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 2026-07-21] LHN - Marking an unread message as unread does not mark the conversation as unread #95103
👋 Friendly reminder that deploy blockers are time-sensitive ⏱ issues! Check out the open `StagingDeployCash` deploy checklist to see the list of PRs included in this release, then work quickly to do one of the following:
Identify the pull request that introduced this issue and revert it.
🤖 Automated deploy-blocker analysis — confidence: High (root cause verified by diffing the production branch against staging).
TL;DR: This is a regression introduced by #94817 (Decompose ReportActionsList: 11, merged 2026-06-30). It removed the block in useUnreadMarker that re-seeds the unread-marker cutoff when the hook instance is reused for a different report. Recommended fix: revert #94817.
What is the root cause of that problem?
useUnreadMarker keeps the marker cutoff in local state unreadMarkerTime, seeded from the report's lastReadTime. Because the report view / hook instance is reused across report switches, there was a block that re-seeded unreadMarkerTime from the current report's lastReadTime whenever reportID changed:
const[trackedReportID,setTrackedReportID]=useState(reportID);if(trackedReportID!==reportID){setTrackedReportID(reportID);setUnreadMarkerTime(reportLastReadTime);// re-seed on report switch}
#94817deleted this block (its only functional change; the author's new test assumes "one mount per report", which is not true when the instance is reused). After deletion, on navigating away and back:
Marking a message unread lowers lastReadTime in Onyx and emits unreadAction_…; the listener updates the marker on the currently mounted report, so it briefly shows. ✅
Navigating away reuses the same hook instance with a new reportID, but unreadMarkerTime is no longer re-seeded — it keeps the old value, and the "push down to newest" branch (useUnreadMarker.ts:139) advances it to the most recent action.
On returning, the cutoff is stale/too-new, so getUnreadMarkerReportAction finds nothing before it → the "New" divider is gone and the conversation no longer reads as unread. ❌
Why this PR (and why it's staging-only) — verification
The re-seed block exists on production (useUnreadMarker.ts:51-54) but is absent on staging (e3b4f3c) — exactly matching "Reproducible in staging: Yes / production: No".
What changes do you think we should make in order to solve the problem?
Revert #94817 — it is a self-contained decompose/refactor and reverting cleanly restores the per-report re-seed.
If a roll-forward is preferred over a revert, re-add the trackedReportID guard so unreadMarkerTime is re-seeded from the current report's lastReadTime whenever reportID changes.
What alternative solutions did you explore? (Optional)
Considered the self-authored isUnread suppression (#91940) as the cause but ruled it out — it is already in production.
Next Steps for Contributor+ team: To accept:@MelvinBot implement [this](https://github.com/Expensify/App/pull/94817) to create a draft PR (revert). To refine:@MelvinBot <your feedback> To reject: Explain why you are rejecting Melvin's proposal.
changed the title [-]LHN - Marking an unread message as unread does not mark the conversation as unread[/-][+][Due for payment 2026-07-13] LHN - Marking an unread message as unread does not mark the conversation as unread[/+]on Jul 6, 2026
The solution for this issue has been 🚀 deployed to production 🚀 in version 9.4.27-2 and is now subject to a 7-day regression period 📆. Here is the list of pull requests that resolve this issue:
@rlinoz@VickyStash 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]. If no checklist is needed for this issue, you can click: [no checklist button]
changed the title [-][Due for payment 2026-07-13] LHN - Marking an unread message as unread does not mark the conversation as unread[/-][+][Due for payment 2026-07-21] LHN - Marking an unread message as unread does not mark the conversation as unread[/+]on Jul 14, 2026
The solution for this issue has been 🚀 deployed to production 🚀 in version 9.4.33-8 and is now subject to a 7-day regression period 📆. Here is the list of pull requests that resolve this issue:
@rlinoz@VickyStash 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]. If no checklist is needed for this issue, you can click: [no checklist button]
Activity
applause-bot commentedon Jul 1, 2026
While Applause doesn't embed attachments for this bug, here are the links for quick view:
https://platform.applause.com/services/links/v1/external/296d56b2dd68bc99ba5cb911039c102f5f7f619530cf043575a91519fee93b54
melvin-bot commentedon Jul 1, 2026
Triggered auto assignment to @carlosmiceli (
DeployBlockerCash), see https://stackoverflowteams.com/c/expensify/questions/9980/ for more details.melvin-bot commentedon Jul 1, 2026
💬 A slack conversation has been started in #expensify-open-source
github-actions commentedon Jul 1, 2026
👋 Friendly reminder that deploy blockers are time-sensitive ⏱ issues! Check out the open `StagingDeployCash` deploy checklist to see the list of PRs included in this release, then work quickly to do one of the following:
MelvinBot commentedon Jul 2, 2026
Proposal
TL;DR: This is a regression introduced by #94817 (
Decompose ReportActionsList: 11, merged 2026-06-30). It removed the block inuseUnreadMarkerthat re-seeds the unread-marker cutoff when the hook instance is reused for a different report. Recommended fix: revert #94817.What is the root cause of that problem?
useUnreadMarkerkeeps the marker cutoff in local stateunreadMarkerTime, seeded from the report'slastReadTime. Because the report view / hook instance is reused across report switches, there was a block that re-seededunreadMarkerTimefrom the current report'slastReadTimewheneverreportIDchanged:#94817 deleted this block (its only functional change; the author's new test assumes "one mount per report", which is not true when the instance is reused). After deletion, on navigating away and back:
lastReadTimein Onyx and emitsunreadAction_…; the listener updates the marker on the currently mounted report, so it briefly shows. ✅reportID, butunreadMarkerTimeis no longer re-seeded — it keeps the old value, and the "push down to newest" branch (useUnreadMarker.ts:139) advances it to the most recent action.getUnreadMarkerReportActionfinds nothing before it → the "New" divider is gone and the conversation no longer reads as unread. ❌Why this PR (and why it's staging-only) — verification
production(useUnreadMarker.ts:51-54) but is absent on staging (e3b4f3c) — exactly matching "Reproducible in staging: Yes / production: No".src/hooks/useUnreadMarker.ts(rest of the diff is tests/JSDoc).ReportUtils.isUnread()andshouldDisplayNewMarkerOnReportActionis already onproduction, so it can't cause a staging-only regression. That logic is the cause of the separate, production-reproducible [$250] Chat - The new message marker isn't displayed when user marks their own message as unread #94794 (own-message marker) — not this issue.Key refs (SHA
e3b4f3c):markCommentAsUnread(writeslastReadTimecorrectly — not the culprit),useUnreadMarker.What changes do you think we should make in order to solve the problem?
Revert #94817 — it is a self-contained decompose/refactor and reverting cleanly restores the per-report re-seed.
If a roll-forward is preferred over a revert, re-add the
trackedReportIDguard sounreadMarkerTimeis re-seeded from the current report'slastReadTimewheneverreportIDchanges.What alternative solutions did you explore? (Optional)
Considered the self-authored
isUnreadsuppression (#91940) as the cause but ruled it out — it is already in production.Related issues
Next Steps for Contributor+ team:
To accept:
@MelvinBot implement [this](https://github.com/Expensify/App/pull/94817)to create a draft PR (revert).To refine:
@MelvinBot <your feedback>To reject: Explain why you are rejecting Melvin's proposal.
rlinoz commentedon Jul 2, 2026
Hey @VickyStash would be able to take a quick look at this one?
VickyStash commentedon Jul 2, 2026
Reverting of #94817 doesn't fix the issue for me.
Going to try to check #94149
VickyStash commentedon Jul 2, 2026
Reverting of #94149 fixes the issue
VickyStash commentedon Jul 2, 2026
@rlinoz Instead of reverting the whole #94149 updates, we can just revert the
useMarkAsReadupdates, wdyt?8 remaining items
rlinoz commentedon Jul 2, 2026
CP request is here https://expensify.slack.com/archives/C07J32337/p1782992082260759
m-natarajan commentedon Jul 2, 2026
This issue is not repro on Windows 10/Chrome Build v9.4.26-4
Recording.996.mp4
[-]LHN - Marking an unread message as unread does not mark the conversation as unread[/-][+][Due for payment 2026-07-13] LHN - Marking an unread message as unread does not mark the conversation as unread[/+]melvin-bot commentedon Jul 6, 2026
Reviewinglabel has been removed, please complete the "BugZero Checklist".melvin-bot commentedon Jul 6, 2026
The solution for this issue has been 🚀 deployed to production 🚀 in version 9.4.27-2 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 2026-07-13. 🎊
For reference, here are some details about the assignees on this issue:
melvin-bot commentedon Jul 6, 2026
@rlinoz @VickyStash 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]. If no checklist is needed for this issue, you can click: [no checklist button]
[-][Due for payment 2026-07-13] LHN - Marking an unread message as unread does not mark the conversation as unread[/-][+][Due for payment 2026-07-21] LHN - Marking an unread message as unread does not mark the conversation as unread[/+]melvin-bot commentedon Jul 14, 2026
The solution for this issue has been 🚀 deployed to production 🚀 in version 9.4.33-8 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 2026-07-21. 🎊
For reference, here are some details about the assignees on this issue:
melvin-bot commentedon Jul 14, 2026
@rlinoz @VickyStash 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]. If no checklist is needed for this issue, you can click: [no checklist button]