Merge branch 'fix/restore-outcome-and-walk-gaps' into fix/sim-flakes-and-open-race

# Conflicts:
#	docs/BUGS.md

crates/pdfs-cli/src/main.rs

@@ -1467,7 +1467,11 @@ fn cmd_sync_restore(yes: bool, device: Option<String>) -> Result<()> {
             items: chosen,
         },
         None => CtlRequest::RestoreSyncFolders { items: chosen },
-    })?)
+    })?)?;
+    // The daemon acks before it works, so the outcome, and any folder it had
+    // to skip, only shows up later.
+    println!("Run `pdfs activity` to see which folders were restored.");
+    Ok(())
 }
 
 /// Expand a leading `~` in a hand-typed path. The daemon requires an absolute
@@ -3446,6 +3450,79 @@ fn cmd_sync_limit(up: Option<u64>, down: Option<u64>) -> Result<()> {
     })?)
 }
 
+/// How long the oldest write may wait to upload before `diagnose` calls it
+/// out. Under a day it is a slow link or a big file; past it, something is
+/// holding the queue.
+const STAGED_STALE_SECS: u64 = 86_400;
+
+/// `diagnose` findings for the queue and the pause, from the `Status` reply.
+/// `paused_for` is how long a timed pause still runs, in seconds.
+fn sync_health(
+    failing_ops: u64,
+    failing_error: Option<&str>,
+    staged_bytes: u64,
+    staged_oldest_secs: u64,
+    paused: bool,
+    paused_for: Option<i64>,
+) -> Vec<(DiagnoseLevel, &'static str, String)> {
+    let mut findings = Vec::new();
+    if failing_ops > 0 {
+        findings.push((
+            DiagnoseLevel::Warn,
+            "  failing writes",
+            match failing_error {
+                Some(error) => format!("{failing_ops} keep failing; the last error: {error}"),
+                None => format!("{failing_ops} keep failing"),
+            },
+        ));
+    }
+    if staged_bytes > 0 && staged_oldest_secs >= STAGED_STALE_SECS {
+        findings.push((
+            DiagnoseLevel::Warn,
+            "  oldest unsent write",
+            format!(
+                "{} old, {} only on this computer",
+                human_secs(staged_oldest_secs),
+                human_bytes(staged_bytes)
+            ),
+        ));
+    }
+    if paused {
+        findings.push((
+            DiagnoseLevel::Warn,
+            "  sync",
+            match paused_for {
+                Some(left) if left > 0 => {
+                    format!("paused for another {}", human_secs(left as u64))
+                }
+                _ => "paused until resumed (`pdfs sync resume`)".to_string(),
+            },
+        ));
+    }
+    findings
+}
+
+/// `diagnose` findings for the synced folders that need someone to look.
+fn folder_health(
+    folders: &[pdfs_core::control::SyncFolderInfo],
+) -> Vec<(DiagnoseLevel, String, String)> {
+    folders
+        .iter()
+        .filter_map(|folder| {
+            let what = match folder.state.as_str() {
+                "error" => "sync error; `pdfs sync list` and the activity feed say more",
+                "conflict" => "conflict copies to look at",
+                _ => return None,
+            };
+            Some((
+                DiagnoseLevel::Warn,
+                format!("  synced folder [{}]", folder.id),
+                format!("{}: {what}", folder.local_path),
+            ))
+        })
+        .collect()
+}
+
 fn unix_now() -> i64 {
     std::time::SystemTime::now()
         .duration_since(std::time::UNIX_EPOCH)
@@ -3782,21 +3859,26 @@ fn cmd_diagnose() -> Result<()> {
     let socket_exists = socket.exists();
     let socket_kind_ok =
         std::fs::symlink_metadata(&socket).is_ok_and(|m| m.file_type().is_socket());
+    // A stale socket file with no listener is the common failure, so the reply
+    // is what decides this rather than the file's existence.
+    let status = control_request(CtlRequest::Status);
     report.finding(
         if socket_exists && !socket_kind_ok {
             DiagnoseLevel::Fail
-        } else if socket_exists {
+        } else if socket_exists && status.is_ok() {
             DiagnoseLevel::Ok
         } else {
             DiagnoseLevel::Warn
         },
         "  control socket",
-        socket.display().to_string(),
+        if socket_exists && socket_kind_ok && status.is_err() {
+            format!("{} (left over; nothing listens on it)", socket.display())
+        } else {
+            socket.display().to_string()
+        },
     );
-    // A stale socket file with no listener is the common failure, so the reply
-    // is what decides this rather than the file's existence.
     let mut daemon_responding = false;
-    match control_request(CtlRequest::Status) {
+    match status {
         Ok(CtlResponse::Status {
             mountpoint,
             online,
@@ -3804,6 +3886,12 @@ fn cmd_diagnose() -> Result<()> {
             tokens_unsaved,
             pending_uploads,
             pending_changes,
+            failing_ops,
+            failing_error,
+            staged_bytes,
+            staged_oldest_secs,
+            paused,
+            paused_until,
             version,
             ..
         }) => {
@@ -3878,6 +3966,27 @@ fn cmd_diagnose() -> Result<()> {
                     format!("{pending} waiting to upload")
                 },
             );
+            // The rest of the daemon's state that the report used to drop:
+            // writes that keep failing, a pause, a folder in error. Each of
+            // them stops changes from reaching Drive while everything above
+            // reads as fine (`docs/BUGS.md` B268).
+            for (level, check, detail) in sync_health(
+                failing_ops,
+                failing_error.as_deref(),
+                staged_bytes,
+                staged_oldest_secs,
+                paused,
+                paused_until.map(|until| until - unix_now()),
+            ) {
+                report.finding(level, check, detail);
+            }
+            if let Ok(CtlResponse::SyncFolders { items }) =
+                control_request(CtlRequest::ListSyncFolders)
+            {
+                for (level, check, detail) in folder_health(&items) {
+                    report.finding(level, &check, detail);
+                }
+            }
             match await_integrity_check() {
                 Ok(CtlResponse::CacheReport {
                     integrity_problems,
@@ -4075,6 +4184,50 @@ mod pause_tests {
 mod diagnose_tests {
     use super::*;
 
+    /// A healthy queue says nothing; writes that keep failing, an old unsent
+    /// write and a pause each say so (`docs/BUGS.md` B268).
+    #[test]
+    fn a_stuck_queue_or_a_pause_is_reported() {
+        assert!(sync_health(0, None, 0, 0, false, None).is_empty());
+        assert!(sync_health(0, None, 4096, STAGED_STALE_SECS - 1, false, None).is_empty());
+        let checks: Vec<&str> = sync_health(
+            2,
+            Some("upload a.txt: quota"),
+            4096,
+            STAGED_STALE_SECS,
+            true,
+            Some(600),
+        )
+        .into_iter()
+        .map(|(level, check, _)| {
+            assert!(matches!(level, DiagnoseLevel::Warn));
+            check.trim()
+        })
+        .collect();
+        assert_eq!(checks, ["failing writes", "oldest unsent write", "sync"]);
+    }
+
+    /// Issue #53: a folder that would not sync, and `diagnose` found nothing.
+    #[test]
+    fn a_synced_folder_in_error_is_reported() {
+        let folder = |id: i64, state: &str| pdfs_core::control::SyncFolderInfo {
+            id,
+            local_path: format!("/home/u/f{id}"),
+            remote_uid: "vol~link".into(),
+            mode: "mirror".into(),
+            state: state.into(),
+            last_sync: 0,
+            pending_mode: None,
+            progress: None,
+            paused: false,
+        };
+        let findings =
+            folder_health(&[folder(1, "idle"), folder(2, "error"), folder(3, "syncing")]);
+        assert_eq!(findings.len(), 1);
+        assert_eq!(findings[0].1.trim(), "synced folder [2]");
+        assert!(findings[0].2.starts_with("/home/u/f2: sync error"));
+    }
+
     #[test]
     fn finding_severity_is_counted_centrally() {
         let mut report = DiagnoseReport::default();

crates/pdfs-core/src/error.rs

@@ -160,13 +160,14 @@ impl CoreError {
     /// telling the user their connection is down beats telling them the
     /// server refused. The SDK's other failures say what went wrong instead
     /// (see [`proton_kind`]).
+    ///
+    /// Callers also hand it this crate's [`Error`] from the cache and the
+    /// database, which is not a transport error at all. Taking those for
+    /// "offline" told the user their connection was down when a pin failed on
+    /// the database (`docs/BUGS.md` B266), so the source chain is searched for
+    /// either type before falling back.
     pub fn from_api(e: &(dyn std::error::Error + 'static), what: &str) -> Self {
-        let kind = match e.downcast_ref::<proton_sdk::ProtonError>() {
-            Some(proton) => proton_kind(proton),
-            // Not the SDK's at all: a timeout around the call, or the socket.
-            None => ErrorKind::Offline,
-        };
-        Self::new(kind, format!("{what}: {e}"))
+        Self::new(api_kind_of(e), format!("{what}: {e}"))
     }
 
     /// Prefix the message while keeping the classification.
@@ -208,7 +209,13 @@ impl From<serde_json::Error> for CoreError {
 
 impl From<Error> for CoreError {
     fn from(e: Error) -> Self {
-        let kind = match &e {
+        Self::new(core_kind(&e), e.to_string())
+    }
+}
+
+/// What this crate's [`Error`] means to a front-end.
+fn core_kind(e: &Error) -> ErrorKind {
+    match e {
         Error::NotLoggedIn
         | Error::ReloginRequired
         | Error::Locked
@@ -216,11 +223,26 @@ impl From<Error> for CoreError {
         | Error::HumanVerificationRequired(_) => ErrorKind::Denied,
         Error::Proton(proton) => proton_kind(proton),
         _ => ErrorKind::Internal,
-        };
-        Self::new(kind, e.to_string())
     }
 }
 
+/// The kind for [`CoreError::from_api`]: the first SDK or crate error in `e`'s
+/// source chain decides it.
+fn api_kind_of(e: &(dyn std::error::Error + 'static)) -> ErrorKind {
+    let mut cause = Some(e);
+    while let Some(err) = cause {
+        if let Some(proton) = err.downcast_ref::<proton_sdk::ProtonError>() {
+            return proton_kind(proton);
+        }
+        if let Some(core) = err.downcast_ref::<Error>() {
+            return core_kind(core);
+        }
+        cause = err.source();
+    }
+    // Not the SDK's at all: a timeout around the call, or the socket.
+    ErrorKind::Offline
+}
+
 /// What a failed SDK call means to a front-end.
 ///
 /// Only a transport failure is the connection. The rest happened on this
@@ -414,6 +436,20 @@ mod tests {
         assert_ne!(e.kind, ErrorKind::Offline);
     }
 
+    /// A pin or a search that fails on the database is not the connection, so
+    /// it must not say "You're offline" (`docs/BUGS.md` B266). An SDK error
+    /// wrapped in this crate's type classifies as the SDK error does.
+    #[test]
+    fn a_database_error_handed_to_from_api_is_not_offline() {
+        let db = Error::Other("database is locked".into());
+        assert_eq!(CoreError::from_api(&db, "pin").kind, ErrorKind::Internal);
+        let wrapped = Error::Proton(api(C::DoesNotExist));
+        assert_eq!(
+            CoreError::from_api(&wrapped, "load node").kind,
+            ErrorKind::NotFound
+        );
+    }
+
     /// `Error` is the plumbing type; converting it must classify as sharply as
     /// the direct path, or the two disagree about the same failure.
     #[test]

crates/pdfs-fuse/src/control.rs

@@ -1526,9 +1526,17 @@ fn handle_control_conn(core: &Core, username: &str, mountpoint: &Path, stream: U
             let core = core.clone();
             std::thread::spawn(move || {
                 match core.restore_sync_folders(&items, device.as_deref()) {
-                    Ok(message) => {
+                    Ok(outcome) => {
                         core.touch_profile();
-                        core.log_activity(ActivityKind::Download, &message, "", true);
+                        if !outcome.complete() {
+                            warn!(skipped = ?outcome.skipped, "restore skipped folders");
+                        }
+                        core.log_activity(
+                            ActivityKind::Download,
+                            outcome.summary(),
+                            outcome.skipped.join("; "),
+                            outcome.complete(),
+                        );
                     }
                     Err(e) => {
                         warn!(error = %e, "restore failed");

crates/pdfs-fuse/src/filesystem.rs

@@ -1434,6 +1434,14 @@ impl ProtonFs {
                     // (`ContentCache::clone_or_copy`).
                     if let Err(e) = ContentCache::clone_or_copy(&pending.path, &path) {
                         let _ = std::fs::remove_file(&path);
+                        // The drain lets go of the pending entry before it
+                        // discards the blob, so a blob that went away under the
+                        // copy is a revision that landed or was replaced: the
+                        // same race the re-check below catches, only earlier.
+                        if self.core.pending_blob(&uid).as_ref() != Some(&pending.path) {
+                            warn!(%uid, attempt, "queued revision went away while opening for write; retrying");
+                            continue;
+                        }
                         error!(%uid, source = %pending.path.display(), error = %e,
                             "copy queued revision into write scratch failed");
                         break 'install Err(Errno::EIO);

crates/pdfs-fuse/src/profile.rs

@@ -388,7 +388,7 @@ impl Core {
         &self,
         items: &[pdfs_core::control::RestoreItem],
         device: Option<&str>,
-    ) -> CoreResult<String> {
+    ) -> CoreResult<RestoreOutcome> {
         if items.is_empty() {
             return Err(CoreError::invalid("nothing to restore"));
         }
@@ -434,10 +434,19 @@ impl Core {
                     continue;
                 }
             };
-            let id = self
-                .db
-                .sync_folder_add(&local.to_string_lossy(), &item.remote_uid, &device.share_id)
-                .map_err(|e| CoreError::internal(format!("db: {e:?}")))?;
+            let id = match self.db.sync_folder_add(
+                &local.to_string_lossy(),
+                &item.remote_uid,
+                &device.share_id,
+            ) {
+                Ok(id) => id,
+                Err(e) => {
+                    // The folders restored before this one still need their
+                    // watches.
+                    let _ = self.sync_tx.send(sync::SyncMsg::Rewatch);
+                    return Err(CoreError::internal(format!("db: {e:?}")));
+                }
+            };
             // `ondemand` is applied through the normal request path, which
             // mounts the FUSE session and handles the folder being busy — the
             // restore does not get its own copy of that logic.
@@ -454,12 +463,35 @@ impl Core {
         for reason in &skipped {
             info!(reason, "restore skipped a folder");
         }
-        let mut message = format!("restored {restored} folder(s)");
-        if !skipped.is_empty() {
-            message.push_str(&format!("; skipped: {}", skipped.join(", ")));
-        }
         info!(restored, skipped = skipped.len(), "restore complete");
-        Ok(message)
+        Ok(RestoreOutcome { restored, skipped })
+    }
+}
+
+/// What a restore did: how many folders it attached, and why it left out the
+/// others.
+pub(crate) struct RestoreOutcome {
+    pub(crate) restored: usize,
+    pub(crate) skipped: Vec<String>,
+}
+
+impl RestoreOutcome {
+    /// The activity entry's headline.
+    pub(crate) fn summary(&self) -> String {
+        match (self.restored, self.skipped.len()) {
+            (n, 0) => format!("restored {n} folder(s)"),
+            (0, k) => format!("restored no folders; skipped {k}"),
+            (n, k) => format!("restored {n} folder(s); skipped {k}"),
+        }
+    }
+
+    /// Whether the restore did all it was asked. A skipped folder is a
+    /// failure the user has to see: the reply went out as "restoring folders"
+    /// before any work, and a restore that skipped every folder used to be
+    /// logged as a success, so nothing anywhere said that nothing would
+    /// download (`docs/BUGS.md` B269).
+    pub(crate) fn complete(&self) -> bool {
+        self.skipped.is_empty()
     }
 }
 

crates/pdfs-fuse/src/sim/daemon.rs

@@ -592,14 +592,22 @@ mod tests {
             daemon.pending(),
             drive.tree()
         );
-        // The bytes kept for the handle went once the trash landed with it closed.
-        let cached: Vec<_> = std::fs::read_dir(dir.join("cache"))
+        // The bytes kept for the handle went once the trash landed with it
+        // closed. The queue empties as the trash's row goes, a moment before
+        // the eviction, so this waits too.
+        let cached = || -> Vec<_> {
+            std::fs::read_dir(dir.join("cache"))
                 .unwrap()
                 .filter_map(|e| e.ok())
                 .filter(|e| e.file_type().is_ok_and(|t| t.is_file()))
                 .map(|e| e.file_name())
-            .collect();
-        assert!(cached.is_empty(), "{cached:?}");
+                .collect()
+        };
+        assert!(
+            wait_until(Duration::from_secs(30), || cached().is_empty()),
+            "{:?}",
+            cached()
+        );
 
         assert!(matches!(daemon.stop(), Ok(MountOutcome::Shutdown)));
         let _ = std::fs::remove_dir_all(&dir);
@@ -1998,6 +2006,21 @@ mod tests {
         daemon.write_new(&path, b"first").unwrap();
         assert!(wait_until(Duration::from_secs(30), || held.reached()));
         std::fs::write(&path, b"second").unwrap();
+        // The kernel sends the release after `close` returns. Attaching the
+        // bytes resets the create's attempts, so a release that came after the
+        // lost reply had it retried at once, and the retry landed before
+        // anything here saw it fail. The bytes it replaced go only once they
+        // are attached.
+        let staged = dir.join("cache").join("staging");
+        assert!(wait_until(Duration::from_secs(30), || {
+            let blobs: Vec<Vec<u8>> = std::fs::read_dir(&staged)
+                .into_iter()
+                .flatten()
+                .filter_map(|e| e.ok())
+                .filter_map(|e| std::fs::read(e.path()).ok())
+                .collect();
+            blobs.iter().any(|b| b == b"second") && !blobs.iter().any(|b| b == b"first")
+        }));
         drop(held);
         assert!(wait_until(Duration::from_secs(30), || {
             daemon

crates/pdfs-fuse/src/sync.rs

@@ -1162,14 +1162,22 @@ impl Core {
         let uids = self
             .rt
             .block_on(self.drive.enumerate_folder_children_node_uids(folder))
-            .map_err(|e| format!("enumerate {folder}: {e}"))?;
+            .map_err(|e| format!("enumerate {folder}: {e}"));
+        let uids = match unlisted_gap(walk, prefix, uids)? {
+            Some(uids) => uids,
+            None => return Ok(()),
+        };
         if uids.is_empty() {
             return Ok(());
         }
         let nodes = self
             .rt
             .block_on(self.drive.enumerate_nodes_light(&uids))
-            .map_err(|e| format!("resolve nodes: {e}"))?;
+            .map_err(|e| format!("resolve nodes: {e}"));
+        let nodes = match unlisted_gap(walk, prefix, nodes)? {
+            Some(nodes) => nodes,
+            None => return Ok(()),
+        };
         // The SDK leaves out a node it cannot decrypt, and says so only in its
         // log. Its name went with it, so all this can record is the folder.
         if nodes.len() < uids.len() {
@@ -1213,10 +1221,14 @@ impl Core {
         let sized: HashMap<String, (i64, Option<String>)> = if stale.is_empty() {
             HashMap::new()
         } else {
-            self.rt
+            let full = self
+                .rt
                 .block_on(self.drive.enumerate_nodes(&stale))
-                .map_err(|e| format!("resolve nodes: {e}"))?
-                .iter()
+                .map_err(|e| format!("resolve nodes: {e}"));
+            let Some(full) = unlisted_gap(walk, prefix, full)? else {
+                return Ok(());
+            };
+            full.iter()
                 .map(|n| {
                     (
                         n.uid.to_string(),
@@ -2034,6 +2046,34 @@ struct RemoteWalk<'a> {
     gaps: &'a mut RemoteGaps,
 }
 
+/// A listing call's result in [`Core::walk_remote`]. A failure below the root
+/// records `prefix` as unlisted and returns `None`, so the walk goes on
+/// without that folder. A failure at the root still fails the pass, since
+/// nothing would be left to compare.
+///
+/// One listing error, after the SDK's own retries, used to fail the whole
+/// walk. Nothing is applied before the walk ends, so on a link that drops a
+/// request now and then a large first sync started over from nothing on
+/// every pass and never got past it (`docs/BUGS.md` B271).
+fn unlisted_gap<T>(
+    walk: &mut RemoteWalk<'_>,
+    prefix: &str,
+    result: Result<T, String>,
+) -> Result<Option<T>, PassAbort> {
+    match result {
+        Ok(value) => Ok(Some(value)),
+        Err(error) if prefix.is_empty() => Err(error.into()),
+        Err(error) => {
+            warn!(
+                folder = prefix,
+                error, "sync: a folder on Drive could not be listed; leaving it out of this pass"
+            );
+            walk.gaps.unlisted(prefix);
+            Ok(None)
+        }
+    }
+}
+
 /// Whether the file at `path` is no longer what the scan saw as `planned`.
 /// A file that is gone has nothing left to lose, so it has not changed.
 fn changed_since_scan(path: &Path, planned: &LocalItem) -> bool {

crates/pdfs-fuse/src/sync/planner.rs

@@ -236,6 +236,9 @@ pub(super) struct RemoteGaps {
     short: HashSet<String>,
     /// Paths seen, but whose entry could not be read.
     unread: HashSet<String>,
+    /// Folders whose listing failed outright, after the SDK's retries. Nothing
+    /// under one is known, so all of it is left out of the pass (B271).
+    unlisted: HashSet<String>,
 }
 
 impl RemoteGaps {
@@ -249,8 +252,13 @@ impl RemoteGaps {
         self.unread.insert(rel);
     }
 
+    /// The listing of `folder` failed, so nothing under it was walked.
+    pub(super) fn unlisted(&mut self, folder: &str) {
+        self.unlisted.insert(folder.to_string());
+    }
+
     pub(super) fn is_empty(&self) -> bool {
-        self.short.is_empty() && self.unread.is_empty()
+        self.short.is_empty() && self.unread.is_empty() && self.unlisted.is_empty()
     }
 
     /// Whether `rel` may be missing from `remote` only because the walk could
@@ -272,7 +280,7 @@ impl RemoteGaps {
         let mut at = parent_rel(rel);
         loop {
             if remote_dirs.contains_key(at) {
-                return self.short.contains(at);
+                return self.short.contains(at) || self.unlisted.contains(at);
             }
             if at.is_empty() {
                 return false;
@@ -295,6 +303,9 @@ impl RemoteGaps {
                     .iter()
                     .map(|rel| format!("{rel} (cannot be read on Proton Drive)")),
             )
+            .chain(self.unlisted.iter().map(|rel| {
+                format!("{rel} (could not be listed on Proton Drive; tried again next pass)")
+            }))
             .collect();
         lines.sort_unstable();
         lines
@@ -844,6 +855,32 @@ mod file_plan_tests {
         assert!(gaps.hides("top.txt", &remote, &dirs));
     }
 
+    /// A folder whose listing failed was walked no further: nothing under it
+    /// reads as deleted on Drive, and a sibling that listed still does
+    /// (`docs/BUGS.md` B271).
+    #[test]
+    fn a_folder_that_failed_to_list_hides_everything_under_it() {
+        let remote = HashMap::from([
+            ("docs".to_string(), folder()),
+            ("pics".to_string(), folder()),
+        ]);
+        let dirs = HashMap::from([
+            (String::new(), uid("root")),
+            ("docs".to_string(), uid("docs")),
+            ("pics".to_string(), uid("pics")),
+        ]);
+        let mut gaps = RemoteGaps::default();
+        gaps.unlisted("docs");
+        assert!(gaps.hides("docs/a.txt", &remote, &dirs));
+        assert!(gaps.hides("docs/sub/deep.txt", &remote, &dirs));
+        assert!(!gaps.hides("docs", &remote, &dirs));
+        assert!(!gaps.hides("pics/gone.jpg", &remote, &dirs));
+        assert_eq!(
+            gaps.report(),
+            ["docs (could not be listed on Proton Drive; tried again next pass)"]
+        );
+    }
+
     #[test]
     fn a_file_whose_size_could_not_be_read_is_hidden() {
         let remote: HashMap<String, RemoteItem> = HashMap::new();

docs/BUGS.md

@@ -68,13 +68,10 @@ original wording.
 
 | Entry | Title | What is left |
 |---|---|---|
+| [B275](#b275--on-a-slow-link-a-files-last-write-can-land-only-in-a-conflict-copy) | On a slow link, a file's last write can land only in a conflict copy | Cause not found; seen once in CI |
 | [B273](#b273--a-restore-proposes-another-computers-paths-when-the-two-share-a-hostname) | A restore proposes another computer's paths when the two share a hostname | Not fixed |
-| [B271](#b271--one-failed-listing-makes-a-large-first-sync-start-over-from-nothing) | One failed listing makes a large first sync start over from nothing | Not fixed |
 | [B270](#b270--one-file-left-in-a-mirror-folder-clears-the-wipe-guard-for-all-the-others) | One file left in a mirror folder clears the wipe guard for all the others | Not fixed |
-| [B269](#b269--a-restore-that-skipped-every-folder-reports-success) | A restore that skipped every folder reports success | Not fixed |
-| [B268](#b268--pdfs-diagnose-says-no-problems-found-while-sync-is-stuck) | `pdfs diagnose` says "No problems found" while sync is stuck | Not fixed |
 | [B267](#b267--a-key-failure-at-a-start-from-a-warm-cache-shows-as-offline-for-good) | A key failure at a start from a warm cache shows as offline for good | Not fixed |
-| [B266](#b266--pin-unpin-search-and-node-loads-report-a-database-error-as-offline) | Pin, unpin, search and node loads report a database error as "offline" | Not fixed |
 | [B265](#b265--address-keys-whose-token-is-for-an-inactive-account-key-never-open) | Address keys whose token is for an inactive account key never open | Needs an SDK change; not reproduced |
 | [B134](#b134--a-file-deleted-online-while-the-drain-checks-its-write-is-kept-as-a-conflict-copy) | A file deleted online while the drain checks its write is kept as a conflict copy | Only with `local_first` off |
 | [B91](#b91--the-daemon-hung-overnight-and-nothing-could-say-what-it-was-waiting-on) | The daemon hung overnight, and nothing could say what it was waiting on | The cause is not diagnosed; the daemon now reports where it is stuck and is restarted |
@@ -88,6 +85,8 @@ original wording.
 
 | Entry | Title | What is left |
 |---|---|---|
+| [B269](#b269--a-restore-that-skipped-every-folder-reports-success) | A restore that skipped every folder reports success | The app shows the result only in the activity feed; no offer to replace or reset the folder in the way |
+| [B268](#b268--pdfs-diagnose-says-no-problems-found-while-sync-is-stuck) | `pdfs diagnose` says "No problems found" while sync is stuck | The sync engine's state is not checked |
 | [B263](#b263--a-mount-that-can-never-succeed-retries-every-5-s-and-tells-only-the-journal) | A mount that can never succeed retries every 5 s and tells only the journal | Errors no retry fixes back off like network errors |
 | [B262](#b262--watching-a-large-synced-folder-holds-up-the-sync-engine-and-a-failed-watch-is-never-retried) | Watching a large synced folder holds up the sync engine, and a failed watch is never retried | The first pass still waits for the watch; a failed watch is not shown in the app |
 | [B259](#b259--the-sync-engine-can-die-or-hang-and-nothing-shows-it-until-a-restart) | The sync engine can die or hang, and nothing shows it until a restart | No pass deadline; an ended engine is reported, not restarted; the trigger in #53 is not known |
@@ -111,6 +110,9 @@ a live account since.
 
 | Entry | Title |
 |---|---|
+| [B274](#b274--opening-a-file-for-writing-fails-with-eio-while-its-queued-revision-lands) | Opening a file for writing fails with EIO while its queued revision lands |
+| [B271](#b271--one-failed-listing-makes-a-large-first-sync-start-over-from-nothing) | One failed listing makes a large first sync start over from nothing |
+| [B266](#b266--pin-unpin-search-and-node-loads-report-a-database-error-as-offline) | Pin, unpin, search and node loads report a database error as "offline" |
 | [B264](#b264--a-mount-that-fails-after-starting-its-workers-leaves-them-running) | A mount that fails after starting its workers leaves them running |
 | [B261](#b261--a-folder-that-has-never-synced-reads-up-to-date) | A folder that has never synced reads "up to date" |
 | [B272](#b272--after-a-folder-is-removed-and-added-again-its-file-events-go-to-the-old-id) | After a folder is removed and added again, its file events go to the old id |
@@ -375,6 +377,56 @@ a live account since.
 
 Newest first.
 
+### B275 — On a slow link, a file's last write can land only in a conflict copy
+
+**Status:** Open
+**Found:** 2026-10-31, CI on #60: `one_client_on_a_slow_link`, profile `one-client-wifi` seed 2,
+failed with `Drive does not hold what the clients wrote, 120s after the links came up`. The
+same seed passed 4 times locally, and #60 changes no daemon logic, so the failure depends on
+timing.
+**Where:** not located; the drain's conflict check in `crates/pdfs-fuse/src/drain.rs` is the
+first suspect
+
+**Repro:** Only the CI run above. The last steps created `x/z/y/a.txt` (37 bytes), wrote 42
+bytes into it at offset 20, and brought every link up. Drive then held the file at its first
+37 bytes, and the full content was in `x/z/y/a (sync-conflict …).txt`. No other client wrote
+to it.
+
+**Cause.** Not known. Nothing was lost, since the conflict copy holds the bytes, but the file
+conflicted with its own create. B183 and B184 found the same symptom after a lost create
+reply.
+
+**Fix.** Not done. Needs a reproduction: run the seed in a loop under load and keep the state
+directory of a failing run.
+
+---
+
+### B274 — Opening a file for writing fails with EIO while its queued revision lands
+
+**Status:** Fixed, not verified live
+**Found:** 2026-10-31, CI: the simulation
+`a_file_whose_upload_was_not_read_back_takes_two_writes_at_once` failed once with
+`Input/output error` on a write, after `copy queued revision into write scratch failed … No
+such file or directory`. It did not fail in 54 local runs.
+**Where:** the write open in `crates/pdfs-fuse/src/filesystem.rs`
+
+**Repro:** Write a file, and open it for writing again at the moment its queued revision
+finishes uploading. The open fails with EIO, and the program sees its save fail.
+
+**Cause.** The open samples the queued revision, then copies its blob into the new handle's
+scratch file. The drain lets go of the pending entry and then deletes the blob when the upload
+lands. A blob deleted between the sample and the copy failed the open, although the re-check
+just after the copy already retries a revision that changed during it.
+
+**Fix.** A failed copy checks whether the pending entry still names that blob. When it does
+not, the revision landed or was replaced, and the open goes around again with the base the
+file has now, as the re-check does. Since the drain drops the entry before the blob, a blob
+that is gone and still named is a real failure and still returns EIO.
+
+**Test:** none deterministic; the simulation above covers the window.
+
+---
+
 ### B273 — A restore proposes another computer's paths when the two share a hostname
 
 **Status:** Open
@@ -416,7 +468,7 @@ folder list has now.
 
 ### B271 — One failed listing makes a large first sync start over from nothing
 
-**Status:** Open
+**Status:** Fixed, not verified live
 **Found:** 2026-10-31, code audit after issue #46: a folder of about 19,900 items restored on a
 second computer sat at `scanning…` with nothing downloaded, and the journal showed
 `RequestTimeout (http 408)`.
@@ -433,8 +485,13 @@ the walk is done. Files that did download keep their baseline rows, so only the
 but on a first sync that is all of it. Progress is also counted only after a folder's full
 listing, so a large folder shows `scanning…` with 0 checked for a long time.
 
-**Fix.** Not done. Record a failed subtree as a gap, as undecryptable nodes already are: no
-deletions under it, and the rest of the pass still applies.
+**Fix.** A listing that fails below the sync root records that folder as a gap in
+`RemoteGaps`, next to the folders whose listing came back short. The planner deletes nothing
+under a gap, the rest of the pass still applies, and the pass report names the folder as
+tried again next pass. A failed listing of the root itself still fails the pass, since there is
+nothing left to apply. The progress count inside one large folder is unchanged.
+
+**Test:** `a_folder_that_failed_to_list_hides_everything_under_it` in `sync::planner`.
 
 ---
 
@@ -461,7 +518,7 @@ choice to the user: delete on the other side, or sync from scratch.
 
 ### B269 — A restore that skipped every folder reports success
 
-**Status:** Open
+**Status:** Partly fixed
 **Found:** 2026-10-31, code audit after issue #46: the journal showed `restore complete
 restored=0 skipped=1` several times, and the user saw nothing.
 **Where:** the `RestoreSyncFolders` and `RestoreDeviceFolders` handler in
@@ -476,15 +533,20 @@ result, `restored 0 folder(s); skipped: …`, is then logged as a successful act
 reason is only in the activity text and the journal. When the database insert fails partway,
 the `?` also skips the `Rewatch` for the folders already restored.
 
-**Fix.** Not done. Log a restore that skipped folders as a failure, name the reason in the app
-and the CLI, and offer to replace or reset the folder in the way. Send the `Rewatch` on every
-exit path.
+**Fix.** `restore_sync_folders` returns what it restored and what it skipped, with the reason
+for each skip. The daemon logs a restore that skipped any folder as a failed activity, with the
+reasons as its detail, and warns in the journal. A database error partway now sends the
+`Rewatch` for the folders already restored before it returns. `pdfs devices join` points to
+`pdfs activity` for the result.
+
+Left: the app shows the result only in the activity feed, and nothing offers to replace or reset
+the folder in the way.
 
 ---
 
 ### B268 — `pdfs diagnose` says "No problems found" while sync is stuck
 
-**Status:** Open
+**Status:** Partly fixed. The sync engine's own state is not checked yet.
 **Found:** 2026-10-31, issue #53: a folder that never synced, and `diagnose` found nothing.
 **Where:** `diagnose` in `crates/pdfs-cli/src/main.rs`
 
@@ -497,8 +559,16 @@ folders or for `Diagnostics`, so a folder in error and a stalled worker are not
 control socket file with no daemon behind it is also reported `[ok]`; only the next line says
 the daemon does not answer.
 
-**Fix.** Not done. Warn on failing or parked writes, a pause, old staged data and folders in
-error, and report the sync engine's state once B259 exposes it.
+**Fix.** `diagnose` warns about writes that keep failing (with the last error), a write that
+has waited a day or more to upload, a pause, and each synced folder in `error` or `conflict`
+(`sync_health` and `folder_health` in `crates/pdfs-cli/src/main.rs`). A control socket file
+that nothing listens on is a warning, not `[ok]`. Parked uploads are left out: a file still
+wearing a download's temporary name is normal. Not done: the sync engine's state, which
+`Diagnostics` reports once B259's change is in.
+
+**Test:** `a_stuck_queue_or_a_pause_is_reported` and `a_synced_folder_in_error_is_reported` in
+`pdfs-cli`. On 2026-10-31 the new `diagnose` against a healthy 3.6.0 daemon still reported
+"No problems found".
 
 ---
 
@@ -525,7 +595,7 @@ report the others as their own state through `Status`.
 
 ### B266 — Pin, unpin, search and node loads report a database error as "offline"
 
-**Status:** Open
+**Status:** Fixed, not verified live
 **Found:** 2026-10-31, code audit after issue #38. B205 fixed the photo upload path only.
 **Where:** `CoreError::from_api` in `crates/pdfs-core/src/error.rs`; its callers in
 `crates/pdfs-fuse/src/lib.rs` (`pin`, `unpin`, the searches, `load node`)
@@ -538,8 +608,11 @@ it was written for SDK calls, where that means a transport error. These callers
 `pdfs_core::Error` values from the cache and the database. A `ProtonError` wrapped in
 `pdfs_core::Error::Proton` also misses the downcast, since the source chain is not walked.
 
-**Fix.** Not done. Use `CoreError::internal` at the database callers, and walk the source
-chain in `from_api` before falling back.
+**Fix.** `from_api` walks the source chain. The first `ProtonError` or `pdfs_core::Error` in it
+decides the kind, the same way `From<Error>` does, and only an error with neither, a transport
+error, reads as offline. That covers every caller at once, the database ones included.
+
+**Test:** `a_database_error_handed_to_from_api_is_not_offline` in `pdfs-core`.
 
 ---