Git pin now works (LiveKitClient 2.15.3 pre-downloaded), but the build box's
cached spec repo is stale: "Unable to find a specification for LiveKitUniFFI
(= 0.0.6)". Both LiveKitUniFFI 0.0.6 and LiveKitWebRTC 144.7559.11 are published
on the trunk, so --repo-update on the real pod install refreshes the repo and
resolves them.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The git-tag pin worked, but the build failed: "could not find compatible versions
… In snapshot (Podfile.lock): LiveKitClient (= 2.0.18) … In Podfile: LiveKitClient
(from git, tag 2.15.3)". Cause: `npx cap sync` runs `pod install` internally with
the PRE-pin Podfile, creating a Podfile.lock pinned to 2.0.18; our injected git-tag
source then conflicts with that lock. Fix: rm the Podfile.lock right after injecting
the pin so the real "Install CocoaPods" step re-resolves against tag 2.15.3.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The LiveKitClient CocoaPod on trunk caps at 2.0.18 (2.1+ is SPM-only), so the
2.15 CallKit audio API was unreachable. BUT the repo still ships a valid podspec
at tag 2.15.3, and its deps (LiveKitWebRTC 144.7559.11, LiveKitUniFFI 0.0.6,
SwiftProtobuf) ARE on trunk. So instead of a risky SPM migration:
- ios-patch.sh injects `pod 'LiveKitClient', :git => <repo>, :tag => '2.15.3'`
into the generated Podfile (after cap sync, before pod install). The NativeCall
podspec's '~> 2.0' is satisfied by 2.15.3. Idempotent; hard-fails if it can't
find the App target so we never silently fall back to 2.0.18.
- Plugin re-adds the CallKit<->LiveKit audio-session coordination, now compilable:
auto-config OFF + engine OFF at load; configure session + enable engine in
didActivate; disable in didDeactivate; request mic permission on connect.
Core Room APIs (connect/disconnect/setMicrophone) verified compatible between
2.0.18 and 2.15.3 against the real source.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Root cause of the intermittent dead-mic / no-audio / late-speaker: LiveKit's
automatic AVAudioSession config races CallKit's activation. The fix needs the
2.15+ audio API, which the build wasn't getting ('~> 2.0' resolved an older 2.x
from a stale spec cache). So:
- NativeCall.podspec: pin LiveKitClient '~> 2.15'.
- codemagic.yaml: 'pod install --repo-update' so the spec repo knows 2.15.x.
- Plugin: disable LiveKit auto audio-session config + keep the engine OFF; in
CXProvider didActivate set the session category and enable the engine; in
didDeactivate disable it. Request mic permission on connect so enabling the
engine in didActivate doesn't block on undetermined permission.
All AudioManager APIs verified against the raw 2.15.3 source (setEngineAvailability,
AudioEngineAvailability.default/.none, audioSession.isAutomaticConfigurationEnabled).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Second failure showed 'AudioManager' has no member 'audioSession' either — even
though the 2.15.3 source has both audioSession and setEngineAvailability. So the
build's CocoaPods is resolving an OLDER 2.x (stale spec repo/cache) that predates
the engine-observer audio API. Rather than keep guessing, drop ALL LiveKit
AudioManager audio-session code and keep the last known-good audio behavior
(LiveKit defaults). The valuable fixes that use only CallKit/AVFoundation stay:
re-ring blip guard, CallKit<->UI mute sync, and the WS reportIncomingCall ring
path. Proper CallKit audio-session coordination is deferred until the pod is
pinned/upgraded to LiveKit 2.15+.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The build failed: 'AudioManager' has no member 'setEngineAvailability' — that API
is only on unreleased/main docs, not in the resolved LiveKitClient 2.15.3. Use the
API that actually ships (per audio.md): disable LiveKit's automatic AVAudioSession
configuration (AudioManager.shared.audioSession.isAutomaticConfigurationEnabled =
false) and configure the session ourselves in CXProvider didActivate. Removed all
setEngineAvailability(.none/.default) calls.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Plugin (rides next build):
- Audio: disable LiveKit auto audio-session config + keep the engine OFF, then
configure the session and start the engine ONLY in CXProvider didActivate
(stop in didDeactivate). Fixes intermittent dead mic / no audio and the "speaker
turns on late" routing. Request mic permission on connect so enabling the engine
in didActivate can't block on undetermined permission (SDK #815).
- Re-ring blip: a cancel push for a call we already ended/known no longer reports
a NEW incoming call (that was the phantom "rings back for a second"); it ends
the known call cleanly, and only reports+ends for a truly unknown (cold) call.
- Mute display: answer/outgoing reflect muted-by-default on the CallKit screen;
setMuted now drives mute THROUGH CallKit so the system screen and the in-app
meeting UI stay in sync.
- reportIncomingCall: new method to ring CallKit from a WebSocket call event — a
2nd path alongside the VoIP push for when the app is open (push can be delayed);
deduped by UUID.
Web (deploys now; the WS ring path activates once the build has the new method):
- onDmCall/onGroupCall call nativeReportIncoming for native incoming calls.
- audioActivated telemetry.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Plugin connects the LiveKit room without enabling the mic (nothing captured/
published until the user taps Mic, which is also when iOS asks permission).
Web: meetMic starts false so the mic button shows muted; on callConnected the
web pushes the muted state to the plugin so builds whose plugin still connects
the mic live are muted too.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The custom call overlay was wrong — the native call must use the app's actual
meeting UI. Fix: a native call now JOINS the mesh room like any participant, so
the caller/callee tiles, roster, mute state and the whole answer/end lifecycle
run through the existing (tested) meeting code. The only native-specific bit is
meetNative=true → the WebView does NOT open its own SFU media connection (the
plugin already owns this identity's one LiveKit connection); mic/hang-up bridge
to the plugin. This fixes, via existing server code, all the reported bugs:
- "no meeting window" → the real meeting window opens on answer/outgoing.
- "caller stuck Ringing after pickup" → mesh peer-join clears the waiting tile
and finishMeetingJoin marks the call answered.
- "call still running after the other side hung up" → mesh leave ends the DM
for both (signaling leaveMeeting); plus an idempotent endDmCallByRoom backup
kicks a stuck peer when the ending side's WebSocket is down.
- "accept on one device doesn't stop the other" → markDmAnswered (fired on mesh
join) emits call-taken to the user's other sockets; deliverLocal fans to all.
- "second-device accept wins / collision" → the other device's ring is dismissed
so it can't double-join the same identity.
home.html: enterMeeting(code, audioOnly, {native, uuid}); skip sfuConnect when
native; toggleMic->plugin; toggleCam blocked (video is the next phase); leave ->
callkitEnd (guarded against the plugin's endCall re-firing).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Root cause of "answered but the call disconnects": native calls carry media
over LiveKit and bypass the mesh, so the server only learned "answered" from a
WebView POST. On a cold/locked answer the app is still launching and that event
was lost, so the 40s unanswered timer fired and cancelled the live call. Fix:
- Plugin: notifyListeners("answerCall", retainUntilConsumed:true) so a killed/
locked pickup isn't lost before the WebView JS attaches.
- Server markDmAnswered: emit call-taken to the callee's OTHER devices (stop the
ring; no teardown) and call-answered to the caller (flip UI to connected).
- Server declineDmCall: ignore a decline once the call is answered, so dismissing
a stale ring on a second device can't kill the live call.
Also adds a UI-only in-app call screen for native calls (caller + callee):
avatar, name, live timer, mute (-> plugin), end (-> CallKit). Native media has
no meeting window of its own; this covers "no meeting window / can't unmute".
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Caller saw the native callee TWICE — the outgoing 'Ringing…' placeholder (never cleared,
since a native callee doesn't send the mesh 'answered') plus a separate id-labelled tile.
Now when a native LiveKit participant joins: remove the __waiting placeholder, stop
ringback, and label the tile from CONTACTS (name + DP) instead of the raw user id.
Also (plugin, needs build): connectRoom disconnects any previous LiveKit connection
before joining, so repeated calls never leave duplicate/stale participants in the room.
WebView fix deploys now (no rebuild); the connectRoom fix rides the next build.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Native (CallKit+LiveKit) participants join the LiveKit room but not our WS mesh, so
peerIdForUid() had no mapping and sfuAttach dropped their track — the web caller stayed
on 'waiting' and never heard the native callee. Now sfuAttach/sfuDetach fall back to
keying the tile+audio by the LiveKit identity ('lk:'+id) when there's no mesh peer, and
stop the ringback. So a native<->web call crosses audio. Served — no rebuild.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Re-ring on disconnect: sendCallCancel now only fires for UNANSWERED calls. An answered
call ends via the WS event on both (awake) sides; a cancel push was re-ringing the
device that just hung up.
- Native-call telemetry (temporary): the plugin fires callConnected/callError on its
LiveKit connection; the WebView reports nc-answer/nc-connected/nc-error/nc-end/
nc-outgoing to /api/push-debug so we can see from server logs whether the native room
actually connects (no device console available). Served — no rebuild needed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Build error 'unable to resolve module dependency: LiveKit' — the LiveKitClient pod
sets no module_name, so CocoaPods names the module after the pod (LiveKitClient). The
SDK compiled/linked fine; only the import statement was wrong. Types (Room, etc.)
unchanged.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Complete the WebView/server side of native LiveKit calls:
- routes.js: /api/calls/answered (markDmAnswered) + /api/calls/end (endCallByRoom) so
the server learns a NATIVE call was answered/ended (native media runs over LiveKit,
bypassing our mesh/WS lifecycle). Additive no-ops for WebView/mesh calls.
- home.html: for native calls the WebView no longer joins the room (one connection per
identity — the plugin holds it). answerCall -> POST /api/calls/answered + clear invite;
endCall -> /api/calls/end (answered) or /api/calls/decline (still ringing). Outgoing
DM/group calls fetch a LiveKit token and hand it to NativeCall.reportOutgoingCall
instead of enterMeeting. Removed the old callHandoff mic-repush.
Server deploys now; the plugin (native LiveKit) needs a Codemagic build. Still gated by
CALLKIT_ENABLED=0 — flip to 1 only after the build is installed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The plugin now carries the call media NATIVELY: on answer it connects the LiveKit Room
from the VoIP payload's url+token and publishes the mic; outgoing calls connect via
reportOutgoingCall(url,token). CallKit stays ACTIVE for the whole call (foregrounds the
app, keeps it alive) — the mic works because LiveKit's audio runs natively and
coordinates with CallKit (unlike WebKit's WebRTC). setMuted -> setMicrophone; end ->
disconnect. Removed the handoff hack.
NOT testable yet: still need (a) WebView to stop joining the room for native calls (one
connection per identity) and drive outgoing via reportOutgoingCall, and (b) server
lifecycle endpoints for native calls (answered/ended), since native media bypasses our
mesh/WS signaling. LiveKit Swift API authored without a local compile — expect a build
iteration or two. Don't build/flip CALLKIT_ENABLED yet.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Increment 1 server side. Move livekitToken() into server/livekit.js (shared by routes.js
and calls.js). calls.js now mints a per-callee LiveKit join token and calls.js/push.js
put {livekitUrl, livekitToken} in the VoIP invite payload, so the native plugin can
connect the LiveKit room immediately on answer — even from a killed state, before the
WebView loads. No behaviour change while CALLKIT_ENABLED=0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Foundation for native LiveKit calling: the LiveKit iOS SDK carries the call media
natively and coordinates its audio engine with CallKit's AVAudioSession (AudioManager
.setEngineAvailability on didActivate/didDeactivate) — the thing WebKit's WebRTC can't
do. With this, the CallKit call stays active (stable ring + lock-screen answer +
background) AND the mic works. Native connection code + server LiveKit token in the VoIP
payload + WebView coordination come next. Don't build yet.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
CRASH: iOS terminates an app that receives a VoIP push without calling
reportNewIncomingCall. My re-ring 'fix' made the cancel path call completion() without
reporting -> crash. Always report then immediately end on cancel (a tiny ring blip is
unavoidable; a crash is worse).
MIC + earpiece: an ACTIVE CallKit call reserves the mic (WebView WebRTC gets a dead mic
+ earpiece routing). So on answer keep CallKit active only long enough to foreground the
app, then end it and fire 'callHandoff'; the WebView forces the loudspeaker and
re-acquires the mic (sfuSetMic off/on, retried while it finishes joining).
MISSED CALL: endDmCallByRoom now sends a plain missed-call banner to the callee when the
call ends unanswered (timeout / caller hung up before pickup); skipped on decline.
Server (missed banner) deploys now; plugin + web handoff need a Codemagic build.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Answering did not open the app: ending the CallKit call immediately (to free the mic)
made iOS cancel the app launch before it foregrounded. Keep the call ACTIVE on answer
— fulfilling an active-call answer is what foregrounds/unlocks the app — and fire the
answerCall event so the WebView joins. didActivate stays a no-op (don't fight WebKit's
mic). The CallKit call is ended later when the WebView call ends.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Pivot away from the native-LiveKit rewrite. Discovery: the 'voip' UIBackgroundMode
already keeps the WebView's call audio alive when backgrounded (confirmed on device),
so background audio is solved WITHOUT native media. The only issue was CallKit
reserving the mic. So use CallKit purely for the incoming RING:
- Plugin CXAnswerCallAction: fulfill, then immediately end the CallKit call
(reportCall endedAt) to RELEASE the mic, and fire answerCall to the WebView after a
~1s beat so iOS tears down the CallKit audio session first. didActivate no longer
reconfigures the session (was fighting WebKit).
- home.html: outgoing calls no longer register with CallKit (WebView-only → mic works);
incoming still rings via CallKit → hands off to the WebView on answer.
Net: native full-screen ring + working mic + background audio + all existing call
features. Needs a Codemagic build; then flip CALLKIT_ENABLED=1 to test.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
MIC BROKEN with CallKit: a CallKit call reserves the microphone, so the WebView's
WebRTC can't capture it — calls are unusable until native LiveKit media lands. Add a
server kill-switch (CALLKIT_ENABLED, default OFF) so CallKit can be flipped without an
app rebuild: /api/meetings/config now returns callkit; setupNativeCall bails when off
(-> WebView calls, mic works); push.js only sends VoIP/CallKit pushes when enabled.
Deploying with the flag unset immediately restores working WebView calls.
RE-RING: a late cancel push for an already-declined call hit the plugin's 'unknown
uuid' path and re-reported a fresh incoming call (~1s re-ring). Track endedCalls and
make a late cancel for an already-ended call a no-op.
Server part deploys now (no rebuild); plugin re-ring fix ships with the native build.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Bug: a killed/backgrounded callee is woken only for the CallKit ring and has no
WebSocket yet, so the existing dm-call active:false (WS-only) never reaches it and
it keeps ringing after the caller hangs up.
Fix: send a 'cancel' VoIP push on call teardown.
- push.js: sendCallCancel() sends a {type:'cancel',callUUID} VoIP push to the user's
ios-voip tokens; invites now carry type:'invite'.
- calls.js: endDmCallByRoom + endGroupCallByRoom fire sendCallCancel to the rung users.
- NativeCallPlugin: on a cancel push, end the reported call (reportCall endedAt); if the
invite was never seen, report-then-end to satisfy iOS's 'report a call per VoIP push'.
Server part deploys now; the plugin part needs the next Codemagic build.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- didReceiveIncomingPush: normalise payload.dictionaryPayload ([AnyHashable:Any])
to [String:Any] before storing/using (the reported build error at :121).
- CXProviderConfiguration(localizedName:) instead of the no-arg init (available on
all deployment targets, avoids an availability edge).
- Make two 'calls[uuid] ?? [:]' bindings explicitly [String:Any] to avoid empty-
literal inference ambiguity.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The native call feature (iOS). Backward-compatible: without the plugin (current
builds) nativeCallOn() is false and every CallKit branch is skipped, so web/older
builds behave exactly as before.
Native (mobile/plugins/native-call, a local Capacitor plugin like audio-route):
- PushKit: registers for VoIP pushes, reports the VoIP token to JS (-> /api/v1/devices
'ios-voip'). On an incoming VoIP push, reports a CallKit incoming call (full-screen
ring, works when the app is force-killed).
- CallKit: answer/decline/end -> events to JS; configures the call AVAudioSession on
didActivate so the WebView's WebRTC audio rides a call-priority session (background).
- Outgoing calls register with CallKit too (reportOutgoingCall) so they get the same
active-call background-audio context.
- NativeCall.podspec (frameworks CallKit/PushKit/AVFoundation); added to mobile deps;
ios-patch.sh now sets UIBackgroundModes = [audio, voip] (voip required for PushKit).
Web bridge (home.html): setupNativeCall() registers the VoIP token, joins on CallKit
answer, leaves/declines on CallKit end; on CallKit devices the in-app call-invite popup
+ WebAudio ring are suppressed (the system rings instead); outgoing calls are reported
to CallKit; call-end events dismiss the CallKit call. calls.js threads a stable call
uuid through the dm-call/group-call WS events + start responses so both sides can match
the CallKit call.
Needs a Codemagic build to compile the plugin; first on-device iteration expected.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
First slice of the native call feature. Backward-compatible: with no VoIP tokens
registered yet it behaves exactly like today's banner push.
- push.js: sendApnsVoip() sends a PushKit VoIP push (apns-push-type 'voip', topic
<bundle>.voip, reusing the same .p8) to wake a killed app for CallKit; and
sendCallNotification() which PREFERS a VoIP push when the user has an 'ios-voip'
token, else falls back to the normal alert/banner push (Android/web/pre-CallKit iOS).
- routes.js: /api/devices now accepts platform 'ios-voip' (the PushKit token, stored
alongside the normal alert token in device_tokens).
- calls.js: each call now carries a stable crypto.randomUUID() (CallKit needs a UUID
to report + later cancel the call); DM and group call notifications route through
PUSH.sendCallNotification instead of the raw banner push.
Next: the native-call Capacitor plugin (PushKit + CallKit + LiveKit iOS SDK).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Notifications were silent despite the payload requesting sound:'default' and correct
device settings. Ship an explicit tone: a generated PCM .wav is bundled into the app
(ios-patch.sh copies it in; add-share-extension.rb adds it to the App target's Copy
Bundle Resources, tolerantly) and the server now sends sound:'notif.wav'. Part of the
consolidated iOS build alongside the background-audio + call-push + AppDelegate fixes.
Only affects iOS-native-app tokens (currently just the one test device); web push
ignores the field. Needs a fresh Codemagic build for the bundled file to exist.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Capacitor/WebView apps get suspended by iOS a few seconds after backgrounding, which
freezes the WebRTC mic + audio pipeline — so when the user minimised the app or locked
the phone during a call, no one could hear anyone. Add UIBackgroundModes=[audio] to
Info.plist (via ios-patch.sh) so iOS keeps the audio session (and the app) alive while
a call is actively playing/recording. Video rendering still pauses in the background
(unavoidable in a WebView) but voice continues. Needs a fresh Codemagic build.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Calls only notified over the chat WebSocket (CHAT.pushToUser), so a CLOSED app
(no live socket) never rang — unlike messages, which also call PUSH.sendToUser.
Add PUSH.sendToUser for both DM (startDmCall -> callee) and group (startGroupCall
-> other members) so APNs/FCM/WebPush alerts a closed device.
To make the alert actionable, add CALLS.replayActiveCalls(userId, ws), invoked
from the chat-hello handler: when a socket (re)connects, re-send any dm-call /
group-call the user is currently being rung into (the original events fire once at
call start and are missed by an app that was closed). Opening the app from the push
then re-surfaces the invite so they can answer within the ring window.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
ROOT CAUSE of iOS push not working: Capacitor 7's default AppDelegate.swift
template does NOT implement application(_:didRegisterForRemoteNotificationsWithDeviceToken:)
or ...didFailToRegisterForRemoteNotificationsWithError:. So when @capacitor/push-
notifications calls registerForRemoteNotifications(), iOS fetches the APNs token
and calls the AppDelegate, but nothing posts .capacitorDidRegisterForRemoteNotifications,
so the plugin never delivers the token to JS. register() 'succeeds' yet neither the
registration nor registrationError event fires — proven by server push telemetry
(register-called logged; no token, no error; device_tokens stayed empty).
inject-push.js adds the two forwarding methods to the CI-generated AppDelegate
(idempotent, tolerant — never fails the build), wired into ios-patch.sh after the
audio patch. Verified against the real Capacitor 7 template: methods land inside
the class, braces balance, both listeners present.
This is the missing piece alongside the earlier aps-environment entitlement fix and
the server APNs config. Needs a fresh Codemagic build to take effect.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
device_tokens stays empty after reinstall+Allow, so the APNs token is never
obtained or never reaches the server, and there's no device console on Windows.
Add a /api/push-debug collector and breadcrumbs through setupNativePush (plugin
presence, permission state, register call, registration event/error, token POST
result) so the failing step is visible in server logs. Temporary — remove once
push is confirmed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The @capacitor/push-notifications plugin does not add the Push Notifications
capability to the CI-generated Xcode project (that's a manual Xcode step), and
add-share-extension.rb only merged the App Group into App.entitlements, assuming
aps-environment was already there. It never was — so on device PushNotifications.
register() failed with 'no valid aps-environment entitlement', no APNs token was
obtained, and device_tokens stayed empty (server had nothing to push to).
ios-patch.sh now creates App/App.entitlements with aps-environment=production
before the share-extension script merges the App Group in. Still requires the App
ID to have Push Notifications enabled (so the profile carries the entitlement) and
the server APNS_* key set (Step 5) for end-to-end delivery.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Root cause of a post-deploy outage: recreating the app container can assign it a NEW
docker network IP. Nginx Proxy Manager caches the app's upstream IP at config-load,
so it kept connecting to the OLD IP (which, after adding the postgres/redis services,
was reassigned to bizgaze-postgres) → 'Connection refused' → site down until nginx
re-resolved. deploy.sh now runs 'nginx -s reload' in the NPM container after verify
(normal + rollback), non-fatal if NPM isn't detected. This makes deploys self-healing
for the app-IP-change case.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Two things:
1. FIX a live regression the async conversion missed: chat.js calls repos via the
lazy repos() helper (not the R. prefix), so my sweep skipped it — effectiveStatus
/ broadcastPresence read `repos().users.byId(userId)` synchronously, but that's a
Promise now, so presence broadcasts always reported status 'active' and dropped
last_seen. Now awaited (effectiveStatus/broadcastPresence async); touchSeen is a
fire-and-forget UPDATE with .catch. Audited all non-R. repo calls — only chat.js
was affected (media.js backfill was already awaited).
2. Swappable pub/sub for multi-instance real-time fan-out (the actual blocker to
running >1 instance — not the DB). server/pubsub.js picks a backend by
PUBSUB_BACKEND (default 'memory'). Local socket delivery is UNCHANGED; publish is
additive — memory = no-op (zero hot-path cost, identical single-instance
behaviour), redis = fan-out to other instances with a self-echo guard. chat.js
pushToUser/broadcastPresence now also publish; each instance subscribes to deliver
remote events to its local sockets. Interface is tiny so Redis is one swappable
file (Postgres LISTEN/NOTIFY or NATS could drop in the same way — never hardwired,
as requested). Dormant redis service added to compose behind the 'scale' profile;
redis dep added; PUBSUB_BACKEND/REDIS_URL documented.
Validated: smoke 22/22 (memory), e2e chat delivery green. NOTE: full multi-instance
also needs distributed presence (isOnline is per-process) + meeting-signaling
sharing — chat/presence fan out via this layer; those are follow-ups.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds a dedicated Postgres 16 service (container bizgaze-postgres, own named volume
bizgaze_pg_data, healthcheck) on the shared NPM network. The app depends_on it
healthy. Inert until DB_BACKEND=pg is set in .env — default stays SQLite, so this
deploy changes nothing functionally; it just makes the engine available. Documented
POSTGRES_PASSWORD / DATABASE_URL / DB_BACKEND in .env.example.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The DM/group thread queries used `(? IS NULL OR created_at < ?)` — an all-NULL
param Postgres can't type ('could not determine data type of parameter') — and an
unaliased FROM-subquery (Postgres requires an alias). Both rewritten to add the
`created_at < ?` clause only when a cursor is given, and alias the subquery `t`.
Portable; sqlite db-smoke still 22/22.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- db/pg.js: the pg backend (prepare/exec/tx/init) — ?→$N translation, BIGINT parsed
as Number (matches sqlite; else expires_at<Date.now() compares string<number),
transactions on one pooled client, init() applies schema.pg.sql. Same interface as
db/sqlite.js, so repos are unchanged.
- repos.js: the ~7 SQLite-only queries rewritten to run on BOTH engines —
audit.add @named→positional; email lookups COLLATE NOCASE→LOWER()=LOWER();
INSERT OR IGNORE→ON CONFLICT DO NOTHING (addMember/poll vote/favorite);
mergeInto's UPDATE OR IGNORE→UPDATE…WHERE NOT EXISTS/NOT IN and INSERT OR
REPLACE→ON CONFLICT DO UPDATE. Re-validated on sqlite: db-smoke still 22/22.
- server.js: boot now `await db.init()` before listening (pg creates tables; sqlite
no-op), so the first request can't hit a missing table.
- db/migrate-sqlite-to-pg.js: one-shot row copy in FK order (bulk insert, TRUNCATE
first so re-runnable). audit_log id left to PG's identity.
- package.json: add pg ^8.13.1.
Next: validate DB_BACKEND=pg smoke against a real Postgres on the server, then merge.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The full sync→async conversion is complete and green on the SQLite backend. Every
DB call across the app now awaits the async adapter, so the identical code runs on
Postgres at cutover.
Converted (this commit finishes Phase 3):
- session.js: currentUser/apiKeyFromReq async → 63 route awaits + WS + static.
- routes.js: all ~250 R.* awaited; DTO helpers (namesFor, avatarsFor, buildMsgDTO,
buildPollDTO, reactionsForMessage, postSystemMessage, pushGroupUpdate,
issueRefreshToken, provisionFromBizgaze) made async; every `.map(x=>buildDTO(x))`
restructured to `await Promise.all(...map(async...))` preserving order; `.filter`
predicates that hit the DB moved to an `asyncFilter` helper; chained
`R.x.y(...).length/.map/.filter` wrapped as `(await R.x.y(...)).method`; stream
upload handlers (recording/transcript/attachment) made async.
- calls.js / signaling.js: all call/meeting fns async; leaveMeeting AWAITS
persistCallHistory + finalizeTranscript BEFORE endCallByRoom (ordering matters —
fire-and-forget would race the map teardown); WS handle()/cleanup() async with
.catch guards.
- static.js: authAttachment(Raw) async (the .some carrier check became a loop),
handleGet async; server.js dispatch catches handler rejections → 500 not a hang.
- media.js backfill, push.js, reminders.js, webhooks.js await their repo calls.
Validation on DB_BACKEND=sqlite: db-smoke 22/22; legacy e2e 80 checks pass with zero
FAILs (throws only at a PRE-EXISTING WS lobby-drift assertion, unrelated). Every
server file `node --check` clean.
Still on the branch — master untouched. Next: Phase 5 (pg backend + ~7 dialect
queries + data migration + Docker Postgres + cutover), then merge.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
On the db-migration branch only; master stays clean + deployable. Foundation
(adapter, pg schema, smoke harness) is already on master and safe.
Done:
- repos.js fully async (Phase 2, validated: node --check clean, no missed transforms).
- session.js currentUser/apiKeyFromReq async.
- Mechanical `await` prefix applied across routes/static/calls/signaling/reminders/
webhooks/push.
Remaining (does NOT compile yet — deterministic to finish):
1. Async cascade: helper fns that now contain `await` must be marked async and their
callers awaited. node --check points to each (namesFor, authAttachmentRaw/
authAttachment in static, the WS handlers in calls/signaling, reminders/webhooks
loops).
2. DTO builders are the real work: namesFor, avatarsFor, buildPollDTO, buildMsgDTO,
recDTO all became async — every `.map(x => buildMsgDTO(...))` etc. must become
`await Promise.all(arr.map(async x => ...))`.
3. Chained calls `R.x.y(...).map/.length/.includes` → `(await R.x.y(...)).method`.
4. Then: node --check all green → node test/db-smoke.js green → e2e → merge to master.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
server/dbx.js selects a backend by DB_BACKEND (default sqlite; pg added at cutover).
server/db/sqlite.js wraps the synchronous node:sqlite instance in the async
interface repos will call — prepare(sql).{get,all,run}, exec(sql), tx(fn), init().
Results come back as resolved Promises so identical repo code runs on synchronous
SQLite (dev/test) and asynchronous Postgres (prod).
tx() gives multi-statement atomicity that stays correct on both engines (sqlite is
single-connection; the pg backend will run it on one pooled client) — needed for the
account-merge transaction in repos.
Verified: get/all/run/tx all work end-to-end; confirmed no code reads
.changes/.lastInsertRowid, so the repo conversion is purely sync->Promise. Unwired —
nothing requires dbx.js yet; prod path untouched.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Covers the DB-backed HTTP paths the async repo conversion touches — auth, users,
messages, attachments, conversations/groups, reactions, mentions, edit/delete,
polls, scheduled meetings (paginated), favorites, audit — asserting current API
shapes. Runs to completion with a pass/fail count and honours DB_BACKEND so it
doubles as the sqlite-vs-pg parity check at cutover. No WS/signaling (in-memory,
not the DB).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
server/db/schema.pg.sql — the full Postgres DDL, every column defined up front (no
ALTER-ordering fragility). SQLite→PG type mapping documented in-file (epoch-ms
INTEGER→BIGINT, 0/1 flags→SMALLINT kept numeric so app code is unchanged, sizes→
BIGINT, audit rowid→GENERATED IDENTITY). Mirrors the three existing FKs and adds a
new idx_messages_attachment (the /files auth scan we cached earlier becomes a keyed
lookup).
Validated against a throwaway Postgres 16: loads with no errors, 24 tables + 44
indexes created. Unwired — nothing uses it yet; the SQLite path is untouched.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Migration prep. Two real fixes surfaced while building a regression harness:
1. db.js: the `guest_emails` and `lobby` ALTER TABLEs sat at lines 241/244, BEFORE
scheduled_meetings is CREATEd (line 299). On a FRESH database the ALTER fails
(no table yet), is swallowed by the try/catch, and the columns are never added —
so a brand-new deploy is missing them and scheduling with guests crashes. Prod
escaped it only by incremental deploy history. Moved both ALTERs to after the
CREATE. (The upcoming Postgres schema defines every column up front, so this
whole class of ordering bug goes away there.)
2. test/e2e.js: /api/meetings returns paginated `{list, pastTotal, page, pageSize}`
now, not a bare array — updated three `.data.find` → `.data.list.find`.
No prod behaviour change (prod already has the columns; ALTERs are idempotent).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The top-bar buttons looked heavy against the navy bar (thick grey X-circle, bold
"Send" pill). Swapped for light SF Symbols on the navy bar: a thin `xmark` for
cancel and a `paperplane.fill` for send (semibold, enables when ≥1 chat is picked).
Icon-only send matches the Teams reference — the radio checks already show what's
selected, so the "(N)" count text is dropped. Native-only — needs a build.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Two things from testing (the in-sheet picker with avatars + multi-select works):
1. Double "Send to": after the extension sent, opening the app ALSO popped the web
"Send to" modal for the same file. Cause: the extension wrote a safety-net
manifest up front, which the app then picked up. Now the manifest represents an
UNSENT share only — written solely when the extension can't send (no token) or a
send fails. A successful in-sheet send clears the staged files and leaves nothing,
so the app never re-offers it. Cancel also clears staged files (no orphans).
2. Layout aligned to the Teams reference:
- Preview strip of thumbnails for what's being shared (image → the image, video →
first frame via AVAssetImageGenerator, else a doc icon).
- Radio selectors on the right — an always-visible empty circle that fills to a
navy check when selected (clearer multi-select than an appear-on-select tick).
- "Recent chats" section header; subtitle under each name (Direct message /
Group · N members).
- Clearer branding: bold white "Share to Biz Connect" on the navy bar.
Native-only — NEEDS A NEW iOS BUILD. Balance + selectors checked.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Three follow-ups on the in-sheet picker:
- Profile photos: rows now show the real avatar (fetched from the conversations
API's `avatar` field with the bearer token, or a data-URL decoded inline),
rendered as a circle; coloured initials as the fallback — matching the app.
- Multi-select: tap toggles a checkmark instead of sending immediately; a "Send (N)"
button in the nav bar sends to every selected chat. Each file is uploaded ONCE and
its attachment id reused across all targets (the server allows the uploader to
reattach the same id), so multi-send doesn't re-upload.
- Branding: navy (#1F3B73) navigation bar with a white "Biz Connect" prompt over the
"Send to…" title and white controls.
Native-only — NEEDS A NEW iOS BUILD. Balance + selectors checked.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replaces the wrong approach (stage + try to bounce into the app, which iOS blocks)
with the one Teams/WhatsApp actually use: the picker and the send happen INSIDE the
share extension, so there's no app-open at all. Tap Share → Biz Connect → pick a
chat → it uploads and sends, right there in the sheet.
How the extension can send without the app: it's a separate process that can't see
the web app's HttpOnly cookie, so:
- server: GET /api/share/token mints a bearer token for the logged-in user.
- web: on every launch the app fetches that token and hands it to the extension via
the App Group (ShareInbox.setAuth writes token+base to the shared UserDefaults).
- extension: reads the token and calls the SAME API the native client uses —
GET /api/messages/conversations to list chats, POST /api/messages/upload for each
file, POST /api/messages to send. Native UITableView picker with search.
Robustness: it still stages the files + writes a manifest first, so if there's no
token yet (user never signed in) or the send fails, the file isn't lost — the app
collects it on next open, exactly as before. On success the manifest is cleared so
the app doesn't re-offer it.
Server + web are live now; the token endpoint is harmless until a build ships the
extension. NEEDS A NEW iOS BUILD for the picker itself.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A Share Extension opening its host app is unsupported on modern iOS (restricted
~iOS 14), so the programmatic bizconnect://share open is silently blocked and the
user just saw a blank flash back to Photos — looking broken even though the files
staged fine.
The extension now shows a small native card after staging: "✓ Ready to send — Open
Biz Connect to choose a chat", with an "Open Biz Connect" button (user-initiated
open has the best chance of working) and a Done button. It still attempts the
auto-open first. Either way the app collects the staged files when next opened, so
the manual path that already works is unchanged — this just removes the "did it
even work?" confusion.
Renamed the local `staged` array to `collected` to free `staged` for the state
flag. Balance + selectors checked. NEEDS A NEW iOS BUILD (native change).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Sravan