changed the title [-]Expense - Tooltip "Try it out" remains in incorrect position when recipients list is filtered[/-][+][$250] Expense - Tooltip "Try it out" remains in incorrect position when recipients list is filtered[/+]on Mar 19, 2026
Please re-state the problem that we are trying to solve in this issue.
The "Try it out" educational tooltip on the Manager McTest item stays at its original screen position when the recipients list is filtered, creating a visible gap between the tooltip and the item it should be anchored to.
What is the root cause of that problem?
The EducationalTooltip measures its target's screen coordinates once during the initial layout and has no mechanism to re-measure when the target element moves due to list filtering.
In BaseEducationalTooltip, the onLayout handler sets setShouldMeasure(true) on the first call, which triggers the effect at line 122 to call show.current() and position the tooltip. On subsequent layout changes (e.g. when filtering moves the item), shouldMeasure is already true, so setShouldMeasure(true) is a no-op — React skips the state update since the value hasn't changed, and the effect never re-runs.
The existing re-measurement paths only cover resize events (line 83) and scroll-stop events (line 97). Filtering changes the list data without emitting a scroll event, so neither path fires. The onLayout handler updates the show.current ref with a fresh measurement function, but nothing actually invokes it after the initial render.
What changes do you think we should make in order to solve the problem?
In BaseEducationalTooltip, call renderTooltip() from the onLayout handler when the tooltip has already been measured and is visible:
onLayout: (e: LayoutChangeEventWithTarget)=>{if(!shouldMeasure){setShouldMeasure(true);}consttarget=e.target||e.nativeEvent.target;tooltipElementRef.current=target;show.current=()=>measureTooltipCoordinate(target,updateTargetBounds,showTooltip);// Re-measure tooltip position when target layout changes after initial measurementif(shouldMeasure&&shouldRender){renderTooltip();}},
renderTooltip() already exists and calls getTooltipCoordinates → measureInWindow on the target element to get its current screen position, then updates the tooltip via updateTargetBounds. Since onLayout fires after the layout engine has computed the new position, the measurement is accurate. There's no infinite loop risk because onLayout only fires on actual layout changes, not React re-renders.
What alternative solutions did you explore? (Optional)
The existing proposal suggests passing index from ListItemRenderer into InviteMemberListItem so the uniqueID prop changes on filtering. While that would work for this specific list, it fixes the problem at the consumer level rather than the tooltip level. My approach fixes it in BaseEducationalTooltip directly, which means:
It handles any reason the tooltip target moves (not just list index changes)
It's a single-file, single-location change
Future uses of EducationalTooltip in other dynamic layouts get the fix for free
Please re-state the problem that we are trying to solve in this issue.
The "Try it out" tooltip is anchored to the Manager row when the recipients list first renders, but after filtering the list, the Manager row moves and the tooltip stays at its old coordinates, leaving a visible gap.
What is the root cause of that problem?
This looks like a regression in the new SelectionList/SelectionListWithSections path rather than in the tooltip positioning math itself.
InviteMemberListItem builds the tooltip uniqueID from sectionIndex and index, and BaseEducationalTooltip already re-runs renderTooltip() when that uniqueID changes. But the new src/components/SelectionList/ListItem/ListItemRenderer.tsx never forwards index or sectionIndex, so InviteMemberListItem falls back to 0 for both values and the tooltip uniqueID stays "0-0" even when filtering moves the row. The older src/components/SelectionListWithSections/BaseSelectionListItemRenderer.tsx did forward those props, so the new sectioned-list renderer dropped the re-measure trigger this tooltip already relies on.
// Prevent unexpected scrolling on mobile Chrome after the context menu closes by ignoring programmatic focus not triggered by direct user interaction.
// Prevent unexpected scrolling on mobile Chrome after the context menu closes by ignoring programmatic focus not triggered by direct user interaction.
What changes do you think we should make in order to solve the problem?
Restore that metadata flow in the new sectioned-list renderer so the existing tooltip re-measure path can run again.
Pass the row's actual index through src/components/SelectionList/ListItem/ListItemRenderer.tsx into InviteMemberListItem.
If needed for parity with the legacy sectioned-list renderer, also thread a section-derived sectionIndex through the new flattened-section path.
Keep EducationalTooltip itself unchanged so its current uniqueID effect handles re-measuring when the filtered row moves.
This keeps the fix scoped to the regression path and avoids turning every EducationalTooltip layout update into a fresh measurement.
What alternative solutions did you explore? (Optional)
We could also call renderTooltip() from BaseEducationalTooltip.onLayout after the first measurement. That would likely fix this issue too, but it broadens the change to every educational tooltip and adds extra measurements for any later layout updates, while the existing uniqueID path already covers the filtered-list case once the new renderer forwards the row metadata again.
⚠️@arlo-owls Your proposal is a duplicate of an already existing proposal and has been automatically withdrawn to prevent spam. Please review the existing proposals before submitting a new one.
changed the title [-][$250] Expense - Tooltip "Try it out" remains in incorrect position when recipients list is filtered[/-][+][Due for payment 2026-04-07] [$250] Expense - Tooltip "Try it out" remains in incorrect position when recipients list is filtered[/+]on Mar 31, 2026
The solution for this issue has been 🚀 deployed to production 🚀 in version 9.3.50-8 and is now subject to a 7-day regression period 📆. Here is the list of pull requests that resolve this issue:
@jayeshmangwani The PR fixing this issue has been merged! The following checklist (instructions) will need to be completed before the issue can be closed. Please copy/paste the BugZero Checklist from here into a new comment on this GH and complete it. If you have the K2 extension, you can simply click: [this button]
[Contributor] The offending PR and associated issue have been commented on, pointing out the bug it caused and why, so the author and reviewers can learn from the mistake.
Link to the comment on the PR: N/A - Couldn’t find the exact PR—this might’ve been there for a while and just showed up during testing.
Link to the comment on the Issue: N/A - Couldn’t find the exact PR—this might’ve been there for a while and just showed up during testing.
[Contributor] If the regression was CRITICAL (e.g. interrupts a core flow) A discussion in #expensify-open-source has been started about whether any other steps should be taken (e.g. updating the PR review checklist) in order to catch this type of bug sooner.
Link to discussion: Not critical - just a tooltip position issue on list refresh, so no need for a new discussion or checklist
[Contributor] If it was decided to create a regression test for the bug, please propose the regression test steps using the template below to ensure the same bug will not reach production again.
[BugZero Assignee] Create a GH issue for creating/updating the regression test once above steps have been agreed upon.
Activity
MelvinBot commentedon Mar 19, 2026
Proposal
Please re-state the problem that we are trying to solve in this issue.
The "Try it out" tooltip on the Manager McTest item in the recipients list stays in its original position when the list is filtered. When the user types "manager" in the search field, the Manager item moves in the list, but the tooltip remains at the old coordinates, causing a visible gap and misalignment.
What is the root cause of that problem?
The
EducationalTooltipmeasures its position once viameasureInWindowon the initialonLayoutand renders as a fixed-position portal detached from the list. When the list is filtered, no re-measurement is triggered because:uniqueIDnever changes:InviteMemberListItemcomputesuniqueID={${sectionIndex}-${index}}(InviteMemberListItem.tsx:116), butListItemRenderernever passesindexorsectionIndexprops (ListItemRenderer.tsx:64-113). Both default to0, souniqueIDis always"0-0". TheuseEffectinBaseEducationalTooltip(line 85) includesuniqueIDin its dependency array and would callrenderTooltip()to re-measure — but sinceuniqueIDnever changes, it never fires.Filtering is not a scroll event: The
shouldHideOnScrollmechanism (BaseEducationalTooltip.tsx:87-114) only responds toCONST.EVENTS.SCROLLINGevents. Filtering changes the list data, not the scroll position, so no scroll event is emitted and the hide/re-show/re-measure cycle never triggers.onLayoutdoesn't callshow.current(): Even ifonLayoutfires at the new position, the handler (BaseEducationalTooltip.tsx:164-172) only updatesshow.current— it does not invoke it. TheuseEffectthat callsshow.current()(line 122-137) depends on[shouldMeasure, shouldRender, shouldSuppressTooltip], none of which change on filtering.What changes do you think we should make in order to solve the problem?
Pass the item's actual
indexfromListItemRenderertoInviteMemberListItemso thatuniqueIDreflects the item's real position in the list. When filtering causes the Manager item to move to a different index,uniqueIDwill change, triggering the existinguseEffectatBaseEducationalTooltip.tsx:85which callsrenderTooltip()to re-measure and reposition the tooltip.Specifically, in
ListItemRenderer.tsx, pass theindexprop (which is already available as a parameter) through to the<ListItem>component. TheInviteMemberListItemalready accepts and usesindex— it just never receives it from its parent.What alternative solutions did you explore? (Optional)
Call
show.current()directly fromonLayout: This would re-measure on every layout event, which is more robust but could cause unnecessary re-measurements and potential flickering.Add
BoundsObservertoEducationalTooltip: The regularBaseTooltipusesBoundsObserver(backed byBoundingClientRectObserver) to continuously track position changes. Adding this toEducationalTooltipwould be the most robust solution but is web-only and adds complexity for a case that theuniqueIDfix handles cleanly.Next Steps for Contributor+ team: Reply with
@MelvinBot implement thisto create a draft PR,@MelvinBot <your feedback>to refine this analysis, or explain why you are rejecting Melvin's proposal.melvin-bot commentedon Mar 19, 2026
Triggered auto assignment to Contributor-plus team member for initial proposal review - @jayeshmangwani (
External)[-]Expense - Tooltip "Try it out" remains in incorrect position when recipients list is filtered[/-][+][$250] Expense - Tooltip "Try it out" remains in incorrect position when recipients list is filtered[/+]melvin-bot commentedon Mar 19, 2026
Job added to Upwork: https://www.upwork.com/jobs/~022034615061209274058
twilight2294 commentedon Mar 19, 2026
Proposal
Please re-state the problem that we are trying to solve in this issue.
The "Try it out" educational tooltip on the Manager McTest item stays at its original screen position when the recipients list is filtered, creating a visible gap between the tooltip and the item it should be anchored to.
What is the root cause of that problem?
The
EducationalTooltipmeasures its target's screen coordinates once during the initial layout and has no mechanism to re-measure when the target element moves due to list filtering.In BaseEducationalTooltip, the
onLayouthandler setssetShouldMeasure(true)on the first call, which triggers the effect at line 122 to callshow.current()and position the tooltip. On subsequent layout changes (e.g. when filtering moves the item),shouldMeasureis alreadytrue, sosetShouldMeasure(true)is a no-op — React skips the state update since the value hasn't changed, and the effect never re-runs.The existing re-measurement paths only cover resize events (line 83) and scroll-stop events (line 97). Filtering changes the list data without emitting a scroll event, so neither path fires. The
onLayouthandler updates theshow.currentref with a fresh measurement function, but nothing actually invokes it after the initial render.What changes do you think we should make in order to solve the problem?
In BaseEducationalTooltip, call
renderTooltip()from theonLayouthandler when the tooltip has already been measured and is visible:renderTooltip()already exists and callsgetTooltipCoordinates→measureInWindowon the target element to get its current screen position, then updates the tooltip viaupdateTargetBounds. SinceonLayoutfires after the layout engine has computed the new position, the measurement is accurate. There's no infinite loop risk becauseonLayoutonly fires on actual layout changes, not React re-renders.What alternative solutions did you explore? (Optional)
The existing proposal suggests passing
indexfromListItemRendererinto InviteMemberListItem so theuniqueIDprop changes on filtering. While that would work for this specific list, it fixes the problem at the consumer level rather than the tooltip level. My approach fixes it in BaseEducationalTooltip directly, which means:EducationalTooltipin other dynamic layouts get the fix for freejainam884 commentedon Mar 19, 2026
Proposal
Please re-state the problem that we are trying to solve in this issue.
The "Try it out" tooltip is anchored to the Manager row when the recipients list first renders, but after filtering the list, the Manager row moves and the tooltip stays at its old coordinates, leaving a visible gap.
What is the root cause of that problem?
This looks like a regression in the new
SelectionList/SelectionListWithSectionspath rather than in the tooltip positioning math itself.InviteMemberListItembuilds the tooltipuniqueIDfromsectionIndexandindex, andBaseEducationalTooltipalready re-runsrenderTooltip()when thatuniqueIDchanges. But the newsrc/components/SelectionList/ListItem/ListItemRenderer.tsxnever forwardsindexorsectionIndex, soInviteMemberListItemfalls back to0for both values and the tooltipuniqueIDstays"0-0"even when filtering moves the row. The oldersrc/components/SelectionListWithSections/BaseSelectionListItemRenderer.tsxdid forward those props, so the new sectioned-list renderer dropped the re-measure trigger this tooltip already relies on.App/src/components/SelectionList/ListItem/InviteMemberListItem.tsx
Lines 104 to 116 in bdd0e98
App/src/components/SelectionList/ListItem/ListItemRenderer.tsx
Lines 64 to 109 in bdd0e98
App/src/components/SelectionListWithSections/BaseSelectionListItemRenderer.tsx
Lines 73 to 123 in bdd0e98
What changes do you think we should make in order to solve the problem?
Restore that metadata flow in the new sectioned-list renderer so the existing tooltip re-measure path can run again.
indexthroughsrc/components/SelectionList/ListItem/ListItemRenderer.tsxintoInviteMemberListItem.sectionIndexthrough the new flattened-section path.EducationalTooltipitself unchanged so its currentuniqueIDeffect handles re-measuring when the filtered row moves.This keeps the fix scoped to the regression path and avoids turning every
EducationalTooltiplayout update into a fresh measurement.What alternative solutions did you explore? (Optional)
We could also call
renderTooltip()fromBaseEducationalTooltip.onLayoutafter the first measurement. That would likely fix this issue too, but it broadens the change to every educational tooltip and adds extra measurements for any later layout updates, while the existinguniqueIDpath already covers the filtered-list case once the new renderer forwards the row metadata again.arlo-owls commentedon Mar 20, 2026
🚫 Duplicated proposal withdrawn by 🤖 ProposalPolice.
melvin-bot commentedon Mar 20, 2026
✅ Contributor details stored successfully. Thank you for contributing to Expensify!
github-actions commentedon Mar 20, 2026
22 remaining items
[-][$250] Expense - Tooltip "Try it out" remains in incorrect position when recipients list is filtered[/-][+][Due for payment 2026-04-07] [$250] Expense - Tooltip "Try it out" remains in incorrect position when recipients list is filtered[/+]melvin-bot commentedon Mar 31, 2026
Reviewinglabel has been removed, please complete the "BugZero Checklist".melvin-bot commentedon Mar 31, 2026
The solution for this issue has been 🚀 deployed to production 🚀 in version 9.3.50-8 and is now subject to a 7-day regression period 📆. Here is the list of pull requests that resolve this issue:
If no regressions arise, payment will be issued on 2026-04-07. 🎊
For reference, here are some details about the assignees on this issue:
melvin-bot commentedon Mar 31, 2026
@jayeshmangwani The PR fixing this issue has been merged! The following checklist (instructions) will need to be completed before the issue can be closed. Please copy/paste the BugZero Checklist from here into a new comment on this GH and complete it. If you have the K2 extension, you can simply click: [this button]
melvin-bot commentedon Mar 31, 2026
Triggered auto assignment to @mallenexpensify (
Awaiting Payment)melvin-bot commentedon Apr 7, 2026
Payment Summary
Upwork Job
BugZero Checklist (@mallenexpensify)
jayeshmangwani commentedon Apr 8, 2026
BugZero Checklist:
[Contributor] The offending PR and associated issue have been commented on, pointing out the bug it caused and why, so the author and reviewers can learn from the mistake.
Link to the comment on the PR: N/A - Couldn’t find the exact PR—this might’ve been there for a while and just showed up during testing.
Link to the comment on the Issue: N/A - Couldn’t find the exact PR—this might’ve been there for a while and just showed up during testing.
[Contributor] If the regression was CRITICAL (e.g. interrupts a core flow) A discussion in #expensify-open-source has been started about whether any other steps should be taken (e.g. updating the PR review checklist) in order to catch this type of bug sooner.
Link to discussion: Not critical - just a tooltip position issue on list refresh, so no need for a new discussion or checklist
[Contributor] If it was decided to create a regression test for the bug, please propose the regression test steps using the template below to ensure the same bug will not reach production again.
[BugZero Assignee] Create a GH issue for creating/updating the regression test once above steps have been agreed upon.
Link to issue: https://github.com/Expensify/Expensify/issues/622208
Regression Test Proposal
Test:
Try it outtooltip next to the Manager McTest itemmanagerin the search input fieldDo we agree 👍 or 👎
Zapier Logs
Run ID: 00040eee-5b59-a4c0-ce80-85c5d9cece44mallenexpensify commentedon Apr 8, 2026
Payment Summary
Contributor+: @jayeshmangwani due $250 via NewDot
^ Test case created
laurenreidexpensify commentedon Apr 15, 2026
New Expensify Approved: @jayeshmangwani $250