@@ -9,6 +9,9 @@ use std::collections::{BTreeMap, HashMap, HashSet, VecDeque};
99use std::sync::Arc;
1010use std::time::{Duration, SystemTime, UNIX_EPOCH};
1111
12+use base64::prelude::*;
13+use base64::Engine;
14+
1215use crux_core::{
1316 macros::effect,
1417 render::{render, RenderOperation},
@@ -58,6 +61,24 @@ const MAX_CONTACTS: usize = 500;
5861/// the budget is dropped exactly like an undecryptable blob would be.
5962const MAX_POLL_TOTAL_BYTES: usize = 64 * 1024 * 1024;
6063
64+/// Cap on the number of events we will attempt to decrypt out of a single poll
65+/// page, independent of the byte budget. Minimum-size blobs (173 bytes decoded)
66+/// would otherwise permit ~388K events within the 64 MiB byte budget, each
67+/// costing a full scalar mult + SHA-512 + AEAD pass — enough to freeze the app
68+/// for tens of seconds per poll. An honest relay caps at 50 events
69+/// (`maxPollEvents`); this is a hostile-relay bound. Set above `MAX_ACK_IDS`
70+/// (1024) so a legitimate large page from one sender still fully processes.
71+const MAX_POLL_EVENTS: usize = 2000;
72+
73+/// Upper bound on a poll response body we will buffer and parse, in bytes.
74+///
75+/// Before the per-blob byte/count budgets in `ingest_poll` ever run, the client
76+/// reads the entire response body into a `Vec<u8>` and deserializes it. A
77+/// multi-gigabyte response from a hostile relay would OOM the app before
78+/// `ingest_poll`'s defenses are consulted. This cap (generously above what 50
79+/// max-size blobs + JSON envelope could cost) bounds the allocation.
80+const MAX_POLL_RESPONSE_BYTES: usize = 64 * 1024 * 1024;
81+
6182/// How long a poll may stay in flight before we assume the shell lost it.
6283///
6384/// `polling` gates the whole long-poll loop, and it is only cleared by a
@@ -73,6 +94,26 @@ const MAX_POLL_TOTAL_BYTES: usize = 64 * 1024 * 1024;
7394/// relay, so leave the margin.
7495const POLL_WATCHDOG_MS: u64 = 90_000;
7596
97+/// How long a send may stay in flight before we assume the shell lost it.
98+///
99+/// `flushing` gates the whole send loop, and it is only cleared by a
100+/// `SendResult`. If the shell ever drops the HTTP effect (a cancelled task, a
101+/// backgrounded app, a bug), `flushing` stays `true` and the outbox is
102+/// permanently stuck — new messages queue up but never send. The watchdog
103+/// releases the guard so the next `StartFlush` can retry.
104+///
105+/// Kept above the shell's 70 s HTTP timeout for the same reason as
106+/// `POLL_WATCHDOG_MS`: a misfire on a merely slow send is survivable (the late
107+/// `SendResult` finds `in_flight` already cleared and pops nothing), but each
108+/// one costs a redundant request.
109+const FLUSH_WATCHDOG_MS: u64 = 90_000;
110+
111+/// How long an auth round-trip may stay in flight before we assume the
112+/// shell lost it. Same rationale as `POLL_WATCHDOG_MS`: `authenticating`
113+/// gates the whole auth flow, and without a watchdog a dropped effect
114+/// leaves it stuck forever. Kept above the 70 s shell HTTP timeout.
115+const AUTH_WATCHDOG_MS: u64 = 90_000;
116+
76117/// Messages per `/messages` request.
77118///
78119/// The relay rejects a longer batch outright (`413 batch_too_large`,
@@ -106,10 +147,18 @@ const K_PROFILE: &str = "profile";
106147const K_CONTACTS: &str = "contacts";
107148const K_CURSOR: &str = "cursor";
108149const K_OUTBOX: &str = "outbox";
150+const K_SEEN_IDS: &str = "seen_ids";
109151fn k_messages(peer: &str) -> String {
110152 format!("messages:{peer}")
111153}
112154
155+/// N7: Cap on the per-peer `seen_ids` dedup set. Without a cap the set grows
156+/// without bound — a peer who sends thousands of messages accumulates
157+/// thousands of IDs, and the `seen_ids` kv blob is rewritten in full on every
158+/// incoming message. Kept well above `MAX_MESSAGES_PER_PEER` (1000) so a
159+/// trim-and-replay within the conversation window still hits the set.
160+const MAX_SEEN_IDS_PER_PEER: usize = 2000;
161+
113162// ---------------------------------------------------------------------------
114163// Effect
115164// ---------------------------------------------------------------------------
@@ -177,6 +226,9 @@ pub enum Event {
177226 LoadedOutbox(#[facet(opaque)] KvData),
178227 #[serde(skip)]
179228 #[facet(skip)]
229+ LoadedSeenIds(#[facet(opaque)] KvData),
230+ #[serde(skip)]
231+ #[facet(skip)]
180232 LoadedMessages(String, #[facet(opaque)] KvData),
181233 #[serde(skip)]
182234 #[facet(skip)]
@@ -210,6 +262,17 @@ pub enum Event {
210262 #[serde(skip)]
211263 #[facet(skip)]
212264 StartFlush,
265+ /// Fires `FLUSH_WATCHDOG_MS` after generation `n`'s flush is issued; a no-op
266+ /// unless that exact send is somehow still in flight — same pattern as
267+ /// `PollWatchdog`.
268+ #[serde(skip)]
269+ #[facet(skip)]
270+ FlushWatchdog(u64),
271+ /// Fires `AUTH_WATCHDOG_MS` after generation `n`'s auth round-trip is
272+ /// issued; a no-op unless that exact auth is somehow still in flight.
273+ #[serde(skip)]
274+ #[facet(skip)]
275+ AuthWatchdog(u64),
213276 #[serde(skip)]
214277 #[facet(skip)]
215278 ChallengeResult(u64, #[facet(opaque)] HttpResult),
@@ -221,7 +284,7 @@ pub enum Event {
221284 PollResult(u64, #[facet(opaque)] HttpResult),
222285 #[serde(skip)]
223286 #[facet(skip)]
224− SendResult(#[facet(opaque)] HttpResult),
287+ SendResult(u64, #[facet(opaque)] HttpResult),
225288}
226289
227290// ---------------------------------------------------------------------------
@@ -247,9 +310,9 @@ pub enum Page {
247310 EditProfile,
248311}
249312
250−/// The five kv keys loaded at startup (`settings`, `profile`, `contacts`,
251−/// `cursor`, `outbox`). `Connect` waits for all of them.
252−const STARTUP_LOADS: u8 = 5;
313+/// The six kv keys loaded at startup (`settings`, `profile`, `contacts`,
314+/// `cursor`, `outbox`, `seen_ids`). `Connect` waits for all of them.
315+const STARTUP_LOADS: u8 = 6;
253316
254317/// The batch a live `/messages` request is carrying: the first `count` items of
255318/// the outbox, all bound for `recipient`.
@@ -275,6 +338,12 @@ pub struct Model {
275338 profile: OwnProfile,
276339 contacts: BTreeMap<String, Contact>,
277340 messages: BTreeMap<String, Vec<StoredMessage>>,
341+ /// Q18: Per-peer dedup ID set, independent of the trimmed conversation.
342+ /// Without this, a replayed blob whose `id` was aged out by `trim_history`
343+ /// (past the 1000-message cap) passes the dedup check and re-enters the
344+ /// conversation. This set is loaded from and persisted to kv alongside
345+ /// messages, so it survives restarts.
346+ seen_ids: BTreeMap<String, HashSet<String>>,
278347 outbox: VecDeque<OutboxItem>,
279348 cursor: i64,
280349 token: Option<String>,
@@ -317,6 +386,12 @@ pub struct Model {
317386 /// items it sent off the head of the outbox, so two overlapping sends would
318387 /// pop each other's messages.
319388 flushing: bool,
389+ /// Which flush the in-flight `SendResult` / `FlushWatchdog` is allowed to
390+ /// belong to. Same pattern as `poll_gen`: bumped every time a flush is
391+ /// issued and every time one is abandoned, so a `SendResult` or watchdog
392+ /// from a superseded send cannot clear the guard or pop items belonging to
393+ /// the replacement.
394+ flush_gen: u64,
320395 /// What the in-flight send will pop off the head of the outbox when it
321396 /// succeeds — see `InFlight`. `None` while nothing is in flight, and also
322397 /// while a send is in flight whose items have since been removed from under
@@ -381,6 +456,7 @@ impl Default for Model {
381456 profile: OwnProfile::default(),
382457 contacts: BTreeMap::new(),
383458 messages: BTreeMap::new(),
459+ seen_ids: BTreeMap::new(),
384460 outbox: VecDeque::new(),
385461 cursor: 0,
386462 token: None,
@@ -392,6 +468,7 @@ impl Default for Model {
392468 authenticating: false,
393469 auth_gen: 0,
394470 flushing: false,
471+ flush_gen: 0,
395472 in_flight: None,
396473 send_batch_budget: None,
397474 loads_pending: 0,
@@ -443,6 +520,24 @@ impl Model {
443520 self.auth_gen
444521 }
445522
523+ /// Open a new flush generation and return it. Same pattern as
524+ /// `next_poll_gen`: a `SendResult` or `FlushWatchdog` whose generation no
525+ /// longer matches is dropped, so a late result from a superseded send
526+ /// cannot clear the guard or pop items belonging to the replacement.
527+ fn next_flush_gen(&mut self) -> u64 {
528+ self.flush_gen = self.flush_gen.wrapping_add(1);
529+ self.flush_gen
530+ }
531+
532+ /// Abandon the in-flight flush, if any: release the guard and retire the
533+ /// generation so a late `SendResult` cannot clear the *replacement* flush's
534+ /// guard or pop its items.
535+ fn abandon_flush(&mut self) {
536+ self.flushing = false;
537+ self.in_flight = None;
538+ self.flush_gen = self.flush_gen.wrapping_add(1);
539+ }
540+
446541 /// Abandon the in-flight auth round-trip (either leg).
447542 fn abandon_auth(&mut self) {
448543 self.authenticating = false;
@@ -489,7 +584,7 @@ fn now_ms() -> i64 {
489584 SystemTime::now()
490585 .duration_since(UNIX_EPOCH)
491586 .map(|d| d.as_millis() as i64)
492− .unwrap_or(0)
587+ .unwrap_or(1)
493588}
494589
495590/// Bare lowercased hostname of a server URL (no scheme/port/trailing dot).
@@ -577,6 +672,16 @@ struct PollEvent {
577672 #[serde(rename = "encryptedBlob")]
578673 encrypted_blob: String,
579674}
675+
676+/// Q17: Deferred delivery-ack accumulator. Collects ack_ids and the max `ts`
677+/// across all `delivery.ack` payloads from one sender in a single poll page,
678+/// so the staleness check and watermark update run once per sender after the
679+/// event loop — immune to relay page ordering.
680+#[derive(Default)]
681+struct PendingAck {
682+ ts: i64,
683+ ids: HashSet<String>,
684+}
580685#[derive(Serialize)]
581686struct SendBatch {
582687 messages: Vec<Envelope>,
@@ -620,6 +725,7 @@ impl App for Skrepka {
620725 KeyValue::get(K_CONTACTS).then_send(Event::LoadedContacts),
621726 KeyValue::get(K_CURSOR).then_send(Event::LoadedCursor),
622727 KeyValue::get(K_OUTBOX).then_send(Event::LoadedOutbox),
728+ KeyValue::get(K_SEEN_IDS).then_send(Event::LoadedSeenIds),
623729 ])
624730 .and(render())
625731 }
@@ -697,10 +803,51 @@ impl App for Skrepka {
697803 }
698804 self.startup_done(model).and(render())
699805 }
806+ Event::LoadedSeenIds(res) => {
807+ match parse_kv::<BTreeMap<String, HashSet<String>>>(res) {
808+ Ok(Some(map)) => model.seen_ids = map,
809+ Ok(None) => {}
810+ Err(()) => {
811+ // Non-fatal: dedup falls back to the conversation-derived
812+ // set, which is weaker but functional.
813+ model.seen_ids = BTreeMap::new();
814+ }
815+ }
816+ self.startup_done(model).and(render())
817+ }
700818 Event::LoadedMessages(peer, res) => {
701819 match parse_kv::<Vec<StoredMessage>>(res) {
702820 Ok(Some(list)) => {
703− model.messages.insert(peer, list);
821+ // D5: Add on-disk message IDs to seen_ids so future
822+ // replays don't pass dedup. Without this, a seen_ids
823+ // kv load failure leaves the set empty, and on-disk
824+ // messages aren't protected against replay.
825+ if let Some(id_set) = model.seen_ids.get_mut(&peer) {
826+ id_set.extend(list.iter().map(|m| m.id.clone()));
827+ }
828+ // Merge: a poll that beat this lazy load may have already
829+ // inserted messages for this peer. Overwriting with the
830+ // on-disk list (which the GET read before the poll wrote)
831+ // would silently destroy the freshly-ingested messages.
832+ // Union by id, keeping both the on-disk history and any
833+ // in-memory additions.
834+ match model.messages.get_mut(&peer) {
835+ Some(existing) => {
836+ let existing_ids: HashSet<&String> =
837+ existing.iter().map(|m| &m.id).collect();
838+ let new_msgs: Vec<StoredMessage> = list
839+ .into_iter()
840+ .filter(|m| !existing_ids.contains(&m.id))
841+ .collect();
842+ existing.extend(new_msgs);
843+ // Re-sort by ts to maintain the invariant.
844+ existing.sort_by_key(|m| m.ts);
845+ trim_history(existing);
846+ }
847+ None => {
848+ model.messages.insert(peer, list);
849+ }
850+ }
704851 }
705852 Ok(None) => {}
706853 Err(()) => {
@@ -810,11 +957,20 @@ impl App for Skrepka {
810957 // would then ack a sequence number B never issued, and B would
811958 // drop every message waiting for us.
812959 model.abandon_poll();
960+ // ...and retire the send that is still in flight *to relay A*,
961+ // for the same reason: its `SendResult` carries A's acceptance,
962+ // and popping items here would eat whatever the new relay's
963+ // send is about to carry. The generation check in `SendResult`
964+ // is what makes the late result a no-op.
965+ model.abandon_flush();
813966 // The old relay's rate limit was the old relay's opinion. A new one is
814967 // entitled to take our sends immediately, and carrying the pause over
815968 // would park the outbox against a limit that no longer applies.
816969 model.flush_paused_until = 0;
817970 model.send_retries = 0;
971+ // D2: Reset poll retries too — the new relay's poll failures
972+ // should start fresh, not inherit the old relay's backoff.
973+ model.poll_retries = 0;
818974 let writes = if self.refuse_write(model) {
819975 Command::done()
820976 } else {
@@ -862,6 +1018,15 @@ impl App for Skrepka {
8621018 }
8631019 }
8641020 Event::AddContact { input, nickname } => {
1021+ // Clear a stale error from a previous failed add.
1022+ model.error = None;
1023+ // A nickname is stored in the contacts kv blob, which is
1024+ // rewritten in full on every change. An unbounded nickname
1025+ // bloats the blob and every subsequent write.
1026+ if nickname.len() > protocol::MAX_DISPLAY_NAME_LEN {
1027+ model.error = Some("nickname too long".into());
1028+ return render();
1029+ }
8651030 match phonemic::try_parse_pubkey(&input) {
8661031 Some(hex) if hex != model.my_pubkey => {
8671032 model
@@ -891,6 +1056,9 @@ impl App for Skrepka {
8911056 if let Some(c) = model.contacts.get_mut(&peer) {
8921057 c.blocked = blocked;
8931058 }
1059+ // Clear a stale error (e.g. "cannot send to a blocked contact")
1060+ // when the user unblocks — the block that caused it is now lifted.
1061+ model.error = None;
8941062 if blocked {
8951063 // Gating *ingest* on the block only stops the acks we have not
8961064 // queued yet. Anything already sitting in the outbox for this
@@ -909,6 +1077,7 @@ impl App for Skrepka {
9091077 Event::DeleteContact { peer } => {
9101078 model.contacts.remove(&peer);
9111079 model.messages.remove(&peer);
1080+ model.seen_ids.remove(&peer);
9121081 model.outbox.retain(|item| item.recipient != peer);
9131082 // The send in flight may be carrying the very items we just
9141083 // dropped; if so its result must pop nothing.
@@ -925,6 +1094,8 @@ impl App for Skrepka {
9251094 // message the user just asked us to forget. Delete the key.
9261095 KeyValue::delete(k_messages(&peer)).then_send(Event::Saved),
9271096 self.persist_outbox(model),
1097+ // Q18: persist the seen_ids map without the removed peer.
1098+ self.persist_seen_ids(model),
9281099 ])
9291100 .and(render())
9301101 }
@@ -933,6 +1104,32 @@ impl App for Skrepka {
9331104 bio,
9341105 photo,
9351106 } => {
1107+ // Clear a stale error from a previous failed save.
1108+ model.error = None;
1109+ // The recipient's `parse_payload` enforces these caps and
1110+ // silently drops oversized profiles. Without these checks the
1111+ // sender's local profile would be saved, broadcast to every
1112+ // contact, and silently dropped by every recipient — a silent
1113+ // divergence between what the sender sees and what contacts see.
1114+ if display_name.chars().count() > protocol::MAX_DISPLAY_NAME_LEN {
1115+ model.error = Some("display name too long".into());
1116+ return render();
1117+ }
1118+ if bio.chars().count() > protocol::MAX_BIO_LEN {
1119+ model.error = Some("bio too long".into());
1120+ return render();
1121+ }
1122+ if photo.as_ref().is_some_and(|p| p.len() > protocol::MAX_PHOTO_LEN) {
1123+ model.error = Some("photo too large".into());
1124+ return render();
1125+ }
1126+ // S6: Validate the photo is valid base64 before broadcasting —
1127+ // the recipient's parse_payload rejects invalid base64, so an
1128+ // invalid photo would be silently dropped by every recipient.
1129+ if photo.as_ref().is_some_and(|p| BASE64_STANDARD.decode(p).is_err()) {
1130+ model.error = Some("photo is not valid base64".into());
1131+ return render();
1132+ }
9361133 model.profile = OwnProfile {
9371134 display_name,
9381135 bio,
@@ -941,8 +1138,8 @@ impl App for Skrepka {
9411138 model.page = Page::Conversations;
9421139 let ts = now_ms();
9431140 let payload = Payload::Profile {
944− display_name: model.profile.display_name.clone(),
945− bio: model.profile.bio.clone(),
1141+ display_name: Some(model.profile.display_name.clone()),
1142+ bio: Some(model.profile.bio.clone()),
9461143 photo: model.profile.photo.clone(),
9471144 };
9481145 // Broadcast the new profile to every contact we haven't blocked.
@@ -1021,6 +1218,11 @@ impl App for Skrepka {
10211218 let gen = model.next_auth_gen();
10221219 req.build()
10231220 .then_send(move |r| Event::ChallengeResult(gen, r))
1221+ .and(
1222+ Time::notify_after(Duration::from_millis(AUTH_WATCHDOG_MS))
1223+ .0
1224+ .then_send(move |_| Event::AuthWatchdog(gen)),
1225+ )
10241226 .and(render())
10251227 }
10261228 Err(_) => {
@@ -1199,22 +1401,90 @@ impl App for Skrepka {
11991401 return render();
12001402 }
12011403 let bytes = resp.take_body().unwrap_or_default();
1202− let parsed: PollResp = serde_json::from_slice(&bytes).unwrap_or_default();
1404+ // Bound the response before parsing: a hostile relay can return
1405+ // a multi-gigabyte body that OOMs the app before `ingest_poll`'s
1406+ // per-blob budgets are even consulted.
1407+ if bytes.len() > MAX_POLL_RESPONSE_BYTES {
1408+ model.error = Some("relay sent an oversized poll response".into());
1409+ return self.backoff_poll(model);
1410+ }
1411+ // Q25: A malformed or unparseable poll response must not be
1412+ // treated as an empty page — `unwrap_or_default()` would adopt
1413+ // cursor 0, forcing a re-fetch from the beginning and wasting
1414+ // bandwidth on already-seen (deduped) messages. Back off instead.
1415+ let parsed: PollResp = match serde_json::from_slice(&bytes) {
1416+ Ok(p) => p,
1417+ Err(_) => {
1418+ model.error = Some("relay sent a malformed poll response".into());
1419+ return self.backoff_poll(model);
1420+ }
1421+ };
1422+ let cursor_before = model.cursor;
12031423 let cmd = self.ingest_poll(model, parsed);
12041424 // The re-poll hangs off the *cursor write*, not off this event.
12051425 // Polling again is what acks the batch we just took — the server
12061426 // deletes everything up to `cursor` — so it must not happen until
12071427 // the batch is durable on our side. `SavedCursor` re-polls.
1208− cmd.and(
1428+ //
1429+ // Arm a watchdog alongside the cursor write: if the shell ever
1430+ // drops the KV effect (suspension, bug), `SavedCursor` never fires,
1431+ // `polling` stays `false`, and the loop is dead until the next
1432+ // foreground `Connect`. The watchdog fires `Poll` if the loop is
1433+ // still idle — same pattern as the poll/send HTTP watchdogs.
1434+ //
1435+ // Skip the cursor write when the cursor did not advance (an empty
1436+ // or fully-deduped page): there is nothing new to persist, so
1437+ // re-poll directly. The watchdog is still armed as a fallback.
1438+ let cursor_advanced = model.cursor != cursor_before;
1439+ cmd.and(if cursor_advanced {
12091440 KeyValue::set(K_CURSOR, json_bytes(&model.cursor))
1210− .then_send(Event::SavedCursor),
1441+ .then_send(Event::SavedCursor)
1442+ } else {
1443+ Command::event(Event::Poll)
1444+ })
1445+ .and(
1446+ Time::notify_after(Duration::from_millis(POLL_WATCHDOG_MS))
1447+ .0
1448+ .then_send(|_| Event::Poll),
12111449 )
12121450 .and(render())
12131451 }
12141452
12151453 // ---------------- outbox ----------------
12161454 Event::StartFlush => self.flush_next(model),
1217− Event::SendResult(Ok(resp)) => {
1455+ Event::FlushWatchdog(gen) => {
1456+ // Only the flush this watchdog was armed for, and only while it
1457+ // is still the current one and still in flight. A watchdog whose
1458+ // send already completed (or was superseded) is a stale timer and
1459+ // must not touch a healthy successor — same pattern as
1460+ // `PollWatchdog`.
1461+ if gen != model.flush_gen || !model.flushing {
1462+ return Command::done();
1463+ }
1464+ // Retire it as we give up on it. The request itself is not
1465+ // cancelled, so if it eventually resolves the generation check
1466+ // in `SendResult` is what stops it from popping the replacement's
1467+ // items.
1468+ model.abandon_flush();
1469+ self.flush_next(model)
1470+ }
1471+ Event::AuthWatchdog(gen) => {
1472+ // Only the auth this watchdog was armed for, and only while it is
1473+ // still the current one and still in flight.
1474+ if gen != model.auth_gen || !model.authenticating {
1475+ return Command::done();
1476+ }
1477+ model.abandon_auth();
1478+ Command::event(Event::Authenticate)
1479+ }
1480+ Event::SendResult(gen, Ok(resp)) => {
1481+ // A result for a send we already abandoned (a watchdog gave up
1482+ // on it, or a server change invalidated it). Dropping it is safe
1483+ // and required: its items were re-queued, so popping by count
1484+ // here would eat whichever messages slid into their place.
1485+ if gen != model.flush_gen {
1486+ return Command::done();
1487+ }
12181488 // The in-flight send is over; `flush_next` refuses to run while
12191489 // this is set, so it must be cleared before *any* branch below.
12201490 model.flushing = false;
@@ -1304,16 +1574,24 @@ impl App for Skrepka {
13041574 // later, under load that has nothing to do with this one — starts
13051575 // at whatever height the last one climbed to.
13061576 model.send_retries = 0;
1307− for _ in 0..batch.map_or(0, |b| b.count) {
1308− model.outbox.pop_front();
1577+ // D3-app: Don't pop if kv is broken — the persist would be
1578+ // refused, and the items would be lost from memory but still
1579+ // on disk, causing duplicate sends on relaunch.
1580+ if !model.kv_load_failed {
1581+ for _ in 0..batch.map_or(0, |b| b.count) {
1582+ model.outbox.pop_front();
1583+ }
13091584 }
13101585 return self.persist_outbox(model).and(self.flush_next(model));
13111586 }
13121587 // A transient failure: a 5xx, a rate limit, a relay restarting.
13131588 self.retry_batch(model, batch)
13141589 }
1315− Event::SendResult(Err(_)) => {
1590+ Event::SendResult(gen, Err(_)) => {
13161591 // A transport error is the same transient failure as a 5xx.
1592+ if gen != model.flush_gen {
1593+ return Command::done();
1594+ }
13171595 model.flushing = false;
13181596 let batch = model.in_flight.take();
13191597 self.retry_batch(model, batch)
@@ -1488,6 +1766,16 @@ impl Skrepka {
14881766 KeyValue::set(k_messages(peer), json_bytes(&list)).then_send(Event::Saved)
14891767 }
14901768
1769+ /// Q18: Persist the per-peer dedup ID set. Written alongside messages so
1770+ /// the set survives restarts and covers IDs that `trim_history` has aged
1771+ /// out of the conversation.
1772+ fn persist_seen_ids(&self, model: &mut Model) -> Command<Effect, Event> {
1773+ if self.refuse_write(model) {
1774+ return Command::done();
1775+ }
1776+ KeyValue::set(K_SEEN_IDS, json_bytes(&model.seen_ids)).then_send(Event::Saved)
1777+ }
1778+
14911779 fn schedule_reconnect(&self, model: &mut Model) -> Command<Effect, Event> {
14921780 let delay = backoff_ms(model.poll_retries);
14931781 model.poll_retries = model.poll_retries.saturating_add(1);
@@ -1518,6 +1806,9 @@ impl Skrepka {
15181806 let Some(peer) = model.active_peer.clone() else {
15191807 return render();
15201808 };
1809+ // Clear a stale error from a previous failed send, so the error display
1810+ // does not persist after the user fixes the issue and sends again.
1811+ model.error = None;
15211812 // Blocking cuts the peer off in both directions (PROTOCOL.md §4). The
15221813 // ingest side drops their messages; the send side must refuse too, or
15231814 // a blocked peer still receives our texts, acks, and the liveness
@@ -1530,6 +1821,15 @@ impl Skrepka {
15301821 if body.is_empty() {
15311822 return render();
15321823 }
1824+ // The recipient's `parse_payload` enforces `MAX_BODY_LEN` and silently
1825+ // drops oversized text. Without this check the message would be stored
1826+ // locally, sent to the relay, and marked "sent" — but never arrive, never
1827+ // get acked, and the sender would see it stuck as "sent" forever with no
1828+ // error.
1829+ if body.len() > protocol::MAX_BODY_LEN {
1830+ model.error = Some("message too long".into());
1831+ return render();
1832+ }
15331833
15341834 // Check the contact cap *before* mutating anything. The old order
15351835 // cleared `compose` and inserted the message into `model.messages`
@@ -1574,14 +1874,21 @@ impl Skrepka {
15741874 .insert(peer.clone(), Contact::new(peer.clone(), String::new(), ts));
15751875 }
15761876
1577− // On first contact, share our profile *before* the text so the
1578− // recipient's first impression is a name and avatar, not a truncated
1579− // @p that gets replaced a moment later. Both are queued behind any
1580− // in-flight items, so the ordering is preserved through the outbox.
1877+ let text = Payload::Text { id, body };
1878+ model.outbox.push_back(OutboxItem::new(
1879+ peer.clone(),
1880+ Arc::new(protocol::serialize_payload(&text, ts)),
1881+ ));
1882+
1883+ // On first contact, share our profile *after* the text. The text
1884+ // auto-creates the contact on the recipient's side; the profile that
1885+ // follows in the same batch then applies to the now-existing contact.
1886+ // (Previously profile was queued before the text, but the recipient's
1887+ // ingest drops profiles from non-contacts, so the profile was lost.)
15811888 if was_new {
15821889 let profile = Payload::Profile {
1583− display_name: model.profile.display_name.clone(),
1584− bio: model.profile.bio.clone(),
1890+ display_name: Some(model.profile.display_name.clone()),
1891+ bio: Some(model.profile.bio.clone()),
15851892 photo: model.profile.photo.clone(),
15861893 };
15871894 model.outbox.push_back(OutboxItem::profile(
@@ -1590,12 +1897,6 @@ impl Skrepka {
15901897 ));
15911898 }
15921899
1593− let text = Payload::Text { id, body };
1594− model.outbox.push_back(OutboxItem::new(
1595− peer.clone(),
1596− Arc::new(protocol::serialize_payload(&text, ts)),
1597− ));
1598−
15991900 Command::all([
16001901 self.persist_messages(model, &peer),
16011902 self.persist_contacts(model),
@@ -1723,6 +2024,7 @@ impl Skrepka {
17232024 };
17242025
17252026 model.flushing = true;
2027+ let gen = model.next_flush_gen();
17262028 model.in_flight = Some(InFlight {
17272029 recipient,
17282030 count: messages.len(),
@@ -1733,7 +2035,12 @@ impl Skrepka {
17332035 .header("authorization", format!("Bearer {token}"))
17342036 .body_json(&SendBatch { messages })
17352037 {
1736− Ok(req) => persist.and(req.build().then_send(Event::SendResult)),
2038+ Ok(req) => persist.and(req.build().then_send(move |r| Event::SendResult(gen, r)))
2039+ .and(
2040+ Time::notify_after(Duration::from_millis(FLUSH_WATCHDOG_MS))
2041+ .0
2042+ .then_send(move |_| Event::FlushWatchdog(gen)),
2043+ ),
17372044 Err(_) => {
17382045 // Not a dead loop: the next StartFlush (from a poll page or a
17392046 // fresh send) retries the head of the outbox.
@@ -1803,8 +2110,20 @@ impl Skrepka {
18032110 return Command::done();
18042111 }
18052112 let prev_cursor = model.cursor;
1806− if page.cursor > model.cursor {
1807− model.cursor = page.cursor;
2113+ // Clamp the relay-supplied cursor: honest cursors are wall-clock-seeded
2114+ // sequence numbers, so anything far in the future is a hostile or buggy
2115+ // relay. Adopting `i64::MAX` would ack (delete) all queued mail on the
2116+ // next poll and persist that poison across relaunches.
2117+ let clamped_cursor = page.cursor.min(now_ms().saturating_add(60_000));
2118+ // Q16: Only advance the cursor when the page actually carried events.
2119+ // An empty page with a cursor higher than ours is suspicious — a
2120+ // compromised relay could poison the cursor to ack (delete) messages
2121+ // the client hasn't seen. When there are no events, there is nothing
2122+ // to acknowledge, so the cursor should not move. (An honest relay's
2123+ // empty-page cursor equals what the client sent, so this is a no-op
2124+ // in the normal case.)
2125+ if !page.events.is_empty() && clamped_cursor > model.cursor {
2126+ model.cursor = clamped_cursor;
18082127 }
18092128 let Some(id) = model.identity() else {
18102129 return Command::done();
@@ -1813,12 +2132,24 @@ impl Skrepka {
18132132 let now = now_ms();
18142133 let mut touched_convos: HashSet<String> = HashSet::new();
18152134 let mut contacts_dirty = false;
2135+ let mut outbox_dirty = false;
18162136 // sender hex -> message ids needing acks
18172137 let mut ack_targets: BTreeMap<String, Vec<String>> = BTreeMap::new();
1818− // Lazily-built per-sender sets of known message IDs for O(1) text dedup.
1819− let mut seen_ids: HashMap<String, HashSet<String>> = HashMap::new();
2138+ // Q17: Per-sender deferred ack accumulation. Collected during the event
2139+ // loop, applied after it — so relay page ordering cannot cause an older
2140+ // ack to be dropped as stale by a newer one processed first.
2141+ let mut pending_acks: HashMap<String, PendingAck> = HashMap::new();
18202142
18212143 let mut budget = MAX_POLL_TOTAL_BYTES;
2144+ let mut events_processed = 0usize;
2145+ // N9: Track whether any event was successfully decrypted and stored.
2146+ // When the budget/count is exhausted mid-page, we ack the page (keep
2147+ // the advanced cursor) only if at least one event was processed —
2148+ // the unprocessed tail is from a hostile/misconfigured relay and
2149+ // losing them is better than livelocking. If zero events were
2150+ // processed (a completely hostile page), restore the cursor so we
2151+ // don't ack mail we never saw.
2152+ let mut events_stored = 0usize;
18222153 for ev in page.events {
18232154 // Reject an oversized blob on its hex length, before it costs us a
18242155 // decode. No conforming sender produces one (the relay caps blobs at
@@ -1826,19 +2157,37 @@ impl Skrepka {
18262157 if ev.encrypted_blob.len() > MAX_BLOB_LEN * 2 {
18272158 continue;
18282159 }
2160+ // Event count cap: the byte budget alone permits ~388K minimum-size
2161+ // events, each costing a full decrypt attempt. A hostile relay can
2162+ // freeze the app for tens of seconds per poll with a page of small
2163+ // junk blobs. Stop processing past this count.
2164+ if events_processed >= MAX_POLL_EVENTS {
2165+ // N9: If we stored at least one event, keep the advanced cursor
2166+ // to ack the page and avoid a livelock. The unprocessed events
2167+ // are hostile junk or beyond our capacity — we can't process
2168+ // them anyway, and re-fetching them forever would wedge the
2169+ // client.
2170+ if events_stored == 0 {
2171+ model.cursor = prev_cursor;
2172+ }
2173+ model.error = Some("relay sent too many poll events".into());
2174+ break;
2175+ }
2176+ events_processed += 1;
18292177 match budget.checked_sub(ev.encrypted_blob.len() / 2) {
18302178 Some(left) => budget = left,
18312179 None => {
1832− // A page this large is a misbehaving relay, not real mail.
1833− // Stop here; the rest of the page is discarded exactly as an
1834− // undecryptable blob would be.
1835− //
1836− // Restore the cursor: it was advanced to `page.cursor` above,
1837− // but we have not processed the remaining events. If we keep
1838− // the advanced cursor, the next poll acks the whole page to
1839− // the relay — including every event we skipped — and the
1840− // relay deletes them. Those messages are permanently lost.
1841− model.cursor = prev_cursor;
2180+ // N9: Budget exhausted. If we stored at least one event,
2181+ // keep the advanced cursor to ack the page — re-fetching
2182+ // an oversized page forever is a livelock. The unprocessed
2183+ // events are beyond our capacity and would be re-served
2184+ // on every poll, wedging the client. Acking them breaks
2185+ // the loop. If zero events were stored (completely hostile
2186+ // page), restore the cursor to avoid acking mail we
2187+ // never saw.
2188+ if events_stored == 0 {
2189+ model.cursor = prev_cursor;
2190+ }
18422191 model.error = Some("relay sent an oversized poll page".into());
18432192 break;
18442193 }
@@ -1852,6 +2201,10 @@ impl Skrepka {
18522201 let Some(parsed) = protocol::parse_payload(&dec.plaintext) else {
18532202 continue;
18542203 };
2204+ // D1: Any event that was successfully decrypted AND parsed counts
2205+ // as "stored" for cursor advancement — the client has processed it
2206+ // even if it was a duplicate/stale/stranger that was dropped.
2207+ events_stored += 1;
18552208 let sender = dec.sender_hex;
18562209
18572210 // Mail from ourselves. A relay cannot forge this — the sender identity
@@ -1903,12 +2256,35 @@ impl Skrepka {
19032256 Contact::new(sender.clone(), String::new(), ts),
19042257 );
19052258 contacts_dirty = true;
2259+ // D6-app: Send our profile to the new contact so they
2260+ // see our display name and avatar, not just our @p.
2261+ let profile = Payload::Profile {
2262+ display_name: Some(model.profile.display_name.clone()),
2263+ bio: Some(model.profile.bio.clone()),
2264+ photo: model.profile.photo.clone(),
2265+ };
2266+ model.outbox.push_back(OutboxItem::profile(
2267+ sender.clone(),
2268+ Arc::new(protocol::serialize_payload(&profile, ts)),
2269+ ));
2270+ outbox_dirty = true;
19062271 }
19072272 let convo = model.messages.entry(sender.clone()).or_default();
1908− // Lazily build the per-sender ID set on first touch.
1909− let id_set = seen_ids.entry(sender.clone()).or_insert_with(|| {
1910− convo.iter().map(|m| m.id.clone()).collect()
1911− });
2273+ // Q18: Use the persistent per-peer dedup set from
2274+ // `model.seen_ids` (loaded from kv at startup) instead of
2275+ // deriving it from the trimmed conversation. Without this,
2276+ // a replayed blob whose `id` was aged out by `trim_history`
2277+ // (past the 1000-message cap) passes the dedup check and
2278+ // re-enters the conversation.
2279+ let id_set = model
2280+ .seen_ids
2281+ .entry(sender.clone())
2282+ .or_default();
2283+ // Also seed from the conversation in case the kv load was
2284+ // missing or partial — covers the upgrade path.
2285+ if id_set.is_empty() {
2286+ id_set.extend(convo.iter().map(|m| m.id.clone()));
2287+ }
19122288 // dedup by id — O(1) via HashSet
19132289 if !id_set.contains(&msg_id) {
19142290 // The conversation is *kept* sorted rather than re-sorted:
@@ -1926,6 +2302,29 @@ impl Skrepka {
19262302 },
19272303 );
19282304 id_set.insert(msg_id.clone());
2305+ // N7: Cap the dedup set so it doesn't grow without
2306+ // bound. The set is rewritten to kv on every message
2307+ // receipt, so an unbounded set makes every receipt
2308+ // slower. Keep the most recent IDs by draining excess
2309+ // from the oldest. `HashSet` has no ordering, so when
2310+ // the cap is exceeded we rebuild from the conversation
2311+ // (which is trimmed to MAX_MESSAGES_PER_PEER and
2312+ // sorted by ts) — the oldest IDs in the set that are
2313+ // no longer in the conversation are the ones to drop.
2314+ if id_set.len() > MAX_SEEN_IDS_PER_PEER {
2315+ let convo_ids: HashSet<&String> =
2316+ convo.iter().map(|m| &m.id).collect();
2317+ let excess = id_set.len() - MAX_SEEN_IDS_PER_PEER;
2318+ let to_remove: Vec<String> = id_set
2319+ .iter()
2320+ .filter(|id| !convo_ids.contains(*id))
2321+ .take(excess)
2322+ .cloned()
2323+ .collect();
2324+ for id in to_remove {
2325+ id_set.remove(&id);
2326+ }
2327+ }
19292328 trim_history(convo);
19302329 touched_convos.insert(sender.clone());
19312330 // Only ack a message we actually stored — a duplicate id
@@ -1934,39 +2333,26 @@ impl Skrepka {
19342333 }
19352334 }
19362335 Payload::DeliveryAck { ack_ids } => {
1937− // Staleness check: drop a replayed ack whose ts predates the
1938− // last one we processed for this contact.
1939− let stale = model
1940− .contacts
2336+ // Q17: Defer the staleness check until after the event loop.
2337+ // Within a single page, a relay can reorder events so that a
2338+ // newer ack (higher ts) is processed before an older one. If
2339+ // we applied the watermark immediately, the older ack would be
2340+ // dropped as stale — even if it acks different message IDs.
2341+ // Instead, accumulate all ack_ids per sender and the max ts,
2342+ // then apply them in a second pass below the loop.
2343+ let prev_ts = pending_acks
19412344 .get(&sender)
1942− .is_some_and(|c| ts < c.last_ack_ts);
1943− if stale {
1944− continue;
1945− }
1946− // The watermark is clamped to *now*, not merely into the skew
1947− // window: `ts` is the peer's to choose, and `last_ack_ts` is
1948− // what every later ack is measured against. A single ack at
1949− // `ts = i64::MAX` would otherwise park the watermark a minute
1950− // into the future, and every honest ack that followed — the
1951− // ones that actually mark our messages delivered — would read
1952− // as stale and be dropped.
1953− if let Some(c) = model.contacts.get_mut(&sender) {
1954− c.last_ack_ts = ts.min(now);
1955− contacts_dirty = true;
1956− }
1957− let ack_set: HashSet<String> = ack_ids.iter().cloned().collect();
1958− if let Some(convo) = model.messages.get_mut(&sender) {
1959− let mut changed = false;
1960− for m in convo.iter_mut() {
1961− if m.outgoing && ack_set.contains(&m.id) && !m.delivered {
1962− m.delivered = true;
1963− changed = true;
1964− }
1965− }
1966− if changed {
1967− touched_convos.insert(sender);
1968− }
2345+ .map(|p| p.ts)
2346+ .unwrap_or(0);
2347+ if ts > prev_ts {
2348+ let entry = pending_acks.entry(sender.clone()).or_default();
2349+ entry.ts = ts;
19692350 }
2351+ pending_acks
2352+ .entry(sender.clone())
2353+ .or_default()
2354+ .ids
2355+ .extend(ack_ids);
19702356 }
19712357 Payload::Profile {
19722358 display_name,
@@ -1993,9 +2379,17 @@ impl Skrepka {
19932379 // A replayed profile at the same `ts` overwrites with identical
19942380 // content, which is a no-op in practice.
19952381 if ts >= c.last_profile_ts {
1996− c.display_name = display_name;
1997− c.bio = bio;
1998− c.photo = photo;
2382+ // D13: Only update fields that are present (Some).
2383+ // An absent field (None) means "don't change".
2384+ if let Some(name) = display_name {
2385+ c.display_name = name;
2386+ }
2387+ if let Some(b) = bio {
2388+ c.bio = b;
2389+ }
2390+ if let Some(p) = photo {
2391+ c.photo = Some(p);
2392+ }
19992393 c.last_profile_ts = ts.min(now);
20002394 contacts_dirty = true;
20012395 }
@@ -2003,6 +2397,42 @@ impl Skrepka {
20032397 }
20042398 }
20052399
2400+ // Q17: Apply deferred delivery acks. Collected during the event loop
2401+ // above, now applied with a single staleness check per sender using the
2402+ // highest `ts` seen — so relay page ordering cannot cause an older ack
2403+ // to be dropped as stale by a newer one processed first.
2404+ for (sender, pending) in &pending_acks {
2405+ let stale = model
2406+ .contacts
2407+ .get(sender)
2408+ .is_some_and(|c| pending.ts < c.last_ack_ts);
2409+ if stale {
2410+ continue;
2411+ }
2412+ // The watermark is clamped to *now*, not merely into the skew window:
2413+ // `ts` is the peer's to choose, and `last_ack_ts` is what every later
2414+ // ack is measured against. A single ack at `ts = i64::MAX` would
2415+ // otherwise park the watermark a minute into the future, and every
2416+ // honest ack that followed — the ones that actually mark our messages
2417+ // delivered — would read as stale and be dropped.
2418+ if let Some(c) = model.contacts.get_mut(sender) {
2419+ c.last_ack_ts = pending.ts.min(now);
2420+ contacts_dirty = true;
2421+ }
2422+ if let Some(convo) = model.messages.get_mut(sender) {
2423+ let mut changed = false;
2424+ for m in convo.iter_mut() {
2425+ if m.outgoing && pending.ids.contains(&m.id) && !m.delivered {
2426+ m.delivered = true;
2427+ changed = true;
2428+ }
2429+ }
2430+ if changed {
2431+ touched_convos.insert(sender.clone());
2432+ }
2433+ }
2434+ }
2435+
20062436 // Queue delivery acks, in batches the recipient will actually accept.
20072437 //
20082438 // `parse_payload` drops a `delivery.ack` carrying more than `MAX_ACK_IDS`
@@ -2025,6 +2455,7 @@ impl Skrepka {
20252455 Arc::new(protocol::serialize_payload(&payload, ack_ts)),
20262456 ));
20272457 }
2458+ outbox_dirty = true;
20282459 }
20292460
20302461 // Persist everything touched.
@@ -2032,10 +2463,16 @@ impl Skrepka {
20322463 for peer in &touched_convos {
20332464 cmds.push(self.persist_messages(model, peer));
20342465 }
2466+ // Q18: Persist the dedup ID set whenever we touched conversations.
2467+ if !touched_convos.is_empty() {
2468+ cmds.push(self.persist_seen_ids(model));
2469+ }
20352470 if contacts_dirty {
20362471 cmds.push(self.persist_contacts(model));
20372472 }
2038− cmds.push(self.persist_outbox(model));
2473+ if outbox_dirty {
2474+ cmds.push(self.persist_outbox(model));
2475+ }
20392476 cmds.push(Command::event(Event::StartFlush));
20402477 Command::all(cmds)
20412478 }
@@ -2108,7 +2545,7 @@ mod tests {
21082545 use crate::model::{MAX_OUTBOX_RETRIES, OUTBOX_TTL_MS};
21092546
21102547 fn with_identity() -> Model {
2111− let id = Identity::from_seed(&[3u8; 32]);
2548+ let id = Identity::from_seed(&[3u8; 32]).unwrap();
21122549 Model {
21132550 secret_key: Some(Zeroizing::new(id.secret_key.to_vec())),
21142551 my_pubkey: id.public_key_hex(),
@@ -2117,7 +2554,7 @@ mod tests {
21172554 }
21182555
21192556 fn peer_hex(seed: u8) -> String {
2120− Identity::from_seed(&[seed; 32]).public_key_hex()
2557+ Identity::from_seed(&[seed; 32]).unwrap().public_key_hex()
21212558 }
21222559
21232560 /// `Command::event` *queues* an event; it does not run it. The runtime drains
@@ -2140,7 +2577,7 @@ mod tests {
21402577
21412578 /// A poll page carrying one payload, encrypted from `sender_seed` to `me`.
21422579 fn page_from(sender_seed: u8, me: &Model, payload: &Payload, ts: i64) -> PollResp {
2143− let sender = Identity::from_seed(&[sender_seed; 32]);
2580+ let sender = Identity::from_seed(&[sender_seed; 32]).unwrap();
21442581 let recipient = Identity::from_secret_bytes(me.secret_key.as_ref().unwrap()).unwrap();
21452582 let json = protocol::serialize_payload(payload, ts);
21462583 let blob = crate::crypto::encrypt(
@@ -2229,7 +2666,8 @@ mod tests {
22292666 /// One full send round-trip: flush the outbox, then answer with `status`.
22302667 fn send_and_answer(app: &Skrepka, m: &mut Model, status: u16) {
22312668 let _ = app.update(Event::StartFlush, m);
2232− let _ = app.update(Event::SendResult(Ok(ok_response(status))), m);
2669+ let gen = m.flush_gen;
2670+ let _ = app.update(Event::SendResult(gen, Ok(ok_response(status))), m);
22332671 }
22342672
22352673 // -----------------------------------------------------------------------
@@ -2258,7 +2696,7 @@ mod tests {
22582696 // The last one exhausts the budget: the head goes, and what was stuck
22592697 // behind it goes out.
22602698 let _ = app.update(Event::StartFlush, &mut m);
2261− let mut cmd = app.update(Event::SendResult(Ok(ok_response(503))), &mut m);
2699+ let mut cmd = app.update(Event::SendResult(m.flush_gen, Ok(ok_response(503))), &mut m);
22622700
22632701 assert_eq!(m.outbox.len(), 1, "the undeliverable item is dropped");
22642702 assert_eq!(m.outbox[0].recipient, behind, "and the queue moves on");
@@ -2377,7 +2815,7 @@ mod tests {
23772815
23782816 let _ = app.update(Event::StartFlush, &mut m);
23792817 let mut cmd = app.update(
2380− Event::SendResult(Ok(response_with_header(429, "retry-after", "2"))),
2818+ Event::SendResult(m.flush_gen, Ok(response_with_header(429, "retry-after", "2"))),
23812819 &mut m,
23822820 );
23832821
@@ -2408,7 +2846,7 @@ mod tests {
24082846 m.poll_retries = 0;
24092847
24102848 let _ = app.update(Event::StartFlush, &mut m);
2411− let mut cmd = app.update(Event::SendResult(Ok(ok_response(429))), &mut m);
2849+ let mut cmd = app.update(Event::SendResult(m.flush_gen, Ok(ok_response(429))), &mut m);
24122850
24132851 assert_eq!(m.outbox[0].retries, 0, "still not the message's fault");
24142852 assert_eq!(
@@ -2421,7 +2859,7 @@ mod tests {
24212859 // A second 429 climbs, rather than restarting at 3 s.
24222860 m.flush_paused_until = 0;
24232861 let _ = app.update(Event::StartFlush, &mut m);
2424− let mut cmd = app.update(Event::SendResult(Ok(ok_response(429))), &mut m);
2862+ let mut cmd = app.update(Event::SendResult(m.flush_gen, Ok(ok_response(429))), &mut m);
24252863 assert_eq!(scheduled_delays(&mut cmd), vec![backoff_ms(1)], "it climbs");
24262864 assert_eq!(m.send_retries, 2);
24272865 }
@@ -2439,7 +2877,7 @@ mod tests {
24392877
24402878 let _ = app.update(Event::StartFlush, &mut m);
24412879 let _ = app.update(
2442− Event::SendResult(Ok(response_with_header(429, "retry-after", "2"))),
2880+ Event::SendResult(m.flush_gen, Ok(response_with_header(429, "retry-after", "2"))),
24432881 &mut m,
24442882 );
24452883 assert!(
@@ -2495,7 +2933,7 @@ mod tests {
24952933
24962934 let _ = app.update(Event::StartFlush, &mut m);
24972935 let mut cmd = app.update(
2498− Event::SendResult(Ok(response_with_header(429, "retry-after", "999999999"))),
2936+ Event::SendResult(m.flush_gen, Ok(response_with_header(429, "retry-after", "999999999"))),
24992937 &mut m,
25002938 );
25012939
@@ -2530,7 +2968,7 @@ mod tests {
25302968 assert!(batches[0].iter().all(|e| e.to == peer));
25312969
25322970 // ...and its success clears the whole batch, not just the head.
2533− let _ = app.update(Event::SendResult(Ok(ok_response(200))), &mut m);
2971+ let _ = app.update(Event::SendResult(m.flush_gen, Ok(ok_response(200))), &mut m);
25342972 assert!(m.outbox.is_empty());
25352973 }
25362974
@@ -2555,7 +2993,7 @@ mod tests {
25552993 assert_eq!(batches[0].len(), 2, "only the run at the head");
25562994 assert!(batches[0].iter().all(|e| e.to == a));
25572995
2558− let mut cmd = app.update(Event::SendResult(Ok(ok_response(200))), &mut m);
2996+ let mut cmd = app.update(Event::SendResult(m.flush_gen, Ok(ok_response(200))), &mut m);
25592997 assert_eq!(m.outbox.len(), 2);
25602998 let batches = sent_batches(&mut cmd);
25612999 assert_eq!(batches[0].len(), 1, "then b's single message");
@@ -2578,7 +3016,7 @@ mod tests {
25783016 let batches = sent_batches(&mut cmd);
25793017 assert_eq!(batches[0].len(), MAX_SEND_BATCH);
25803018
2581− let _ = app.update(Event::SendResult(Ok(ok_response(200))), &mut m);
3019+ let _ = app.update(Event::SendResult(m.flush_gen, Ok(ok_response(200))), &mut m);
25823020 assert_eq!(m.outbox.len(), 20, "the rest stays queued for the next batch");
25833021 }
25843022
@@ -2640,7 +3078,7 @@ mod tests {
26403078 let sent_bytes: usize = first[0].iter().map(|e| e.encrypted_blob.len()).sum();
26413079
26423080 // ...and the relay says that was too big.
2643− let mut cmd = app.update(Event::SendResult(Ok(ok_response(413))), &mut m);
3081+ let mut cmd = app.update(Event::SendResult(m.flush_gen, Ok(ok_response(413))), &mut m);
26443082
26453083 assert_eq!(
26463084 m.send_batch_budget,
@@ -2661,7 +3099,7 @@ mod tests {
26613099 assert!(shrunk > 0, "but not an empty one");
26623100
26633101 // The relay takes the smaller batch, and exactly it is popped.
2664− let _ = app.update(Event::SendResult(Ok(ok_response(200))), &mut m);
3102+ let _ = app.update(Event::SendResult(m.flush_gen, Ok(ok_response(200))), &mut m);
26653103 assert_eq!(m.outbox.len(), 5 - shrunk, "exactly the batch that landed is gone");
26663104 assert_eq!(
26673105 m.send_batch_budget,
@@ -2766,7 +3204,7 @@ mod tests {
27663204 let parsed = protocol::parse_payload(item.envelope_json.as_bytes()).unwrap();
27673205 match parsed.payload {
27683206 Payload::Profile { display_name, .. } => {
2769− assert_eq!(display_name, "Three", "and it is the latest one");
3207+ assert_eq!(display_name, Some("Three".to_string()), "and it is the latest one");
27703208 }
27713209 _ => panic!("expected a profile"),
27723210 }
@@ -2809,12 +3247,12 @@ mod tests {
28093247 );
28103248 assert_eq!(m.outbox.len(), 2, "the in-flight item is left alone");
28113249
2812− let _ = app.update(Event::SendResult(Ok(ok_response(200))), &mut m);
3250+ let _ = app.update(Event::SendResult(m.flush_gen, Ok(ok_response(200))), &mut m);
28133251 assert_eq!(m.outbox.len(), 1, "the sent one is popped");
28143252 let parsed = protocol::parse_payload(m.outbox[0].envelope_json.as_bytes()).unwrap();
28153253 match parsed.payload {
28163254 Payload::Profile { display_name, .. } => assert_eq!(
2817− display_name, "New",
3255+ display_name, Some("New".to_string()),
28183256 "and the newer profile survives to be sent"
28193257 ),
28203258 _ => panic!("expected a profile"),
@@ -2855,7 +3293,7 @@ mod tests {
28553293 assert_eq!(m.outbox.len(), 1);
28563294
28573295 // The send lands. It must pop nothing.
2858− let _ = app.update(Event::SendResult(Ok(ok_response(200))), &mut m);
3296+ let _ = app.update(Event::SendResult(m.flush_gen, Ok(ok_response(200))), &mut m);
28593297 assert_eq!(m.outbox.len(), 1, "the friend's message is untouched");
28603298 assert_eq!(m.outbox[0].recipient, friend);
28613299 }
@@ -2996,8 +3434,8 @@ mod tests {
29963434 .insert(peer.clone(), Contact::new(peer.clone(), String::new(), 0));
29973435
29983436 let profile = |name: &str| Payload::Profile {
2999− display_name: name.into(),
3000− bio: String::new(),
3437+ display_name: Some(name.to_string()),
3438+ bio: Some(String::new()),
30013439 photo: None,
30023440 };
30033441 let page = page_from(9, &m, &profile("Far"), i64::MAX);
@@ -3094,8 +3532,11 @@ mod tests {
30943532 let (m, peer) = ingest_one_text_from_an_unblocked_peer();
30953533 assert_eq!(m.messages[&peer].len(), 1);
30963534 assert_eq!(m.messages[&peer][0].body, "hi");
3097− assert_eq!(m.outbox.len(), 1, "one delivery.ack is queued");
3535+ // D6-app: A profile broadcast is queued alongside the delivery.ack
3536+ // so the new contact sees our display name and avatar.
3537+ assert_eq!(m.outbox.len(), 2, "one delivery.ack + one profile queued");
30983538 assert_eq!(m.outbox[0].recipient, peer);
3539+ assert_eq!(m.outbox[1].recipient, peer);
30993540 }
31003541
31013542 /// A peer sending `ts = i64::MAX` used to pin itself to the top of the
@@ -3276,6 +3717,28 @@ mod tests {
32763717 assert!(m.error.is_some());
32773718 }
32783719
3720+ /// A stale error from a previous failed add is cleared on the next attempt,
3721+ /// so the AddContact form does not show an old error after the user fixes
3722+ /// the input and tries again.
3723+ #[test]
3724+ fn add_contact_clears_a_stale_error_on_success() {
3725+ let app = Skrepka;
3726+ let mut m = with_identity();
3727+ m.error = Some("invalid public key".into());
3728+
3729+ let peer = peer_hex(9);
3730+ let _ = app.update(
3731+ Event::AddContact {
3732+ input: peer.clone(),
3733+ nickname: "Bob".into(),
3734+ },
3735+ &mut m,
3736+ );
3737+
3738+ assert!(m.error.is_none(), "a successful add clears the stale error");
3739+ assert!(m.contacts.contains_key(&peer));
3740+ }
3741+
32793742 #[test]
32803743 fn send_text_enqueues_message_profile_and_outbox() {
32813744 let app = Skrepka;
@@ -3367,6 +3830,30 @@ mod tests {
33673830 assert!(vm.contacts.iter().any(|c| c.pubkey == peer && c.blocked));
33683831 }
33693832
3833+ /// Unblocking clears a stale "cannot send to a blocked contact" error, so the
3834+ /// chat error display does not persist after the user lifts the block.
3835+ #[test]
3836+ fn unblocking_clears_the_stale_send_error() {
3837+ let app = Skrepka;
3838+ let mut m = with_identity();
3839+ let peer = peer_hex(9);
3840+ let mut contact = Contact::new(peer.clone(), "Bob".into(), 0);
3841+ contact.blocked = true;
3842+ m.contacts.insert(peer.clone(), contact);
3843+ m.error = Some("cannot send to a blocked contact".into());
3844+
3845+ let _ = app.update(
3846+ Event::SetBlocked {
3847+ peer: peer.clone(),
3848+ blocked: false,
3849+ },
3850+ &mut m,
3851+ );
3852+
3853+ assert!(!m.contacts[&peer].blocked);
3854+ assert!(m.error.is_none(), "unblocking clears the stale error");
3855+ }
3856+
33703857 /// Gating ingest stops the acks we have not queued yet — but the ack for the
33713858 /// message that made the user hit "block" is already in the outbox, and an ack
33723859 /// is exactly the "we are online and reading you" signal the block exists to
@@ -3486,7 +3973,7 @@ mod tests {
34863973 assert!(m.flushing);
34873974
34883975 let resp = crux_http::testing::ResponseBuilder::ok().body(Vec::new()).build();
3489− let _ = app.update(Event::SendResult(Ok(resp)), &mut m);
3976+ let _ = app.update(Event::SendResult(m.flush_gen, Ok(resp)), &mut m);
34903977
34913978 assert!(!m.flushing, "flushing must be cleared on the success path");
34923979 assert!(m.outbox.is_empty(), "the sent item is dropped from the outbox");
@@ -3508,7 +3995,7 @@ mod tests {
35083995
35093996 // Its success must both drop the item and let the *next* one go out.
35103997 let resp = crux_http::testing::ResponseBuilder::ok().body(Vec::new()).build();
3511− let _ = app.update(Event::SendResult(Ok(resp)), &mut m);
3998+ let _ = app.update(Event::SendResult(m.flush_gen, Ok(resp)), &mut m);
35123999 assert_eq!(m.outbox.len(), 1);
35134000 assert!(m.flushing, "the second item is now in flight");
35144001 }
@@ -3651,6 +4138,7 @@ mod tests {
36514138 Event::LoadedProfile(Ok(None)),
36524139 Event::LoadedContacts(Ok(None)),
36534140 Event::LoadedCursor(Ok(None)),
4141+ Event::LoadedSeenIds(Ok(None)),
36544142 ]
36554143 .into_iter()
36564144 .enumerate()
@@ -3664,7 +4152,7 @@ mod tests {
36644152 assert_eq!(connects(&mut cmd), 0, "and none of them connects");
36654153 }
36664154
3667− // The fifth — and only the fifth — emits Connect.
4155+ // The sixth — and only the sixth — emits Connect.
36684156 let mut cmd = app.update(Event::LoadedOutbox(Ok(None)), &mut m);
36694157 assert_eq!(m.loads_pending, 0);
36704158 assert_eq!(connects(&mut cmd), 1, "the last load connects, exactly once");
@@ -3754,7 +4242,7 @@ mod tests {
37544242 );
37554243
37564244 // The recovery is a relaunch, which re-runs the loads and clears the latch.
3757− let sk = Identity::from_seed(&[3u8; 32]).secret_key.to_vec();
4245+ let sk = Identity::from_seed(&[3u8; 32]).unwrap().secret_key.to_vec();
37584246 let _ = app.update(Event::IdentityLoaded(sk), &mut m);
37594247 assert!(
37604248 !m.kv_load_failed,
@@ -3861,8 +4349,8 @@ mod tests {
38614349 );
38624350
38634351 let profile = Payload::Profile {
3864− display_name: "Mallory".into(),
3865− bio: "spam".into(),
4352+ display_name: Some("Mallory".into()),
4353+ bio: Some("spam".into()),
38664354 photo: Some("aGk=".into()),
38674355 };
38684356 let page = page_from(9, &m, &profile, now_ms());
@@ -3943,8 +4431,8 @@ mod tests {
39434431 let peer = peer_hex(9);
39444432
39454433 let profile = Payload::Profile {
3946− display_name: "Mallory".into(),
3947− bio: String::new(),
4434+ display_name: Some("Mallory".into()),
4435+ bio: Some(String::new()),
39484436 photo: Some("aGk=".into()),
39494437 };
39504438 let page = page_from(9, &m, &profile, now_ms());
@@ -4156,7 +4644,7 @@ mod tests {
41564644
41574645 // One page holding more messages from one sender than fits in a single ack.
41584646 let count = protocol::MAX_ACK_IDS + 5;
4159− let sender = Identity::from_seed(&[9u8; 32]);
4647+ let sender = Identity::from_seed(&[9u8; 32]).unwrap();
41604648 let me = Identity::from_secret_bytes(m.secret_key.as_ref().unwrap()).unwrap();
41614649 let events = (0..count)
41624650 .map(|i| {
@@ -4313,6 +4801,10 @@ mod tests {
43134801 // send_text block check
43144802 // -----------------------------------------------------------------------
43154803
4804+ // -----------------------------------------------------------------------
4805+ // send_text body length check
4806+ // -----------------------------------------------------------------------
4807+
43164808 /// Blocking cuts a peer off in both directions. `send_text` must refuse to
43174809 /// enqueue a message to a blocked contact — otherwise the block only stops
43184810 /// their incoming mail while we keep sending them ours.
@@ -4341,4 +4833,195 @@ mod tests {
43414833 // and resend, and losing their draft would be surprising.
43424834 assert_eq!(m.compose, "hello?");
43434835 }
4836+
4837+ /// A text body over `MAX_BODY_LEN` would be stored locally and sent to the
4838+ /// relay, but the recipient's `parse_payload` silently drops oversized text.
4839+ /// The sender would see it stuck as "sent" forever with no error.
4840+ #[test]
4841+ fn send_text_refuses_an_oversized_body() {
4842+ let app = Skrepka;
4843+ let mut m = with_identity();
4844+ let peer = peer_hex(9);
4845+ m.active_peer = Some(peer.clone());
4846+ m.compose = "x".repeat(protocol::MAX_BODY_LEN + 1);
4847+
4848+ let _ = app.update(Event::SendText, &mut m);
4849+
4850+ assert!(m.outbox.is_empty(), "nothing is queued");
4851+ assert!(m.messages.get(&peer).is_none_or(Vec::is_empty), "nothing stored");
4852+ assert!(!m.compose.is_empty(), "compose is preserved");
4853+ assert!(m.error.as_ref().is_some_and(|e| e.contains("too long")));
4854+ }
4855+
4856+ /// A stale error from a previous failed send must be cleared on the next
4857+ /// attempt, so the error display does not persist after the user fixes the
4858+ /// issue and sends again successfully.
4859+ #[test]
4860+ fn send_text_clears_a_stale_error_on_success() {
4861+ let app = Skrepka;
4862+ let mut m = with_identity();
4863+ let peer = peer_hex(9);
4864+ m.active_peer = Some(peer.clone());
4865+ // Simulate a prior failure that left an error behind.
4866+ m.error = Some("message too long".into());
4867+
4868+ m.compose = "hello".into();
4869+ let _ = app.update(Event::SendText, &mut m);
4870+
4871+ assert!(m.error.is_none(), "a successful send clears the stale error");
4872+ assert!(!m.outbox.is_empty(), "the message was queued");
4873+ }
4874+
4875+ // -----------------------------------------------------------------------
4876+ // SaveProfile field length checks
4877+ // -----------------------------------------------------------------------
4878+
4879+ #[test]
4880+ fn save_profile_rejects_an_oversized_display_name() {
4881+ let app = Skrepka;
4882+ let mut m = with_identity();
4883+
4884+ let _ = app.update(Event::SaveProfile {
4885+ display_name: "x".repeat(protocol::MAX_DISPLAY_NAME_LEN + 1),
4886+ bio: String::new(),
4887+ photo: None,
4888+ }, &mut m);
4889+
4890+ assert!(m.outbox.is_empty(), "no broadcast queued");
4891+ assert!(m.profile.display_name.is_empty(), "local profile not updated");
4892+ assert!(m.error.as_ref().is_some_and(|e| e.contains("display name")));
4893+ }
4894+
4895+ #[test]
4896+ fn save_profile_rejects_an_oversized_bio() {
4897+ let app = Skrepka;
4898+ let mut m = with_identity();
4899+
4900+ let _ = app.update(Event::SaveProfile {
4901+ display_name: "Alice".into(),
4902+ bio: "x".repeat(protocol::MAX_BIO_LEN + 1),
4903+ photo: None,
4904+ }, &mut m);
4905+
4906+ assert!(m.outbox.is_empty(), "no broadcast queued");
4907+ assert!(m.profile.bio.is_empty(), "local profile not updated");
4908+ assert!(m.error.as_ref().is_some_and(|e| e.contains("bio")));
4909+ }
4910+
4911+ #[test]
4912+ fn save_profile_rejects_an_oversized_photo() {
4913+ let app = Skrepka;
4914+ let mut m = with_identity();
4915+
4916+ let _ = app.update(Event::SaveProfile {
4917+ display_name: "Alice".into(),
4918+ bio: String::new(),
4919+ photo: Some("x".repeat(protocol::MAX_PHOTO_LEN + 1)),
4920+ }, &mut m);
4921+
4922+ assert!(m.outbox.is_empty(), "no broadcast queued");
4923+ assert!(m.profile.photo.is_none(), "local profile not updated");
4924+ assert!(m.error.as_ref().is_some_and(|e| e.contains("photo")));
4925+ }
4926+
4927+ // -----------------------------------------------------------------------
4928+ // AddContact nickname length check
4929+ // -----------------------------------------------------------------------
4930+
4931+ #[test]
4932+ fn add_contact_rejects_an_oversized_nickname() {
4933+ let app = Skrepka;
4934+ let mut m = with_identity();
4935+ let peer = peer_hex(9);
4936+
4937+ let _ = app.update(Event::AddContact {
4938+ input: peer.clone(),
4939+ nickname: "x".repeat(protocol::MAX_DISPLAY_NAME_LEN + 1),
4940+ }, &mut m);
4941+
4942+ assert!(!m.contacts.contains_key(&peer), "no contact created");
4943+ assert!(m.error.as_ref().is_some_and(|e| e.contains("nickname")));
4944+ }
4945+
4946+ // -----------------------------------------------------------------------
4947+ // FlushWatchdog unsticks a wedged send loop
4948+ // -----------------------------------------------------------------------
4949+
4950+ /// `flushing` is only cleared by `SendResult`. If the shell drops the HTTP
4951+ /// effect, the outbox is permanently stuck. The watchdog releases the guard
4952+ /// so the flush retries.
4953+ #[test]
4954+ fn the_flush_watchdog_unwedges_a_send_that_never_resolved() {
4955+ let app = Skrepka;
4956+ let mut m = with_identity();
4957+ m.token = Some("t".into());
4958+ let peer = peer_hex(9);
4959+ m.outbox.push_back(queued(&peer, "stuck"));
4960+
4961+ let _ = app.update(Event::StartFlush, &mut m);
4962+ assert!(m.flushing, "the send is in flight");
4963+ assert!(m.in_flight.is_some());
4964+
4965+ // The send never comes back. The watchdog fires, clears the stuck guard,
4966+ // and retries the flush — the message goes out again.
4967+ let mut cmd = app.update(Event::FlushWatchdog(m.flush_gen), &mut m);
4968+
4969+ assert!(m.flushing, "a fresh send is in flight (the retry)");
4970+ assert!(m.in_flight.is_some(), "in_flight is set for the retry");
4971+ let batches = sent_batches(&mut cmd);
4972+ assert_eq!(batches.len(), 1, "the message is re-sent");
4973+ }
4974+
4975+ /// A watchdog that fires after the send already completed is a no-op.
4976+ #[test]
4977+ fn a_stale_flush_watchdog_leaves_a_completed_send_alone() {
4978+ let app = Skrepka;
4979+ let mut m = with_identity();
4980+ m.token = Some("t".into());
4981+ m.outbox.push_back(queued(&peer_hex(9), "hi"));
4982+
4983+ let _ = app.update(Event::StartFlush, &mut m);
4984+ let _ = app.update(Event::SendResult(m.flush_gen, Ok(ok_response(200))), &mut m);
4985+ assert!(!m.flushing);
4986+ assert!(m.outbox.is_empty());
4987+
4988+ // The watchdog for the completed send fires late.
4989+ let _ = app.update(Event::FlushWatchdog(m.flush_gen), &mut m);
4990+ assert!(!m.flushing, "a completed send is not re-armed");
4991+ assert!(m.outbox.is_empty());
4992+ }
4993+
4994+ /// A late `SendResult` from a superseded send is dropped by the generation
4995+ /// check. The watchdog abandoned the first send and retried under a new
4996+ /// generation; the late result's generation no longer matches, so it cannot
4997+ /// pop the retry's items or clear its guard. The retry's own `SendResult`
4998+ /// is what eventually pops them.
4999+ #[test]
5000+ fn a_late_send_result_after_the_watchdog_is_dropped_by_generation() {
5001+ let app = Skrepka;
5002+ let mut m = with_identity();
5003+ m.token = Some("t".into());
5004+ let peer = peer_hex(9);
5005+ m.outbox.push_back(queued(&peer, "stuck"));
5006+
5007+ let _ = app.update(Event::StartFlush, &mut m);
5008+ let first_gen = m.flush_gen;
5009+ // The watchdog fires and retries the flush under a new generation.
5010+ let _ = app.update(Event::FlushWatchdog(first_gen), &mut m);
5011+ assert_eq!(m.outbox.len(), 1, "the item is still queued (re-sent)");
5012+ assert_ne!(m.flush_gen, first_gen, "the generation was bumped");
5013+ assert!(m.flushing, "the retry is in flight");
5014+
5015+ // The original send's result arrives late. Its generation no longer
5016+ // matches, so it is dropped — it cannot pop the retry's items or
5017+ // clear the retry's guard.
5018+ let _ = app.update(Event::SendResult(first_gen, Ok(ok_response(200))), &mut m);
5019+ assert!(m.flushing, "the retry's guard is untouched");
5020+ assert_eq!(m.outbox.len(), 1, "the late result popped nothing");
5021+
5022+ // The retry's own result arrives and pops the items.
5023+ let _ = app.update(Event::SendResult(m.flush_gen, Ok(ok_response(200))), &mut m);
5024+ assert!(m.outbox.is_empty(), "the retry's result pops the items");
5025+ assert!(!m.flushing, "and the guard is released");
5026+ }
43445027}