mirror of
https://github.com/permissionlesstech/bitchat.git
synced 2026-07-27 14:45:23 +00:00
5e69a19cb3b0f9326ed500b53a43ed259e6d6407
1022
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 jackandClaude Fable 5
|
5e69a19cb3 |
Document the V3 transport architecture and remaining roadmap
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> |
||
|
jackandClaude Fable 5
|
5c2903ec7c |
Extract mesh-ping diagnostics state into a pure engine-confined tracker
First slice of the feature-module direction: BLEMeshPingTracker owns the outstanding-probe map and the per-link inbound response budget as pure state (register/resolve/expire/reset), so the security invariants — a pong only resolves against the probed peer, the budget keys on the ingress link because claimed senders are forgeable, panic reset drops probes and budget together — are now unit-tested without queues or radios. The transport keeps only packet I/O, timers, and main-actor delivery around it. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> |
||
|
jackandClaude Fable 5
|
cbfdb9696e |
Wire gateway/bridge/panic features to capability ports, not BLEService
App wiring discovered mesh-only features by casting the Transport to the concrete BLEService class in nine places. Those surfaces are now three capability protocols — BluetoothStateReporting, PanicResettingTransport, and MeshBridgingTransport — discovered with as? like any optional capability, so the bootstrapper, panic flow, and lifecycle coordinator no longer name the concrete transport at all. A future second mesh transport picks up gateway/bridge wiring and the panic lifecycle by conforming, and the remaining Transport god-protocol requirements can migrate to the same pattern. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> |
||
|
jackandClaude Fable 5
|
87cbf5a03c |
Unify the message and collections queues into one serial engine queue
The old model ran a concurrent message queue over a second concurrent collections queue whose barrier flags served as the real mutual exclusion — every field carried an ownership comment, and correctness lived in per-site discipline. The message queue is now a single serial engine queue that owns all mesh protocol state; the collections queue, its 98 sync/async hops, and every barrier flag are gone. Cross-thread callers go through onEngine, which documents and (in debug) enforces the transport's sync-edge order: main and test threads may block on the engine, the engine may block on bleQueue and the crypto/identity queues, and nothing may block the other way. The debug trap caught two latent inversions the leaf-lock structure had been masking: the verified-announce rebind path re-resolved the ingress link through the engine from inside its bleQueue critical section (it now receives the already-resolved link), and the noise session-generation closures sync-re-entered the engine from the noise manager's queue while their own engine slot was blocked on it (they now touch engine state directly, which the held slot makes exclusive). BLE throughput is orders of magnitude below what one serial queue sustains; the full suite runs at identical speed. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> |
||
|
jackandClaude Fable 5
|
d2a47d1ade |
Stop bleQueue maintenance and status paths from blocking on collectionsQueue
The traffic-burst tracker becomes a lock-backed monitor (written by the receive pipeline, read by scan-duty adaptation and announce pacing on bleQueue), the status-log peer summary and topology refresh read the already lock-backed registry directly, and the stalled-fragment reap moves to an async collections hop with the gossip resync request inside it. bleQueue no longer sync-waits on the collections queue anywhere. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> |
||
|
jackandClaude Fable 5
|
0bde6d21f7 |
Move BLE link egress/ingress buffers to bleQueue ownership
pendingPeripheralWrites, pendingNotifications, and pendingWriteBuffers were collectionsQueue-guarded, but every producer and drain already runs on bleQueue next to the CoreBluetooth objects they feed — each access paid a cross-queue barrier for state that never leaves the radio thread, and the notification drain even invoked peripheralManager.updateValue from the collections queue. They are now bleQueue-confined like the link state store: CB delegate callbacks and drains touch them directly, and the few engine-side entry points hop to bleQueue (the direction the transport's sync-edge order already allows). This clears most bleQueue-to-collectionsQueue sync edges ahead of merging the collections queue into the message queue. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> |
||
|
jackandClaude Fable 5
|
f718097c74 |
Make peer registry and local announce state lock-backed
The main actor answered isPeerConnected/peerNickname/currentPeerSnapshots and flipped runtime capability bits by blocking on collectionsQueue behind whatever transport work was in flight. Peer state now lives in a lock-backed BLEPeerRegistryStore (every registry mutation is a single whole-transition method, so readers never observe a torn state), and the runtime capability bits move into BLELocalIdentityStateStore next to the identity they ride announces with. No transport entry point called from the main actor blocks on a transport queue for peer state anymore. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> |
||
|
|
0152196ac2 |
Deflake CI, and make the flake class unrepeatable (#1491)
* Deflake two iOS-sim CI tests with unbounded timing assumptions Both failed on main-adjacent CI runs during this session's PRs. Neither was a product bug; both asserted things about real time that a loaded runner is under no obligation to honour. **NetworkReachabilityGateTests: a wall-clock upper bound.** test_monitor_duplicateUpdatesDoNotPostponeOfflineCommit slept 500ms for real, then asserted total elapsed time was under 1.4s to prove a duplicate mid-window had not restarted the 1.0s debounce. One CI run took 3.75s. No wall-clock bound can separate "deadline preserved" from "runner is slow", because Task.sleep and the asyncAfter flush are both real time and neither is bounded above. The deadline property was already covered deterministically one level down: test_debounce_duplicateObservationsPreservePendingDeadline drives ReachabilityDebounce with injected timestamps and checks pendingRemaining directly. So the monitor test now asserts only what needs a real monitor — that a duplicate still yields exactly one committed false through the debounce — with an injected clock for the arithmetic and a generous liveness budget. Renamed to say what it actually checks. No coverage lost, and it runs in 0.14s instead of ~3.8s because the real sleep is gone. **NoiseEncryptionServiceTests: injected timeouts that also arm during setup.** #1483 diagnosed and fixed exactly this in the quarantine-restore test, but two sibling tests kept the shape. Their injected ordinaryResponderHandshakeTimeout (0.04 and 0.06) also arms during the establishSessions setup handshake, where bob is the responder — so a preempted runner fires it mid-setup, tears down the half-open responder, and message 3 gets answered as a fresh initiation. The CI failure named setup's own `#expect(finalMessage == nil)` seeing a 96-byte message 2, which is precisely the signature #1483 recorded. Raised both to 1.0s, matching #1483's remedy: the scenarios still need the responder timeout to fire, and it still does, just with room for setup to complete first. Thin 1-second waitUntil budgets in the file now use the shared TestConstants.longTimeout; every one of them backs a positive assertion, so waitUntil still returns the moment the condition holds and nothing gets slower in the passing case. No assertion weakened in either file. Verified 3x sequentially and 3x with all 18 cores saturated. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * Raise two more starvation-prone test deadlines Both surfaced on the CI runs for this PR and #1487, both in tests neither PR touches, both the same shape as the two already fixed here: a deadline sized for the work rather than for a runner executing many suites at once. VoiceNotePlaybackControllerTests waited 5s for a @MainActor Task that playback schedules for the session acquire and its failure path. When that Task is not scheduled in time the helper reports *two* failures — the wait, and the `!isPlaying` the un-run failure path has not reset yet — which reads like a playback bug rather than a starved scheduler. That is exactly what CI showed. GeoRelayDirectoryTests waited 10s for work the directory runs in Task.detached(priority: .utility); utility priority competes with every other suite. The retry-scheduling case timed out at exactly 10.06s with the retry never scheduled, reading like a missing retry rather than a starved background task. Raised to 30s each with the reasoning recorded at the helper. Both helpers return as soon as their condition holds, so nothing slows down when tests pass — only the genuine-failure case takes longer to report. Verified 3x with all cores saturated: both suites clean. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * Make the flake class unrepeatable, not just fixed Raising four deadlines fixed the four tests that happened to fire. The class was still there: fourteen separate waitUntil helpers, most defaulting to 1.0s, plus wait call sites with literal budgets. The fifth instance would have landed the same way, on someone else's unrelated PR. The rule, now written down in TestConstants.settleTimeout: **a wait deadline is not a latency budget.** It exists so a genuine hang eventually fails the suite, so size it for the worst-case scheduler, never for how long the operation should take. Waits return as soon as their condition holds, so a generous deadline is free in the passing case and only extends genuine failures. - Every wait helper now defaults to TestConstants.settleTimeout (30s), and the literal wait call sites below the floor were converted too — sixteen sites across ten files. - TestTimingHygieneTests enforces it by scanning the test sources: wait defaults and wait call sites must be at least minimumSettleTimeout, and no test may assert an upper bound on elapsed wall-clock time (the assertion that started this, which cannot separate correct behaviour from a slow machine). - Both rules waive per line with "test-timing-ok: <reason>", accepted on the line or in the comment block above it so the reason has room to be a sentence. One legitimate use so far: a NEGATIVE wait in NoiseCoverageTests asserting a promotion has *not* completed within 50ms, where a long deadline would only make the suite slow while still passing. Injected production timeouts are deliberately not matched — the Noise handshake timeouts are the behaviour under test, and short values are correct there. Verified the guard actually fails: a canary file with both banned shapes was flagged with file and line, and the suite went green again once it was removed. Full suite passes with all 18 cores saturated. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * Close the guard's blind spot: named short timeouts A fifth flake landed on CI while the first guard was in review — GossipSyncBoardTests timing out at 1.03s — and the guard did not catch it, because the deadline was `TestConstants.shortTimeout` rather than a literal. A literals-only scan cannot see a short value behind a symbol, which is the more common way it is written: 81 wait sites used shortTimeout (1s) or defaultTimeout (5s), every one of them below the floor. Rather than guess which of those were safe to raise, all 81 were converted and the suite timed. Runtime went 15s -> 72s, which located the genuine negative waits precisely: ten tests that assert something does *not* happen and therefore always run their deadline out. Measurement instead of a heuristic, since a mis-guess in either direction is invisible — too short reintroduces the flake, too long silently costs a minute a run. Those 28 sites now use `TestConstants.negativeWaitWindow`, a named constant whose doc explains the inverted reasoning: for a negative wait, starvation can only make the assertion *more* likely to hold, so short is correct, and the name states the polarity instead of leaving a bare literal that reads like the mistake. Suite is back to 15.3s. The guard now also rejects `timeout: TestConstants.shortTimeout` and `defaultTimeout`, while accepting `negativeWaitWindow` by name. Re-verified with a canary carrying every banned shape — literal default, both named constants, and an elapsed-time upper bound. All four were flagged with file and line; the suite went green again on removal. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * Delete shortTimeout now that nothing may use it The hygiene guard bans `TestConstants.shortTimeout` at every wait site and the last users were converted, so Periphery correctly flagged the constant itself as dead and failed CI. Remove it from both TestConstants copies; the banned-name entry stays so the symbol cannot quietly return. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: jack <jackjackbits@users.noreply.github.com> Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
|
14e7b428d9 |
Add a real security policy (#1489)
* Add a real security policy Two drive-by template PRs (#1118, #1482) tried to fill this gap with unedited boilerplate. This is the actual policy: private vulnerability reporting (now enabled on the repo) as the channel, honest expectations for a volunteer project, and a scope section that separates the properties the app promises from the documented design behaviors that keep getting reported as vulnerabilities. Closes #1081 Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Address Codex review: name the Nostr envelope format precisely 'Gift-wrapped' reads as NIP-59, and the scope section is exactly where a researcher calibrates expectations. Say what it is: bitchat's own private-envelope scheme, explicitly not NIP-17/44/59. 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> |
||
|
|
c671e3df66 |
Keep the composer focused after sending with the return key (#1490)
* Keep the composer focused after sending with the return key On iOS, return sends and then drops the keyboard, so every message in a back-and-forth costs an extra tap to reopen it. sendMessage() is the onSubmit handler; reasserting the FocusState there keeps the keyboard up between messages. macOS already refocuses on appear and is unaffected. Fixes #457 Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Address Codex review: refocus only on return-key submission The reassert moves from sendMessage() into the TextField's onSubmit, so the send button no longer reopens a deliberately dismissed keyboard on iOS and never moves focus on macOS. 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> |
||
|
|
132120a88e |
Close the three holes the #1486 review found (#1488)
* Fix the three follow-ups from the #1486 review Three defects shipped with the censorship-resilience merge, all confirmed against main: 1. Source-manifest verification silently accepted added files. shasum -c checks only the files the manifest lists, and the Xcode project compiles every source file present in the tree — so a hostile mirror could pass verification by adding a file rather than modifying one. The manifest header and VERIFYING-A-BUILD.md now require the completeness check (git status --porcelain, or a path diff for tarballs) alongside the hash check. 2. A relay removed while Tor was bootstrapping reconnected anyway. dropRelays never subtracted from pendingTorConnectionURLs, and a custom relay passes the allow-list filter, so draining the pending queue resurrected a relay someone had explicitly deleted. 3. Turning Tor off mid-bootstrap read as 'network may be blocking tor'. shutdownCompletely left the detached 75s poll loop running, which then stamped bootstrapDidStall over the clean shutdown state; and the stall handler guarded on torEnforced, which is compile-time true in release, instead of the runtime preference. The poll loop is now generation-fenced (shutdown, dormancy, and restart each invalidate it) and the handler consults persistedTorPreference(). Both app-side fixes carry regression tests proven to fail pre-fix. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Address Codex review: ignored files and manifest placement git status --porcelain omits ignored paths, and .gitignore covers build/ — a planted bitchat/build/Evil.swift would compile via the synchronized group while the documented check stayed silent. The checkout check now uses --ignored. The downloaded manifest also has to live outside the tree, or it trips the completeness checks itself; the doc now says so and references it at /tmp throughout. 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> |
||
|
|
c1ce9029d8 |
Keep working when the network is hostile, and make the app verifiable (#1486)
Tier 3 of a protest-hardening review. The BLE mesh is already properly fenced off from network reachability and needs nothing here; every gap is on the internet side, or in how someone gets a build they can trust. Say when Tor is blocked. The bootstrap poll loop simply ended at its 75-second deadline, leaving isStarting true with no further state, so a network that blocks Tor was indistinguishable from a slow one and the UI said "starting tor…" indefinitely. TorManager now exposes bootstrapDidStall and posts .TorBootstrapDidStall, and the app reports that mesh messaging still works while internet delivery is paused. It is cleared on each new start or restart, so a later attempt can report again. Let relays be added by hand. The four built-in relays are well-known clearnet hostnames, which is four names for a censor to block and no recourse short of a new build. NostrRelaySettings persists up to eight additional relays, normalized, .onion accepted, with a settings editor. They join the same target set as the built-ins and are subject to the same activation policy. Removal reconciles against the previous set: the teardown path iterates the current targets, so without that a removed relay's socket and queued sends would linger — covered by a test that fails without it. The merged list is cached rather than recomputed, because allowedRelayList consults it once per candidate URL and would otherwise read UserDefaults inside that loop. Stop stranding people who denied location. The activation gate required location permission or a mutual favorite, but teleporting into a geohash requires neither, so someone with no permission and no favorites could sit in a channel that never connected while Tor and the relays stayed suppressed and nothing said why. Being in a location channel is now a third arm of the gate, in both the activation service and the relay manager's copy of the policy, and leaving the channel closes it again. Stop burning the Tor timeout when Tor is off. GeoRelayDirectory awaited Tor readiness unconditionally, but with the preference off TorManager has been shut down, so every refresh spent the full bootstrap deadline and the directory froze on its cached copy. It now keys on the preference, not on live readiness: Tor wanted but unavailable must still skip the fetch rather than fall back to clearnet. Say what turning Tor off costs. The toggle's copy described it as hiding your IP "for location channels", understating both scope and consequence. It now names private messages too, and while the toggle is off the settings screen states that every relay can see the device IP. Make builds verifiable. There was no release verification of any kind: no signatures, no checksums, no documented procedure. Post-takedown that is the acute gap, because mirrors appear and people install whatever they can find during a shutdown. source-manifest.yml publishes a per-tag SHA-256 manifest with a provenance attestation, self-checking before it publishes, and docs/VERIFYING-A-BUILD.md explains how to verify source and states plainly that compiled builds from anywhere but the App Store cannot be verified. It also records the gaps honestly: no published signing key, no reproducible build, no non-GitHub mirror. docs/TOR-INTEGRATION.md was substantially stale — it documented a torrc, SOCKSPort and ControlPort that the in-process Arti client does not use, and claimed there are no user-visible settings — so it is rewritten, including the deferred gap below. Deferred: no Tor bridges or pluggable transports. arti-client is built without pt-client or bridge-client and bootstraps from stock config, so in a country that blocks Tor outright there is still no circumvention path — only a clear report that there isn't. Closing it needs the Rust features, bridge config through the FFI, and an xcframework rebuild under the pinned toolchain with a provenance update, which is its own change. Co-authored-by: jack <jackjackbits@users.noreply.github.com> Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
|
c079d2ab5d |
Harden what a locked or seized device gives away (#1484)
* Harden what a locked or seized device gives away The realistic compromise for many of the people this app is built for is not interception but a phone taken, and often unlocked under coercion. Content encryption is in good shape; these are the gaps around it. Hide notification previews, by default. Notification content is rendered by the system on the lock screen, so it was readable without unlocking: DM alerts carried the sender's nickname and the full message body, and geohash alerts put the geohash in the title. Alerts now state that a DM, mention, or location-channel activity arrived and withhold the rest until the app is opened. userInfo still carries the routing peer ID and deep link, neither of which the system displays, so taps land where they did. A settings toggle restores full previews for anyone who wants them. Default-on is the deliberate part: a phone face-up on a table should not narrate conversations, and someone who wants previews can say so. Cover the window on willResignActive, so the snapshot iOS stores for the app switcher shows a placeholder rather than an open conversation. Opaque rather than blurred, because blurred large text stays partly legible and the snapshot goes to disk. Added synchronously from a UIKit notification with queue: nil, since the capture follows shortly after and an OperationQueue hop or a SwiftUI state change can lose that race. Panic wipe already deleted snapshots already on disk; this stops new ones from being worth deleting. Bound media by age as well as size. The 100 MB quota only ever considered incoming files, so outgoing media had no lifetime at all and a received photo could outlive its conversation indefinitely. A launch-time sweep now deletes managed media older than seven days, incoming and outgoing, with the same exemptions quota eviction honors: in-flight live captures and files reserved by a delivery or deletion in progress. Make /clear tell the truth on the mesh timeline. It recorded an echo watermark and left the gossip archive on disk for up to 6 hours, so someone who cleared before a police stop had deleted nothing. Clearing now erases the archive too. The watermark still matters: it suppresses pre-clear messages this device hears again from peers. The cost is that the device stops serving recent public backlog until it hears fresh traffic, which is a fair reading of what clearing a timeline means. Documented in PRIVACY_POLICY.md and the privacy assessment, including a new section on what is deliberately NOT addressed: there is still no duress mechanism of any kind (no decoy passphrase, no wipe-on-failed-auth, no app lock), macOS gets no file-protection classes, and media is not sealed at the app layer. The duress question is a product decision as much as an engineering one, since in some jurisdictions destroying data on demand is itself an offence and hiding may protect someone better than destroying, so it is called out rather than guessed at. Three findings from the audit that prompted this work turned out to be already fixed on main and are not included: keychain accessibility is AfterFirstUnlockThisDeviceOnly with a retrying migration, the panic media wipe uses a two-location durable marker transaction, and panic already discards staged share-extension content. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * Inject the previews preference instead of reading shared defaults CI caught a real problem this introduced. `NotificationServiceTests` already asserted the full-preview title and body, and both it and the new redaction tests read `NotificationPrivacySettings` from `UserDefaults.standard` in the same process. Whichever ran second depended on the other's cleanup, so it passed locally and failed in CI: XCTAssertEqual failed: ("🔒 new dm") is not equal to ("🔒 DM from Alice") Fixed at the source rather than by ordering or serialization. `NotificationService` now takes a `hidePreviewsProvider`, defaulting to the real preference, so each test states which behavior it asserts. The pre-existing test asks for previews shown and gains a redacted counterpart; the redaction tests no longer touch the shared store. `NotificationPrivacySettings` also gained store-injecting accessors so the default-value and round-trip assertions can use an isolated suite rather than mutating preferences other tests read. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> --------- Co-authored-by: jack <jackjackbits@users.noreply.github.com> Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
|
229a41557e |
docs: correct inaccurate privacy and metadata claims (#1485)
* docs: correct inaccurate privacy and metadata claims Several documented guarantees did not match the implementation. These matter more than ordinary doc drift: someone deciding whether to carry this phone to a protest reads these sentences as the threat model. - Peer IDs were described as "short ephemeral IDs derived per session" that "rotate periodically" and "prevent tracking". They are the first 8 bytes of the Noise static key fingerprint, stable across sessions and reboots, and replaced only by a panic wipe. Corrected in the whitepaper (§3, §8), IdentityModels, and BitchatProtocol, whose header notes claimed "no persistent identifiers in protocol headers" while every header carries exactly one. - "No plaintext message content is ever written to disk" was false for accepted media, which is stored unsealed under the platform's data-protection class. Narrowed to what actually holds. - Padding was described as applying to all packets but fragments. Only noiseEncrypted and noiseHandshake frames are padded; the pad bytes equal the pad length rather than being random; and because that length must fit one byte, a frame needing over 255 bytes of padding is emitted unpadded. Documented in the whitepaper (§4.1) and MessagePadding. - The gossip archive window is 6 hours in production, not the 15 minutes claimed in PRIVACY_POLICY.md and the privacy assessment. The 15-minute figure is the struct default that BLEService overrides. - The privacy assessment credited iOS BLE address randomization without noting that stable app-layer identifiers defeat it. The whitepaper's future-work list now names the changes these corrections imply: rotating on-air identity, padding for non-Noise types, and making the announce neighbor list optional. No behavior change; comments and documentation only. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * docs: correct the same claims in the README The README repeats two of the claims corrected elsewhere in this PR, and it is the document people actually read before deciding to trust the app. - "no persistent identifiers" is the inverse of what the mesh does; it now points at the whitepaper's identity and metadata sections. - "end-to-end encryption with forward secrecy" holds for live Noise sessions but not for sealed store-and-forward mail, which the whitepaper already flags as its main cryptographic trade-off. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> --------- Co-authored-by: jack <jackjackbits@users.noreply.github.com> Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
jackandGitHub
|
934b2cd2d3 |
Harden iOS-sim CI: destination fallback + Noise reconnect test determinism (#1483)
Two CI-robustness fixes for the iOS-simulator job seen on main: 1. The destination picker no longer dies on runner images with placeholder/unavailable simulator lists (macos-26-arm64 20260720.0258): it filters error/unavailable rows, falls back to xcrun simctl's available-device list, and as a last resort creates a simulator from the newest installed iOS runtime's own supported iPhone device types. Each path logs which route was taken. 2. Deflakes NoiseEncryptionServiceTests' quarantine-restore test: the injected 20ms responder timeout also armed during the SETUP handshake, so a loaded runner tore down the half-open responder session before the scenario began (production quarantine/restore verified correct — the failure signature was a fresh msg2 to a session-less peer). Timeout raised to 1s and the fixed 100ms sleep replaced with a bounded waitUntil on the restore condition. No assertion weakened; 10/10 sequential + 3/3 under full-core CPU load. |
||
|
jackandGitHub
|
a1711bd399 |
Retry unacknowledged DMs after Noise replacement (#1462)
DMs sent through an established Noise session are retained in the durable MessageRouter outbox until an authenticated delivery/read ack, and retried under the same message ID when the peer's replacement handshake authenticates — fixing DMs silently lost into stale local sessions after a remote app restart. Mesh ack handling is peer-scoped end-to-end (alias-scoped delivery status, peer-bound markDelivered, PeerMessageKey retry state), so a colliding message ID from another conversation can no longer clear or promote foreign state; bridge-drop dedup keys are recipient-scoped with hashed persistence. Includes review fix: acks arriving after a relaunch (durable outbox restored, conversation not) now clear the peer-scoped router entry unconditionally — only the UI status transition is gated on conversation presence — so a delivered message can no longer re-send to the attempt cap and be marked failed despite delivery. Regression test simulates the relaunch through real store persistence; the dead allowedPeerIDs parameter (unscoped-tombstone trap) is removed. |
||
|
jackandGitHub
|
a4d294015a |
Keep DMs canonical across transport aliases (#1461)
Joins the two authenticated aliases of one account — the full noise-key conversation ID used by the Nostr path and its derived 16-hex mesh short ID — into one canonical conversation, migrating messages and handing off an open DM sheet (short ID while connected, stable ID offline). Delivery/read-ack outbox removal is scoped to the authenticated peer's aliases with per-peer tombstones, so a colliding message ID queued for a different peer survives, including across protected-data cold-load recovery. Geohash DMs never enter the alias path (their conversation keys carry no noise key). |
||
|
jackandGitHub
|
660632ef6b |
Keep open DMs alive across Bluetooth settings (#1460)
Presenting the root Bluetooth alert while the people/DM sheet was up forced SwiftUI to dismiss the sheet, whose binding setter called endConversation() — destroying the open DM. The root alert is now gated behind a modal-presentation guard (scene active, no competing modal), a second copy presents from inside the sheet, and the sheet-dismissal setter only ends the conversation when the dismissal is genuinely explicit. Includes review fix: the voice-recording error alert (mic permission denied inside an open DM) destroyed the DM through the exact same path — it now participates in the same guard and gating pattern at both root and sheet level. |
||
|
jackandGitHub
|
c72bb4ca2e |
Make private-media deletion transactional (#1468)
Receiver-side private-media deletion (per-bubble delete and /clear) becomes a write-ahead transaction: a deletion journal in BLEPrivateMediaReceiptStore is the atomic commit point, materialization (tombstones + payload unlinks) is idempotent and retried on recovery, path reservations in BLEIncomingFileStore prevent delete racing an in-flight arrival, and overlapping /clear operations are serialized with panic-generation invalidation. Integrates with the receipt quarantine (a pending journal entry outranks quarantine; materialization never resurrects a quarantined ID). Review fixes included: (1) /clear no longer deletes outgoing media mirrored into another conversation (alias protection now mirrors the incoming path, with a regression test that fails pre-fix); (2) explicit delete of legacy incoming media actually unlinks the decrypted payload when unreferenced — gated on pending-delivery/reservation state, restoring main's delete semantics safely instead of leaving plaintext for quota cleanup; (3) refused deletions surface a localized system message in the affected chat instead of failing silently (30-locale key). Full local suite 1876 green. |
||
|
jackandGitHub
|
d6bd4f0681 |
Retry confirmed private media after reconnect (#1467)
Sender-side retry of private media whose local BLE fragment completion never received a remote delivery receipt: the encrypted file packet is retained in memory (bounded: 8 packets / 4MB / 120s / 2 retries per message / 2 per reconnect) and re-sent on peer reconnect or re-authentication, gated at four boundaries on capability bit 9 plus the exact Noise session generation, so legacy or bit-8-only peers can never receive a retry and the receiver's stable-ID ledger dedups anything re-sent. Remote delivery/read receipts release retention; limits and expiry end in a visible failed status. Includes review fix: a dropped policy-resolution callback can no longer wedge retries for a peer — the pending resolution is cleared on peer disconnect (with a regression test proving recovery). |
||
|
jackandGitHub
|
fb451bc6d0 |
Persist authenticated private-media delivery receipts (#1466)
Adds a durable receiver-side ledger (BLEPrivateMediaReceiptStore) mapping a deterministic private-media message ID — hash-bound to sender, recipient, and entropy-bearing filename — to the stored file before UI delivery and before the Noise-encrypted delivery ACK, so sender retries and relaunches cannot create duplicate bubbles or files. Receipts are unforgeable without the Noise session; capability bit 9 rides the existing announce bitfield (no wire-format change; rolling upgrade safe both directions). Includes review fixes: corrupt receipt records are quarantined per-record (bytes preserved at <id>.json.corrupt, only that ID fail-closed — previously one bad record silently disabled ALL inbound private media forever); the panic reset now reaches BLEService's own store instance via completePanicReset with a production-wiring test; and both content.delivery.reason.* strings ship with full 30-locale coverage. |
||
|
jackandGitHub
|
a9ceddab21 |
Deliver live DM transport events synchronously (#1465)
Removes the nested-Task deferral when typed transport events reach ChatViewModel: didReceiveTransportEvent now dispatches to @MainActor *Synchronously coordinator methods inline within the single notifyUI main-actor hop, fixing a real ordering race where a delivery-status update emitted right after a message could apply before the message existed in the ConversationStore and be dropped. Adds SynchronousMessageTransportEventDelegate so BLEService.deliverTransportEvent returns a documented accepted/unconfirmed verdict for .messageReceived (consumed by the private-media receipt layer above). Intra-main-actor only — no cross-queue sync added; the bleQueue-never-blocks-on-main invariant is preserved. |
||
|
jackandGitHub
|
e3e97d51ec |
Preserve early Noise ciphertext across reconnect promotion (#1464)
Companion to #1463 (they were built as a unit): adds a bounded deferral buffer (4/peer, 32 global, byte-budgeted, responder-timeout lifetime) for ciphertext that arrives before XX message 3 or before BLE installs generation-bound transport state, retried exactly once after the serialized authentication/restore callback. Reclassifies malformed/forged/replayed/oversized/rate-limited ciphertext as drop-only, so attacker bytes can no longer evict an established session (on main, a 1-byte forgery evicts the session via the old clearSession + re-handshake path). Serializes Noise packet reception as messageQueue barriers. Honors #1463's timeout-restore deferral through the same-generation ready path (verified by negative test: without the gate, parked DMs drain under discarded keys). Full local suite 1789 tests green. |
||
|
jackandGitHub
|
78a81e5b57 |
Make ordinary Noise reconnects atomic and race-safe (#1463)
Replaces destroy-then-rebuild Noise re-handshakes with an atomic reconnect protocol: prepared XX message-1 handoff tokens (claim-once, invalidated by crossed inbound initiations), receive-only quarantine of the established transport while an inbound replacement proves identity (promote on success, rollback+cooldown on failure/timeout), deterministic lower-peerID-wins crossed-initiator resolution, and a per-link-epoch BLE revalidation policy that re-proves cached sessions inside the same bleQueue critical section as the link rebind. On main, a fresh msg1 simply destroys an established session and rekey spans two non-atomic barriers. Includes the review fix: timeout-restores defer outbound queue draining until the convergence retry completes (restore reason plumbed end-to-end), so DMs are never drained under keys a restarted peer already discarded — with a deterministic interleaving test. Identity-mismatch restores drain immediately. Full local suite 1768 tests green. |
||
|
jackandGitHub
|
e9275cb3d8 |
Relabel private Nostr envelopes honestly in docs; harden legacy envelope validation (#1480)
Split-out safe half of #1437 (the kind-1402 wire migration stays held for Android coordination). Docs (README/WHITEPAPER/PRIVACY_POLICY/privacy-assessment) now describe the actual proprietary DM construction — kind 1059 gift wrap carrying XChaCha20-Poly1305 with a 24-byte nonce, base64url v2: framing, and an HKDF that borrows the nip44-v2 info label but is not the NIP-44 key schedule — instead of claiming NIP-17/44/59. Hardens the existing legacy inbound path: 64 KiB ciphertext cap before decode (~46x the largest producible legacy envelope), outer kind/recipient-tag/signature binding, tagless kind-13 seal binding, unsigned kind-14 inner binding, inner tags restricted to the two shapes deployed clients emit (verified against every historical iOS release and a fixture frozen from Android production), SecRandomCopyBytes failure now throws, non-UTF-8 plaintext now throws instead of returning empty. Adds frozen cross-platform fixtures with hash-pinned generators; interop-reviewed with no rejection surface for deployed clients. |
||
|
jackandGitHub
|
55f824a11f |
Make nearby notes tests parallel-safe (#1471)
Injects the location-notes enabled getter and settings-change publisher into NearbyNotesCounter (defaults reproduce the prior wiring exactly), so tests no longer mutate the persistent UserDefaults key shared across parallel test worker processes. No production behavior change. Un-stacked from the Codex chain (both files were byte-identical between main and the parent branch) and rebased onto the post-#1479 deflaked test suite. |
||
|
jackandGitHub
|
2d96fd99a1 |
Encrypt private media before BLE fragmentation (#1434)
Closes the last cleartext private-content path over BLE: private DM images/voice were sent as plaintext signed fileTransfer packets, TTL-relayed across the mesh, so every relay saw the full bytes. Now the complete BitchatFilePacket is encrypted as a single Noise AEAD message (inner type 0x20, matching Android) and the opaque ciphertext is fragmented. Adds an authenticated in-session capability proof (0x21 TLV: capabilities + Ed25519 key), TOFU-style downgrade pinning, a per-send consent dialog for the signed-cleartext fallback to legacy peers, and a cancellation/admission registry so cancel/delete cannot race a deferred cleartext send. Android wire constants (0x20 / 0x21 / capability bit 8) confirmed shipping. The 256-fragment preflight cap applies only to the directed fileTransfer migration fallback; encrypted media to capable peers uses the full receiver ceiling. Rebased over #1428/#1349: identity reads go through BLELocalIdentityStateStore; the session-bound authenticated signing-key check and the announce-path TOFU pin are kept as complementary checks. Full local suite green (1744+197 tests). |
||
|
jackandGitHub
|
10886428ca |
Deflake GeoRelayDirectoryTests under constrained runners (#1479)
Raises the file-local waitUntil default from 1s to 10s (returns early on success, so zero cost on healthy runs) and fixes a latent race in the observers test: it waited on the fetch-probe request count, which increments before handleFetchSuccess settles isFetching/lastFetchAt on the MainActor, so the next trigger notification posted into that gap was silently swallowed by the guard. Now waits on the .geoRelayDirectoryDidRefresh notification posted at the end of the synchronous success handler. Negative checks assert against captured baselines. No assertion weakened; 15/15 x 9 local runs including under 2x-ncpu CPU load. |
||
|
jackandGitHub
|
bcb21f2116 |
Require review before importing shared content (#1430)
Replaces the share-extension handoff that auto-sent shared text/URLs into the current channel with a validated, single-slot 24h envelope that shows the destination and a bounded preview and requires an explicit tap to copy into the composer — it never sends. Rejects malformed/oversized/control-character/non-HTTP(S)/expired payloads (including bidi-override U+202E), and clears on consume/cancel/expiry/panic. Rebased onto main with Persian strings added for all new keys (30-locale coverage verified), plus a panic-recovery clear so an envelope staged during an interrupted wipe cannot survive relaunch. |
||
|
jackandGitHub
|
b96a41054e |
Pin announce signing keys to stop mesh identity spoofing (#1349)
TOFU-pins the Ed25519 signing key per noise key so a self-consistent announce that reuses a victim's peerID+noise key with the attacker's signing key/nickname is rejected instead of overwriting the registry and persisted identity. Complements #1432: that PR's signed-LEAVE verification reads the stored signing key this PR protects from announce-path poisoning. Rebased onto main as a single commit; swift build --build-tests green, 40/40 targeted tests incl. the spoofing e2e cases, and #1432's suites 50/50 unaffected. |
||
|
jackandGitHub
|
d326cecb63 |
Fix BLE identity state races (#1428)
Converts BLEAnnounceThrottle to a lock-backed class and moves myPeerID/myPeerIDData/myNickname into a lock-backed BLELocalIdentityStateStore snapshot, so announces can never observe a split peer-ID/wire-ID during panic rotation. Serializes sendAnnounce onto the messageQueue barrier and moves the panic-reset pendingNoiseSessionQueues clear onto collectionsQueue (the queue every other mutation site uses). Rebased onto main after #1431/#1432; integrates with #1431's panic-suspend structure (guard retained in both the outer sendAnnounce and the deferred worker). |
||
|
jackandGitHub
|
b59ad97dd6 |
perf: verify Nostr inbound signatures once, off the main actor (#1352)
Moves inbound Nostr Schnorr verification and NIP-17 gift-wrap decryption off the main actor into per-relay bounded AsyncStream pipelines (parallel across relays, ordered per relay), dedups before crypto, and removes the redundant second re-verify in NostrInboundPipeline/GeoPresenceTracker. Adds a panic-wipe generation counter so in-flight decrypts are discarded after a wipe. Rebased onto main; #1451's 256 KiB pre-parse byte cap and tag limits are preserved and still enforced before any crypto/allocation. |
||
|
jackandGitHub
|
58ccb30575 |
Route GeoRelay updates through reviewed data (#1436)
Replaces the scheduled workflow that pushed upstream georelay CSV directly to main unreviewed with a validator-backed automation that resolves an immutable upstream commit, validates against the reviewed baseline, and opens a non-auto-merged PR (or tracking issue) instead of writing to main. Adds a mirrored strict client-side validator in GeoRelayDirectory (all-or-nothing, size/row/host/coord caps, streamed 512KiB cap, >=50% baseline overlap) that fails safe to last-known-good, and retargets runtime refresh to bitchat's reviewed copy. |
||
|
jackandGitHub
|
400ac4c904 |
Avoid conversation index rebuilds at retention cap (#1435)
Replaces ConversationStore's message-ID->physical-index map with logical indexes plus a head indexOffset, so cap-eviction advances the offset instead of rebuilding the whole dictionary on every steady-state append (~23.7x measured ingest throughput at cap). Adds a steady-state benchmark floor and a 1,200-op differential stress test against an O(n) reference model. |
||
|
jackandGitHub
|
565d2b7773 |
Make developer cleanup artifact-only (#1433)
Rewrites just clean/nuke to remove only ignored build artifacts (.DerivedData/.build), never git-checkout or rm tracked project files (previously could destroy uncommitted work). Adds a CI guard (check-just-clean-safety.sh) against reintroducing source-mutating cleanup, plus README/Local.xcconfig modernization. |
||
|
jackandGitHub
|
6b71ad2a64 |
Run iOS simulator tests in CI and repair coverage reporting (#1429)
Adds an ios-tests CI job that runs the UIKit/CoreBluetooth-conditional suite on the first available iPhone simulator (serial), and fixes the coverage-summary step ordering so the profile isn't invalidated by the serial benchmark rebuild. |
||
|
jackandGitHub
|
16324c819f |
Bind Noise sessions to claimed peer identities; require signed leaves (#1432)
Adds remote-static-key->peerID binding at Noise handshake completion (closes a mesh impersonation/MITM hole where a peer could complete a handshake under another peer's ID). Also hardens LEAVE handling to require a verified signature and suppresses relay of unverifiable leaves. |
||
|
|
cd727c6867 |
Make panic wipe deterministic and device-bound (#1431)
* Make panic wipe deterministic and device-bound * Scope install markers to iOS * Harden panic recovery and service shutdown * Invalidate queued BLE ingress during panic * Harden panic keychain and media cleanup --------- Co-authored-by: jack <jackjackbits@users.noreply.github.com> Co-authored-by: jack <jack@deck.local> |
||
|
GitHub Action
|
fb8fe39713 | Automated update of relay data - Sun Jul 26 06:54:16 UTC 2026 | ||
|
|
ca18843bb0 |
Add Persian (fa) localization and an in-app language picker (#1443)
* Add Persian (fa) localization and an in-app language picker Persian was the one notable gap in the 29-language catalog. Translate all 381 strings (plus the share extension) with proper plural substitutions, and register fa in knownRegions. Settings gains a LANGUAGE section: a picker over every language the bundle ships (native names via Locale), backed by an AppleLanguages override so users can run bitchat in a language different from the device's. Localization resolves at process start, so the picker surfaces a restart note. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Remove unused AppLanguageSettings.currentOverride (Periphery) AppInfoView reads the override via @AppStorage, so the accessor was dead code and failed the Periphery CI scan. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> Co-authored-by: jack <jackjackbits@users.noreply.github.com> Co-authored-by: jack <212554440+jackjackbits@users.noreply.github.com> |
||
|
|
593fd7d737 |
Harden public intake bounds against untrusted growth (#1451)
* Bound public rate-limit buckets against attacker-keyed growth. Keep the NIP-13 PoW sender bypass, skip content-bucket minting on sender reject, and evict idle/oldest entries at a hard cap. Co-authored-by: Cursor <cursoragent@cursor.com> * Stop Cashu-looking text from skipping long-message guards. Oversized public content always collapses and takes the plain formatting path so remote tokens cannot force layout/regex DoS. Co-authored-by: Cursor <cursoragent@cursor.com> * Cap teleported geohash participant markers. Bound the set with FIFO eviction, clear it on channel switch, and prune markers that leave the visible participant list. Co-authored-by: Cursor <cursoragent@cursor.com> * Bound untrusted Nostr relay frames and event tags. Reject oversized inbound messages before JSON parse, cap tag arrays/values at decode, and stop logging raw tag contents. Co-authored-by: Cursor <cursoragent@cursor.com> * Fail soft when Noise handshake state is unexpectedly missing. Replace the initiator startHandshake force unwrap with a guard that throws invalidState instead of crashing. Co-authored-by: Cursor <cursoragent@cursor.com> * Cap geohash nickname cache from remote Nostr events. FIFO-evict at capacity, clear on channel switch, and prune nicknames that leave the visible participant list. Co-authored-by: Cursor <cursoragent@cursor.com> * Avoid overlapping exclusive access in the rate limiter. Make bucket helpers static so inout dictionary updates do not conflict with a mutating call on self. Co-authored-by: Cursor <cursoragent@cursor.com> * Fix rate-limiter tests for mutating allow under #expect. Call allow outside the macro so Swift Testing does not capture an immutable copy of the struct. Co-authored-by: Cursor <cursoragent@cursor.com> * Drop unused WebSocket data helper; reset rate limiter on panic wipe. dataWithinInboundLimit replaced the unbounded path, and panic clear should not leave public intake buckets behind. Co-authored-by: Cursor <cursoragent@cursor.com> --------- Co-authored-by: Cursor <cursoragent@cursor.com> |
||
|
|
733098bb63 |
fix: use country-level resolution for low-precision geohashes (#1044)
* fix: use country-level resolution for low-precision geohashes (#887) * fix: skip admin fallback when country exists but is duplicate Prevents mixed labels like "United Kingdom and Scotland" for single-country geohashes. Admin fallback now only triggers when no country is available from the placemark. * fix: migrate stale low-precision bookmark names so country-first logic applies Users who bookmarked a <=2-char geohash before the country-first resolver kept the old administrativeArea cache (e.g. "England" for `gc`) because resolveBookmarkNameIfNeeded bails when bookmarkNames[gh] is non-nil. Add a one-shot, versioned migration that drops cached names for <=2-char geohashes on load; the next LocationChannelsSheet .onAppear re-resolves them via the fixed logic. Higher-precision entries are untouched. * Fix low-precision geohash country names --------- Co-authored-by: jack <212554440+jackjackbits@users.noreply.github.com> Co-authored-by: jack <jackjackbits@users.noreply.github.com> |
||
|
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. |
||
|
|
b205a55523 |
Consolidate duplicate BLE links after restore; duplicate fragment-stream suppression; onChange coalescing (#1426)
* Consolidate duplicate BLE links after restore; suppress duplicate fragment streams; onChange coalescing
Field evidence (two-phone test after BLE state-restoration relaunches):
both phones held 2-3 simultaneous same-role connections to each other —
one side received every packet from three distinct centrals all bound to
the same peer — so every PTT voice frame arrived 3x, and a 41KB voice
file went out as TWO complete independent 89-fragment streams (different
assembly ids), both fully reassembled and the duplicate only dropped at
the very end by messageID dedup. 2-3x airtime/battery on all traffic
plus doubled reassembly memory.
Root causes and fixes:
- Duplicate links stayed unbound forever: announces (the only packet
that binds a link to a peer) went through the per-peer duplicate-link
collapse, so a peer's second/third link never received the announce it
needed to become bound — and unbound links pass the collapse untouched,
so every broadcast sprayed down all of them. Direct announces now
bypass the collapse and reach every live link (relayed announces keep
it); once bound, the existing collapse dedups all traffic.
- Same-role link retirement: a verified direct announce now consolidates
our central-role connections to that peer — keep the link the announce
arrived on (or the most recently bound one), cancel other connected
links bound to the same peer. One connection per role per peer is the
normal dual-role topology; only same-role duplicates are touched, only
links already announce-bound are retired (never pre-announce links),
at most one retirement per peer per rebind-cooldown window, and the
peer keeps a live link either way. Directness stays forgeable (TTL is
unsigned), so a replay could nominate the survivor — bounded by the
cooldown and by DM routing's existing canDeliverSecurely gate. A
rotation rebind now also cancels other stale links still bound to the
rotated-away ID, so the ghost identity retires promptly.
- Deterministic preferred-link collapse: when several bound links to one
peer are candidates, collapse now keeps the peer's most recently bound
link (the reverse-mapped one) instead of dictionary order; links
without a discovered characteristic are excluded from fanout (they
cannot be written to, and could silently eat a peer's collapsed copy).
links(to:) now reports all bound peripheral links, and removing one
duplicate no longer clobbers the reverse map of the survivor.
- Duplicate fragment streams: a transferId-less resend (gossip-sync
replay, spool) of file content already being fragmented out to a
covering audience is dropped at the outbound scheduler (broadcast
covers everyone; directed covers its recipient). App-initiated sends
carry an explicit transferId the progress UI tracks and always run.
A peer that asks after the stream completes still gets a resend.
- didUnsubscribeFrom no longer flaps a peer that is still live on other
links (the far side retiring its duplicate arrives as an unsubscribe);
didDisconnectPeripheral bookkeeping is skipped for self-retired links
by removing the store entry before cancelling.
Also: guard the DeliveryStatus onChange state write in TextMessageView/
MediaMessageView (unconditional per-row writes under a message storm
tripped SwiftUI's "tried to update multiple times per frame" warning).
The restore-path bgRemaining=∞ log was already fixed on main (#1425
review follow-up
|
||
|
|
f8ab0a7dc0 |
Fix restore-path main↔bleQueue deadlock and courier drop amplification (#1425)
* Fix restore-path main↔bleQueue deadlock and courier drop amplification A device froze permanently in a two-phone test. Debugger stacks showed an ABBA deadlock: the main actor was in bleQueue.sync (delivery-ack send → broadcastPacket → readLinkState) while bleQueue was in main.sync (captureBluetoothStatus reading backgroundTimeRemaining). The load that lined the two edges up came from a courier-drop amplification storm: drop dedup was in-memory only while the outbox driving 120s re-deposits is persisted, so every relaunch republished the same undelivered DM as a fresh 24h relay drop and every gateway relaunch re-fetched the whole backlog — ~20 copies of one DM delivered in 40ms, each triggering decrypt + delivery + ack + handshake work. Fixes, in rank order: - Edge B (P0): captureBluetoothStatus no longer main.syncs from bleQueue; backgroundTimeRemaining is sampled on main and cached behind a lock. Invariant documented: bleQueue must NEVER sync-dispatch to main. - Edge A (P0, defense in depth): sendDeliveryAck / sendReadReceipt / sendPrivateMessage / sendNoisePayload / triggerHandshake hop to messageQueue like sendMessage, so no main-actor call path reaches readLinkState's bleQueue.sync. - Drop dedup (P1): publishedDropKeys and seenDropEventIDs persist across relaunches (new BridgeDropDedupStore, entries expire with the 24h NIP-40 drop window; wiped on panic) — one drop per message ID per 24h regardless of relaunch count. - Receiver dedup (P1): openCourierEnvelope dedups on the inner private message ID before delivery, so a duplicate copy costs one decrypt and never re-delivers, re-acks, or re-triggers a handshake. - Handshake gating (P2): queued acks initiate a Noise handshake only for reachable peers; mail from absent/rotated identities no longer turns each copy into a mesh-wide handshake flood (the ack stays queued and flushes when a session eventually establishes). - Outbox (P3): re-enqueueing a queued message ID carries over its depositedCourierKeys so resends stop re-burning the same courier slots. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Review fixes: offline-drop durability, gateway handoff retry, restore-log freshness, coalesced persist Adversarial review of the storm/deadlock PR surfaced four issues: - Offline blackhole (must-fix): a deposit made while relays were down persisted its dedup key even though the drop only sat in the in-memory pending queue — app killed before reconnect meant the relaunch lost the drop but the persisted key blocked every re-deposit for 24h. The persisted snapshot now excludes keys still pending; they become durable only when flushPendingDrops actually publishes them. - Gateway handoff: seen-event IDs were consumed before the deliverToPeer handoff; a failed handoff (peer walked away) permanently dropped the event for a single-gateway island. deliverToPeer now reports whether the handoff was attempted, and a failure releases the seen slot so a relaunch or backlog redelivery retries. - Restore-path logs: central/peripheral-restore captures logged the init sentinel bgRemaining=∞. The cache is now seeded in init's main-thread branch and restore captures route through the sampler, which refreshes the cached budget before logging. - Persist cost: the dedup record was a full JSON encode + atomic write on the main actor per mutation (once per event during a backlog re-fetch). Writes now coalesce behind a 1s window, flushed immediately on background/terminate; panic wipe stays immediate. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Remove BoundedIDSet.remove, orphaned by the ExpiringIDSet migration The drop-dedup sets that needed slot release moved to ExpiringIDSet; remaining BoundedIDSet users only insert and check. Periphery caught it. 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> |
||
|
|
304460ee83 |
Nearby notes: tap-to-reveal (consent-gate the location subscription) + subscription hygiene (#1422)
* Nearby notes: tap-to-reveal before any relay REQ, plus shared subscription pool The nearby-notes counter used to open a live building-precision (precision-8) geohash REQ to the closest geo relays whenever the mesh public timeline was visible — a passive location side-channel with no opt-in. Now nothing subscribes until one explicit act reveals the counter for the session: tapping the new "check for notes left here" line on the empty mesh timeline (static, no network), opening the notices sheet's geo tab, or a successful /drop. The app-info setting stays the default-ON kill switch and still gates /drop. Subscription hygiene alongside: - LocationNotesManager.deinit now unsubscribes its live REQ (hopping to the main actor like the timer teardown) instead of only invalidating timers. - New refcounted LocationNotesPool dedupes the counter's and the notices sheet's identical 9-cell kind-1 REQs into one shared manager per geohash; both callers release-and-reacquire instead of retargeting in place, and the sheet releases its ref on dismissal. The reveal affordance is localized across all 29 catalog locales, and new NearbyNotesCounterTests cover the no-REQ-before-reveal contract, the 9-cell filter, NIP-40 expiry handling, single unsubscribe on deactivate, and pool refcounting. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * Review fixes: permission-gate the hint, exclude building cell from pre-reveal sampling, explicit-act reveal only Fixes the verified review findings on the tap-to-reveal PR: - Permission dead-end: the "check for notes left here" hint now renders only when location permission is already authorized (it never prompts). Previously a location-denied install could tap it, flip the sticky revealed flag, and get nothing for the rest of the session because retarget() guards on authorization. Predicate lives on NearbyNotesCounter.offersRevealHint(permissionState:) and is reactive to the published permission state. - Building-cell sampling: background geohash sampling subscribed geo-sample-<gh> for every regional level including the precision-8 building cell, pre-reveal — contradicting the claim that nothing building-precision hits a relay before the explicit act. GeoChannelCoordinator now excludes the building level until NearbyNotesCounter.revealed (injectable publisher for tests); coarser levels keep the nearby-conversation hint and participant counts working, and bookmarks stay exempt (bookmarking is explicit). - Implicit reveal: opening the notices sheet no longer reveals — the sheet auto-lands on the geo tab whenever a location channel is selected, so browsing a remote geohash and opening notices revealed the LOCAL building subscription. reveal() now fires only on the person actively picking the geo segment, and only when the sheet has a geo scope (the empty-mesh hint tap and /drop stay as before). - Subscription hygiene: switching the sheet geo → mesh releases the pooled notes manager (the REQ was left streaming behind the mesh board); switching back re-acquires from the pool. The dismissal release stays balanced — liveGeoManager is nil after the tab-switch release. - VoiceOver: the hint button exposes the plain localized action text instead of the decorated "* 📍 … *" label. - Removed LocationNotesManager.setGeohash: zero callers, and calling it on a pooled instance would corrupt the pool's keying and refcounts. New tests: hint permission gate, explicit-geo-tab reveal contract, building-cell sampling exclusion before/after reveal, pool release/re-acquire round trip. Full suite (1467) green; iOS simulator build green. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Deflake peer-snapshot binding waits: longTimeout for the multi-hop pipeline CI failed once on initialization_bindsPeerSnapshotsIntoAllPeers: the snapshot -> allPeers binding crosses the mock transport's unstructured Task, UnifiedPeerService.updatePeers, a receive(on: main), and another Task { @MainActor } in bindPeerService — all contending with every parallel worker. On the failing runner the whole suite took 10.1s (usually ~4.4s locally), so the positive 5s defaultTimeout wait lost the race. That's exactly the case TestConstants.longTimeout documents; passing runs return as soon as the condition holds and never pay it. Not caused by the tap-to-reveal changes: nothing on that pipeline was touched, and 20 full parallel-suite loops each on the branch and on origin/main reproduce zero failures locally. The two sibling waits on the same pipeline in this file get the same timeout. 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> |
||
|
|
b7c6f42b3a |
Close forged-directness link-rebind DoS on DM routing (#1421)
* Gate connected-link trust on a secure session; deny forged direct announces the connected shortcut
Residual gap after the rotation-heal containment (#1401): "verified
direct" announces prove the signature but not directness — TTL is
unsigned — so a malicious connected peer can replay a victim's fresh
announce with its TTL restored. When the victim has no live link, the
replayer's link rebinds to the victim's ID and reads as "connected",
and MessageRouter's connected fast-path then trusts it outright: every
DM stalls on a Noise handshake the replayer can never complete and is
silently lost while showing "sent".
Router-level trust gate (no wire change):
- Transport gains canDeliverSecurely(to:) — BLE answers with an
established Noise session; Nostr keeps its prompt-delivery predicate;
the protocol default forwards to canDeliverPromptly for transports
without a forgeable link layer.
- MessageRouter.sendPrivate only trusts a connected link outright when
it can deliver securely; otherwise it still sends (kicking the
handshake on a genuine link) but retains a copy and hands a sealed
copy to couriers, like the reachable path. flushOutbox gets the same
gate so a flush over an insecure link resends instead of dropping the
retained copy.
Presence hardening (defense in depth): a "direct" announce arriving on
a link already bound to a different peer no longer shortcuts the
claimed peer into "connected" — only real link state does. Genuine
first-contact and direct announces are unaffected; only the ambiguous
heal path loses the forgeable shortcut.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
* Harden the insecure-link outbox bound; document the second-replay presence gap and the courier metadata tradeoff
Review follow-ups on the canDeliverSecurely gate:
- flushOutbox no longer counts connected-but-insecure flushes toward the
maxSendAttempts drop: the message was actually transmitted over a live
link, so a peer whose Noise handshake stalls across reconnect flapping
must not burn through the cap and lose the store-and-forward copy the
gate exists to preserve. Retention stays bounded by the 24h outbox TTL
and the per-peer FIFO cap; acks still clear it. Attempt-counting stays
for reachable-only (heuristic) sends. Regression test: >8 connected-
insecure flushes keep the retained copy, drop callback never fires.
- Document the known second-replay presence gap: linkBoundToOtherPeer
reads the binding before rebindLinkAfterVerifiedDirectAnnounce steals
the link, so once a first replay has rebound a link to an absent
victim's ID, a second replay marks the victim connected. Presence
display only — DMs stay on the retain+courier path via the router
gate. Not closed at the announce layer because the post-rebind state
is indistinguishable from a legitimate rotation/reconnect heal (a
supported, field-verified flow). Covered by a two-announce test.
- Note the accepted courier-spray metadata tradeoff at the connected-
insecure send: nearby verified peers receive a sealed copy (they learn
a DM exists, never its content), cleared on ack.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* Promote a healed rotation to connected; normalize the secure-delivery probe; retain Codable properties in Periphery
Codex review follow-ups on
|
||
|
|
c8479c15e3 |
PTT hardening: per-peer burst keys + live-capture storage quota (#1420)
* PTT hardening: burst-ID collision hijack + live-capture quota bypass Fix C — burst-ID collision hijack: inbound live-voice assemblies and the finished-burst registry were keyed by the sender-chosen burst ID alone, so an attacker who observed a public burst ID could race a START and capture the real talker's frames (packets from the true sender were dropped as "collisions"). Assemblies now key on (peerID, scope, burstID): a colliding START from another peer opens its own capped assembly and can never divert the victim's frames; finalized-note absorption matches on the same triple, preserving the sender/scope binding. Live capture files also gain the peer ID in their name so colliding bursts land on distinct paths (still rejected by burstID(fromVoiceFileName:), so live names remain unabsorbable). Fix B — live-burst files bypassed the incoming-media quota: progressive voice_live_*.aac captures were written via raw FileHandle without ever touching BLEIncomingFileStore.enforceQuota, growing disk unbounded outside the 100 MB LRU accounting. The coordinator now takes an injected BLEIncomingFileStore, reserves pttMaxBurstBytes before opening a capture, and sweeps orphaned voice_live_* partials from previous sessions at startup. enforceQuota gains an `excluding:` set so in-flight partials are never LRU-evicted mid-stream; existing callers are unaffected. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * Review fixes: pattern-guard live captures in quota, scope in capture names Quota-eviction gap: BLEFileTransferHandler enforces the quota for every finalized arrival via enforceQuota(reservingBytes:) with no exclusion set, so a file landing at quota could LRU-evict an in-flight live capture — unlinking the inode under the coordinator's open FileHandle and leaving a dead bubble. Protection is now layer-independent: enforceQuota itself skips voice_live_* names (they still count toward usage), and the excluding: parameter is gone since the pattern guard covers its only caller. Orphaned partials are still reclaimed by the coordinator's startup sweep, which the quota deliberately never touches. Same-peer cross-scope truncation: makeIncomingURL omitted the scope, so one peer running a DM burst and a public burst with the same burst ID mapped both assemblies onto one path — and FileManager.createFile truncates, so each START corrupted the other capture. Names now mirror the assembly key: voice_live_<burstHex>_<peerID>_<dm|mesh>.aac. The sweep and quota guard match on the voice_live_ prefix and burstID(fromVoiceFileName:) still rejects every live name, so live captures remain unabsorbable; sameBurstIDCoexistsAcrossScopes now asserts both files survive with intact contents. Nits: the coordinator's file operations route through the store's injectable FileManager instead of FileManager.default, and the TestEnvironment.isRunningTests branch in init is replaced by a sweepsOnInit parameter (tests sharing the real application-support directory pass false for hermeticity). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * Promote kept live captures off the voice_live_ name at finalize Codex P1 follow-up: when a burst finalizes with frames but its .m4a note never arrives, the voice_live_*.aac capture stays behind as the bubble's replayable audio — yet the startup sweep deletes every voice_live_* file and the quota guard skips them forever, so a kept fallback was both quota-immune for the rest of the session and doomed at the next launch. finalize now promotes the capture to a plain voice_ name (same suffix) and republishes the row pointing at it; the finished-burst registry tracks the promoted URL so a late note still absorbs and deletes it. voice_live_ is thereby scoped to genuinely in-flight captures: the sweep never touches a referenced fallback, and promoted files age out of the quota like any finalized media. If the move fails the live name is kept — exactly the pre-promotion behavior. Premise correction, verified while tracing the reference model: chat rows are NOT persisted across restarts (ConversationStore is in-memory; the gossip archive replays MessageType.message packets only), so today's sweep never orphaned a persisted row — the fix removes the in-session quota immunity and makes the invariant hold if row persistence ever lands. Crash case stays deliberate and documented: a mid-burst partial from a dead process is swept because no surviving row can reference it. Tests: finalize-as-fallback promotes the file, repoints the row, and survives a later coordinator's startup sweep; promoted fallbacks are LRU-evictable by the quota; the sweep still removes true orphans while sparing promoted names; existing burst tests updated for the promoted paths. 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> |
||
|
|
84d315e62d |
Bridge dedup: content-derived stable mesh message ID (#1419)
* Bridge dedup keys on a content-derived stable mesh message ID Public mesh messages carry no message ID on the BLE wire, so every non-origin device minted a fresh UUID and the bridge's m-tag dedup only matched on the origin device: duplicate bridged rows, misattributed "across the bridge" counts, redundant downlink rebroadcasts, and an m-tag spoof vector (an attacker could claim a victim's message ID). Every device now derives the same stable ID from the signed wire fields (sender ID + ms timestamp + trimmed content, SHA256/32 hex) via the new MeshMessageIdentity — zero BLE wire change. The bridge event's m tag carries the origin coordinates ["m", senderIDHex, timestampMs] and receivers recompute the key from those plus the event's own content instead of trusting a claimed ID; old-format/absent tags fall back to the event ID as before. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * Fix mixed-version message loss: m tag leads with the derived stable ID The previous layout (["m", senderIDHex, timestampMs]) broke v1.7.0 receivers: their parser takes m[1] unconditionally as the timeline dedup key whenever the tag has >= 2 elements, so every bridged message from a new-version sender keyed on the CONSTANT sender hex and inject-dedup dropped all but the first. The tag is now ["m", <derived stable ID>, senderIDHex, timestampMs]: old parsers get a per-message-unique m[1] (exactly today's semantics), while the new parser recomputes the ID from elements 2-3 plus the event's own content and never trusts element 1, keeping the recompute-don't-trust property. Also: - Soften the overstated security claim in MeshMessageIdentity and the BridgeService classify comment: forging a chosen ID onto different content is infeasible, but all three hash inputs are cleartext on the radio, so identical-content front-running by a radio-local attacker remains possible (no worse than the unbridged mesh). - Fix the stale archivedEchoKeys rationale: re-synced copies of others' messages now carry the derived stable ID (insert-by-ID catches them); the content key remains for echo--prefixed archive rows + self echoes. - Tests: old-parser semantics on the new tag (m[1] per-message-unique and equal to the derived ID) and a forged-m[1] event that cannot pre-poison a genuine message's dedup slot. 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> |
||
|
|
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> |