Pull the ~300 LOC URL → PlayInput dispatch out of audio_play into a
new audio/play_input.rs:
- PlayInput enum (Bytes / SeekableMedia / Streaming)
- PlayInputContext struct holding the precomputed inputs (url, gen,
duration_hint, format/cache hints, optional reused chained bytes)
- select_play_input() async — handles all four branches: reused
chained bytes, psysonic-local://, ranged HTTP with format sniff,
and the legacy non-seekable streaming fallback
- url_format_hint() — the conservative URL→extension allowlist
helper, used to be inline in audio_play
audio_play is now focused on the orchestration above the source layer:
ghost-command guard, gapless pre-chain detection, bumping generation,
loudness/replay-gain math, fade-in setup, building the actual rodio
Source pipeline, sink swap with crossfade, spawning progress task.
audio/commands.rs: 980 → 676 LOC. play_input.rs: 385 LOC. Behaviour
preservation hinges on identical analysis-seed spawning, format-hint
fallback chain, range-detection logic, and gen-cancel checks — all
moved verbatim. Smoke test focus: psysonic-local playback, manual skip
during ranged HTTP, format-hint-less servers (legacy fallback path),
preload cache hit replay, manual skip onto pre-chained track.
Lift set_window_decorations + linux_webkit_apply_smooth_scrolling +
set_linux_webkit_smooth_scrolling out of app_api/core.rs into
app_api/platform.rs (~49 LOC). These are platform-tweak Tauri
commands (Linux WebKitGTK settings + Linux native-decoration toggle)
— logically separate from runtime control, telemetry, cli bridge,
or wire-UA setup.
This is the "platform" slice from cucadmuh's plan for splitting
core.rs's mixed concerns. Together with perf and cli_bridge, core.rs
is now down to runtime control (greet, exit_app), logging (3 cmds),
and UA setup — 56 LOC, focused.
Lift the four cli_publish_* Tauri commands (player_snapshot,
library_list, server_list, search_results) out of app_api/core.rs
into app_api/cli_bridge.rs. Each is a thin pass-through to
`crate::cli::write_*_response` — the renderer-side counterpart to the
file-based IPC layer in cli/exchange.rs.
This is the "cli-bridge" slice from cucadmuh's plan for splitting
core.rs's mixed concerns. Together with the earlier perf split,
core.rs is now down to runtime-control + platform + logging slices —
one possible follow-up but each is small enough to leave as-is for now.
app_api/core.rs: 122 → 99 LOC.
Lift PerformanceCpuSnapshot struct + the Linux /proc/stat parsers
(parse_proc_stat_line / read_total_jiffies / collect_proc_stats) +
the performance_cpu_snapshot Tauri command (~110 LOC) out of
app_api/core.rs into app_api/perf.rs. Self-contained CPU-usage
telemetry; nothing else in app_api references the helpers.
This is the "telemetry" slice cucadmuh's plan called out for splitting
core.rs's mixed runtime/platform/snapshot concerns. Other slices
(cli-bridge, runtime-control, platform window/scrolling) can follow.
app_api/core.rs: 235 → 122 LOC. lib.rs registration unchanged —
performance_cpu_snapshot is re-exported through `pub(crate) use
perf::*` in app_api/mod.rs.
Move the per-generation progress + ended-detection task (~205 LOC)
out of audio/commands.rs into audio/progress_task.rs. The task is now
a sibling submodule that both audio_play (commands.rs) and
audio_play_radio (radio_commands.rs) import as
`super::progress_task::spawn_progress_task`, replacing the previous
`super::commands::spawn_progress_task` cross-import that radio was
using as a workaround.
audio/commands.rs: 1185 → 977 LOC. Cleanup: dropped now-unused
AtomicU32 and AudioCurrent imports from commands.rs.
What's left in commands.rs is now just audio_play (~760 LOC) and
audio_chain_preload (~195 LOC) — the playback orchestrator proper.
Splitting those further is a bigger structural job than file moves.
Pull audio_preload (background fetch + analysis seed for the next
track in the queue) out of audio/commands.rs into
audio/preload_commands.rs (~67 LOC). It's a self-contained
fetch-and-cache flow — distinct from audio_chain_preload (which
constructs the gapless source chain) and audio_play (which starts
playback) so it makes sense to live alongside them rather than inside
the same file.
audio/commands.rs: 1241 → 1185 LOC. lib.rs invoke_handler updated.
Pull audio_pause / audio_resume / audio_stop / audio_seek (~210 LOC)
out of audio/commands.rs into audio/transport_commands.rs. They mutate
state.current on an already-running sink and coordinate radio
warm/cold resume — distinct concern from playback startup
(audio_play / audio_chain_preload / audio_preload).
audio/commands.rs: 1447 → 1240 LOC. Imports trimmed: TryLockError,
radio_download_task, RADIO_BUF_CAPACITY are no longer pulled in here;
they followed the transport block to its new home. lib.rs
invoke_handler updated.
The remaining commands.rs is now focused on the playback orchestration
proper (track + chain preload + initial preload + the shared
spawn_progress_task helper).
Pull audio_play_radio (~165 LOC) out of audio/commands.rs into
audio/radio_commands.rs. Live-radio playback differs from main track
playback: no gapless chain, no seek, no replay-gain, no preload —
collecting it on its own makes both modules easier to reason about.
The shared spawn_progress_task helper (still used by audio_play /
audio_chain_preload / audio_resume in commands.rs and now by radio)
moves from private to `pub(super)` so radio_commands can call it.
A follow-up could lift it into helpers.rs proper.
audio/commands.rs: 1614 → 1447 LOC. lib.rs invoke_handler updated.
Cleanup: dropped now-unused imports (`super::sources::*`,
`RadioLiveState`, `RadioSharedFlags`) from commands.rs.
Pull the six configuration setters out of audio/commands.rs into
audio/mix_commands.rs (~170 LOC):
- audio_set_volume (with replay-gain-aware ramp)
- audio_update_replay_gain (resolves cache-backed loudness, computes
gain, ramps sink, emits NormalizationStatePayload)
- audio_set_eq (10-band gains + pre-gain)
- audio_set_crossfade
- audio_set_gapless
- audio_set_normalization
These are pure AudioEngine state mutations + (for normalization) an
ipc emit. They don't drive playback; collecting them in one place
makes commands.rs more focused on the playback orchestrators.
lib.rs invoke_handler updated. audio/commands.rs: 1784 → 1614 LOC.
Pull autoeq_entries + autoeq_fetch_profile out of audio/commands.rs
into audio/autoeq_commands.rs (~52 LOC). They proxy autoeq.app +
GitHub raw content via Rust to bypass WebView CORS — pure HTTP-fetch
flow with no playback state coupling, so they don't need to live next
to the playback orchestrator.
lib.rs invoke_handler updated to register them under
`audio::autoeq_commands::*`. audio/commands.rs: 1830 → 1784 LOC.
Hotspot J first slice: split the device-listing + device-selection
commands out of audio/commands.rs (1912 LOC monolith) into a new
audio/device_commands.rs (~95 LOC). Moved:
- audio_canonicalize_selected_device
- audio_list_devices_for_engine (kept pub for cli/exchange.rs callers)
- audio_list_devices
- audio_default_output_device_name
- audio_set_device
audio/mod.rs re-exports audio_default_output_device_name +
audio_list_devices_for_engine from the new submodule. The four
#[tauri::command] device handlers are registered in lib.rs run() under
their new path `audio::device_commands::*`.
audio/commands.rs goes from 1912 → 1830 LOC and drops its
`use super::dev_io::*` since playback/radio/EQ don't touch device
enumeration. Behaviour-preserving — same Tauri commands, same dev_io
helpers, same selected_device + stream_handle mutations.
Pull the Linux-only single-instance D-Bus stack out of cli/mod.rs into
a new cli/linux_forward.rs:
- tauri_identifier (reads identifier from embedded tauri.conf.json)
- single_instance_bus_name + single_instance_object_path (D-Bus path
derivation matching tauri-plugin-single-instance)
- linux_bus_name_has_owner (zbus NameHasOwner wrapper)
- linux_is_primary_instance_running (pub)
- LinuxPlayerForwardResult enum (pub)
- linux_try_forward_player_cli_secondary (pub) — the heavy lifter:
forwards argv via D-Bus ExecuteCallback, then for list/search
commands reads the response file and prints to stdout
The whole submodule is `#[cfg(target_os = "linux")]` gated at the mod
declaration so non-Linux builds don't see it at all (cleaner than the
previous per-item gating).
Public surface preserved: lib.rs and main.rs keep importing
cli::linux_is_primary_instance_running and cli::LinuxPlayerForwardResult
unchanged via the linux-only pub use re-export.
cli/mod.rs: 696 → 550 LOC. Behaviour-preserving — same D-Bus protocol,
same response-file polling, same stdout output strings.
Pull the JSON response-file layer out of cli/mod.rs into a new
cli/exchange.rs:
- cli_*_path helpers (snapshot, library, server, search, audio_device)
— XDG_RUNTIME_DIR-aware path resolvers
- write_cli_snapshot
- write_library_cli_response, write_server_list_cli_response,
write_search_cli_response, write_audio_device_cli_response
- read_*_cli_response_blocking pollers (private to cli)
- print_*_cli_stdout wrappers (private to cli; thin "JSON or human"
switches around the presenter functions)
Externally-visible names (cli::write_cli_snapshot, the four
write_*_cli_response, the path helpers) preserved via `pub use
exchange::*` re-export at the cli root. Internal-only functions
demoted to `pub(super)`.
cli/mod.rs: 895 → 696 LOC. Behaviour-preserving — same file paths,
same atomic write-via-tempfile pattern, same poll intervals + ready
predicates.
Pull the JSON→stdout formatter layer out of cli/mod.rs into a new
cli/presenters.rs:
- print_library_human, print_server_list_human, print_search_human
- print_info_human + its helpers sorted_kv, value_inline
- print_audio_devices_human (kept as `pub` and re-exported from mod.rs)
Each takes a `serde_json::Value` and writes to stdout. No mutation,
no IO besides println, no inward calls back into mod.rs. cli/mod.rs
imports them via `use presenters::{...}` so the existing call sites
(print_*_cli_stdout wrappers, run_info_and_exit) read unchanged.
cli/mod.rs: 1151 → 895 LOC. Behaviour-preserving — same output
strings, same column separators, same scope handling for search.
Convert the 1485-LOC cli.rs into a cli/ module directory and lift the
parse-only layer into its own submodule. cli/parse.rs now owns:
- the CliCommand / PlayerCliCmd / RepeatCliMode / SearchCliScope /
MixCliMode enums
- the CliActionRegistry parser (reads shortcutActions.ts at startup)
- all wants_* flag helpers (--version, --info, --logs, --tail, --json,
--quiet, --follow, logs_tail_lines)
- parse_cli_command + parse_player_cli_at + parse_repeat_mode
cli/mod.rs keeps the rest for now: print_*_human presenters, exchange
write_/read_ functions, run_info_and_exit, run_tail_and_exit, Linux
single-instance + IPC forwarding, describe_cli_command,
emit_player_cli_cmd, handle_cli_on_primary_instance. Those are the
next split candidates (presenters / exchange / linux_forward).
Public surface preserved via `pub use parse::*;` so main.rs and lib.rs
keep importing cli::wants_version, cli::parse_cli_command etc. unchanged.
cli/mod.rs goes from 1485 → 1151 LOC. Behaviour-preserving — same
parsing rules, same compile-time include of shortcutActions.ts, same
registry-driven verb resolution.
Move tray-related state holders out of lib.rs into a new
src-tauri/src/tray_runtime.rs:
- TrayState / TrayTooltip type aliases
- TrayPlaybackState newtype
- TrayMenuItems / TrayMenuItemsState
- TrayMenuLabels (+ Default impl) / TrayMenuLabelsState
- tray_state_icon helper
The actual tray builder + Tauri commands (try_build_tray_icon,
set_tray_tooltip, set_tray_menu_labels, toggle_tray_icon) already
lived in lib_commands/sync/tray.rs and continue to reach the types
through the existing super::* re-export chain.
lib.rs goes from 549 → 486 LOC. Behaviour-preserving — same managed
state, same Tauri State<T> registrations in run().
Move ~440 LOC of analysis-queue plumbing out of lib.rs into a new
src-tauri/src/analysis_runtime.rs:
- AnalysisBackfillQueueState/Shared + worker loop + lazy-init
- AnalysisCpuSeedQueueState/Shared + worker loop + lazy-init
- submit_analysis_cpu_seed
- emit_analysis_queue_snapshot_line + analysis_queue_snapshot_loop
- WaveformUpdatedPayload (only used internally)
External callers keep their existing paths via `pub(crate) use
analysis_runtime::*` re-export at the lib.rs root. Direct accesses to
the static globals from app_api/analysis.rs are replaced with a new
prune_analysis_queues(keep) function so the statics stay private.
lib.rs goes from 999 → ~540 LOC. sync_cancel_flags stays put (sync
domain, separate concern). Tray-related types stay too — they're a
separate split candidate.
Behaviour-preserving: same workers, same queues, same events, same
return shapes. cargo check clean.
Both count_mono_frames_from_audio_bytes and decode_scan_pcm started
with the same ~25 LOC of Symphonia setup (Cursor → MediaSourceStream →
probe → track-select with two fallbacks → codec_params → decoder).
Extract that into open_decode_session(bytes) returning a DecodeSession
struct (format + decoder + track_id + timeline_hint).
Behaviour delta: count_mono_frames_from_audio_bytes previously failed
silently if the decoder couldn't be built (.ok()?). Now it logs the
same `[analysis] decoder make failed: …` line that decode_scan_pcm
already emitted. Same return value (None), only an extra debug-print.
Net: -38 +32 in compute.rs. Both call sites now read as a single line
of setup followed by their loop.
Pull the duplicated `subsonic UA + timeout client builder` and
`stream → .part → rename → cleanup` flow out of cache/offline.rs,
sync/device.rs, and sync/batch.rs into a new lib_commands/file_transfer
module.
- subsonic_http_client(timeout): standard UA + single timeout. Used by
the 3 simple-timeout sites (offline track, sync_track, batch sync).
download_update / download_zip / fetch_netease_lyrics keep their own
builders since they need separate connect+overall timeouts or extra
headers.
- stream_to_file(response, dest): chunked HTTP body → file (moved from
cache/offline.rs, where it was the de-facto shared helper anyway).
- finalize_streamed_download(response, dest, part): stream to .part
then rename, with best-effort cleanup of .part on any failure.
Behaviour delta: the offline + device single-track flows now also clean
up an orphan .part file when the final rename fails — previously only
the batch sync did this. Strictly safer; no observable change on the
success path.
downloads.rs (download_update / download_zip) keeps its inline progress
chunk loops — those emit per-chunk progress events with flow-specific
payload shapes and intervals; sharing them would force a callback API
that's heavier than the duplication it removes.
Net delta: -53 LOC across the 3 call sites.
Lift the duplicated `dir_size` walker and the empty-directory upward-prune
loop out of cache/offline.rs and cache/hot.rs into a shared
cache/fs_utils.rs module. Behaviour-preserving: same boundary semantics,
same skip-on-error policy, same `remove_dir`-only pruning (never
remove_dir_all).
Net delta: -76 LOC across cache/, single source of truth for cache cleanup.
* feat(song-info): show absolute file path on Navidrome via native API
Subsonic's `getSong.view` returns at most a relative path (or none on
Navidrome), so the Path row in the Song Info modal stayed empty for
most users. Feishin and the Navidrome web client surface the full
server-side path by hitting Navidrome's native `/api/song/{id}` instead.
Added an `nd_get_song_path` Tauri command that logs in to the native
API with the active server's credentials, fetches the song, and returns
the `path` string. Wired it into `SongInfoModal`: the Subsonic
`getSong` call still drives the rest of the dialog, and the native call
runs in parallel only when the active server's identity is
"navidrome". When it returns a path, that absolute path replaces the
relative Subsonic value; native-API failures are silent and the modal
falls back to whatever Subsonic provided.
No token cache yet — the modal is opened occasionally enough that one
fresh login per call is fine.
Closes discussion #479.
* docs: changelog entry for PR #504
Logs the Navidrome native-API absolute file path support for the
Song Info dialog in v1.46.0 "## Added".
* docs: settings contributor entry for PR #504, drop competitor mention
Adds the song-info absolute path bullet to Psychotoxical's contributors
list in Settings, and rewords the existing CHANGELOG entry so neither
text references competing clients.
* fix(preview): keep preview sink volume in sync with player slider (#498)
The Rust preview sink had its volume set once at audio_preview_play and
then never updated. audio_set_volume only ramps the main sink, so slider
movements during a preview had zero effect on the preview level. With
the default loudness normalization (-4.5 dB pre-analysis attenuation)
applied at start, even a 100% slider gives 1.0 × 0.596 × MASTER_HEADROOM
≈ 53% — matching the user-visible "fixed at around 50%" symptom.
- Add audio_preview_set_volume Rust command that updates the preview
sink if one is active (clamp + master headroom mirror the path used
in audio_preview_play).
- Extract the preview-volume calculation in previewStore into
computePreviewVolume() so startPreview and the new sync path share
one formula (slider value, plus the LUFS pre-analysis attenuation
the engine already applies to the main sink).
- Subscribe to playerStore at module level: when volume changes and a
preview is active, push the recomputed value to Rust. Auth /
normalization tweaks during preview are intentionally not synced —
preview is short and that case is rare.
Reported by netherguy4.
* docs: changelog entry for PR #502
Logs the preview-volume-slider sync fix in v1.46.0 "## Fixed".
* fix(tray): resume rendering on second-launch restore (#497)
The single-instance plugin callback that handles a second launch (e.g.
via desktop / start-menu shortcut while the main window is hidden in
the tray) was missing the RESUME_RENDERING_JS injection that the tray-
icon restore path already does.
When the main window is closed to the tray, PAUSE_RENDERING_JS sets
window.__psyHidden = true, marks <html data-psy-native-hidden="true">,
and zeroes --psy-anim-speed. A global CSS rule then pauses every
animation under <html>. Page wrappers across the app use
.animate-fade-in (animation: fadeIn ... both) which starts at
opacity: 0 — so when a route mounts after navigation the new wrapper
stays frozen at opacity: 0 and the page looks blank.
Restoring via the tray icon worked because that path injects
RESUME_RENDERING_JS before show(); only the second-launch path was
missing it. Mirror the same eval here so both restore paths are
consistent.
Reported by netherguy4.
* docs: changelog entry for PR #501
Logs the tray second-launch restore-rendering fix in v1.46.0 "## Fixed".
* feat(composer): Browse by Composer page (issue #465)
New library section listing every artist credited as composer on at
least one track, with a detail page showing all works they're credited
on in that role. Targeted at classical-music libraries where the
"recording artist" tag carries the orchestra and the "composer" tag
carries Bach / Mozart / Chopin.
Hits Navidrome's native /api/artist?_filters={"role":"composer"} for
the listing and /api/album?_filters={"role_composer_id":"…"} for the
works grid — Subsonic getArtist only follows AlbumArtist relations and
returns 0 albums for composer-only credits, so the native API is the
only path that works. Requires Navidrome 0.55+ (uses
library_artist.stats role aggregation); on older / pure-Subsonic
servers the page shows a one-line capability banner.
- Two new Tauri commands: nd_list_artists_by_role +
nd_list_albums_by_artist_role, generic over participant role so
conductor / lyricist / arranger pages are trivial to add later.
- Composers grid: text-only compact tiles (name + participation count
pulled from stats[role].albumCount). No avatars — composer libraries
carry no useful imagery and the listing endpoint exposes no image
URLs anyway.
- ComposerDetail: hero with Last.fm bio (via getArtistInfo2) plus the
full work grid, with a graceful fallback when the artist has no
external info synced.
- Sidebar entry default off (Feather icon) — opt-in for the niche
classical use case.
- nd_retry backoffs widened from [500] to [300, 800, 1800] — helps
every nd_* call survive intermittent TLS-handshake-EOF errors that
some reverse-proxy setups produce when keep-alive pools churn.
- Distinguishes "server can't do this" (HTTP 400/404/422/501) from
transient errors so the capability banner only fires when the server
actually rejects the request shape; everything else gets a retry
button.
- i18n in all 8 supported locales.
* fix(composer): address review feedback on detail page + role queries
- Re-fetch ComposerDetail when music-library scope changes; previously
the album grid stayed stale until navigation while the list refreshed.
- Thread library_id through nd_list_artists_by_role and
nd_list_albums_by_artist_role so role queries respect the active
Navidrome library, matching the Subsonic musicFolderId already piped
through libraryFilterParams().
- Fix CachedImage cache-key mismatch on ComposerDetail: a Last.fm header
image was stored under the Subsonic cover-art key, aliasing cache
entries and risking cross-source pollution.
- Consolidate the two contradictory composer-imagery comments in
Composers.tsx into a single accurate one (the older one referenced an
Images toggle that was never implemented).
- Align openLink toast duration with ArtistDetail (1500ms -> 2500ms).
* fix(composer): keep bio across scope changes, add share, degrade gracefully
Three remaining items from the latest review pass on the composer flow.
1. Bio survives a music-library scope change.
The previous fix added musicLibraryFilterVersion to the load effect,
but that effect also did setInfo(null) while the getArtistInfo effect
still depended on [id] alone — so a scope bump on the open page
wiped the bio without re-fetching it. Move the info reset into the
bio effect (keyed on id) and out of the load effect: the album grid
still refreshes on scope change; the Last.fm header image and
biography survive untouched, since both are library-independent.
2. Composers join the share pipeline as a first-class entity kind.
Extend EntityShareKind with 'composer' (and isEntityKind), branch
applySharePastePayload to validate via getArtist (same id pool) and
navigate to /composer/:id, and wire a Share button into
ComposerDetail. A pasted composer link now opens the composer view
instead of the artist view, matching what was copied. i18n added in
all 8 locales (sharePaste.composerUnavailable, openedComposer;
composerDetail.shareComposer, unknownComposer).
3. Partial server failure no longer hides the works.
If getArtist rejects but ndListAlbumsByArtistRole succeeds, the page
used to show full "not found" despite having data to display. Switch
the not-found gate to require both empty (`!artist && !albums`) and
render a degraded header (placeholder name, no Wikipedia / favourite
/ share / Last.fm image) when only metadata is missing.
* fix(composer): right-click share copies a composer link, not an artist link
The context menu opened from a composer card / row uses type='artist'
because every composer-action (radio, favourite, rating, add-to-playlist)
is identical to the artist counterpart — they share an id space and a
backend representation. Sharing was the one exception: the "Share Link"
entry produced a 'psysonic2-' string with k='artist', so a paste opened
/artist/:id even though the user came from /composers.
Add an optional shareKindOverride to openContextMenu (default: undefined,
preserves existing behaviour) and have the artist-typed branch consult
it when calling copyShareLink. Composers.tsx now passes 'composer' on
both right-click sites; nothing else changes downstream because the
override only affects the share kind.
* polish(composer): show Last.fm avatar even without server metadata
Two minor follow-ups from the latest review.
- ComposerDetail: drop the `&& artist` guard on the header-avatar render
path. info?.largeImageUrl can resolve through getArtistInfo(id) without
ever needing the SubsonicArtist record, so the previous gate hid a
perfectly good Last.fm portrait whenever getArtist failed but the
bio fetch succeeded. Replace artist.name with displayName so the
alt / aria-label degrade to the localised "Composer" placeholder
instead of empty strings.
- copyEntityShareLink: doc comment now mentions composer alongside
track / album / artist.
* fix(composer): derive Last.fm cache key from route id, not from artist record
Follow-up to the previous polish: the avatar render path no longer
requires `artist` to be populated, but the cache-key gate still did. So
when getArtist failed but getArtistInfo returned a Last.fm portrait, the
key fell through to coverKey — which is empty without an artist record,
re-creating the very aliasing bug the earlier Subsonic-vs-Last.fm fix
was meant to close.
Switch the Last.fm branch to the route id (same id namespace as the
SubsonicArtist record), so the key stays stable whenever Last.fm art is
shown, independent of getArtist succeeding.
* docs: CHANGELOG + Contributors entry for composer browsing (PR #487)
* fix(analysis): prune stale backfill jobs and limit prefetch window
Drop pending backfill and cpu-seed jobs that are no longer in the active playback queue, and add debug counters for pruned work. Limit loudness backfill scheduling to the current track plus the next five tracks to prevent runaway queue growth in dev sessions.
* chore(analysis): remove unused loudness prefetch parameter
Drop the now-unused incoming-tracks parameter from the loudness prefetch helper and update internal call sites to match the current queue-window scheduling logic.
* docs(changelog): document analysis queue control fix (#480)
Add a short 1.46.0 Fixed entry describing stale backfill pruning, the current+5 loudness backfill window cap, and debug prune counters for diagnostics.
* docs(contributors): add cucadmuh entry for PR #480
Logs the analysis-queue prune + loudness backfill window cap in the
Settings → System → Contributors list.
* fix(player): freeze main seekbar during track preview
Preview pauses the main sink in Rust while isPlaying stays true in the
store, so WaveformSeek's interpolation rAF must not advance progress.
* fix(audio): recover output after sleep and stalled streams
Add platform-specific post-sleep recovery hooks for Windows and Linux, and add a watchdog that reopens the output stream when playback is active but sample progress stalls, so audio can recover without restarting the app.
* fix(ui): remove card hover lift and smooth artwork zoom
Remove vertical hover translation from album and artist cards, and move image fade transition out of inline styles so cover zoom uses CSS timing consistently.
* fix(player): prevent seekbar jump after preview ends
Reset interpolation anchor timing when preview freeze state changes so the main seekbar does not momentarily jump forward before resyncing.
* fix(audio): reduce false watchdog recoveries and add diagnostics
Arm stalled-output recovery only after long poll gaps that suggest sleep/resume, and add detailed watcher logs for arm/clear/trigger paths to diagnose unintended stream reopens.
* chore(ui): drop card GPU hints and clarify macOS sleep scope
Remove translateZ and will-change hints from album and artist cover images to avoid per-card compositing overhead on software-composited Linux paths, and document why post-sleep recovery hooks currently target only Windows and Linux.
* docs(audio): document intentional Win32 callback pointer lifetime
Add inline rationale for the two Box::into_raw pointers in Windows suspend/resume registration so future maintenance does not treat the process-lifetime pointers as accidental leaks.
* docs(changelog): summarize playback stability updates for PR #476
Add a high-level changelog entry for preview seekbar fixes, sleep/wake audio recovery hooks and watchdog diagnostics, and card-hover stability adjustments from PR #476.
* docs(contributors): add cucadmuh entry for PR #476
Logs the post-sleep audio recovery, preview-seekbar fixes and card
hover stability work in the Settings → System → Contributors list.
* chore: upgrade dependencies and migrate playback to rodio 0.22
Bump npm and Rust crates; adapt symphonia decoding, ringbuf 0.5, lofty tags,
and discord-rich-presence usage. Use native rodio Player/MixerDeviceSink and
cpal device descriptions; drop the unused cpal patch. Align Vite 8 build
targets and chunking; remove redundant dynamic imports and fix hot-cache debug
logging imports.
* perf(build): lazy-load routes and restore default chunk warnings
Lazy-load all routed pages with React.lazy to shrink the main bundle; wrap root
Routes in Suspense for lazy Login. Drop chunkSizeWarningLimit override so Vite
uses the default 500 kB threshold.
* fix(windows): tray double-click without spurious menu; clean unused import
Disable tray menu on left mouse-up on Windows so a double-click to hide the
main window does not immediately reopen the context menu (tray-icon default
menu_on_left_click). Gate std::fs in app_api/core behind cfg(linux) for
/proc-only code so Windows builds stay warning-free.
* fix(sidebar): preserve new-releases read state under storage cap
When merging seen album ids, keep the current newest sample first so the
500-id localStorage limit does not truncate freshly marked reads and bring
back the unread badge.
* fix(audio): hot-cache replay, analysis no-op skips, playback source UI
Retain stream_completed_cache across audio_stop so end-of-queue replay can
use RAM promote or disk hot file instead of re-ranging HTTP.
Add cpu_seed_redundant_for_track gate before file/bytes seeds and local-file
spawn; emit analysis:waveform-updated only on Upserted. Ranged/legacy promote
checks generation after await before filling the slot.
Frontend: promote on same-track and cold resume; set currentPlaybackSource on
resume, queue undo restore, and gapless track switch so cache/stream icons stay
accurate. Import tauri::Manager for try_state in audio_play.
* fix(ts): narrow activeServerId for hot-cache promote calls
promoteCompletedStreamToHotCache expects a string; bind non-null server ids
in repeat-one, playTrack prev/same-track, and cold resume paths so tauri
production build (tsc) succeeds.
* fix(player): handle same-track hot-cache promote promise chain
Add .catch for promoteCompletedStreamToHotCache → runPlayTrackBody so sync
throws and unexpected rejections do not surface as unhandled in DevTools;
reset defer-hot-cache prefetch and isPlaying on failure.
* chore(nix): sync npmDepsHash with package-lock.json
* chore(release): finalize 1.46.0 CHANGELOG with PR #463 links
Document the release with full GitHub PR #463 on every subsection so
entries stay attributable if sections are reordered. Fix ContextMenu
lines where dynamic imports were accidentally merged onto one line.
* docs(contributors): credit cucadmuh for #463
* feat(linux): optional native GDK for Nix gdk-session
Introduce PSYSONIC_ALLOW_NATIVE_GDK so main skips the default GDK_BACKEND=x11
pin when the Nix gdk-session wrapper sets the flag. Remove GDK_BACKEND from
the npm tauri:dev script so it does not override nix develop defaults.
* fix(ui): portal server switch menu above sidebar
Main column stacks below the sidebar (layout z-index), so an in-tree dropdown
could never win over the left nav. Render the menu via createPortal to
document.body with fixed coordinates, matching the library scope picker.
* feat(perf): add mainstage probe controls and cut WebKit repaint load
Add a dedicated performance probe surface for mainstage/home toggles and wire Linux CPU diagnostics to isolate expensive UI paths. Tune waveform drawing and Home artwork clipping/windowing so visible content loads immediately while reducing WebKit compositor pressure during playback.
* fix(perf): stop hero rotation when section is off-screen
Gate hero auto-rotation and backdrop crossfade by real viewport visibility using the actual scrolling ancestor. This prevents periodic 10-second CPU spikes from hidden hero updates while preserving normal behavior when the hero is visible.
* fix(perf): isolate player progress updates from mainstage diagnostics
Add probe toggles for PlayerBar waveform and live progress UI updates to confirm playback progress churn as the main CPU driver.
Restore Home artwork quality defaults and keep visual-degradation modes opt-in via debug flags only.
* fix(hero): resume background and autoplay after viewport return
Re-check hero visibility on focus/visibility changes and add a short recovery poll while off-screen so missed scroll/RAF events cannot leave hero animation paused.
Co-authored-by: Cursor <cursoragent@cursor.com>
* fix(perf): decouple playback progress from mainstage compositing pressure
Throttle audio progress delivery and route live seekbar timing through a lightweight progress channel to cut focus-time WebKit CPU spikes.
Add focused diagnostics in Performance Probe and restore hero/waveform behavior so visuals remain stable while profiling.
* fix(debug): open performance probe with Ctrl+Shift+D
Replace logo-triggered opening with a keyboard shortcut and keep logo purely decorative to avoid accidental probe activation.
* docs(changelog): document experiment/performance probe and playback work
Add an [Unreleased] section for the performance probe, throttled audio
progress IPC, snapshot-based live UI updates, WaveformSeek scheduling
over the same canvas bar renderer, Hero/Home rail fixes, and Linux/Nix
GDK dev ergonomics.
* perf(linux): add WebKit probe, throttle progress IPC, snapshot playback UI
Ship Performance Probe (Ctrl+Shift+D), Rust-throttled audio:progress, a
playback progress snapshot channel with coarse Zustand timeline commits,
Linux /proc CPU readout for the probe, Hero and Home rail artwork fixes,
Tracks SongRail windowing parity, MPRIS cleanup, gated perf counters, and
WaveformSeek paused-seek correctness. Documented in CHANGELOG for PR #452.
* docs(changelog): fold perf work into 1.45.0 and refresh date
Drop the separate 1.45.1 heading; keep PR #452 notes under 1.45.0 Added and
set the section date to 2026-05-04. Restore the safety preface before the
versioned sections.
* docs(changelog): order 1.45.0 Added entries by PR number
Sort the 1.45.0 release notes so subsections follow ascending PR id (390
through 452), with PR #452 last.
---------
Co-authored-by: Cursor <cursoragent@cursor.com>
* fix(audio): frame-align gapless-off track-separation silence
The 500 ms silence prepended between tracks when gapless playback is
disabled and the previous track ended naturally was built with
`Zero<f32>::new(ch, sr).take_duration(500ms)`. Rodio's `TakeDuration`
computes its sample count via integer-nanosecond division
(`1_000_000_000 / (sr * ch)`), which truncates: at 44.1 kHz / 2 ch
this emits 44103 samples = 22051.5 frames, half a frame short.
That half-frame leak shifts the next source's L/R parity in the
device frame stream. Multiple users have reported the next track
playing only on the right channel — exactly when gapless is OFF and
the previous track ended naturally (manual skip and album-first-play
bypass the silence prepend, which matches the reproducer report).
Replace with `SamplesBuffer::new(ch, sr, vec![0; frames * ch])`
where `frames = sr / 2`. Frame-aligned by construction, same
audible effect.
* docs(changelog): add #439 mono-channel fix entry
* chore(credits): add #439 to Psychotoxical contributions
* docs(changelog): strip @ from non-contributor mention in #435 entry
Plain-text 'zunoz on Discord' instead of '@zunoz' so GitHub does not
attribute the requester as a contributor on subsequent merges.
* feat(shortcuts): unify action-driven shortcut and CLI routing
Centralize shortcut action metadata in one TypeScript registry and route keyboard, global shortcut, mini-window, and CLI inputs through shared runtime handlers.
Keep CLI as an abstract transport layer by emitting player-command payloads without depending on shortcut definitions.
* feat(shortcuts): generate CLI action help from shortcut registry
Move no-arg player commands and their descriptions into the central action registry so CLI parsing and --player help are derived dynamically from one source of truth. Also route runtime action execution through the registry and remove duplicated shortcut runtime handling.
* feat(shortcuts): add new input actions and hidden F1 help binding
Add the requested input actions (search, advanced search, sidebar, mute, equalizer, repeat, now playing, lyrics, favorite current track) to the central shortcut action registry and wire runtime handlers for sidebar/equalizer toggles. Keep Help bound to F1 by default while hiding it from Settings input lists, and backfill persisted keybindings with new defaults so F1 works for existing users.
Requested by @zunoz (Discord community).
The squash-merge of PR #419 was performed against an outdated PR base
that predated several main-side refactors and features. The resulting
squash inadvertently re-introduced files that had already been removed
(`src-tauri/src/audio.rs` monolith, `app-icon.png`) and reverted main's
content for ~20 files (`src-tauri/src/lib.rs` decompose, `src/App.tsx`
animation-pause, `src/components/AlbumRow.tsx` headerExtra, etc).
This commit:
* Restores all collateral-damage files to their pre-#419 main state
(5662dafe), including the audio module split, lib decompose,
Cargo.toml `windows` dep, animation-pause-on-blur logic, and
`reducedAnimations` toggle plumbing in WaveformSeek/Settings.
* Keeps the genuine #419 work intact: tri-state duration toggle,
position counter, persistent Now Playing collapse, animated EQ
indicator (in QueuePanel.tsx, 8 locales, authStore, components.css).
* Merges authStore.ts so both `reducedAnimations` (from #426) and
`queueNowPlayingCollapsed` (from #419) coexist.
* Adds the `.eq-bars.paused` CSS rule manually since components.css
needed a 5662dafe base + the single #419 addition.
* Fixes a latent type error in OverlayScrollArea.tsx (`useRef<T>(null)`
is `RefObject<T>` / read-only under current `@types/react`; widened to
`useRef<T | null>(null)` so the existing `wrapRef.current = el`
assignment compiles).
Verified locally with `cargo check` and `npm run build` — both green.
* feat(queue): add ETA display, equalizer indicator and collapsible now playing
* deleted endsAt and showDuration strings, changed eta update to 30s
* feat(queue): ETA tooltip, persistent Now Playing collapse, EQ bar pause, remove redundant Play icon
* feat(queue): fold ETA into existing total/remaining toggle as third mode
The standalone ETA span next to the track counter is removed; instead the
clickable duration label in the queue header now rotates through three
modes per click: total → remaining → eta → total. Counter (N/M) stays
where it was.
ETA mode keeps the live-feel treatment from the original PR (accent
colour while playing, muted at 50% opacity when paused). The other two
modes use plain accent.
i18n: queue.etaTooltip removed (no longer a separate descriptive label),
queue.showEta added as the action tooltip ('Show estimated end time')
in all 8 locales — matches the showRemaining / showTotal pattern.
* docs(changelog): add #419 queue UX improvements entry
Adds the [1.45.0] / Added entry for this PR's queue panel refinements
(position counter, tri-state duration toggle including ETA, collapsible
Now Playing section, animated EQ indicator).
---------
Co-authored-by: Psychotoxical <171614930+Psychotoxical@users.noreply.github.com>
* fix(audio): promote WASAPI render thread to MMCSS Pro Audio on Windows
Wraps the outermost audio source in a `PriorityBoostSource` that calls
`AvSetMmThreadCharacteristicsW("Pro Audio")` on its first sample. The
cpal output-stream callback runs `Source::next` on the WASAPI render
thread, which is otherwise normal-priority and gets preempted under
WebView2 / DWM / GPU pressure — producing the audible click/stutter
reported in issue #334. No-op on Linux/macOS (PipeWire/rtkit and
CoreAudio promote their audio threads externally).
* fix(build): repair Windows compile after audio split + lib decompose
Two pre-existing build breakers on Windows that surfaced after the
`use super::*;` cleanup (9cc74a7) and the lib-decompose refactor
(cfeec22):
- `audio/device_watcher.rs` lost the `output_enumeration_includes_pinned`
import. Add it cfg-gated to `not(target_os = "linux")` since it's only
called inside the non-Linux pinned-device fallback path.
- `lib_commands/sync/tray.rs::is_tiling_wm()` is `#[cfg(target_os = "linux")]`,
but its Tauri command wrapper `is_tiling_wm_cmd()` is unconditional.
Add a `#[cfg(not(target_os = "linux"))]` stub returning `false`.
No behavior change on Linux. Restores `cargo build` on Windows + macOS.
* perf(ui): pause cosmetic animations when window loses OS focus
Companion to the WASAPI MMCSS fix in this branch — issue #334
reported audible audio stutter on low-end laptops, partly traced
to WebView2 still compositing CSS animations + waveform canvas at
full rate while the user has alt-tabbed into another app.
Existing `data-app-hidden` / `data-psy-native-hidden` flags pause
animations only when the window is fully hidden (minimized or
behind a tray). When the window is visible-but-unfocused — the
common "music in the background while I work" case — every infinite
keyframe + 60fps rAF loop keeps running.
Add a parallel `data-app-blurred` flag driven by `window.focus`/
`window.blur`, plus matching `__psyBlurred` on the global window
object for imperative checks. The CSS rule that pauses every
`animation-play-state` on hidden gets a third selector for blurred.
WaveformSeek's two 60fps rAF loops (animated styles + settings demo)
extend their existing `document.hidden || __psyHidden` short-circuit
to also include the new blurred flag, falling back to the same
400ms polling path that already exists for hidden.
Verified visually on the dashboard: alt-tab → `<html
data-app-blurred="true">`, mesh blobs in the fullscreen player +
marquee + waveform 60fps loop all pause. Click back into the
window → blob/marquee/waveform resume immediately.
Cross-platform: focus/blur events fire reliably on all targets, so
the optimization applies to Windows, Linux (incl. WebKitGTK with
software compositing where the savings are largest), and macOS.
* chore: remove stale app-icon.png from repo root
* perf(ui): add 'Reduce animations' toggle that caps animated seekbar styles to 30 fps
Off by default. When on, animated seekbar styles (pulsewave, particletrail,
liquidfill, retrotape) skip every other rAF and advance the animation timer by
a doubled delta, so wave speed is unchanged while draws are halved. Aimed at
low-end Windows GPUs where the 60 fps shadow-blur loop drives ~40 % WebView2
GPU usage even when focused.
Toggle lives directly under the seekbar style picker in Settings → Appearance.
i18n in all 8 supported languages.
* fix(preview): sync audio start, ring animation, and download timeout
Three coupled fixes for the track-preview engine:
1. Audio sync. `Sink::try_seek` was running on a worker thread after
`sink.append(source)`, so the sink began playing position 0 while
the seek was still iterating to the mid-track target. With the
30 s `take_duration` cap counting wall-clock from append, audio
could only become audible ~25% into the preview window. The seek
now runs on the bare source before append, then `take_duration`
wraps it — playback starts at the seek position with the cap
measured from there.
2. Ring animation gating. The CSS progress-ring animation was
bound to `is-previewing` (set on click), so the ring sprinted
ahead of any download/decode/seek warmup and didn't reset
cleanly when switching from one preview to another. Added an
`audioStarted` flag in `previewStore` that flips on
`audio:preview-start` from the engine; CSS animation is now
gated on `audio-started` instead. `is-previewing` still drives
tooltip/icon for instant click feedback. Same SVG is reused for
a 25%-arc rotating loading spinner while waiting for audio,
with a 150 ms delay so cached/short previews don't flash.
3. Download timeout. The shared `audio_http_client` caps at 30 s,
which aborts mid-download on multi-hundred-MB uncompressed
files (e.g. 18-min Hi-Res WAV ~600 MB). The preview engine now
builds a dedicated client with a 5 min timeout for the bytes
fetch. Watchdog still bounds the playback window at 30 s once
the audio actually starts.
Touches `audio/preview.rs`, `previewStore.ts`, `components.css`
plus the eight tracklist/player-bar components that render the
preview button.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* docs(changelog): add preview audio sync fix for PR #423
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Five lib_commands submodules either had `use super::*;` duplicated
during the split (core.rs, offline.rs, mini.rs — leftover line 3) or
didn't reference the parent scope at all (navidrome.rs, bandsintown.rs).
Removing them silences all five `unused import` warnings; cargo check
is now warning-free.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Split the large lib command module into nested app_api, cache, sync, and ui submodules while preserving command signatures and behavior. This keeps command responsibilities isolated and makes further extraction safer.
Move audio engine functionality from a single large file into focused Rust modules while preserving behavior and command surface. This keeps boundaries clearer for future extraction and the upcoming lib.rs split.
playback_state is only used for the Windows tray tooltip icon and the Linux
now-playing menu row; compiling it on macOS triggered an unused_variables
warning in CI.
* feat(ui): unify queue toggle handle behavior
Show the queue toggle in the header when the queue is collapsed and use a seam-aligned drag handle when it is open. Hide the seam handle while the main content is actively scrolling to reduce accidental interactions.
* feat(ui): add adaptive header search collapse behavior
Collapse header search to a magnifier when top controls get crowded and expand it as an overlay only while active. Use measured header space with hysteresis to avoid flicker and keep neighboring controls stable.
* chore(ui): remove leftover search prototype artifacts
Drop an unused icon import from live search and remove an unused header container-type rule left from an earlier layout experiment.
* feat(ui): persist sidebar and queue visibility state
Save left sidebar collapse and right queue open/closed visibility in local storage helpers so both panel modes are restored after app restart.
* feat(ui): unify player overflow menu behavior
Use a single overflow menu for click and wheel interactions, with a volume-only mode that keeps the same layout and volume controls as the full menu.
* feat(ui): add wheel seek controls to waveform
Apply 10-second wheel seek steps with trailing 1-second debounce and keep the waveform preview stable so the playhead moves smoothly during rapid scroll input.
* fix(now-playing): stabilize narrow dashboard layout
Switch now-playing responsiveness to container-based breakpoints and prevent stacked widgets from overlapping when width is constrained.
* fix(search): reduce collapse jitter and avoid header overlap
Add a short collapse-state cooldown to prevent threshold flicker and hide conflicting header controls while collapsed search expands as an overlay.
* fix(i18n): localize player overflow controls across locales
Replace hardcoded player overflow labels with translation keys and add the missing keys for all shipped locale files.
* fix(search): keep advanced control clickable in collapsed mode
Prevent focus loss on the advanced search button in collapsed overlay mode so its click handler consistently runs.
* fix(i18n): restore queue translation in offline library
Use the existing queue.appendToQueue key for the offline enqueue button tooltip and label instead of a missing key and hardcoded English text.
* fix(ui): apply overlay scrollbar to right-panel text tabs
Switch now-playing content, lyrics, and info panes to OverlayScrollArea and harden tour-item layout so long concert metadata stays within panel bounds.
* fix(ui): add unread indicator for new releases and guard sidebar drag clicks
Track unread new-release IDs per server/library scope and clear the badge when opening the New Releases page. Also prevent click-through navigation after sidebar drag release and keep related i18n/responsive sidebar-adjacent refinements in this snapshot.
* fix(ui): stabilize live dropdown layering and unread reset flow
Render the topbar Live dropdown via a portal so it consistently overlays sidebar layers. Rework new-releases unread tracking to handle library scope baselines, ignore stale refresh races, and mark items as seen after a 5-second stay on the New Releases page.
* feat(ui): add localized New badges for recently added albums
Show a theme-consistent New badge on album cards and album detail for albums created within the last 48 hours. Localize the badge label across all supported locales and centralize recency logic in a shared utility to avoid duplication.
* fix(album): prevent tracklist jump when entering multiselect
Move bulk selection actions from the tracklist body into the album toolbar next to the track filter. Keep selection controls stable in the header area so enabling multiselect no longer shifts the tracklist content downward.
* fix(tray): add playback-state badge and finalize queue handle tooltip
Show play/pause/stop icons in the Linux tray now-playing entry and persist state safely in Tauri managed state.
Also switch the queue-resize handle tooltip to the dedicated localized key across all locales.
* fix(header): prioritize search collapse before Live/Orbit labels
Make topbar compression deterministic by collapsing search first and compacting Live/Orbit labels only in sustained low-space mode.
Add sticky hysteresis-based header compact state to prevent oscillation while resizing.
* fix(ui): stabilize header compaction and show tray state icons
Prevent topbar flicker in the narrow-width range by tightening compact-mode thresholds, gating on real overflow, and removing width transitions from live search.
Also include playback state icons in tray tooltip text across platforms while preserving tooltip length limits.
* fix(tray): keep tooltip iconization Windows-only
Revert Linux tray tooltip/title fallback attempts and keep state icons only in Windows tray tooltips, while Linux continues to show playback state in the now-playing menu entry.
* fix(ui): restore queue resize response after overlay scroll interactions
Hide the queue handle while scrolling on both the main route viewport and the now-playing viewport, and clear stale thumb-drag state before starting queue resize.
Also ignore inactive/faded overlay thumbs in resizer suppression so horizontal pointer transitions no longer leave the queue seam unresponsive.
* docs(changelog): summarize ui-refinements branch features
Document the branch-level feature additions in 1.45.0 as separate changelog sections and group remaining branch-local fixes under a single polish entry.
* docs(changelog): add PR #397 references for ui-refinements
Attach PR metadata to the new 1.45.0 ui-refinement sections and the polish entry so release notes map directly to the merged branch discussion.
* feat(tray): show now-playing track in tray icon tooltip
Mirrors the current track to the tray tooltip ("Artist – Title") on every
play/pause/track change, falling back to "Psysonic" when nothing is playing.
The cached value survives tray hide/show via a new TrayTooltip state, and
is truncated to 127 chars for Windows NOTIFYICONDATA.szTip.
Closes#383
* feat(tray): linux fallback — disabled menu item shows now-playing track
AppIndicator on Linux has no hover-tooltip API, so the tooltip command was
a silent no-op there. Adds a disabled menu entry at the top of the tray
right-click menu that mirrors the same text. Updated in lockstep with the
Win/macOS tooltip via set_tray_tooltip.
* i18n(tray): localize tray menu labels
Adds a `tray` namespace in all 8 locales (en/de/fr/nl/zh/nb/ru/es) for
Play/Pause, Next/Previous, Show/Hide, Exit, and the Linux-only
"Nothing playing" placeholder.
Rust side: TrayMenuLabels state holds the cached translations and
TrayMenuItems holds the live MenuItem handles. New set_tray_menu_labels
command pushes new strings + applies them via set_text without rebuilding
the tray icon.
Frontend pushes the labels on mount and on every i18n.on('languageChanged').
* feat(player): preview-active state on play button (ring + stop icon)
Checkpoint: play button mirrors the inline preview button from tracklists
during preview playback — hollow circle, accent ring depleting over the
preview duration, Square (stop) icon. Click still resumes main playback,
which the Rust audio engine cancels the preview for.
i18n key player.previewActive in all 8 locales for tooltip + aria-label.
* feat(player): show preview track in player bar + smart stop semantics
The player-bar info cell (cover, title, artist) now mirrors the previewing
track during preview playback, with a small accent "Preview" pill above
the title and an accent top-border on the bar. Rating, fullscreen hint
and album/artist link clicks are suppressed while previewing — they
target the queued track, not the preview.
Stop semantics for the two transport buttons during preview:
- Big play button (Square+ring visual): stops preview, main auto-resumes
if it was playing before. Matches the tracklist preview-button behaviour.
- Small Stop button: new audio_preview_stop_silent Rust command — stops
preview AND leaves main paused, so "Stop = silence" actually goes silent.
previewStore now stores the full PreviewingTrack (id, title, artist,
coverArt) — the seven startPreview call sites pass it through.
i18n key player.previewLabel in all 8 locales.
* feat(audio): rust preview engine with secondary sink
Adds a parallel rodio Sink on the existing OutputStream for 30s
mid-track previews. Two new Tauri commands (audio_preview_play,
audio_preview_stop) plus three events (audio:preview-start /
-progress / -end). The main sink is paused with Sink::pause() and
auto-resumed on preview end iff it was playing beforehand.
* feat(playlists): migrate suggestion preview to rust audio engine
Replaces the HTML5 <audio> path with the new rust preview engine.
previewStore mirrors the engine's start/progress/end events so any
tracklist row can render preview UI from a single source of truth.
Spacebar redirects to stopPreview while a preview plays, hardware
mediakeys are silently dropped (Q5), and tray clicks cancel the
preview before forwarding the original action.
* feat(albums): inline play + preview buttons in tracklist rows
Track number stays static on hover instead of swapping to a play
icon — the dedicated Play and Preview buttons in the title cell
take over click-to-play and click-to-preview. Active+playing rows
keep the eq-bars (also on hover), active+paused rows fall back to
the static accent-coloured number.
Pilot for the wider rollout to other tracklists.
* feat(tracklists): roll out inline play + preview buttons
Mirrors the AlbumTrackList pilot across the remaining track-row
based lists: PlaylistDetail main tracks, Favorites, ArtistDetail
top tracks, RandomMix (both genre-mix and filtered-songs lists).
Track number stays static, the dedicated Play + Preview buttons
in the title cell take over click-to-play and click-to-preview.
* feat(settings): track preview toggle + configurable position and duration
Adds an opt-out switch and two sliders to Settings → Audio:
start position (0-90 % of track length, default 33 %) and preview
duration (5-60 s, default 30 s). The progress-ring animation
follows the duration via a CSS variable so the visual matches the
engine's auto-stop. Disabling the feature hides every inline
preview button via a single root-level data attribute, no per-row
conditional rendering required.
i18n keys added in all 8 locales.
* fix(audio): cancel preview when main playback (re)starts
audio_play, audio_play_radio, audio_resume and audio_stop did not
know about the parallel preview sink, so clicking Play on a track
that was currently being previewed left the preview running on top
of the freshly started main playback. New helper clears the resume
flag, bumps the preview generation, drops the sink and emits an
'interrupted' end event before any of those commands touches the
main sink.
* feat(settings): per-location track preview toggles
Splits the single trackPreviewsEnabled toggle into a master + 6 per-location
sub-toggles (suggestions, albums, playlists, favorites, artist, randomMix).
Master remains the kill switch; sub-toggles are only honoured when master
is on. Each tracklist container is marked with `data-preview-loc="<id>"` and
hidden via scoped CSS when the matching root attribute is "off".
startPreview now takes a location argument so the store can guard logic too.
i18n added in all 8 locales.
* fix(contextmenu): use ChevronsRight for Play Next to distinguish from preview
Introduce a dedicated --logs mode for CLI log viewing with --tail <lines> and
-f/--follow streaming from the normal/debug log channel, keep user-facing CLI
output free of timestamped log formatting, and update bash/zsh completions for
the new logs flags.
* feat(player): persist queue panel visibility across restarts
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* chore(audio): drop unused resolved_loudness_gain_db binding
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* fix(loudness): target sync, -14 pre-analysis ref, queue UI, and reseed
- Front: coalesce loudness refresh by target LUFS; replay-gain IPC dedupe keys
include norm target and effective pre-attenuation so TGT changes apply.
- Rust: placeholder gain before integrated LUFS uses pivot at -14 LUFS; UI gain
from effective trim; reseed loudness after delete when waveform cache would skip.
- Pre-analysis: store attenuation relative to -14 LUFS; engine and UI use an
offset for other targets; migrate legacy absolute values on rehydrate.
- Queue/Settings: Loudness/TGT labels vs value buttons; styles; i18n for help.
* fix(i18n): simplify loudness pre-analysis helper copy
Remove reference-target wording from loudness pre-analysis helper text and keep
only the effective adjustment shown for the current LUFS target in all locales.
The Discord Rich Presence module deliberately swallows every IPC error so
the app stays clean when Discord is not running. The downside: when the
status genuinely stops working (Discord client glitches, app id rejected,
socket pipe stuck), nothing is logged anywhere and we have no signal.
Add debug-build logging at every step of the IPC handshake and every send:
- try_connect: separate failure logs for new() vs connect(), success log
with the app id.
- discord_update_presence: log the rendered details/state on send, log
set_activity failures with the underlying error.
- discord_clear_presence: log clear success and clear failures.
- The pause-path clear_activity inside discord_update_presence likewise
logs failures.
All log lines are wrapped in #[cfg(debug_assertions)] so they compile out
of release builds — no runtime cost or log noise for end users.
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Persist waveform v4 as paired curves per bin (peak and mean absolute value),
invalidate cache rows with insufficient data shape, and render the seekbar as
0.7 * mean + 0.3 * peak for a denser, more stable contour.
Thanks to @peri4ko for proposing this waveform improvement.
Serialize heavy PCM seeding through a dedicated queue with optional priority
for the current track. Emit waveform-updated once per completed seed, fix
Lucky Mix waveform refresh tokens, redact Subsonic URLs in logs, and align
hot-cache prefetch with the queued path.
Six independent paths reach analysis_cache::seed_from_bytes for the same
track_id during a cache-miss play under LUFS:
- track_download_task legacy stream capture-complete
- ranged_download_task HTTP buffer full
- audio_preload bytes-ready
- psysonic-local file-read complete
- spawn_analysis_seed_from_in_memory_bytes (gapless reuse / preloaded /
stream cache)
- analysis_enqueue_seed_from_url (frontend backfill triggered by
refresh:miss while a playback path is mid-seed)
Whenever a track is streamed for the first time with loudness on, two of
these fire in parallel and Symphonia + EBU R128 decode the same buffer
twice — ~30 s of duplicate CPU per track, plus a redundant SQLite write
of an identical row. Reproduced on multiple cache-miss tracks (KjP… ranged
+ backfill, mPdz… preloaded bytes + backfill).
Coordinate at the funnel:
- New SeedFromBytesOutcome::SkippedConcurrent variant.
- SEEDS_IN_FLIGHT static (Mutex<HashSet<String>>) tracks track_ids whose
analysis is currently running. First caller into seed_from_bytes
wins the slot via HashSet::insert; later callers return SkippedConcurrent
immediately and let the winner publish the analysis:waveform-updated
event.
- SeedInFlightGuard (RAII) releases the slot on every exit path including
panics, so a crashing analysis cannot leak the marker.
- enqueue_analysis_seed in lib.rs distinguishes the SkippedConcurrent log
from the genuine "seed result" log so the frontend backfill case
(when it arrives second) doesn't read like a backfill failure.
Existing call-site match arms in audio.rs already use Ok(_) => {} for
non-Upserted outcomes — no changes needed there. enqueue_analysis_seed in
lib.rs already gates the waveform-updated emit on outcome == Upserted, so
the new variant is silent there too.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Psychotoxical
cucadmuh
Kveld.
Maxim Isaev