Tune SQLite and remove blocking calls from async/scan paths:
- Enable WAL + synchronous=NORMAL + busy_timeout + foreign_keys on each
pooled connection, so the five scanners, web server, and retention no
longer contend on the default rollback journal / FULL fsync.
- Run DB work in axum handlers via spawn_blocking (db::run_blocking) so
synchronous rusqlite calls no longer block tokio worker threads.
- Deliver notifications on a dedicated task fed by a bounded channel; the
blocking Pushover HTTP call runs in spawn_blocking, so a slow or
unreachable Pushover can never stall device discovery.
- Make get_db_connection() return Result instead of panicking, so pool
exhaustion surfaces as a 500 rather than crashing the process.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add ActiveStatusCell/PassiveStatusCell wrappers in the scanners common
module so the five per-scanner status.rs files reduce to a single static;
replace parse_parameter_bool/int/string with one generic parse_parameter
over FromStr; extract the shared LIMIT/OFFSET paging clause into
db::apply_paging; and drop a no-op for-loop in the ARP sender.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Collapse the five scanner status models into two shared shapes,
ActiveScannerStatus and PassiveScannerStatus, mirroring the backend's
active/passive vocabulary. Replace the five near-identical per-scanner
detail card files with a single scanner_status_cards.dart (two shared
resolvers plus a config list), and rebuild the combined home card to
iterate a list of scanners with two shape resolvers instead of five
copy-pasted resolve methods.
Extract two reusable mixins:
- PeriodicRebuild: the shared once-a-second "rebuild to refresh elapsed
text" timer used by the scanner cards and the stale indicator.
- PaginatedListState: the shared pagination state, page-size/page-count
getters, cancel-token-aware fetch orchestration, and disposal used by
the device and notification lists.
No behaviour change; ~900 lines removed. Tests and analyzer pass.
The ARP and SNMP scanners reported every online sighting, so a device
seen on multiple IPs or via duplicate ARP replies was counted more than
once. Fold the dedup into ActiveStatus::record_scan, which now takes the
device slice and reports the number of distinct MAC addresses.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The list endpoints now return a total count alongside the page so the
front-end can show how many pages exist and offer a last-page jump.
Backend: add count(is_new) and count_devices(...) (sharing a WHERE-builder
with list_devices so page and count can't drift), wrap both list responses
in {items, total_count} structs, and register them with utoipa.
Front-end: parse the wrapper shape (dropping the fetch-one-extra trick),
add a Last-page button and a responsive "Page X of Y" / "X / Y" label to
the shared PaginationBar, and track the total in both lists. Notifications
re-sync the count on every fetch and decrement it locally on mark-read/
unread removals so the count stays accurate without a re-fetch.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wide layout makes the leading device-type icon column header tappable to
sort; narrow layout gains a "Device Type" option in the sort sheet. The
backend already whitelisted device_type as a sort column.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The device detail chart's time-range selector (Today / Last week / ...)
rendered as a SegmentedButton that overflowed on narrow phone layouts.
Reuse the responsive FilterSelector widget so the same control is used
for both the list filters and the chart: segmented pills on wide layouts,
a compact dropdown combo box on phones.
To keep the two controls consistent, FilterSelector now renders a
SegmentedButton (instead of ChoiceChips) on wide layouts.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace the notifications list's wholesale redraw with a SliverAnimatedList
driven by a GlobalKey, keeping `_items` in lockstep with the animated state.
- Background refreshes (poll, pull-to-refresh, resume, route pop, mark-all)
reconcile against the fetched page: departed rows slide out, newly fetched
rows slide in at the top with a theme-coloured arrival highlight, and
surviving rows stay put (with in-place read-state recolouring under "All").
- Filter/page changes and the initial load reset the list (fresh key) so the
new dataset appears instantly without per-row animation.
- Read/unread removals are owned by the list: buttons play a slide/fade exit,
while swipes let Dismissible animate and then reconcile, avoiding double
animation and the disposed-widget race.
Add the arrival highlight overlay to NotificationCard and cover the new
behaviour with widget tests (swipe-out, flash-in, external removal, in-place
"All" mark, filter reset).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Changing pages in the notifications or devices list gave no cue that
the next page was loading. Render an indeterminate progress bar at the
shell level, pinned flush against the bottom of the page body (above
the nav bar on phones, the screen bottom on wide layouts), driven by a
shared paginationLoading notifier the lists set while fetching. The
pagination bar keeps disabling its buttons during the fetch.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wrap compact (phone) device rows in a Card instead of a bare
Material+Divider so they get the same rounded, spaced look as the
notifications list. Drop the row divider on narrow layouts and reduce
the phone page size to 5 to account for the taller card rows.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
On narrow layouts the devices list's filter chips (Not registered /
Registered / All) competed for horizontal space with the Sort and Filter
icon buttons and overlapped. Introduce a reusable FilterSelector that
keeps the chips on wide layouts but collapses to a compact dropdown
button on phones, and use it for both the devices and notifications
lists for consistency.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
On phone-width layouts (< Breakpoints.medium) the notification and device
lists now request fewer items per page so the list and its pagination bar
fit on screen together on common current phones. Notifications use 4 items
and devices 6 on phones; wider layouts keep 5 and 10 respectively. The
initial fetch is deferred to didChangeDependencies so the page size can read
the screen width from MediaQuery.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The devices list already supported pull-to-refresh; mirror that on the
notifications list by wrapping its CustomScrollView in a RefreshIndicator
with AlwaysScrollableScrollPhysics. Keep the existing list visible during
a refresh (_isLoading = _items.isEmpty) instead of flashing the skeleton,
matching the devices list behaviour.
Add widget tests covering pull-to-refresh for both lists.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
run_android_emulator.sh already boots the emulator and launches the app; it
delegated the boot step to run_android.sh. Inline that logic as a
boot_emulator() helper and drop the now-redundant script.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The pushover config section is now only required when notifications.method
is "pushover". Validation at startup rejects the missing-section case so a
misconfiguration fails fast instead of erroring on every notification.
Updates the sample TOML, README and nix module for consistency.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace the _goToPage wrapper with an optional scrollToTop flag on
_fetchPage, so there is a single fetch entry point. The pagination bar
passes scrollToTop: true; all other callers keep the current behaviour.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The notifications and devices lists kept their scroll offset when paging,
so a new page would open partway down. Attach a ScrollController to each
CustomScrollView and animate back to the top whenever the page changes
via the pagination bar.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
SliverList recycles State objects by index, so when an expanded
notification was marked as read and dropped from the filtered list, the
next notification inherited the expanded state. Give each card a
ValueKey(id) so its expand/collapse state is matched by notification,
not by list position.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Tapping a notification's body already toggled expand/collapse, but the
action area below it wasn't tappable. Wrap the whole card in an InkWell
so tapping anywhere (body or action row) toggles, while the buttons
still handle their own taps.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace the vanilla Flutter splash with the OOTT wordmark (Barlow
Condensed Bold) on the Gruvbox-dark background, matching the in-app
AppBar badge. Generated via flutter_native_splash, covering legacy
Android, the Android 12+ SplashScreen API, and iOS (light + dark).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Three issues prevented `run_android_emulator.sh` from launching the app:
- gradlew's shebang was a hardcoded Nix-store bash path (copied verbatim
from the nixpkgs Flutter SDK template) that breaks once that store path
is garbage-collected, surfacing as a misleading "ProcessException: No
such file or directory". The dev-shell now normalizes the gitignored
android/gradlew shebang to the portable "#!/usr/bin/env sh" on entry.
- `flutter run -d android` never matched: -d resolves a device by id/name,
not platform. Resolve the concrete id (e.g. emulator-5554) from
`flutter devices --machine` instead.
- Install raced the boot ("device is still booting"); wait for
sys.boot_completed=1 before launching.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Rename android-emulator.sh to run_android.sh and run.sh to run_web.sh,
add run_android_emulator.sh to launch the app on the emulator, and
update CLAUDE.md commands accordingly.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Several settings were required at parse time even though they had a
sensible default, causing the backend to panic on startup when omitted:
- notifications.notify_when_not_seen_for now defaults to "1w"
- the whole [networking] section is now optional (no mandatory fields)
- scanner duration/timeout fields (ARP, mDNS, SSDP, SNMP) now fall back
to their defaults when the section is present but the field is omitted.
serde only applies a field default when marked #[serde(default)], so
the per-field attributes were added and the Default impls now share the
same default functions as the single source of truth.
Updates the sample TOML and README accordingly and adds tests covering
each defaulting case.
Frontend usability and aesthetics pass across web and mobile:
- Add shared layout tokens (theme/dimens.dart: Insets + Breakpoints) and
replace magic-number spacing and per-file breakpoint consts.
- Route both themes through buildAppTheme (theme/theme_builder.dart): explicit
useMaterial3 and a shared branded textTheme (Barlow Condensed for
display/headline/title styles); logo now reads its style from the theme.
- Move screen titles into the shared AppBar (route-derived) and drop the
redundant in-body headers; upgrade Settings buttons to M3 FilledButton.
- Add reusable EmptyState and skeleton loaders; Devices empty state links to
scanner status, notifications get filter-aware messages.
- Add a first-run welcome intro and Save-disabled helper text in Settings.
- Make device Filter/Sort adaptive: bottom sheet on phones, dialog on wide.
- Collapse the device-list filter chips and Sort/Filter buttons into one row;
rename the home Scanners card title and align its style.
Tests: add EmptyState/skeleton widget tests and Settings first-run cases;
update tests for the FilledButton swap and relocated titles. 89 passing,
dart analyze clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Introduce a headless `flutter test` suite (83 tests) covering models,
utilities, every API endpoint, and the five screens, with shared helpers
and fixtures so new tests stay terse.
API endpoint methods are statically-dispatched extensions on the
BackendAPI singleton, so they cannot be mocked via `implements`. Mock at
the Dio HTTP-adapter layer (http_mock_adapter) instead, enabled by two
small @visibleForTesting seams: BackendAPI.dioForTesting swaps the
singleton's Dio, and BackendReachability.forceOnlineForTesting() forces a
deterministic online state so polling widgets load.
Add frontend/run_tests.sh and document it in CLAUDE.md.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The google_apis android-36 image always exposes a phantom AT hardware
keyboard that cannot be removed via hw.keyboard=no, the lid switch, or
any emulator flag. With hw.keyboard=no the host keyboard is not forwarded
into the guest, and Gboard hides the on-screen keyboard for non-password
fields while it detects that phantom keyboard, leaving plain text fields
(e.g. Settings > Base URL) untypable.
Enforce hw.keyboard=yes in the AVD's config.ini on every dev-shell entry
so the host keyboard is forwarded to all fields. Applied idempotently so
it also repairs a pre-existing AVD; the changed hardware config makes the
emulator cold-boot once for the setting to take effect.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The Android SDK lives in the read-only Nix store, so the Gradle build
cannot auto-install missing components. Pin everything the Flutter
Android build needs in the dev shell's composeAndroidPackages:
platforms 33-36 (app targets 36; plugins pin 34 and 35), build-tools
35.0.0 (AGP 8.11.1), NDK 28.2.13676358 (flutter.ndkVersion), and
cmake 3.22.1 (plugin native builds).
Also redirect the Android Gradle Plugin to the Nix-patched aapt2 from
the SDK via android.aapt2FromMavenOverride, written to the user-global
~/.gradle/gradle.properties on shell entry (the project's tracked
gradle.properties must not contain a machine-specific store path).
AGP's Maven-downloaded aapt2 cannot run on NixOS.
Ignore the generated frontend/android/build/ directory.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Pin the Android SDK composition (platform 36 + google_apis x86_64 system
image + emulator) via composeAndroidPackages so the emulator image is
always available and reproducible across machines. Auto-create the
"oott_api36" AVD idempotently on dev-shell entry, and add
frontend/android-emulator.sh to launch it.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Recompute the storage estimates around the deduplication window as the
governing cap across all four default-enabled scanners (ARP, mDNS, SSDP,
DHCP), using the documented defaults rather than the old single-ARP
720-scans/day worst case. Present typical (not worst-case) figures, add a
per-scanner event breakdown, rework the tuning levers around retention and
disabling scanners, and note that the SNMP scanner is disabled by default
and excluded from the figures.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Fall back to "0.0.0.0" and 3000 when these fields are omitted from the
[web_server] section (api_key remains required). Update the sample TOML
and README to reflect the new defaults; the Nix module already used them.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Make the [log] section and its level field optional, falling back to
"warn" when omitted. Update the sample TOML, Nix module and README to
reflect the new default.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace the README config table's "Sample value" column with
"Default value" reflecting the real defaults from settings.rs:
mark options with no built-in default as Required and list the
code defaults for the scanners, retention and deduplication.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Reorganize into What is OOTT?, Getting started (Docker + mobile apps),
Deeper dive (Nix flake, config options, HTTPS), and Things to keep in
mind (storage, privileges/ports). Inline a minimal oott.toml and the
compose file in the Docker quickstart, and document the missing
web_server.* options.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds a repo-root release.sh that prompts for the version, bumps it in
Cargo.toml/pubspec.yaml and the About release date, refreshes lockfiles,
runs tests, commits/pushes, tags, creates the GitHub release, and builds
the Docker image, pausing for confirmation between sections.
Release notes live in RELEASE_NOTES.md (newest first, one section per
version); the script extracts the section for the version being cut.
Seeds the file with the v0.1.0 notes and adds gh to the Nix dev shell.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The status screen was a non-scrolling Column, so it overflowed
whenever the offline banner consumed vertical space or the viewport
was too short to fit all scanner cards. Wrap it in a
SingleChildScrollView, matching the other screens.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The version string was duplicated across six places. Collapse it to two
ecosystem sources of truth and derive the rest:
- backend/src/web_server.rs: omit the OpenAPI info.version so utoipa fills
it from CARGO_PKG_VERSION (backend/Cargo.toml); add a test pinning this.
- nix/package.nix: read the version from backend/Cargo.toml via fromTOML.
- nix/frontend.nix: read the version from frontend/pubspec.yaml by splitting
into lines (a whole-file regex triggers catastrophic backtracking in Nix's
regex engine).
- frontend/lib/about/about.dart: read the version at runtime via
package_info_plus instead of a hardcoded constant.
Also drop frontend/pubspec.lock.json: it is unreferenced (Nix's
autoPubspecLock generates its own JSON from pubspec.lock) and was going stale.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
When the same scanner sees the same device (same MAC and IPv4) again within
a configurable window (default 1 minute), only one device_events row is
recorded, keeping the events table from filling with near-identical rows.
Device last_seen updates and notifications are unaffected.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Break the 372-line frontend API client into focused files under
lib/utils/api/: error mapping (api_error.dart), Dio setup behind a
buildDio factory (dio_config.dart), and device/scanner/notification
endpoints as extensions in part files. Collapse the five near-identical
scanner-status methods via a generic _getModel helper and de-duplicate
the pagination logic via a shared _paginate helper.
The BackendAPI singleton and dioErrorToUserMessage remain importable
from oott_api.dart unchanged, so no call sites are affected.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The five scanners (ARP, SNMP, mDNS, SSDP, DHCP) duplicated their
persist-and-notify pipeline, device enrichment, and status state
machines across same-family files. Extract the shared logic so a change
lands in one place instead of three to five.
- scanners/common/pipeline.rs: single record_sighting() persist+notify
path, replacing the per-scanner match blocks. ARP/SNMP now use the
same merge rules as the passive scanners (keep a stored hostname,
never overwrite a known IP with an empty one).
- scanners/common/enrichment.rs: build_device() for vendor/device-type
lookup with the privacy-MAC service fallback.
- scanners/common/{active,passive}_status.rs: the two status state
machines plus their tests, written once. Each scanner status.rs is now
a thin wrapper over its own static.
- utils/network::format_mac(): replaces three identical copies.
- web_server/scanner_status.rs: Active/Passive response types and two
handler helpers, replacing five near-identical structs+handlers. JSON
field names are unchanged so the frontend is unaffected; only OpenAPI
schema names change.
27 files changed, ~900 lines net removed. Build, clippy and all 113
tests (4 new) pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace the lifetime "devices seen since start" counter in the mDNS, SSDP
and DHCP scanners with a rolling count of distinct devices (deduped by
MAC) seen within the last hour.
Each scanner's status now tracks a MAC -> last-seen-time map; the snapshot
prunes entries older than 60 minutes and reports the remaining count. The
API field name (devices_seen) is unchanged, so only its meaning and the
frontend labels ("N devices in the last hour") are updated.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The ARP and SNMP scanner status endpoints and front-end cards now expose
the number of devices found by the most recent successful scan, following
the existing mDNS device-count pattern. The count persists across the
running/waiting transitions, and for SNMP a failed poll keeps the last
good count rather than overwriting it.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add an ArpScanner Default impl (30m/1m/10m, enabled) and mark the
arp_scanner field with serde default, so the [arp_scanner] section can
now be omitted entirely and fall back to code defaults — matching the
pattern used by the SNMP scanner. Previously these three durations were
mandatory and the backend would not start without them.
Reconcile the documentation to the canonical 30m/1m/10m: fix the README
NixOS example (was 15m/20m/30m) and options table (was 15m), and note
that the section is optional in both the README and sample TOML. The
NixOS module and sample TOML already used these values.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
