mirror of
https://github.com/permissionlesstech/bitchat.git
synced 2026-07-26 17:05:21 +00:00
ca4c4973008a6e252ab38bdf91e61dce6f024ba6
10
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 jack
|
b081c98dba | Harden panic recovery and service shutdown | ||
|
jackandGitHub
|
820c933958 |
Harden PTT audio and the 1.7.1 release (#1423)
Centralize PTT and voice audio-session ownership, harden courier/bridge/outbox delivery and recovery, correct location and delivery-state races, add privacy/release metadata, and ship reproducible universal Arti slices with Release CI coverage. Validated by the full iOS suite, repeated audio/fragment/performance regressions, BitFoundation tests, strict lint and dead-code analysis, universal iOS Release builds, and iOS/macOS archives. |
||
|
|
dd6b624cae |
Quality pass on the 1.7.0 batch: fix confirmed bugs, bump to 1.7.1 (#1418)
* Quality pass on the 1.7.0 batch: fix confirmed bugs, bump to 1.7.1 Post-merge review of PRs #1400–#1417 (push-to-talk, mesh bridging, DM store-and-forward, empty-mesh liveliness, geo-notes). Fixes the confirmed, well-scoped findings; deeper architectural/security items are tracked separately. - PTT hot-mic leak: releasing the mic during VoiceCaptureSession.start()'s 150ms retry pause left the mic live and streaming for up to 120s, because cancel() no-op'd once `completed` was set. Bail after the sleep if the hold was released, and make cancel() always tear down a late-started capture. - Bridge courier depositDrop reported success and burned the dedup slot before the drop was actually published (evicted/compose-fail = lying 📦 "carried" with no retry). Only consume publishedDropKeys on durable accept; add BoundedIDSet.remove() to release evicted/failed slots (uses the dead dedupKey). - Blocked senders resurfaced via archived "heard here earlier" echoes, the one path that bypassed the live block filter — filter at seed time. - A late optimistic .sent clobbered the router's .carried state; extend ConversationStore.shouldSkipStatusUpdate to a full precedence guard (sending < sent < carried < delivered < read). - Read receipts were permanently burned when the router dropped them (marked sent then dropped). sendReadReceipt/routeReadReceipt now return Bool; only record as sent on a successful route, else retry on the next read scan. - MessageRouter.cleanupExpiredMessages() had no production caller, so DMs to a peer that never reconnects sat on .sending until relaunch — run it in the 120s bridge sweep. - Sightings tally now rolls over at midnight while idle; wave notification action localized across all 29 locales; bridged anon#tag uses suffix(4) like everything else; makeThrowawayIdentity delegates to NostrIdentity.generate(); .swiftlint.yml excludes .claude worktrees. - Add regression tests: carried→sent no-downgrade, carried→delivered upgrade, evicted pending drop stays retryable. - Bump MARKETING_VERSION to 1.7.1. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * Fix CI: adjust delivery-status benchmark and drop now-dead addSystemMessage The stricter no-downgrade guard (delivered/carried never regress to sent) broke two things the earlier commit didn't catch locally (perf tests are skipped in the default run, and Periphery runs only in CI): - PerformanceBaselineTests delivery benchmarks alternated sent <-> delivered assuming both directions apply; the delivered -> sent half is now correctly skipped, so the pass measured 0 updates. Alternate two delivered timestamps instead — every update is real, no downgrade. - Routing the geoDM "not in a location channel" error into the thread removed the only caller of ChatPrivateConversationContext.addSystemMessage, leaving it (and its mock) dead per Periphery. Drop the protocol requirement, the mock impl, and the now-vacuous systemMessages.isEmpty assertions (the invariant is compile-time enforced: the context can no longer emit a public system line). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * Address review findings: read-receipt dedup, carried-vs-sending, block-time echo purge - Read receipts: claim the receipt in sentReadReceipts synchronously before spawning the routing task (chat open runs two read scans in one MainActor stretch, so the async insert let every unread message route twice), and release the claim when the route fails so the retry-on-failed-route behavior is preserved. - Delivery status: extend the no-downgrade guard so the `.sending` stamp a pre-handshake resend emits can no longer clobber carried/delivered/read (the 📦 indicator survived `.sent` but not `.sending`). - Archived echoes: blocking a peer now purges their carried public messages from the gossip archive at block time (UnifiedPeerService and /block), while the fingerprint-to-peerID mapping is still known — the seed-time filter can't resolve offline non-favorite strangers and stays only as defense-in-depth. New Transport hook (default no-op) + GossipSyncManager.removePublicMessages with immediate persist. - Bridge courier: an envelope that can't encode within the drop size caps fails identically on every attempt; consume the dedup slot so the 120s retry sweep stops re-running Noise sealing on it. - MeshSightingsTracker: cache the day-key DateFormatter instead of building one per call. Tests: double-markAsRead dedup + failed-route retry, carried→sending no-downgrade matrix, block-time purge (manager + service wiring), oversize-drop slot consumption. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Also skip the sent→sending downgrade in the delivery-status guard Codex review follow-up: sendPrivateMessage without an established Noise session emits `.sending` asynchronously, so it can land after the message already reached `.sent` and visibly walk "Sent" back to "Sending...". Treat `.sending` as weaker than `.sent` too — the status was already truthful. `.failed` → `.sending` stays allowed so a retry after a real failure remains visible. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Retain Codable properties in Periphery scan (same fix as #1421) The noiseKey assign-only false positive fired persistently on this branch (twice, including a rerun) despite the baselined USR. Byte-identical to the fix on fix/announce-replay-link-steal so the branches merge cleanly in either order. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: jack <jackjackbits@users.noreply.github.com> Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com> |
||
|
|
ef848857b7 |
Remove dead code found by full Periphery audit; add scan config + advisory CI (#1410)
Periphery 3.7.4 audit of both schemes (macOS + iOS, intersected so platform-specific code is never touched), with test targets indexed and the share extension built. 277 dead declarations removed or demoted: dead forwarding wrappers (ChatViewModel+Nostr/+PrivateChat), removed- feature remnants (autocomplete command suggestions, back-swipe tuning, MediaSendError, GeohashParticipantTracker), unused Tor dormancy bindings, assign-only properties, unused parameters (renamed to _), and redundant public accessibility. 13 orphaned localization keys deleted across all 29 locales (old pre-#1392 location-notes UI, app_info warnings). Two real tests were flagged as unused because they never ran: Swift Testing methods missing @Test (NostrProtocolTests. testAckRoundTripNIP44V2_Delivered, NotificationStreamAssemblerTests. testAssemblesCompressedLargeFrame). Re-armed both; they pass. Deliberately kept, now recorded in .periphery.baseline.json: iOS-only code invisible to the CI macOS scan, C FFI signatures, keep-alive NWPathMonitor reference, InboundEventKey.eventID (dedup semantics), wifiBulk capability bit (reserved for Wi-Fi bulk work, used by BitFoundation package tests), and the String secureClear cluster (exercised by package tests). New: .periphery.yml config and an advisory Dead Code CI job (mirrors the SwiftLint precedent from #1361) that fails on findings not in the committed baseline. Verified: full macOS app suite, BitFoundation (119) and BitLogger (13) package tests green; periphery scan --strict exits clean. Co-authored-by: jack <jackjackbits@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
|
78a291ab77 |
Public-mesh push-to-talk: signed live voice bursts in the mesh channel (#1406)
* Live push-to-talk voice for DMs: stream while you talk, voice note as fallback Holding the mic in a DM now streams AAC frames live over the Noise session (walkie-talkie style, ~0.5s mouth-to-ear at one hop) while recording the same audio as a normal voice note. On release the note ships through the existing fileTransfer pipeline; receivers that heard the live stream absorb it silently into the same bubble (matched by the burst ID embedded in the file name), so reliability comes for free and nobody sees duplicates. Protocol: - NoisePayloadType.voiceFrame = 0x08 carrying VoiceBurstPacket (burstID + seq + START/data/END/CANCELED, length-prefixed AAC frames) - 210-byte burst-content budget keeps each Noise packet inside the 256-byte padding bucket: one BLE frame, never the fragment scheduler - fire-and-forget: frames are dropped (never queued) without an established session; live is only offered when the peer is mesh-reachable Receive: - ChatLiveVoiceCoordinator assembles bursts (jitter-ordered, 0.5s gap skip, 3s idle end, flood/size caps), persists progressively as ADTS .aac so even a partial burst is a replayable bubble - live autoplay only when the conversation is on screen, app active, and the new app-info "live voice messages" toggle is on (also gates live sending) - one-playback-at-a-time via a shared ExclusivePlayback slot Capture: - PTTCaptureEngine taps AVAudioEngine, dual-encodes: live AAC frames + the finalized .m4a (same 16kHz/mono/16kbps settings as VoiceRecorder) - VoiceRecordingViewModel now drives a pluggable VoiceCaptureSession; the composer HUD shows a pulsing LIVE treatment when streaming Includes the push-to-talk design doc, 6 new localization keys across all 29 locales, and unit tests for framing, packetizer budget, ADTS output, codec round-trip, and the assembly/absorb lifecycle. Public-mesh PTT (MessageType 0x29) lands separately on top of this. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Public-mesh push-to-talk: signed live voice bursts in the mesh channel Extends live PTT from DMs to the public mesh timeline. Holding the mic in the mesh channel now broadcasts the burst live as signed voiceFrame packets (MessageType 0x29) while the finalized voice note still ships on release — new clients hear you as you speak and absorb the note silently into the live bubble; old clients (and late joiners) keep receiving the note exactly as before, so mixed-version meshes lose nothing. Wire/relay: - MessageType.voiceFrame = 0x29: ephemeral signed broadcast, never gossip-synced (SyncTypeFlags maps it to no bit), never padded (padding to the 512 block would push every ~490-byte signed packet into fragmentation) - RelayController treats voiceFrame like media fragments: dense-graph TTL clamp contains the sustained ~15 pkt/s per-talker stream, tight 8-25 ms jitter keeps multi-hop latency inside the receiver's 350 ms jitter buffer - inbound gate mirrors public messages: broadcast-only, 30 s freshness cap, packet signature verified against the claimed sender's announce before any audio reaches the UI App: - ChatLiveVoiceCoordinator gains burst scopes: public bubbles land in the mesh timeline, autoplay only while that timeline is on screen, and the finalized-note absorb is scope-bound (a public note can't replace a DM burst or vice versa) - floor courtesy: while someone talks live in the public channel the composer mic tints red and pulses, with an accessibility value naming the talker ("%@ is speaking", localized in all 29 locales); holding still works — a decentralized mesh has no floor arbiter, the tint just discourages talk-over Tests: relay policy (sparse cap + dense clamp), public bubble + talker indicator lifecycle, note absorption into the mesh store, and scope-binding rejection; full suite green (1382 tests). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * PTT follow-ups from review + field test: peer-ID normalization, toggle gates inbound, drop-path diagnostics Codex review fixes (#1403): - makeVoiceCaptureSession normalizes the selected peer with toShort() before the reachability/session checks and binds the send target to that same routing ID — a conversation selected under the stable 64-hex Noise key no longer silently falls back to a classic note while the short-ID session is established - the live-voice toggle now gates inbound bursts too: off means classic-notes-only in both directions (no live bubble, partial file, or early notification; the finalized note still arrives), with a test Field-test diagnostics (first device run: DM frames decrypted but no bubble appeared, with no log evidence of which guard dropped them): - coordinator logs undecodable frames (size + hex prefix) and blocked drops - makeAssembly logs directory/file-handle failures instead of returning nil silently - PTTLiveVoiceSession logs capture start and finish (packet/frame/duration counts); PTTCaptureEngine logs engine start success/failure with the input format; BLEService.sendVoiceFrame logs no-session drops Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Fix iPhone live-capture failure: dead input unit (AURemoteIO -10851, 0 Hz) Field testing showed the phone's live capture failing at mic enable with AURemoteIO -10851 and an input format of 0 Hz / 2 ch — an input unit bound to an earlier (playback-only or settling) audio session. The Mac, which has no session lifecycle, captured fine, which is why public bursts from the Mac worked while phone-side sends degraded from working (first hold) to sporadic to dead across holds. Three layers of defense: - PTTCaptureEngine recreates its AVAudioEngine on every start(), after the session is configured, so the input unit binds to the session that is active now; a dead input (0 Hz or 0 channels) is now a distinct, logged error instead of a silent setup failure - PTTLiveVoiceSession retries the capture start once after a 150 ms route-settle pause - VoiceRecordingViewModel falls back to the classic VoiceRecorder within the same hold if the live engine still cannot start — a route glitch now costs the live stream, never the voice note Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Blue mic when the hold will stream live The mic button now shows readiness at a glance — and doubles as a build marker for device testing: - blue: holding will stream live (DM peer reachable with an established Noise session, or the public mesh channel) - accent (orange in DMs): holding records a classic voice note (no session yet, peer unreachable, or live voice toggled off) - red states unchanged (recording, floor busy) Refactors capture-backend selection into a single liveVoiceTarget() so the indicator and makeVoiceCaptureSession can never disagree. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Revert the blue live-ready mic to the normal accent color The build-verification marker did its job; idle mic color goes back to the accent. The LIVE recording HUD remains the signal for whether a hold is streaming. Keeps the liveVoiceTarget() refactor so backend selection stays in one place. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Leave a trace on every mic press and every inbound-frame drop Field testing read "tap does nothing" as breakage: the mic start is async (permission check + engine spin-up), so releasing before recording begins has always been a silent cancel — for classic voice notes too. Every press now logs which backend it chose and, for quick presses, that it released before recording started. Also logs the two remaining silent drops: inbound voice frames rejected by the live-voice toggle (the one unlogged guard left in the receive path) and the classic-note fallback now includes the toggle state. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Fix mic hold dying instantly in DMs: sheet swipe gesture starved the composer Field logs showed every DM mic hold ending 3-10 ms after it began, on both platforms, while public-channel holds worked — the private sheet wraps its entire content (composer included) in a high-priority swipe-right-to-leave DragGesture, and a high-priority ancestor drag cancels the mic button's press-and-hold within milliseconds. Same starvation mechanism as the DM image-reveal bug (#1402), hitting a drag instead of a tap. The swipe-to-leave gesture now lives on the message list only, so the composer's gestures (mic hold, text field, buttons) are out of its reach and the swipe still works where users actually swipe. Also stops touching the capture engine when a hold cancels before the engine ever started: probing inputNode on a never-started engine instantiates its input unit against whatever session is active and spams benign-but-alarming AURemoteIO -10851 errors into field logs. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Reorder app info sheet: usage first, then settings, then reference New section order: HOW TO USE, then the adjustable bits (appearance, voice, network), then the reference material (features, privacy, symbols legend). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * App info: flow HOW TO USE into one paragraph; "list" and "person" wording The six how-to-use bullets now read as a single comma-separated paragraph (same instruction strings, legacy bullet prefix stripped at render). Two wording updates across all 29 locales: the people icon opens the "list" (not "sidebar"), and you tap a "person's" name (not a "peer's") to start a DM. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: jack <jackjackbits@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
|
eacd8f0750 |
Live push-to-talk voice for DMs (streams while you talk, voice note as fallback) (#1403)
* Live push-to-talk voice for DMs: stream while you talk, voice note as fallback Holding the mic in a DM now streams AAC frames live over the Noise session (walkie-talkie style, ~0.5s mouth-to-ear at one hop) while recording the same audio as a normal voice note. On release the note ships through the existing fileTransfer pipeline; receivers that heard the live stream absorb it silently into the same bubble (matched by the burst ID embedded in the file name), so reliability comes for free and nobody sees duplicates. Protocol: - NoisePayloadType.voiceFrame = 0x08 carrying VoiceBurstPacket (burstID + seq + START/data/END/CANCELED, length-prefixed AAC frames) - 210-byte burst-content budget keeps each Noise packet inside the 256-byte padding bucket: one BLE frame, never the fragment scheduler - fire-and-forget: frames are dropped (never queued) without an established session; live is only offered when the peer is mesh-reachable Receive: - ChatLiveVoiceCoordinator assembles bursts (jitter-ordered, 0.5s gap skip, 3s idle end, flood/size caps), persists progressively as ADTS .aac so even a partial burst is a replayable bubble - live autoplay only when the conversation is on screen, app active, and the new app-info "live voice messages" toggle is on (also gates live sending) - one-playback-at-a-time via a shared ExclusivePlayback slot Capture: - PTTCaptureEngine taps AVAudioEngine, dual-encodes: live AAC frames + the finalized .m4a (same 16kHz/mono/16kbps settings as VoiceRecorder) - VoiceRecordingViewModel now drives a pluggable VoiceCaptureSession; the composer HUD shows a pulsing LIVE treatment when streaming Includes the push-to-talk design doc, 6 new localization keys across all 29 locales, and unit tests for framing, packetizer budget, ADTS output, codec round-trip, and the assembly/absorb lifecycle. Public-mesh PTT (MessageType 0x29) lands separately on top of this. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * PTT follow-ups from review + field test: peer-ID normalization, toggle gates inbound, drop-path diagnostics Codex review fixes (#1403): - makeVoiceCaptureSession normalizes the selected peer with toShort() before the reachability/session checks and binds the send target to that same routing ID — a conversation selected under the stable 64-hex Noise key no longer silently falls back to a classic note while the short-ID session is established - the live-voice toggle now gates inbound bursts too: off means classic-notes-only in both directions (no live bubble, partial file, or early notification; the finalized note still arrives), with a test Field-test diagnostics (first device run: DM frames decrypted but no bubble appeared, with no log evidence of which guard dropped them): - coordinator logs undecodable frames (size + hex prefix) and blocked drops - makeAssembly logs directory/file-handle failures instead of returning nil silently - PTTLiveVoiceSession logs capture start and finish (packet/frame/duration counts); PTTCaptureEngine logs engine start success/failure with the input format; BLEService.sendVoiceFrame logs no-session drops Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Fix iPhone live-capture failure: dead input unit (AURemoteIO -10851, 0 Hz) Field testing showed the phone's live capture failing at mic enable with AURemoteIO -10851 and an input format of 0 Hz / 2 ch — an input unit bound to an earlier (playback-only or settling) audio session. The Mac, which has no session lifecycle, captured fine, which is why public bursts from the Mac worked while phone-side sends degraded from working (first hold) to sporadic to dead across holds. Three layers of defense: - PTTCaptureEngine recreates its AVAudioEngine on every start(), after the session is configured, so the input unit binds to the session that is active now; a dead input (0 Hz or 0 channels) is now a distinct, logged error instead of a silent setup failure - PTTLiveVoiceSession retries the capture start once after a 150 ms route-settle pause - VoiceRecordingViewModel falls back to the classic VoiceRecorder within the same hold if the live engine still cannot start — a route glitch now costs the live stream, never the voice note Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: jack <jackjackbits@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
|
34bae4a89d |
Refactor Voice Messages (#1075)
* Extract voice-related code to a dedicated VM * Minor optimization of permission request + fix ui bug When isPreparing is being set to true before permission is granted, the UI is shown for a fraction of a second and it’s even worse if the permission is being asked constantly if the user drags the finger even when the alert is shown * Remove dead code * Remove unnecessary delegate conformance * `actor VoiceRecorder` + other minor improvements * Centralized opening system settings + open for mic access * Better voice-recording state handling with Enum * Remove manual timer management * Simplify formatting + avoid jumping UI w/ countdown * Add `requestingPermission` state to avoid repetitive calls * Improve guarding of the states after awaits * Fix potential “actor reentrancy across awaits” issue * Remove short-circuiting on .idle + guard before error * Fix playbackLabel’s formatting for duration vs remainder --------- Co-authored-by: jack <212554440+jackjackbits@users.noreply.github.com> |
||
|
evgeniy.chernomortsev
|
5a66f03400 |
fix(build): exclude allowBluetoothHFP on iOS Simulator
allowBluetoothHFP is not available on iOS Simulator, causing build failures. Use conditional compilation to exclude this option when building for simulator while keeping it for device builds. |
||
|
|
0812cafd55 |
Improve mesh media throughput (#845)
* Improve mesh media throughput * Reserve media slots atomically * Prioritize small fragment trains --------- Co-authored-by: jack <jackjackbits@users.noreply.github.com> |
||
|
|
40e54a5120 |
Add voice notes and images over BLE mesh (audio + image only) (#823)
* Add BLE file transfer support and media UX * Gracefully disable mac attachment pickers in sandbox * Tighten spacing above media message bubbles * Reduce vertical padding between chat rows * Restore iOS file importer for attachments * Copy imported files before sending to preserve access * Allow file transfers from connected but unverified peers * Raise BLE notification buffer cap for large file transfers * Revert "Raise BLE notification buffer cap for large file transfers" This reverts commit b624523af843475db84e4a846db8dcbe824ae408. * Add guard to drop oversized BLE notification assemblies * Let BLE assembler accept large frames up to hard cap * Add detailed logging for BLE fragment assembly * Log incomplete BLE frames for debugging * Stop dropping partial BLE frames while assembling notifications * Fix compressed BLE file transfers * Enable mac attachment importers * Allow mac microphone access * Permit mac media library access * Describe microphone usage * Harden attachment transfer bookkeeping * Display recording milliseconds * Restore mac photo picker access * Use Photos picker on mac * Allow long-press reblur on images * Reblur images via swipe * Lowercase image preview buttons * Keep processed images for outgoing messages * Use save panel for mac image export * Fix image attachment detection * Allow user-selected write access * Align mac image JPEG encoding * Revert unsupported JPEG option * Strip metadata in mac image encoding * Normalize mac JPEG color space * Target image byte size across platforms * Fix CFMutableData handling * Preserve packet version when signing * Use unique transfer identifiers * Stub file transfer methods in mock * Stub file transfer methods in mock * Hide absolute paths in media messages * Resolve image/voice path handling * Fix cleanupLocalFile lookup * Restore BLE broadcasts when notify buffer is saturated * Guard peer map reads on BLE message path * Drop attachment ceilings to 1 MiB and bump release version * Reset BLE assembler on stalled fragment trains * Fix binary protocol test fixtures * Fix critical issues from PR #681 review Critical fixes: - BinaryProtocol: Return nil for unknown versions (prevents buffer underflows) - Add BinaryProtocol.Offsets struct to centralize magic numbers - Replace magic offset calculations with named constants Security/Privacy: - FileAttachmentView: Use url.lastPathComponent instead of url.path (prevents exposing full system paths) Documentation: - Fix compression algorithm documentation (zlib, not LZ4) All tests passing. * Fix UI freeze when receiving voice notes Problem: AVAudioPlayer initialization in VoiceNotePlaybackController.init() was running synchronously on main thread during view creation, blocking UI for 50-200ms per voice note. Solution: - Remove eager preparePlayer() call from init - Load duration asynchronously on background queue - Player is only prepared when playback is actually requested via ensurePlayerReady() This prevents UI freezes when voice notes appear in the chat. * Fix memory leaks and post-playback freeze Fixes: 1. Post-playback freeze: audioPlayerDidFinishPlaying now dispatches to main thread before updating @Published properties (Swift concurrency violation) 2. Unbounded waveform cache: Implement LRU eviction with 20-entry limit - Track last access time for each cached waveform - Evict oldest entry when cache is full - Prevents unlimited memory growth as voice notes accumulate 3. Audio buffer memory leaks: Wrap computeWaveform in autoreleasepool - AVAudioPCMBuffer allocations are autoreleased - Pool ensures buffers are freed promptly 4. Image processing memory: Add autoreleasepool around compression loops - Each jpegData() call creates temporary objects - Inner pool per iteration prevents memory spikes during quality search Memory should now remain stable during extended use. * Eliminate disk I/O from SwiftUI view rendering path Critical performance fix for UI freezes when receiving media: Problem: mediaAttachment(for:) was called during every SwiftUI render, performing synchronous disk I/O on main thread: - FileManager.fileExists() called 2-6x per message (checking subdirs) - applicationFilesDirectory() creating directories on every call - With multiple media messages, this meant 20-100+ disk ops per render Solution: 1. Remove fileExists checks - construct URLs directly - Files are validated during playback/display (fail gracefully if missing) - Sender determines subdirectory (outgoing vs incoming) 2. Cache applicationFilesDirectory() result - Static cache prevents repeated FileManager.url() calls - Directory created only once 3. Remove redundant playback.replaceURL() in VoiceNoteView.onAppear - Controller already initialized with correct URL This eliminates ALL disk I/O from the view rendering hot path. * Cache Nostr identity derivation to prevent crypto during view rendering Critical performance fix: Problem: formatMessageHeader() called deriveIdentity(forGeohash:) during every SwiftUI render for every media message. Each call performed: - Keychain I/O (getOrCreateDeviceSeed) - HMAC-SHA256 computation - Up to 10 secp256k1 key validations (elliptic curve crypto) With multiple media messages, this resulted in 100s of milliseconds of blocking crypto on main thread per render cycle. Solution: Add thread-safe cache for derived identities - Check cache before expensive crypto operations - NSLock protects concurrent access - Identity is deterministic per geohash, so caching is safe This eliminates crypto from the hot rendering path. * Cache geohash identity in ChatViewModel to prevent crypto during rendering Additional optimization for location channels (voice notes are mesh-only, but this helps with text message rendering in geohash channels): - Add cachedGeohashIdentity to avoid deriveIdentity calls during rendering - Check cache before falling back to crypto derivation - Reduces main thread crypto work in location channels * Make voice note loading completely lazy with deferred initialization Aggressive performance optimization to prevent UI freezes: Problem: Even with async loading, creating 10+ VoiceNotePlaybackController instances simultaneously (when scrolling past multiple voice notes) spawned 20+ concurrent background tasks, potentially starving main thread. Solution - Ultra-lazy loading: 1. VoiceNotePlaybackController.init() now does ZERO work - No duration loading - No player creation - Instant initialization 2. Duration loaded on-demand via public loadDuration() method - Called from VoiceNoteView.onAppear after 150ms delay - Reduced priority: .utility instead of .userInitiated - Guard prevents duplicate loading 3. Waveform loading also deferred 150ms - Gives UI time to settle after message appears - Prevents task storms when multiple voice notes appear This spreads the work over time instead of all at once. * Ensure /clear and panic triple-tap delete media files Fix: /clear command and panicClearAllData() now properly delete media files 1. /clear (triple-tap on chat): - Deletes outgoing media (voice notes, images, files) - Conservative: only our sent media, preserves received media - Runs in background to avoid UI freeze 2. panicClearAllData() (triple-tap on bitchat/ header): - Deletes ALL media files (incoming + outgoing) - Removes entire files directory and recreates structure - Ensures complete data wipe for emergency scenarios Both operations run async on .utility queue to prevent blocking UI. * Fix infinite render loop and apply all security fixes CRITICAL BUG FIX - Infinite Render Loop: Root Cause: Duplicate view identity in ContentView.swift:368 ForEach(messageItems) { item in // Already uses item.id via Identifiable messageRow(...) .id(item.id) // ❌ REDUNDANT modifier caused identity re-evaluation loop } When @Published properties updated, SwiftUI re-evaluated .id() → appeared as 'new' identity → triggered re-render → infinite loop. Caused UI freezes, keyboard failures, and 100% CPU usage. Fix: Remove redundant .id() modifier - ForEach already has stable identity. PERFORMANCE FIXES: 1. Waveform Cache Deadlock (Waveform.swift) - Removed nested queue.async(barrier) on cache hits - Was causing task saturation and potential deadlocks 2. Async Send Pattern (ContentView.swift) - Clear input immediately, defer actual send to next runloop - Prevents blocking current event handler 3. Proper Swift Concurrency (VoiceNoteView.swift) - Switch from .onAppear + DispatchQueue to .task - Cleaner async/await pattern for loading 4. Remove Redundant objectWillChange (ChatViewModel.swift) - @Published already triggers updates automatically - Explicit send() was causing double update cycles SECURITY FIXES (C1-C5, H1-H2): C1. Path Traversal Protection (BLEService.swift) - Unicode normalization, null byte removal - Replace ALL path separators, reject dotfiles - Validate paths don't escape directory C2. Integer Overflow (BitchatFilePacket.swift) - Use UInt64 for TLV parsing, safe Int conversion C3. MIME Validation (BLEService.swift) - Whitelist: JPEG, PNG, GIF, WebP, M4A, MP3, WAV, OGG, PDF - Magic byte validation for all types - Lenient on M4A (platform variations) C4. Compression Bomb (BinaryProtocol.swift) - Ratio validation <= 50,000:1 - Defense-in-depth with 1MB size cap C5. TOCTOU Race (ChatViewModel.swift) - Direct removeItem without fileExists check H1. File Size Validation (ChatViewModel, ImageUtils) - Check attributes BEFORE Data(contentsOf:) - Prevents memory exhaustion H2. Metadata Stripping (ImageUtils.swift) - Remove ALL metadata keys from JPEG encoding - Only compression quality set - Protects GPS/EXIF/device info privacy RESULT: ✅ No render loops ✅ Works with Xcode debugger ✅ Voice notes display properly ✅ All security vulnerabilities fixed ✅ 164 tests passing Production ready. * Complete all translations to 100% and fix auto-extraction - Mark non-localizable strings with Text(verbatim:) to prevent extraction - Update UI strings to lowercase per style guide (open, save, close, recording) - Add complete translations for all 29 languages (194/194 strings at 100%) - Remove empty/duplicate entries (@, bitchat/, Open, Recording %@) - Add proper localization comments for all user-facing strings * macOS: Focus message input on launch instead of nickname field * Remove debug print statements from sendMessage * Optimize voice note codec to 16 kHz / 20 kbps for smaller file sizes - Reduce sample rate from 44.1 kHz to 16 kHz (telephony standard) - Lower bitrate from 32 kbps to 20 kbps - Results in ~37% file size reduction (~150 KB/min vs 240 KB/min) - Increases max voice note length from 4.4 to 7 minutes over 1 MiB BLE limit - Maintains excellent voice quality using native AAC-LC codec * Fix critical security issues in fragment reassembly and file cleanup Fragment Reassembly Race Condition (CRITICAL): - Wrap all incomingFragments/fragmentMetadata access in collectionsQueue.sync - Prevents concurrent modification crashes from multi-threaded access - Minimizes lock contention by doing heavy work (reassembly/decode) outside locks - Add upper bound check: reject fragments with total > 10,000 (DoS prevention) - Add cumulative size validation before storing fragments (memory DoS prevention) File Cleanup Path Traversal (CRITICAL): - Use NSString.lastPathComponent to extract filename safely - Prevents directory traversal attacks via malicious filenames - Add path prefix validation before file deletion - Now checks both incoming and outgoing directories (fixes disk leak) Additional Protections: - Fragment assemblies now limited by both count (128) and cumulative bytes (1MB) - Explicit checks for "." and ".." filenames in cleanup - Defense-in-depth: multiple validation layers * Fix post-rebase compilation errors - Remove duplicate NostrIdentityBridge and Bech32 from NostrIdentity.swift (now in separate files) - Add caching to NostrIdentityBridge.deriveIdentity() for performance - Remove duplicate NotificationStreamAssembler from BLEService.swift - Remove duplicate function declarations in BLEService.swift - Remove duplicate DeliveryStatusView and PaymentChipView from ContentView.swift - Fix PeerID type conversions throughout (use .id for String, PeerID(str:) for wrapping) - Update ContentView body to use main's simple VStack structure - Fix NostrIdentityBridge instance method calls - Remove privateChatView (replaced with sheet-based UI in main) Build and tests passing (137/139 tests pass). * Fix remaining compilation issues after rebase - Fix PhotosUI import order (must be after platform imports) - Fix Data.WritingOptions.atomic reference - Add identity derivation caching to NostrIdentityBridge - Fix all remaining PeerID type conversions in ChatViewModel - Fix ContentView body structure to use main's VStack layout - Fix PaymentChipView API usage (now uses PaymentType enum) Build and tests now passing. * Add proper availability checks for PhotosPickerItem PhotosPickerItem requires iOS 16+ / macOS 13+ but canImport(PhotosUI) succeeds on older macOS versions. Add compiler version check to ensure PhotosPicker code only compiles when actually available. This fixes CI build failures on older macOS environments. * Limit PhotosPicker to iOS only to fix CI PhotosPickerItem has SDK availability issues on macOS in CI. Change PhotosPicker from canImport(PhotosUI) to os(iOS) only. macOS users can still import images via file importer (.fileImporter). This is actually cleaner as macOS file picker is more familiar to users. Fixes CI build failures. * Convert new tests to Swift Testing * Fix compilation issue * Add the missing `fileTransfer` case * Explicitly list all Enum cases to get compile-time errors * Revive lost `NotificationStreamAssembler` changes * Allow file fragments to account for protocol overhead * Simplify PR: Focus on audio+image, fix EXIF stripping, remove file transfers - Fix critical EXIF privacy issue in iOS image processing - Both iOS and macOS now use CGImageDestination for metadata stripping - Shared encodeJPEG function ensures no GPS, camera, or metadata leaks - Remove file transfer functionality to simplify PR scope - Deleted FileAttachmentView - Removed sendFileAttachment from ChatViewModel - Removed file picker UI from ContentView - Simplified attachment dialog to image only (iOS) or voice only - Keep focused media features: - Voice recording and playback - Image sending with progressive reveal - Binary protocol for media transfer * Improve camera UX: Direct camera access with camera icon - Change paperclip icon to camera icon for clearer affordance - Open camera directly on tap (no confirmation dialog) - Add CameraPickerView wrapper for UIImagePickerController - Remove PhotosPicker in favor of direct camera access - Images still processed through ImageUtils with EXIF stripping - Accessibility: Added 'Take photo' label * Full-screen camera with photo library option - Change to fullScreenCover for immersive camera experience - Add action sheet with 'Take Photo' and 'Choose from Library' options - Renamed ImagePickerView to support both camera and library sources - Both options open full-screen for better UX - Updated accessibility label to 'Add photo' (more accurate) * Fix camera white bars with overFullScreen presentation - Changed modalPresentationStyle from .fullScreen to .overFullScreen - This should eliminate white bars at top/bottom on notched devices - Explicitly set showsCameraControls and cameraOverlayView for camera mode * Gesture-based photo access: Tap for library, long-press for camera UX improvements: - Tap camera icon → Photo library (common use case) - Long press camera icon (0.3s) → Direct camera (quick photos) - Removed action sheet entirely for cleaner flow - Power users can long-press for instant camera access This is more discoverable and eliminates an extra step in the UI. * Simplify camera presentation to reduce frame errors - Changed back to standard .fullScreen presentation - Removed overFullScreen which was causing frame dimension errors - Let iOS handle safe areas automatically (white bars are intentional) - Reduces gesture gate timeout warnings Note: White bars on notched devices are iOS default behavior for UIImagePickerController. This respects safe areas for status bar and home indicator. True edge-to-edge would require custom AVFoundation camera implementation. * Force dark mode on camera/picker for black safe area bars - Set overrideUserInterfaceStyle = .dark on UIImagePickerController - Changes white bars to black (much better looking) - Camera controls and photo library also appear in dark mode - Consistent dark appearance regardless of system settings * Optimize sheet presentation for camera UI - Force .large detent for maximum height - Hide drag indicator for cleaner look - Use ignoresSafeArea to give camera full space - Should show complete flash button and controls * Fix P1: Add DoS protections to PeerID fragment handler Critical security fix addressing Codex review feedback: The PeerID overload of handleFragment (which is actually called by CoreBluetooth) was missing key safety checks that existed in the String overload: 1. Added total <= 10000 check to prevent unbounded fragment counts 2. Added cumulative size check against FileTransferLimits before storing each fragment 3. Prevents memory exhaustion DoS attacks via malicious fragment streams This ensures the actually-used code path has proper bounds checking. * Fix decompression size limit to support max-sized file transfers Root cause: BinaryProtocol.decode() was rejecting decompressed payloads larger than maxPayloadBytes (1 MB), but TLV-encoded file transfers are slightly larger due to metadata overhead. Fixes: - Changed decompression limit from maxPayloadBytes to maxFramedFileBytes - This accounts for TLV overhead (~50 bytes) + binary protocol headers - Now allows ~1.12 MB decompressed payloads (1 MB + overhead budget) The failing test was: - Creating 1 MB file content - TLV encoding adds ~50 bytes (1,048,627 total) - Compression reduces to ~1,084 bytes (highly repetitive data) - During decode, decompression was rejecting the 1,048,627 byte output - Now correctly allows it since 1,048,627 < 1,179,760 (maxFramedFileBytes) All 154 tests now pass including 'Max-sized file transfer survives reassembly' * Fix critical thread-safety crash in PeerID fragment handler CRITICAL: The PeerID version of _handleFragment was accessing incomingFragments dictionary without collectionsQueue synchronization, causing crashes when multiple BLE threads processed fragments concurrently. Crash stack trace pointed to line 3431 (dictionary subscript) with: 'doesNotRecognizeSelector' - classic concurrent mutation crash. Fix: - Wrapped ALL incomingFragments/fragmentMetadata access in collectionsQueue.sync(flags: .barrier) - Matches the thread-safe pattern used in String version - Separate cleanup into its own barrier block after reassembly - Prevents concurrent dictionary mutations from multiple BLE threads This is the same pattern as the String version (line 1128) which didn't crash. * Remove Localizable.xcstrings formatting noise The Localizable.xcstrings file had massive formatting-only changes (spacing: 'key' vs 'key :') that added 50K+ lines to the PR diff. This was just Xcode reformatting with no actual string changes. Reverted to main's version to keep PR focused on actual code changes. * Add macOS photo picker support - Added MacImagePickerView with NSOpenPanel for macOS - macOS shows photo.circle.fill icon (no camera hardware) - Opens native file picker for images (.png, .jpeg, .heic) - Images processed through ImageUtils with EXIF stripping - Simple sheet with Select/Cancel buttons Cross-platform photo sharing now works: - iOS: Tap for library, long-press for camera - macOS: Tap for file picker * Add Localizable strings for camera and voice features Xcode auto-generated localization strings for new UI elements: - Camera/photo picker labels - Voice recording UI strings - Media attachment descriptions These are legitimate new strings needed for the audio+image feature, not just formatting changes. --------- Co-authored-by: jack <jackjackbits@users.noreply.github.com> Co-authored-by: islam <2553451+qalandarov@users.noreply.github.com> |