CRITICAL (5):
- Hub panic recovery now calls h.Stop() after 3 panics (ws/hub.go)
- Ring buffer EventsSince returns non-nil empty slice for current seq (ws/ringbuffer.go)
- PTT event listener stores unsubscribe handle to prevent leak (ptt.ts)
- verifyTotp respects config.allowSelfSigned instead of hardcoding (api.ts)
- ptt_listen_for_key uses spawn_blocking to avoid thread pool starvation (ptt.rs)
HIGH - Server (13):
- TOTP rate-limit checked after body decode; counters reset on success
- TOTP enable returns 409 if already enabled (must disable first)
- Global search pre-computes accessible channel IDs for FTS WHERE clause
- DeleteAccount queries roles by name instead of hard-coded IDs
- BackupToSafe uses absClean in VACUUM INTO
- Voice camera slot uses atomic EnableCameraIfUnderLimit DB method
- readPump snapshots voiceChID before unregister for TOCTOU safety
- Voice join sets state after token send; rollback takes broadcast flag
- Updater download uses probe pattern instead of overflow write
- Webhook checks Authorization header before reading body
- Storage.Save adds fsync and fixes double-close
- Default WS origin denies cross-origin (was: accept all)
HIGH - Client (6):
- WS reconnect uses generation counter to discard stale events
- AudioPipeline uses generation counter against stale worklet callbacks
- Screenshare mute state preserved across reconnect (not full leave)
- handleVoiceToken uses iterative loop instead of unbounded recursion
- store.ts re-entrancy guard with pending update queue
- Notification AudioContext cleaned up on logout
Reviewed by 4 parallel agents across Server Core, Server Realtime,
Client & Tauri, and Security. 55 total findings; 24 CRITICAL+HIGH
fixed here, 31 MEDIUM+LOW tracked in vault backlog (T-265–T-295).
- Video grid: sync stream type attribute on updates, add screenshare data attribute
- Dispatcher: handle voice_token messages, improve video track event handling
- LiveKit session: add video track publication support
- Hub: stale client timeout cleanup, improved voice state management
- Voice join/leave: context propagation, better error handling
- Livekit webhook: structured event handling with room/participant data
- Server DB: voice query improvements, new test coverage
- WS integration tests: expanded coverage for voice and LiveKit flows
- Gitignore: add internal dev tools directory, owncord-server.exe
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Fix race condition: handleDisconnected no longer nulls the room during
initial connect, allowing the retry loop to complete all 3 attempts
- Fix TLS proxy port: default to 443 instead of 8443 when server host
has no explicit port (servers behind nginx/reverse proxy)
- Fix cert store key: strip :443 suffix so LiveKit proxy fingerprint
lookup matches ws_proxy's stored key format
- Add node_ip config option for LiveKit WebRTC ICE candidates (required
for remote users behind NAT)
- Add resolved URL to all connection error/retry/reconnect logs for
easier debugging
- Add diagnostic logging to resolveLiveKitUrl showing which path was
taken (direct/proxy/passthrough)
Security fixes (from multi-reviewer code review):
- Add DM participant auth checks to channel_focus, typing, and REST
message endpoints — prevents unauthorized access to DM channels
- Fix TOCTOU race in GetOrCreateDMChannel using IMMEDIATE transaction
- Validate YAML credentials before LiveKit config interpolation
- Add CSS variable injection prevention in custom theme loader
- Validate localStorage JSON before unsafe type casts
LiveKit stability:
- Track remote mic audio elements for cleanup on abnormal disconnect
- Remove duplicate token refresh timer scheduling
- Add .catch() to all floating applyMicMuteState promises
- Clear reconnectAc after async post-connect work completes
- Fix double cmd.Wait() race in LiveKit process Stop()
- Reorder voice_join guards: validate channel before livekit==nil check
- Add startup warning for external LiveKit webhook CIDR mismatch
DM system fixes:
- Emit dm_channel_close WebSocket event from REST close handler
- Re-open DM for caller when channel already exists
- Fix unread count incrementing for own messages and active DMs
- Reset channelBeforeDm after Back navigation (stale state bug)
New feature:
- Voice call duration timer in VoiceWidget (MM:SS / HH:MM:SS elapsed)
- Accent color restored on app startup (was only applied in settings)
Test infrastructure:
- Add DM tables to all test schemas (hubTestSchema)
- Inject test LiveKit client in voice handler tests (fixes 28 failures)
Task A: REST API endpoints in api/dm_handler.go — POST /api/v1/dms
(create/get DM channel), GET /api/v1/dms (list open DMs), DELETE
/api/v1/dms/{channelId} (close DM). Routes registered in router.go.
Task B: WebSocket DM routing — handleChatSend, handleChatEdit,
handleChatDelete, and handleReaction all check channel type and use
participant-based auth for DM channels instead of role permissions.
DM messages delivered via SendToUser to each participant (bypassing
channel-subscription model). Auto-reopens DM for recipient on new
message with dm_channel_open event. New broadcastToDMParticipants
helper. New WS event types: dm_channel_open, dm_channel_close.
Task C: Ready payload includes dm_channels from GetUserDMChannels.
Also adds GetDMParticipantIDs helper to db/dm_queries.go.
- Deafen now also mutes the local microphone (privacy fix)
- Mute/deafen zero the audio pipeline GainNode to guarantee silence
when the replaced sender track bypasses LiveKit's track disable
- Send voice_leave on failed auto-reconnect to prevent ghost states
- Gate voice_join on LiveKit availability (reject if h.livekit == nil)
- IP-restrict /api/v1/livekit/webhook to admin CIDRs
- Warn at startup when LiveKit API keys are auto-generated (ephemeral)
- Fix connecting guard: move pending-join dispatch outside finally block
- Add volume slider + mute button overlay on remote video tiles
- Document token refresh limitation for 4h+ sessions
- Fix TS2306 in rnnoise-worklet test (ts-expect-error for worklet import)
Server fixes:
- voice_leave: broadcast voice_leave even on DB error so peers don't see ghost users (H1)
- migrate: record migration inside transaction for atomicity (H2)
- password: use init() with panic for dummyHash to catch bcrypt init failures (M1)
- password: replace //nolint:errcheck with explicit _ = discard (M2)
- handlers: clarify edit permission comment, fix error message wording (M3)
- auth_handler: distinguish duplicate username (400) from DB error (500) (M4)
Client fixes:
- media: revert YouTube oEmbed to browser fetch — no need to disable cert verification (C1)
- messages.store: fix prependMessages cap to keep newest messages, not oldest (H5)
- ChannelSidebar: ref-count globalDragAc to prevent multi-instance teardown race (H6)
- attachments: replace console.error with project logger (M6)
- ws: clarify lastSeq reset comment to match actual behavior (M5)
CRITICAL:
- Add AuthMiddleware + rate limiting to /livekit/* proxy route (was unauthenticated)
- Remove well-known default LiveKit credentials from source; auto-generate unique
random keys on first run so voice works out of the box securely
- Reject the old "devkey"/"owncord-dev-secret-key-min-32chars" in NewLiveKitClient
HIGH:
- Add 5s timeouts to RemoveParticipant/ListParticipants SDK calls (goroutine leak)
- Fix config.yaml default file permissions from 0644 to 0600
- Fix voice store desync on unexpected LiveKit disconnect (phantom UI state)
- Add in-flight guard to handleVoiceToken (race on rapid channel switch)
- Fix lightbox listener leak on rapid reopen (orphaned mousemove/mouseup/keydown)
- Fix allTracked WeakRef set unbounded growth in media-visibility
- Replace debug console.log with createLogger in embeds.ts
- Stop persisting password in Windows credential blob (only token needed)
- Detach existing audio elements before attaching to prevent double playback on reconnects
- Remove webAudioMix to eliminate Web Audio overhead compounding with multiple participants
- Use participant.setVolume() for full 0-200% per-user volume range
- Add pli_throttle and active_loopback_prevention to LiveKit server config
- Bump version to 1.3.0
Server: suppress errcheck on deferred Close() calls, discard
resp.Body.Close error, remove unused voiceSpeakersPayload type
and buildVoiceSpeakers func.
Client: add unit tests for os-motion, livekitSession, and
safe-render to bring coverage from 74.16% to 76.06% (threshold 75%).
- Add LiveKit health check guard to voice_join (reject if process not running)
- Log webhook DB cleanup errors instead of swallowing
- Reject camera enable if CountActiveCameras query fails
- Add LiveKit health status to /metrics endpoint
- Add 13 unit tests for livekit.go, livekit_process.go, livekit_webhook.go
Server:
- Extract image width/height on upload via image.DecodeConfig (header-only)
- Add width/height columns to attachments table (migration 007)
- Include dimensions in AttachmentInfo JSON (optional, backward-compatible)
Client:
- Reserve exact space for images using server-provided dimensions
- Fallback min-height: 200px for external/old images, cleared on load
- Remove premeasureAll() — was caching wrong heights for unloaded images
- Smart per-type height estimates: 32px dividers, 42/72px text, +220px
per image attachment, +320px for YouTube embeds
- Batch ResizeObserver corrections to single RAF with anchor-based scroll
preservation (topmost visible item stays in place)
- Fenwick tree for O(log n) offset lookups (replaces O(n) linear scans)
- CSS contain: layout style on .msg-image to isolate layout shift
Server:
- Fix YAML injection in LiveKit config generation (quote values)
- Revert token TTL to 4h (no server-side JWT revocation)
- Derive LiveKit publish permissions from user role (prevent SFU bypass)
- Add CAS guard for webhook/voice_leave race condition
- Add voice_leave broadcast to rollbackVoiceJoin (prevent ghost state)
- Limit webhook body to 64KB (prevent memory abuse)
- Add rate limit to voice_token_refresh handler (1/60s)
- Add LiveKit health check endpoint (GET /api/v1/livekit/health, 503 on degraded)
- Add voice_token_refresh WS handler for client-initiated token refresh
- Consolidate voice quality constants (single source of truth)
- Fix video limit TOCTOU race (count from DB instead of LiveKit API)
- Raise default voice_max_video from 10 to 25 (Discord parity)
- Add CountActiveCameras DB query
- Non-blocking broadcast send, circuit breaker, exponential backoff
- Close send channel before context cancel in serve.go
- Guard voice mute/deafen for active channel
- Delete orphaned message on attachment link failure
- Redact query string from proxy logs (prevent token leak)
- Use instance-level HTTP client for health checks (no redirect following)
- Set cmd.WaitDelay to prevent goroutine leak on Windows
- Log buildJSON marshal errors
Client:
- Refactor livekitSession.ts from singleton module to LiveKitSession class
- Share single AudioContext for all analysers (was 1 per participant)
- Extract createRoom() helper (DRY)
- Add token refresh timer (3.5h interval, re-arms on failure)
- Skip setSpeakers if unchanged (sort in-place, no allocations)
- Distinguish user-initiated leave from connection error in retry
- Add YouTube videoId validation (prevent iframe src injection)
- Add try/finally to disableCamera
- Wrap store subscription callbacks in try/catch
- Track and cancel initial scroll RAF on cleanup
- Add 5s timeout + encodeURIComponent to YouTube oEmbed fetch
- Clean raw mic stream on RNNoise suppressor failure
- Full voice cleanup on logout via cleanupAll()
Tests:
- Add 7 new server tests (webhook parsing, voice guards, quality fallback)
- Fix 2 pre-existing test failures (mute/deafen invalid payload)
Architecture:
- Add subscribeSelector to store.ts for selective state subscriptions
- Create reconcileList utility for DOM list patching without rebuild
- Create shared createContextMenu utility (dedup 3 files)
- Convert all 20 subscribe() calls to subscribeSelector across 11 files
- Split renderers.ts (1131L) into 7 focused files by concern
- Split ConnectPage.ts (838L) into ServerPanel + LoginForm + shell
- Fix ineffective (s) => s selector in ChannelSidebar
GIF visibility:
- Add media-visibility.ts with IntersectionObserver + canvas snapshots
- GIFs auto-pause after 10s, play/pause button overlay on hover
- Freeze GIFs on scroll-away, window blur, and minimize
- Wire into media.ts, attachments.ts, embeds.ts renderers
Tests: 46 new tests (1073 total), all passing
Net: -1166 lines across client codebase
Camera fixes:
- Optimistic setLocalCamera(true) before async setCameraEnabled for instant
button highlight, with revert on failure
- Only call checkVideoMode when camera-relevant state changes, not on every
speaking poll tick (100ms)
- VideoGrid.addStream updates existing cells in place instead of
destroy+recreate to prevent black frame flashes
- VideoModeController tracks localTileAdded to avoid redundant addStream calls
Security & hardening (from prior session review):
- Settings store key allowlist prevents arbitrary key writes
- Certificate fingerprint validates SHA-256 colon-hex format
- CredentialData Debug impl redacts token and password
- LiveKit config file written with 0600 permissions
- Token TTL reduced from 24h to 4h
- Null-check on client.user before token generation
- Thread-safe getChannelID/trySendMsg helpers on Client
- Warn on default dev LiveKit credentials
- Devtools feature-gated behind cfg(feature = "devtools")
- Updater uses configure_client for self-signed cert acceptance
- Clamp voice sensitivity input to 0-100 range
- Clear lastConnectToken/Host on logout
- Fix animation frame leak in VoiceAudioTab mic meter
- Add http://localhost:* and ws://localhost:* to CSP connect-src
(WebView2 was blocking LiveKit signal connection)
- Add connection retry (3 attempts, 2s delay) for LiveKit server
startup race condition ("could not find any available nodes")
- Remove broken TURN TLS config from generated livekit.yaml
- Send direct LiveKit URL instead of proxy path (localhost is
treated as secure context in Chromium/WebView2)
- Set LiveKit server host from API config for URL resolution
- Add reverse proxy at /livekit/* that forwards to LiveKit server
- Server sends relative URL "/livekit" in voice_token; client
resolves to wss://server:port/livekit using known server host
- Fix API secret minimum length (32 chars required by LiveKit)
- Pass TLS config to LiveKit process manager for TURN certs
- livekit_webhook.go: HTTP handler for LiveKit participant_joined/left
events with JWT verification; syncs stale voice state on crash recovery
- Mount webhook at POST /api/v1/livekit/webhook in router
- Speaker detection handled client-side via LiveKit SDK events
- Replace VoiceConfig STUN/TURN/MediaPort fields with LiveKit
API key, secret, URL, and binary path
- Add livekit_process.go: companion process manager with
auto-restart and config generation
- Add livekit.go: token generation, room service client,
participant management, video track counting
- Add LiveKit server SDK dependency (server-sdk-go/v2)
- Update config tests for new VoiceConfig fields
- Fix SDP signaling race condition: add per-client negoMu to serialize
renegotiateParticipant / handleVoiceOffer / handleVoiceAnswer so
concurrent OnTrack goroutines don't race through rollback
- Fix handleVoiceLeave triple-fire: early return when clearVoice()
returns zeros so ICE callbacks don't re-enter and corrupt state
- Fix SQLite SQLITE_BUSY errors: add busy_timeout=5000 pragma and
SetMaxOpenConns(1) for file-based databases
- Fix deafen bypass: new remote audio elements now respect localDeafened
state so late-arriving streams are muted immediately
- Add debug-level logging for SDP negotiation, track fan-out, ICE
candidates, voice state changes, room lifecycle, and participant
add/remove
- Bump version to 1.1.1
Audio and video tracks from the same user were both using stream ID "user-{id}",
causing the client's duplicate stream check to silently drop the video track.
Changed to "user-{id}-audio" and "user-{id}-video" so each gets its own stream.
- setupOnTrack now handles both audio and video tracks via kind branch
instead of returning early on non-audio tracks
- RTP forwarding works identically for both kinds; speaker detection
only runs for audio tracks
- handleVoiceCamera enforces MaxVideo limit before DB update, rejecting
with VIDEO_LIMIT error when the cap is reached
- Added TestHandleVoiceCamera_MaxVideoEnforced test
- CRIT-2: Prevent HTTP header injection in Content-Disposition via mime.FormatMediaType
- HIGH-1: Call hub.GracefulStop() on server shutdown to clean up WebSocket/voice state
- HIGH-2: Remove double send on serveErr channel that caused goroutine leak
- HIGH-3: Handle GetUserByID error after registration to prevent nil panic
- HIGH-4: Return nil,nil on sql.ErrNoRows in GetAttachmentByID (matching codebase convention)
- HIGH-5: Add voiceDone channel to bound RTP goroutine lifetime on voice teardown
- Fix handleVoiceICE to return VOICE_ERROR when no PC (consistent with other voice handlers)
- Update tests to match corrected behavior
Voice was broken over NAT due to multiple issues across the audio pipeline:
- Fix GainNode silence: WebView2 silences remote WebRTC streams routed through
Web Audio createMediaStreamSource→GainNode→createMediaStreamDestination.
Replaced with direct HTMLAudioElement playback for remote audio.
- Fix NAT traversal: Add Google public STUN server (stun.l.google.com:19302)
so remote clients can discover their public IP for ICE connectivity.
- Fix signaling race: Catch createOffer InvalidStateError when server
renegotiation offer arrives before client's initial offer is sent.
- Fix device switch: Use replaceTrack() instead of removeTrack+addTrack
to avoid SDP renegotiation. Safe rollback on failure (stop old last).
- Fix speaking flicker: setSpeakers skips local user (VAD is sole authority).
- Fix VAD sample rate: Force 48kHz AudioContext instead of system default
(192kHz) which spread FFT bins too wide for voice frequency detection.
- Fix CSP for WASM: Add wasm-unsafe-eval to script-src for RNNoise.
- Fix clearAuth leak: leaveVoice() called before resetVoiceStore().
- Fix ICE rate limit: Separate limit for ICE candidates (50/s vs 20/s).
- Fix stale ICE errors: Silently drop voice_ice with no PeerConnection.
- Fix audio play() race: Deferred to queueMicrotask after DOM attachment.
Debugging infrastructure:
- Logs tab: Copy All button, Voice Diagnostics panel with live session
state, Probe Audio Levels (measures actual signal at 3 pipeline points),
Test Direct Playback button, Copy Diagnostics button.
- Client logging: WebRTC (PeerConnection lifecycle, ICE candidates with
type/address, track events, negotiation), VAD (start/threshold/destroy),
Audio (device acquisition with settings, device changes), noise suppression
(WASM load timing, worklet vs fallback path), voiceSession (remote stream
parsing failures, deafen state, audio element playback events).
- Server logging: SFU init config, voice room mode transitions/track
lifecycle/close, RTP forwarding with packet counts and first-packet
detection, 5s no-packet warning, track fan-out counts, subscriber
transceiver state, ICE candidate details, voice credentials issued.
- Logger: Error objects now serialize .message and .stack instead of {}.
Per-user volume right-click now works on voice user rows in sidebar.
RNNoise ML noise suppression with AudioWorklet + ScriptProcessor fallback.
Tests: 7 new test cases (replaceTrack, setSpeakers skip-local, clearAuth).
Server:
- Add POST /api/v1/uploads and GET /api/v1/files/{id} endpoints
- Add CreateAttachment DB method for file upload records
- Allow empty message content when attachments are present
- CORS headers on file serving for WebView2 compatibility
Client — File uploads & attachments:
- Clipboard paste (Ctrl+V) and attach button (+) for file uploads
- Preview bar above input with thumbnail, spinner, and remove button
- Images fetched via Tauri HTTP plugin as base64 data URIs (bypasses
self-signed cert rejection in WebView2)
- Three-layer image cache: memory → IndexedDB → network
- In-flight deduplication prevents duplicate concurrent fetches
- Image lightbox with click-to-zoom, scroll wheel zoom, pan, keyboard shortcuts
Client — URL previews & embeds:
- URLs in messages rendered as clickable links
- YouTube embeds with thumbnail, play button, video title via oEmbed API
- Generic link previews with OG metadata (title, description, image)
- Fetched via Tauri HTTP plugin with Facebook crawler User-Agent
- YouTube title cache and OG metadata cache prevent re-fetch on re-render
- Links open in default browser via tauri-plugin-opener
Client — Voice & audio fixes:
- Mute uses replaceTrack(null) for reliable RTP-level muting in WebView2
- Deafen also mutes mic; undeafen/unmute unmutes both
- Muted users show crossed mic icon, deafened show crossed mic + headphone
- Re-apply mute state after input device switch
Client — UX improvements:
- Disable browser context menu globally (only custom menus show)
- Emoji search now matches by keyword names (smile, heart, fire, etc.)
- Emoji picker closes on click outside
- User bar status text moved below username
- Messages sorted chronologically (oldest first, newest at bottom)
- Scroll to bottom on initial load with deferred retries for layout shifts
- Image attachments constrained to 400x350px with click-to-lightbox
When switching voice channels, the old PC's OnICEConnectionStateChange(closed)
fires asynchronously after the new PC is set. Both server and client had the
same race: the stale callback saw "voice is active" and called handleVoiceLeave,
closing the new session.
Server: setupICEMonitor now compares the PC reference before acting on events
Client: joinVoice now calls leaveVoice(false) to clean up old session first
Add comprehensive tests for db, storage, updater, and ws packages
covering edge cases, error paths, and voice handler functions.
New test files for attachment queries and WebSocket coverage boost.
- Exclude voiceSession.ts from coverage (browser API dependency, same
pattern as audio.ts/vad.ts/webrtc.ts)
- Remove stale TestHub_Register_CleansUpOldVoiceState test that tested
old duplicate-login behavior removed in b53c729
jevb