Ticket-admitted room lines never reach the conversation pane #939

Activity

1. Birfy

   closed this as completedin #940on Oct 24, 2026
2. Birfy commented on Oct 24, 2026

   

   Birfy

   on Oct 24, 2026

   Collaborator

   More actions

   Triage: a separate defect from #938, and it is the display side, not the turn side.

   #938 is a face room where a friend's attested plain line opens no turn (hooks.js:1113 plain, groups.js:837 quiet) — fix it and the face-room Bot answers, but nothing here changes. #939 is a ticket-admitted room where the line arrives (it is in room.log) and never reaches the conversation pane. Neither fix makes the other unnecessary.

   Verified on main c812e5d6

   • applyInboundRoomShare (core/src/host/room-membership.js:2034) appends the inbound line to room.log (:2183-2200); it posts nothing into a Session.
   • startInboundTurn (core/src/host/groups.js:716) opens the member's session, runs the turn, and appends the reply with appendRoom (:737-741) — also room.log only.
   • linesVisibleTo (room-membership.js:302) is the only reader of room.log, and it is host-side (:417). Nothing mirrors it into the pane's source.

   So of the two ways the report can be right, exactly one is true, and the first commit must say which: (a) the pane is fed by the room Session and the inbound line was never posted there — the fix is a mirror; or (b) the pane is fed from room.log in the client and the payload strip is the bug — the fix is the payload. Do not do both, and name it with line numbers.

   Scope

   1. In: after a ticket admit, each home's pane shows the other home's lines (host → guest, guest → host).
   2. In: the guest's own lines keep showing once (they are not the defect).
   3. Out: "the guest also sees the host's lines from before the join". That is a catch-up/replay question — how much history a join carries is a product decision, and folding it in would make this unclosable. Its own issue if wanted.
   4. Out: the member count. If the roster is wrong that is the A guest's room shows none of the host's Bots: the join envelope carries no roster, and the mirror records at most one remote #904 family; say so there rather than here.

   Acceptance

   1. Two homes, ticket admit, host writes → the guest's pane contains that line (the sidebar preview is not enough).
   2. Guest writes → the host's pane contains it.
   3. Exactly one line per message: delivery and the turn must not both write it, and a second poll of the same eventId adds nothing.
   4. The author shown is the person who typed it (the Inbound room-share envelopes let the payload decide the inviter and the author, not the transport #909/verifiedAuthor rule), not the Bot that carried the envelope.
   5. A host-level test (not the browser) pins whichever of (a)/(b) is the real seam: the inbound line and the guest Bot's reply both reach the pane's source.

   Written-by: Triage
3. Birfy

   added

   agent-readyMaintainer-approved spec: an agent may take it. Only maintainers set this; never in a templateMaintainer-approved spec: an agent may take it. Only maintainers set this; never in a template

   area:socialFriends, invites, grants, Bot SocialFriends, invites, grants, Bot Social

   bugSomething isn't workingSomething isn't working

   and removed

   agent-readyMaintainer-approved spec: an agent may take it. Only maintainers set this; never in a templateMaintainer-approved spec: an agent may take it. Only maintainers set this; never in a template

   on Oct 24, 2026
4. Birfy commented on Oct 24, 2026

   

   Birfy

   on Oct 24, 2026

   Collaborator

   More actions

   Superseded one minute later: #940 merged at 07:49:11Z (main c7286584) and closed this issue at 07:49:12Z, so the analysis above is not a live task. It stays as the acceptance record only — if the pane seam regresses, those five items are what to test. agent-ready has been removed from this closed issue.

   Written-by: Triage
5. HugoXK

   added a commit that references this issue on Oct 24, 2026

   Show ticket-seat lines in the room session (ThinkFlowLab#940)

   ...

   This issue will close once commit c728658 is merged into the 'main' branch.

   c728658