SamandClaude Fable 5.1
520b6dc2aa
- the party-leader crown on your target's portrait only ever appeared or vanished when you re-targeted them: `PLAYER_FLAGS_CHANGED` was one of eighteen events the migrated stock windows register that nothing here produced, and the census row naming it said "TargetFrame.lua — the AFK/DND badge", which sent the first look at the wrong window — 1.12 draws no AFK/DND badge on any unit frame and build 5875 has no `UnitIsAFK`/`UnitIsDND` at all (ours are a deliberate invention, 2043; the `<AFK>`/`<DND>` the era shows is the chat line's `arg6`), and the event's sole FrameXML consumer is `TargetFrame.lua:88-95` re-running `UnitIsPartyLeader("target")`, whose descriptor leg is bit `0x1` of this very dword; the handler `0x5ee990` XOR-diffs old against new at `0x5ee9b8` and fires 407 at `0x5eea3d` with no bit test in between and ABOVE the local-player GUID gate at `0x5eea93`, once per unit token naming that GUID through `0x515e50`'s fan-out (`arg1` = the token, no `arg2`, and a zero token count skips the fan-out so a stranger nobody targets signals nothing — free here, since `fire_transitions` only ever runs per token); so the trigger is the RAW dword, `UnitState::player_flags`, the `flags`/`UNIT_FLAGS` pattern one field over, because firing off the two bits the struct decodes would under-announce AFK, DND, GM, PvP-desired, helm, cloak and both play-time bits, and `prev.is_some()` keeps a unit's first snapshot silent as its CREATE runs no notify pass; reading the whole handler found an over-fire that was inherited, not invented: benilla fired `PLAYER_UPDATE_RESTING` on ANY flags delta because wow-re's rested-XP note said in as many words that `0x5eeaf2` fires "on every flags change, not only the resting bit", and the bytes say otherwise — `0x5eead0` `f6 45 fc 20 / 74 26` is `test byte [ebp-4],0x20 ; je 0x5eeafc` and the fire sits inside that arm — so every helm toggle, leadership pass and AFK flip had been announcing a resting change to the whole UI; narrowed to the resting bit's own edge, and the same handler's `0x5eeb65 test ah,0x30` arm becomes `PLAYTIME_CHANGED`, whose consumer `PlayerFrame_UpdatePlaytime` already had all three verbs it reads; the claim is corrected in wow-re by a four-derivation round that confirmed all three arms plus the CREATE question; three tests pin it — the producer with a control on `0x400`, a bit no field decodes; the over-fire regression on the real feed, where a helm bit moving fires neither self event; and the crown itself on the stock `TargetFrame.lua`, leadership passing to a held target lighting it with no re-target and the same event for `arg1 = "player"` not — a fixture that first failed on its opening assertion because `UnitIsPartyLeader`'s second leg is a raw GUID compare with no zero guard, so a default `guid: 0` equalled the zeroed leader of an empty group (verified 1.12: it is why `UnitIsPartyLeader(nil)` answers 1 while solo); the unproduced-events census goes 18 → 16 per window, 31 → 29 whole (2078)
- the weapon swing trail darkens with the scene instead of burning at its authored colour: 2076 shipped it unlit and said why — the draw callback turns lighting ON (EGxRs `0x0e = 1` at `0x6c6847`) over vertex format 7, position + packed colour with no normal, and the state bit alone does not say what fixed-function produces from that; the bytes now do: `0x5a1e30` sets DIFFUSE- and AMBIENTMATERIALSOURCE both to `D3DMCS_COLOR1` off the format's diffuse bit, so the vertex colour IS the material; the normal is explicitly disabled rather than stale (`0x592a60` replaces the format mask instead of OR-ing it, `0x59c100` issues `glDisableClientState(GL_NORMAL_ARRAY)` when the bit is clear, and D3D's format-7 descriptor carries −1 in the normal column), so there is no N·L and the diffuse half contributes nothing; and the global `D3DRS_AMBIENT` (EGxRs id 6) has zero writers image-wide and defaults to 0, an earned zero — leaving `out = (Σ enabled lights' ambient) × authoredColour + inherited emissive`, which is why this does not take the lit emitters' `clamp(ambient + diffuse·max(N·L, 0))` term, a sun the reference does not have; `weapon_trail.rs` multiplies the trail's RGB by `WowLighting.ambient`, folded on the CPU into the vertex colours because a per-draw constant rides the vertex stream on this lane, drawing as `EffectLighting::None` and at the authored colour until the first lighting resolve; two named approximations of the class the fishing line already carries — the scene ambient rather than a per-model light collector, so an interior trail takes the exterior ambient, and no emissive, since `EMISSIVEMATERIALSOURCE` is set nowhere and reproducing it would mean inventing a constant; the other half of 2076's open list was a deviation benilla does not have: it claimed the combat-over-combat fast path (`0x5fe43c` returning at `0x5fe48b`, before the trail latch's only read at `0x5fe48e`) was unmodelled and cost trails starting one play early — both wrong, because `0x60d835` arms a one-slot field with a plain `mov` so a second proc OVERWRITES the first and firing on every request would draw superseded arms with superseded colours on 23 of the 34 type-8 kits, and because the driver's request loop already fast-paths it (0406, `select::is_combat_anim` = `0x5fcc10`'s byte-decoded set) and `continue`s without playing, so the edge never rises; a negative asserted about our own code from having written it rather than looked, shipped to wow-re where it drove a build-it verdict for something already built — now pinned by a test that names both mechanisms; and `benilla-extract charprocs` prints `CharParamOne` beside every trail row: wow-re's note justified "not read by this arm" with "every shipped type-8 row carries 20.0 there", and over the 34 kits the column takes six values (22 at 20, six at 15, two at 25, two at 0, one at 17, one at 22 — Sinister Strike's kit 399 is a 15); nothing here reads it either way, but a census that hides the one column a note asserts about cannot catch the note being wrong, and wow-re's re-measure agreed value for value and found the column live at four sites in `0x60d7c0` for other proc types (2079)
- four options that ticked and did nothing — Show Tutorials, Tip of the Day, Profanity Filter, Disable Spam Filter — get their backing, and two of the reasons they were dead were wrong: none of the four was a row in our window at all; they sat in the transcribed `UIOptionsFrameCheckButtons` declaration table beside thirteen more whose features benilla lacks, and these four are exactly the ones where the feature existed and only the setting was missing; `SHOW_TUTORIALS` is a bare `{ index = 28 }`, which an earlier census read as "an index-only entry with no store at all" — wrong, the store is the three special-case arms `UIOptionsFrame_Load` (l.239), `_Save` (l.318) and `_SetDefaults` (l.645) keep for the key, `TutorialsEnabled()` reading the acknowledged bank and `ResetTutorials()`/`ClearTutorials()` writing it, verbs benilla has had since 1976, so the row was one `func` pair away, and the reference's `~=` guard comes with it because without it a Defaults click on a page where tutorials are already on would `ResetTutorials()` and re-arm every popup the player has dismissed; that census moves out of prose into a test checked both directions, the posture the CVar defaults already hold — a name the reference's table carries that nothing registers and no listed blocker fails, and a listed name that has since been registered fails too, twelve rows listed today with the feature each waits on; the two chat filters are PCRE over shipped DBCs, not word lists: `ChatProfanity.dbc` (2289 rows off `patch.MPQ` — `dbc.MPQ`'s copy is 1512) feeds the masker `0x4a1a60` and `SpamMessages.dbc` (28 gold-seller URL patterns; `dbc.MPQ` has none) feeds the predicate `0x4a1ca0`; Blizzard PATCHED PCRE's escape table (`.rdata 0x812330`) so `\<`/`\>` alias onto `\b`, and an unpatched engine makes them literals and silently kills the whole ASCII half of the profanity list while looking correct; CASELESS folds ASCII only, so pattern and subject are both ASCII-lowercased and matched case-sensitively rather than through `(?i)`, which would match `É` where the reference matches nothing; the masker is in place and length-preserving, one token per byte cycling `!@#$%^&*` from a process-global index never reset for the session, patterns in list order not text order — wow-re's oracle transcript reproduced against the real list, "you are a twat and a faggot" → "you are a &*!@ and a !@#$%^", the later word taking the earlier characters because its row (507) precedes the other's (530); two corrections to what this tree believed: `spamFilter` is armed for every chat type EXCEPT `CHAT_MSG_FILTERED` (`0x49aad2`'s `jne` skips the disarm, which the older note read backwards, so our comments claimed the filter bit only on the one type it is exempt from), and `profanityFilter` gates INSIDE the masker at `0x4a1a66`, covering all thirteen call sites rather than the guild MOTD alone; `CHAT_MSG_FILTERED` (`0x5B`) comes with them — it was on the "never `SMSG_MESSAGECHAT` wire" list and is the server's "your message was filtered" notice `ChatFrame.lua` renders at l.1405, dropped with a warning before and delivered now; the chokepoint runs spam, then mask, then fire, in the reference's order, since a dropped line must not consume mask indices; named rather than faked: the other twelve masker sites (mail, guild, item text, the send path) and their `dl = 1` memo, and the whisper leg's `CMSG_CHAT_FILTERED` report whose body is a wow-re DEFERRED with no vmangos handler; each row sits in 1.12's own seat — Profanity Filter closes the Basic Options Display column, Show Tutorials and Show Loading Screen Tips open the Help box, Disable Spam Filter is the last link of the Advanced chat chain and value-inverted, checked ⇔ `spamFilter == "0"` (2077)
- the loading screen shows the tip of the day: `GameTips.dbc` is 74 records in the 5875 enGB chain, off `patch.MPQ`, ids unsorted (the file opens on 396 and closes on 412), so the index the client persists is a record index in load order, never an id; `gameTip` holds the NEXT index, not the one on screen — the client reads it, shows that row and stores `read + 1`, so on-disk values run 1..=74 and never 0, and the reference's own `Config.wtf` reading `SET gameTip "34"` while showing row 33 is exactly what an off-by-one anchors on; the wrap is a clamp, not a modulo (`if (i < 0 || i >= count) i = 0` at `0x46b682`), the walk is sequential, and both guards — `showGameTips` off and the roster's level-zero flag — jump past the `CVar::Set` as well as the draw, so turning tips off freezes the cursor rather than burning through the table invisibly; it draws only on the glue→world transition, because the setter `0x406630` has exactly one caller inside `EnterWorld` and neither `SMSG_TRANSFER_PENDING` arm reaches it, so the pick hangs off the loading screen's world-entry raise and a portal or worldport screen carries no tip; placement is the reference's: `s = 515/(a·1024)` with `a` = aspect × 0.75 (not W/H, which inflates the wrap width by a third at 16:9, and not the FrameXML-length idiom, which is off by 1.25 at 4:3), the block at `(0.5 − s·0.5, 0.1)`, left-justified, body `0xd7c8c8c8` with an opaque black shadow at `(+0.001, −0.001)`, laid out once per raise between the backdrop and the progress bar; the one term deliberately not applied is `yoff = (1 − a)·0.5`, which wow-re left unreconciled (`0x406a60` already applies the letterbox as a viewport and `yoff` is the same term added again inside it, a possible double count on a display narrower than 4:3) — benilla's loading screen is a 4:3 content box by construction and at 4:3 the reference's own `yoff` is zero, so the tip sits at the reference's layout on the one aspect where the ambiguity cannot bite; the `|cffffd100Tip:|r ` prefix and the trailing CR-LF are in the data and the escapes are interpreted, so the gold "Tip:" is markup and the line is built through the markup tokenizer; the cursor persists through the VM's own CVar table (`write_host_cvar`), since the config file is composed from that table and a host write stopping at the knob would reset it every launch — a live run reads `74 game tips` → `tip 0 of 74` → `tip 1 of 74` with `gameTip = "2"` left in the config, and an earlier live run is what caught a `GameTipSetting` the test harness initialised and the real app did not, a panic no unit test could see (2077)
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
189 lines
13 KiB
TOML
189 lines
13 KiB
TOML
[package]
|
|
name = "benilla-app"
|
|
version.workspace = true
|
|
edition.workspace = true
|
|
license.workspace = true
|
|
description = "The Bevy app as a library: everything but the build-id stamp, which lives in the `benilla` launcher shim so a commit dirties ~30 lines instead of this crate (decision 0993)"
|
|
|
|
[features]
|
|
# **The dev/player seam** (decisions 0026 + 1173). Default-ON, so every ordinary `cargo run`,
|
|
# capture, probe and gate builds exactly what it built before. `--no-default-features` compiles
|
|
# the instruments — the debug panel, the perf HUD, the inspector, the capture harness and the
|
|
# probe fleet — physically OUT: not dormant behind a runtime flag, absent. That is the whole
|
|
# requirement a player client has of them, and a runtime toggle cannot meet it.
|
|
#
|
|
# The rule the seam enforces is one-directional: **dev may see anything; nothing may depend on
|
|
# dev.** It is enforced by `cargo build -p benilla --no-default-features` in `scripts/gates.sh`,
|
|
# not by anyone remembering it — 0026 adopted the rule in June 2026 with no failing build behind
|
|
# it, and twenty-four couplings had accumulated by August (1173).
|
|
default = ["dev"]
|
|
#
|
|
# The three benilla crates below are taken with `default-features = false` and their `dev`
|
|
# re-exported here (1175). Their `dev` is default-on too, so without the round trip a
|
|
# `--no-default-features` build gets the feature back through unification — and the binary
|
|
# carries the build machine's source-tree path after all. Same trap as the shim's, one level down.
|
|
dev = ["benilla-assets/dev", "benilla-formats/dev", "benilla-world/dev", "dep:bevy_egui"]
|
|
# Opt-in Tracy profiling: `cargo run --features tracy` streams per-system `info_span!` markers (on the
|
|
# hot systems) + render-pass CPU timings to a Tracy capture (decision 0026 — the dev/player seam;
|
|
# a `docs/howto/running.md` this once cited never landed).
|
|
tracy = ["bevy/trace_tracy"]
|
|
# Opt-in Chrome-trace profiling: `cargo run --features trace_chrome` writes a `trace-<ts>.json` of
|
|
# every system's per-run span next to the working dir. Tracy's offline twin — it needs no GUI and no
|
|
# live capture, so a per-system cost question is answerable from a terminal (and from a run someone
|
|
# else recorded). Aggregate it with `scripts/tracesum.py`.
|
|
trace_chrome = ["bevy/trace_chrome"]
|
|
|
|
[dependencies]
|
|
anyhow.workspace = true
|
|
# OS clipboard access for the UI EditBox copy/cut/paste chords (chat input). Already in the tree via
|
|
# bevy_egui; pinned to the same 3.x so there's one copy. Read on the main thread (the UI key feed is
|
|
# a NonSend system) — required for the macOS NSPasteboard path. Covers macOS, Windows, and Linux/BSD
|
|
# under X11; the Wayland backend is `smithay-clipboard` below (decision 0702).
|
|
arboard = "3"
|
|
benilla-ui = { path = "../benilla-ui" }
|
|
bevy.workspace = true
|
|
wgpu.workspace = true
|
|
# Optional: the dev overlays are its only consumers (the panel, the perf pill, the inspector
|
|
# card), so a player build does not link it at all (`dev` feature, decision 1173).
|
|
bevy_egui = { version = "0.39", optional = true }
|
|
# Zero-copy POD→bytes for the shared light storage buffer (one per-frame write_buffer). Already in
|
|
# the tree transitively; pinned to the same version as the dev-dep below.
|
|
bytemuck = { version = "1", features = ["derive"] }
|
|
# `f16` decode for the booth-target dump (`WOW_BOOTH_DUMP`): the booth render targets are
|
|
# `Rgba16Float` (decision 0804) and the PNG wants display bytes. Already in the tree via
|
|
# wgpu/bevy_image; declared so the call site is an honest dependency.
|
|
half = "2"
|
|
# The shipped FrameXML tree, compiled into the binary (decision 1175 §2): 61 files, ~2.8 MB, so the
|
|
# client cannot boot without its interface on a machine that has never seen our source. A proc
|
|
# macro plus a no-dependency runtime shim — it adds nothing to the graph at runtime, and it is the
|
|
# crate everyone uses for this. `assets/ui` is OUR content (MIT/Apache), not Blizzard's.
|
|
include_dir = "0.7"
|
|
# The 1.12 chat filters are PCRE over shipped DBCs (decision 2077; `text_filter`): 2289 profanity
|
|
# expressions and 28 spam ones, compiled once at startup and run over every chat line. `RegexSet` is
|
|
# exactly the shape that needs — one automaton over the whole list for the "can anything match at
|
|
# all" pass — and the crate is already in the tree transitively.
|
|
regex = "1"
|
|
# `getrusage` — whole-process CPU time (all threads), the perf probes' `cpu_ms`/`cpu_pct` metric
|
|
# (`perf::process_cpu_secs`). Wall-clock fps is unusable as a regression instrument on a machine
|
|
# running parallel sessions' builds; CPU-per-frame is, and it is also the number the Mac report
|
|
# is about ("250 % CPU at 59 fps"). Was dev-only; since 1857 the output backend needs it in every
|
|
# build too (`mach_absolute_time`, the host clock its meters read; `thread_policy_set`, the render
|
|
# thread's realtime policy).
|
|
libc = "0.2"
|
|
# Resample varying-size layer BLPs to a uniform size for the terrain texture array; `png` so we can
|
|
# encode the cursor image into data an NSImage can load (macOS hardware cursor).
|
|
image = { version = "0.25", default-features = false, features = ["png"] }
|
|
# The net↔ECS bridge: a background IO thread streams typed `SessionEvent`s to the ECS over a
|
|
# channel, and outbound `ClientCommand`s back. crossbeam's Sender/Receiver are `Send + Sync + Clone`
|
|
# (std's mpsc Sender is `!Sync`, so it can't be a Bevy resource without a mutex — the very thing this
|
|
# bridge deletes). The canonical Bevy background-thread→channel→system idiom. See decision 0006.
|
|
crossbeam-channel = "0.5"
|
|
benilla-assets = { path = "../benilla-assets", default-features = false }
|
|
benilla-formats = { path = "../benilla-formats", default-features = false }
|
|
benilla-protocol = { path = "../benilla-protocol" }
|
|
benilla-world = { path = "../benilla-world", default-features = false }
|
|
# The CVar config diff (`benilla/config.toml`, decision 0954): serde derive on the persisted
|
|
# shape, toml as its format. (serde_json left with the era manifest — 0978.)
|
|
serde = { version = "1", features = ["derive"] }
|
|
toml = "0.8"
|
|
# ECS-native physics: collider storage + broadphase BVH + shape-casts, with a thin custom kinematic
|
|
# character controller on top (decision 0009). Pinned 0.6: it targets Bevy 0.18 (verified single,
|
|
# unified bevy 0.18.1 in the tree); 0.7.0 jumps to Bevy 0.19 and would duplicate bevy. avian wraps
|
|
# parry under the hood, so adopting it lets us drop the *direct* parry3d dep once the bespoke
|
|
# CollisionWorld is gone. See docs… (decision 0007 superseded by 0009).
|
|
avian3d = "0.6"
|
|
# The glyph shaper + rasterizer for `FontString` text (decision 0068 §2, probe-verified in
|
|
# `probes/text-glyph`). Pinned to 0.16 — the exact version already in the tree transitively
|
|
# (`bevy_text`'s own dependency, see `Cargo.lock`) — so this is the *same* `cosmic-text` the decision
|
|
# already commits to, not a second copy. Used as a standalone shaper/rasterizer (`FontSystem` +
|
|
# `Buffer` + `SwashCache`) in `ui_text.rs`, never `bevy_text`'s `Text2d`/ECS integration (that path
|
|
# doesn't feed our custom z-ordered `UiQuad` stream).
|
|
cosmic-text = "0.16"
|
|
# The font database `cosmic-text` 0.16 uses (its `Source`/`ID`/`FaceInfo` appear in `ui_text.rs`'s
|
|
# public-in-crate types). Pinned to the exact version `cosmic-text` 0.16 depends on so there's one
|
|
# `fontdb` in the graph.
|
|
fontdb = "0.23"
|
|
# kira renders the mix; the device it plays through is OURS (`sound::output`, decision 1857) —
|
|
# kira's own `cpal` backend is not enabled on any target. On macOS that took cpal out of the graph
|
|
# entirely: it could neither report a lost default device nor rebuild on one, and it discarded the
|
|
# IO-cycle timestamp the deadline meters need. Off macOS cpal is back (1920) but as OUR device
|
|
# layer under the same backend, not as kira's — the ring, the render thread and the meters are the
|
|
# same code on every platform.
|
|
kira = { version = "0.12", default-features = false, features = ["mp3", "wav", "pcm", "adpcm"] }
|
|
# The realtime-audio allocation tripwire (decision 1857): `assert_no_alloc(|| …)` around the output
|
|
# IO callback and the render pass aborts a DEBUG build on any allocation inside them, so a
|
|
# realtime-safety regression fails a test or a smoke instead of reaching an ear. Release builds
|
|
# compile it out (`disable_release`, the crate's default) and carry no allocator wrapper.
|
|
assert_no_alloc = "1.1"
|
|
# kira's own decoder, named directly for one trait: the mixer feeds streams through its own
|
|
# `MediaSource` so kira's per-stream decode thread runs OUR code and can promote its QoS (decision
|
|
# 1109 — a bare-spawned decode thread starves under the world-entry burst and zero-fills the mix).
|
|
# No codecs of its own (kira's features enable those); must stay the version kira resolves — one
|
|
# symphonia in the graph, or the trait impl won't line up.
|
|
symphonia = { version = "0.5", default-features = false }
|
|
# The mix tap's audio-thread → writer-thread ring (decision 1112): the tap effect must never
|
|
# block or allocate on the render path. kira's own SPSC ring crate; kept on the version kira
|
|
# resolves so the graph carries one rtrb.
|
|
rtrb = "0.3"
|
|
mint = "0.5"
|
|
# The gossip greeting is a weighted random DRAW per menu-open, not a fixed pick (wow-re
|
|
# `gossip-npctext-law.md` — the reference re-rolls in its gossip handler), so the app needs a uniform
|
|
# source. Pinned to 0.8: the version `benilla-srp` already depends on, so this adds no crate.
|
|
rand = "0.8"
|
|
|
|
[dev-dependencies]
|
|
# The WMO portal-cull audit's raw-face instruments (`WOW_PIN_COL`): the pin probe compares the
|
|
# filtered walking-collision set against the group file's FULL face list with per-face MOPY flags,
|
|
# which only the raw parser exposes — benilla-formats' gathers are exactly the filtered views.
|
|
benilla-wmo = { path = "../benilla-wmo" }
|
|
|
|
# Linux/BSD: the Wayland clipboard (decision 0702). We build with bevy's `wayland` feature, so on a
|
|
# Wayland session winit gives us a native Wayland window and `arboard` — X11-only without its
|
|
# non-default `wayland-data-control` feature — finds no X11 CLIPBOARD owner and every paste fails.
|
|
# smithay-clipboard is the right crate for a windowed app: it drives the *standard* `wl_data_device`
|
|
# off our own `wl_display`/seat, which every compositor supports. arboard's own Wayland feature would
|
|
# not do: it speaks `wlr-data-control`, a privileged manager protocol GNOME/Mutter deliberately
|
|
# doesn't implement. `raw-window-handle` (pinned to the 0.6 already in the tree via bevy_window, so
|
|
# this adds no crate) is only needed to read the `wl_display` out of bevy's `RawHandleWrapper`.
|
|
[target.'cfg(all(unix, not(any(target_os = "macos", target_os = "android"))))'.dependencies]
|
|
raw-window-handle = "0.6"
|
|
smithay-clipboard = "0.7"
|
|
|
|
# macOS: drive a native NSCursor for a crisp hardware cursor. winit drives the cursor via AppKit
|
|
# cursor-rects, which our continuous Metal render loop doesn't honor on mouse-move (custom cursor /
|
|
# hide revert to the arrow); we bypass it with NSCursor + NSWindow::disableCursorRects. See
|
|
# docs/knowledge/camera-control.md.
|
|
[target.'cfg(target_os = "macos")'.dependencies]
|
|
# The audio device layer (decision 1857, `sound::output::coreaudio`): the HAL output unit through
|
|
# coreaudio-rs's `AudioUnit`, and the AudioObject property/listener API through the objc2 CoreAudio
|
|
# bindings — the two crates cpal itself stood on, now used directly. The versions are the ones the
|
|
# graph already carries; the feature sets are the ones cpal proved to compile against.
|
|
coreaudio-rs = { version = "0.13", default-features = false, features = ["audio_unit", "core_audio"] }
|
|
objc2-core-audio = { version = "0.3", default-features = false, features = ["std", "AudioHardware"] }
|
|
objc2-core-audio-types = { version = "0.3", default-features = false, features = ["std", "CoreAudioBaseTypes"] }
|
|
objc2 = "0.5"
|
|
objc2-app-kit = { version = "0.2", features = [
|
|
"NSApplication",
|
|
"NSResponder",
|
|
# NSApplicationActivationPolicy (bgwin's Accessory demotion) is declared in this header.
|
|
"NSRunningApplication",
|
|
"NSWindow",
|
|
"NSView",
|
|
"NSImage",
|
|
"NSCursor",
|
|
"NSEvent",
|
|
] }
|
|
objc2-foundation = { version = "0.2", features = ["NSArray", "NSData", "NSGeometry"] }
|
|
|
|
# Everywhere else: the same device layer over cpal (decision 1920, `sound::output::cpal`). 1857
|
|
# owns the mix-ahead ring, the render thread and the meters on every platform; only the last
|
|
# hop — find the default device, open a stream on it, hand back one buffer per cycle — is
|
|
# platform code, and off macOS that hop is ALSA/PipeWire, WASAPI or the host's own API. cpal is
|
|
# the layer that already speaks all of them; on macOS it is still not in the graph, because
|
|
# there its defects (a no-op error callback on the default device, a pinned `DefaultOutput`
|
|
# unit, the discarded IO-cycle timestamp) are exactly what 1857 went direct to escape. On an
|
|
# unsupported target cpal compiles to its Null host, which reports no device — the same silent
|
|
# posture the old `stub.rs` had, without a second file to rot.
|
|
[target.'cfg(not(target_os = "macos"))'.dependencies]
|
|
cpal = "0.17"
|