Submatter/plans/Plan-workspace-communications/Epic-workspace-chat/Task-chat-realtime-delivery.md
varutasu ff01a5f097
Some checks are pending
CI / Lint, type-check, test (push) Waiting to run
Deploy homelab (CT 116) / Deploy to CT 116 (push) Waiting to run
Make workspace Chat durable and ready for source-built deployment
2026-10-29 19:30:20 +00:00

4.1 KiB

kind slug title plan_slug epic_slug status priority tenant_id owner cursor_todo_id updated_at
task chat-realtime-delivery Deliver scalable realtime chat events workspace-communications workspace-chat in_progress P1 global unassigned null 2026-10-29

Task summary

Broadcast committed chat events over authenticated WebSockets and support horizontal fan-out and reconnect recovery.

Description

Use HTTP/tRPC for commands and Postgres as the source of truth. Consume the outbox contract and atomic writes established by the persistence task; do not create a second message transaction. A worker claims and publishes normalized events through Redis to authenticated WebSocket subscribers. Each channel uses its own (channel_id, sequence) cursor, and clients reconnect by replaying missed rows from Postgres per authorized channel. Reuse Hocuspocus/Yjs only for presence, typing, or drafts rather than canonical history.

Subtasks

  • Finalize one versioned envelope for message, reply, tombstone, channel lifecycle, membership, and read-state events using the existing outbox contract.
  • Add subscriptions/replay using workspace AND context AND channel access; recheck at delivery, close revoked subscriptions, and test that removed access never receives subsequent payloads. No same-workspace-only shortcut or unsupported-object fallback.
  • Claim and publish outbox events through Redis with retry and idempotent consumption semantics.
  • Add per-channel reconnect catch-up and cross-device read-state invalidation.
  • Add heartbeat, backpressure, graceful-shutdown, outage, ordering, and authorization tests.

Go-live reconciliation — 2026-10-29

Operator reports 14/14 live acceptance checks passed on CT 116 via wss://chat.stillwell.cloud/chat, including two-member messaging/replies, reconnect, unread, edit/delete recovery and revocation. CT 102 ledger is reported at 24 through 0023, with Chat tables owned by echodo. Evidence is in docs/Current Progress.md (operator-supplied; not rerun here). Local regression tests cover projection/redaction fixes and hardened configured-origin ticket checks. Next deployment must rebuild web and retest; keep task in progress pending full checks/final review and unverified outage/multi-instance criteria.

Earlier deployment checkpoint — 2026-10-29 (superseded above)

  • Extended the existing collab service with Chat secret/port wiring, dual-listener liveness checks, standalone publication, web build URL, and Traefik TLS example. Public build variables now participate in Turbo build inputs.
  • Added scripts/chat-deployment.test.mjs to the root test command; all five configuration checks pass. This is not live transport acceptance.
  • CT 116 access blocked on trusted SSH host-key verification; no migrations or deployment performed. Production schema and disposable-Postgres, Redis outage, reconnect/revocation and two-browser gates remain open. See docs/deployment-production.md and docs/Current Progress.md.

Owner or assignee

unassigned

Status

in_progress

2026-10-26: Implement signed, scoped short-lived subscription tickets and metadata-only WebSocket invalidation hints. Recheck current Postgres authorization at subscribe, delivery, and heartbeat; reconnect must recover edits/tombstones independently of creation-sequence catch-up. Do not reuse the legacy decode-only Yjs identity as Chat authentication.

Estimation

L

Acceptance criteria

  • A message is visible through the query API even when Redis is unavailable.
  • Multiple realtime instances deliver committed events to local subscribers.
  • Clients recover each subscribed channel from its own sequence without duplicates or assumed cross-channel ordering.
  • Subscription authorization matches public/private/direct/archived rules and is rechecked against current workspace/channel access.
  • Durable mark-read events reconcile unread badges on another connected device.
  • pnpm lint && pnpm type-check && pnpm test pass; pnpm build passes for changed server entry points.
  • Epic: ./Epic-workspace-chat.md
  • Plan: ../Plan-workspace-communications.md