ludic-dev vkgen reads vk.xml into runtime/native/vk_api.ludic (constants, every struct's
<Struct>_sizeof and <Struct>_<field> offsets, one extern per command) and vk_thunks.ll.
Every size and offset was compiled against the SDK's C headers; `ludic-dev test` checks the
tracked files against the registry wherever the Vulkan SDK is installed.
vk_win.ll (vulkan-1.dll) and vk_mac.ll (libvulkan.1.dylib, MoltenVK) open the loader at run
time, so a program built with Vk.* starts on a machine without Vulkan. ludicc and ludic
build link both for any program that uses Vk.*. The seeds are regenerated for the new
compiler.
vk_probe reports what a machine's Vulkan can do; vk_compute dispatches a Slang compute
shader and reads the picture back, clean under the validation layer on an RTX 3070 Ti and
on an M4 Pro through MoltenVK.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
http_win.ll implements http.ll's hs_* contract over WinHTTP: the request is a
record built on the game thread, the whole exchange runs on a worker thread,
TLS uses the system's certificate checks, and headers are answered from the
kept request handle. ludicc links it with winhttp instead of refusing Http.* on
Windows.
win32.ll decodes the splash and App.set_icon images with WIC (ole32): the
splash is a topmost borderless window at the artwork's size in points times the
display scale, composited over its background colour; the icon becomes an
HICON set on the window now or when win_open makes one.
Verified on the PC: examples/library/http.ludic prints its expected output, a
real GET to https://git.workshopsoft.io/ returns 200 with its body and
Content-Type, and the bundled Maroon Lake shows its splash centred at launch and
its icon in the title bar.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The CLI's shell commands go through shell(), which is run() on POSIX and a
scratch script handed to Git for Windows' bash on Windows (exit codes read
directly there). compile_app links through `ludicc -o` on Windows, ludic run
starts the .exe, and ludic-dev build and ensure_ludicc assemble
selfhost/ludicc.win.seed.ll, which ludic-dev reseed now writes beside the
macOS seed. ludic bundle makes build/<name>/ with a GUI-subsystem exe carrying
the .ico beside `app icon` as an llvm-rc resource, game.lpak and packs.index;
ludicc gains --gui and --link, and quotes its whole link line for cmd.exe.
Verified: ludic-dev test 135/135, selfhost-test 32/32; on the PC a checkout
bootstraps from the Windows seed, ludic-dev build makes the toolchain, and
`ludic bundle` makes build/Maroon Lake/, which runs from its own folder.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The arrow Key constants folded to the codes of w/s/a/d, which is what the
per-frame key reports for an arrow, but cocoa.ll has always stored an arrow in
the held-key set under 128-131. So Input.key_down(Key.Up) read the W bit and
Input.move_i's arrow half never moved anything. Input.key keeps its WASD alias,
so games comparing it with 'w' (snake, chronorift) still take the arrows.
New example input_arrows.ludic, checked by ludic-dev test.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
audio_win.ll implements audio.ll's snd_* contract over XAudio2 from IR: a
source voice per clip, restart on play, LoopCount for loops, SetVolume,
SetFrequencyRatio and a balance pan through SetOutputMatrix, with the COM
slots taken from the SDK's xaudio2.h. Clips are RIFF WAVE read through
lp_pak_open (weakly referenced), so a packed sound loads directly. ludicc links
it with xaudio2 and ole32, and silences xinput.lib's importeddllmain warning.
Verified: ludic-dev test 135/135, selfhost-test 32/32; on the PC a harness
loads, plays, pans, loops and stops clips, and Maroon Lake windowed reports
"sound: 31 of 31 clips loaded".
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
win32.ll implements cocoa.ll's contracts over user32 and xinput: the message
pump, the held-key set and frame key in Ludic codes, the mouse with raw input
for cursor mode 2, clip-based cursor modes released on focus loss, XInput pads,
and the software framebuffer present. win32_gl.ll puts the WGL 4.1 core context
on that window (vsync, borderless full screen); gl_win.ll's context code is now
lgl_wgl_core, shared with the headless hidden window, whose win_gl_* stubs move
to gl_win_nowin.ll. audio_win.ll links Audio.* silently for now.
Verified: macOS fixpoint, ludic-dev test 135/135, selfhost-test 32/32; on the
PC gl_triangle renders identically headless and in a real window, and Maroon
Lake builds windowed and runs a playtest to frame 240 in the desktop session.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The UCRT's remove() deletes files only, so a tree removed bottom-up left every
directory behind. emit_win defines remove over DeleteFileA, falling back to
RemoveDirectoryA. Found by Maroon Lake's selftest26 ("1 left behind"), which now
passes on Windows with the rest of that game's self-tests.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
gl_win.ll creates a hidden-window WGL 4.1 core context and carries gl.ll's
float/memory helpers, with ldexp and QueryPerformanceCounter in place of the
libSystem calls. glgen now also writes gl_thunks_win.ll: the same 478 thunks,
calling through pointers that @lgl_win_load fills from wglGetProcAddress (and
opengl32.dll for GL 1.1). ludicc links the pair against opengl32/gdi32/user32
on a Windows target.
Verified: gl_api.ludic and gl_thunks.ll regenerate byte-identically, ludic-dev
test 135/135, selfhost-test 32/32, and headless gl_triangle on an RTX 3070 Ti
matches the macOS frame to within one level per channel.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
--target <triple> (default: the host, from OS=Windows_NT) selects the Windows
runtime. emit_win.ludic defines the POSIX names the backend already calls over
the UCRT and Win32 in IR, so rename replaces an existing file, ftell is 64-bit,
and fopen is binary. Known folders follow %APPDATA% / %LOCALAPPDATA% / %TEMP%,
and the driver speaks cmd.exe, writes .exe outputs and finds LLVM's clang.
Verified: macOS three-stage fixpoint, ludic-dev test 135/135, and on Windows
the Mac-emitted Windows IR and the Windows-built compiler's IR are identical
through two generations.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`ludic bundle` builds an AppIcon.icns and macOS reads it out of the .app, so a
shipped game has an icon. `ludic build` produces a bare executable: no bundle, no
CFBundleIconFile, and so no icon at all - macOS draws the generic green "exec"
tile. That is the build a developer runs every day, which is why "the game has no
icon" can be true for months while the bundle is perfect. It was here: the .app's
icns validated against iconutil, the plist was right, the signature was right,
and NSWorkspace rendered the artwork - and the binary beside it still had the
exec tile.
App.set_icon(path) takes the bytes through lp_pak_open, so a packed path and a
loose one both work and this does not repeat Audio.load's trick of taking a
filesystem path only. NSData copies them, so the buffer goes straight back. A
missing or undecodable image leaves the existing icon alone rather than clearing
it; a bundled app is unaffected; headless links no AppKit and compiles it away.
Verified the Cocoa sequence against an ObjC twin doing the same message sends:
before, a bare binary's applicationIconImage is the generic 128x128 tile; after,
it is the 1024x1024 artwork.
Backend change, so selfhost/ludicc.seed.ll is reseeded: the bootstrap fixpoint
and the C-free rebuild from the seed both pass.
They were the macOS layout on every platform - the comment above them even said
"macOS/BSD layout" - so a game built on Linux wrote its saves to
~/Library/Application Support, a directory that means nothing there. Naming the
right directory is the entire reason a program calls these instead of joining a
path itself.
macOS save/config ~/Library/Application Support/<app>
cache ~/Library/Caches/<app>
Linux save $XDG_DATA_HOME or ~/.local/share/<app>
config $XDG_CONFIG_HOME or ~/.config/<app>
cache $XDG_CACHE_HOME or ~/.cache/<app>
macOS is unchanged to the byte, deliberately. save_dir and config_dir stay the
same directory there: Apple's home for a config file that is not an
NSUserDefaults plist is Application Support too, and a shipped game's settings
must not move out from under it. On Linux XDG separates the two and so does this.
An XDG variable that is set but EMPTY falls back to the default, which is what
the spec says and what an exported-but-unset variable looks like from a shell.
uname is read once and remembered rather than per call, because save_dir is
called on every save.
Verified on both branches. macOS by running it; the Linux branch by building the
probe's IR with the cached platform flag pinned, which exercises the emitted XDG
code exactly - unset, set, and set-but-empty all resolve as the spec says. The
suite's own case asserts the HOST's convention, so a macOS box covers the Apple
branch and CI covers XDG.
This is a backend change, so selfhost/ludicc.seed.ll is reseeded with it: the
bootstrap fixpoint (gen2 == gen3) and the C-free rebuild from the seed both pass.
Three things a game could not say, each of which had been worked around.
Audio.play_at(id, gain:, pitch:, pan:) fires a one-shot with its own gain,
pitch and stereo position. Audio.volume and Audio.pitch are global - they are
the options screen - so a game placing a sound in the world was fighting them,
and distance attenuation was simply not expressible. The backend already took
volume and rate per call; this adds setPan: alongside them and stops routing
through the master state.
ludic.render3d gains outline_model(model, mat, width, r, g, b): a rim around
something that is not an Actor. The outline pass walked the actor list and
stopped, so instanced scatter - a forest - could not be highlighted at all. It
is a queue flushed by the same pass, which is what gets the depth test right
when the caller does not control pass order.
And terrain_coast(cx, cz, margin, fall), the other way to make an island:
the sea around the survey's own edge rather than cut out of the middle of it.
terrain_island measures a radius from a centre, which drowns two thirds of a
real survey to make an island of the rest; this measures inward from the
boundary, so everything the data covers stays land and the coast is where the
data runs out. Both modes gained a strand - the last few metres of height
either side of the water line compressed, which stretches a cliff plunge out
into beach and shallows.
132 regression tests and the self-host fixpoints pass.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`ludic build` produces a program. Double-clicked it opens a Terminal window, it
wears the generic executable icon, it calls itself whatever the file is called,
and it carries none of its assets. `ludic bundle` produces an application.
Everything it needs is in package.ludic, so the command takes no arguments: an
Info.plist and PkgInfo from `app` lines, an .icns built by sips and iconutil at
all ten sizes macOS asks for from a single source PNG, the asset pack in
Contents/Resources, and an ad-hoc signature - which is not optional on Apple
silicon, where an unsigned binary is killed rather than warned about. The bundle
identifier falls back to the package path reversed, so a project that never
thinks about it still gets a defensible one instead of two apps sharing a key
Launch Services hangs the Dock, saved state and permissions off.
A bundled game is moved to ~/Library/Application Support/<name> before main,
because Finder starts a .app with its working directory at "/" where no save
could ever be written. Reads come out of the pack, writes land somewhere real
and per-user, and the game's save code needs no change and no platform
knowledge.
The splash is the other half of looking like an application. A game that loads
165 MB spends a visible moment doing it with nothing on screen, which from the
outside is indistinguishable from a launch that failed. splash_show puts a
borderless window up from the same constructor that mounts the pack - before
main, so it appears while the process is still starting rather than after the
slow part it exists to cover - and reads the artwork out of the pack like any
other asset. It turns the run loop enough times to be mapped and composited
there and then; once composited the backing store survives a busy main thread,
so it stays up for the whole load.
Nothing hides it automatically. Only the game knows when its first real frame is
ready, and a splash that vanishes before that leaves the same black gap it was
covering, so the game calls App.splash_hide(). Headless there is no splash and
the call lowers to nothing.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
A Ludic game opened its assets by a path relative to the working directory, so
`build/mygame` ran only from the project root and there was nothing to give
anyone but "the binary, and also this whole tree". A game is not one file, but
shipping it has to be.
`ludic pack` writes a .lpak: a header, a name-sorted entry table, a name heap
and the blobs, 16-byte aligned. Entries are stored rather than compressed - PNG,
JPEG and glTF binary arrive compressed already, and a decompressor on the load
path would spend CPU to make the file no smaller.
The reader is spliced into the compiler at the one place every asset in a Ludic
program comes through: file_open. gltf_load, tex_load, Audio.load, Fs.read_text
and the renderer's own shader loads all bottom out there, so routing it through
@lp_pak_open reaches every one of them without any of them knowing. A read of a
packed path becomes an fmemopen over the mapped bytes; everything else is the
fopen it always was. Fs.exists and Fs.size consult the packs too, so a game that
guards a load with Fs.exists keeps finding its assets once they are packed.
The pack is mmap'd rather than read: 165 MB of textures costs one syscall at
boot and pages in only what is touched. @lp_pak_boot runs from
@llvm.global_ctors, before main, so a pack is mounted before the game's first
line - and it reads packs.index, a plain list, so the mount order is explicit
and a later pack shadows an earlier one.
Without a packs.index nothing mounts and every open goes to the filesystem
exactly as before, which is every `ludic run` during development. Packing is a
shipping step and is invisible until you ship.
The compiler reproduces itself byte-exactly and the C-free bootstrap from the
seed still holds.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`Gl.*` binds the whole OpenGL 4.1 core API — every entry point of the
platform gl3.h with every GL_* constant, generated by `ludic-dev glgen`
with per-call ABI thunks. Windowed builds get an NSOpenGLContext on the
existing window at Retina resolution; headless builds render into an
offscreen CGL context, so a program that uses Gl.* renders and
screenshots identically under the test harness. It links gl.ll, the
thunks and OpenGL.framework only when used; every other build stays
byte-identical.
packages/ludic.render3d is a physically based renderer written on that
surface: HDRI image-based lighting, GPU-generated terrain with scanned
PBR materials, CDLOD, cascaded shadows, glTF with skinning, instanced
vegetation with impostors, procedural grass, water, SSAO, and an HDR
pipeline with bloom, auto-exposure and ACES.
It also carries this session's work on it: the terrain at half its cost
(10.3 -> 5.4 ms of frame), the streaming hitch that got worse the longer
you played, a resize that emptied the world, and the packaging that lets
a game use the renderer from its own repository — `ludic assets`, the
material manifest shipping with the package, and shader lookup falling
back to the install root. See changes/ for each, with its numbers.
The camping game that drove all of it has moved out to its own
repository, Maroon Lake; examples/rendering/smooth.ludic stays as the
renderer's example here.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`ludic help` ended with a section titled "contributing to the toolchain itself",
listing bootstrap, reseed, docs-gen and release tasks. None of that is available
to someone who installed the language — those tasks need the repository — so the
shipped tool was advertising work its user cannot do, in a namespace they have to
read past to find `new` and `run`.
The tasks move to a second program, dev.ludic -> bin/ludic-dev, built from a
checkout and excluded from every release artifact. `ludic` keeps the project and
package commands and nothing else; `ludic dev …` now explains where the tasks
went instead of failing as an unknown command.
What this shook out: the two programs share prelude/build/project/pkg, so the
helpers each had accreted in whichever file first needed them — cc(),
ensure_ludicc, the string functions, title_case, cmd_version — moved to where
both can see them. The argument-shift indirection added for the `dev` namespace
is gone with the namespace, so commands read argv directly again.
`ludic-dev test` asserts the split rather than trusting it: the staged install
must build a project, and `ludic dev build` there must fail while naming
ludic-dev. install.sh keeps building older tags, whose bootstrap goes through
main.ludic.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The engine runtime and the bundled ludic.* packages belong to the toolchain,
not to the project, so the compiler has to know where the toolchain lives. It
derived that from the directory the binary sits in, which is `bin` — so an
in-repo build only found runtime/native/cocoa.ll when the caller set
LUDIC_HOME=., and an installed compiler had no way to find it at all.
ludic_home() derives it properly instead: $LUDIC_HOME when set, else the
binary's directory with a trailing `bin/` stripped, else a $PATH scan for
argv[0] (an install is invoked by bare name, which carries no directory). Both
layouts that exist then have the same shape — ~/.ludic/{bin,runtime,packages}
and a repo checkout — so the same rule serves both and LUDIC_HOME becomes an
override rather than a requirement.
`import "ludic.core/…"` also falls back to $LUDIC_HOME/packages, after the
project's own ludic_modules/, so a project that has not fetched its own copy
gets the packages that shipped with the toolchain instead of a symlink farm.
The `ludic` multi-call name is dropped from the compiler: that name now belongs
to the CLI, and --run is the flag it drives.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Each clang invocation the task runner makes emitted
warning: overriding the module target triple with arm64-apple-macosx15.5.0
four times per `x build`, with nothing for anyone to act on. `cc()` now passes
-Wno-override-module, the same flag ludicc already passes for its own link.
The comment in selfhost/main.ludic explaining that flag had it backwards — it
claimed "our IR carries an explicit target triple", when the emitted IR names
no triple at all, which is precisely why clang substitutes the host's and
warns. Corrected. No IR changes, so the seed is untouched.
Also adds .claude/launch.json entry for previewing the generated docs locally.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
- `[a, b, c]` list literals (E_LIST → emit_list); static_type learns
slice-element, `new T`, list, string and literal kinds
- `x op= y` lowers through the same path as `x = x op y` (emit_bin_vals):
fixed `*=`/`/=` use the Q16.16 64-bit paths, string `+=` concatenates,
int→long widens; unary `-` keeps a fixed operand's type (arith_ty)
- one `unescape()` table for "strings", 'chars' and `interpolation`;
`'\''`, `'\\'`, `'\"'` no longer read as 0; unterminated char literals
and unexpected characters are errors instead of silently skipped
- every diagnostic is `file:line: error: msg` (g_parse_file / g_err_file,
Node.file + Node.line set by node()); tok_desc() in expectation errors;
duplicate `function` names and unknown `phase` names are reported in
source terms (phase_id used to default unknown phases to Overlay)
- interpolation holes skip braces inside string literals
- hand-IR preludes move from the user `@fn_` prefix to `@lp_` so a user
`is_ws` / `str_eq` / `path_join` no longer collides at link time
- `@ClearColor(expr)` accepts any constant expression; `Os.pid()` added
(docs page + inventory); `str_starts()` in support/str
- main.ludic: `else if` flag ladder, char literals, stale script comments
- examples/lang/operators.ludic covers all of the above; os.ludic covers
Os.pid; docs pages for Os.pid and the Overlay phase; ten changesets
- reseeded: selfhost/ludicc.seed.ll is the new compiler's own fixpoint
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Closes the two open issues and lands the pending unreleased batch:
- #90: `Sprite { atlas: 1 }` routes esys_sprite through atlas_draw_ex
(scale/flip/tint), so cell / cell_span / strip ids of any size draw
through the engine sprite-render system. examples/library/sprite_atlas
is the pixel-readback regression.
- #91: `become` from an @On(Event) listener / global handler / plain
function no longer segfaults the compiler; it emits @L_scene_leave()
(a dispatch on the live scene id) so the leaving scene's on-exit runs.
UI_* handles are readable from any code (widget table built on first
use). examples/library/scene_menus covers it.
- fix: a windowed `ludicc -o` build that reaches the audio runtime only
through the atlas/Assets preload import now links audio.ll +
AVFoundation (the audio backend link was gated on a game-level
Audio.* call, so any windowed game declaring Sprite failed to link).
- the hand-written "Unreleased" CHANGELOG section is converted to
changesets under changes/ so `x release` generates it.
- plus the batch: engine-driven retained UI + UiClicked event, Overlay
phase, TileSkin tilemap-render system, Key.* constants, Font/Ui/File
namespaces, Sprite.strip, prefabs, managers, countdown fields,
enum-typed machines, layer @Queries, ludic.prefs / ludic.dungeon
packages, Ai.seek pathing, Solids.solid2, cursor confine (mode 3)
fix, shooter centre-aim fix, reserved-word function diagnostic.
Verified: x test (124/124), x test-tools, check-impl, check-vocabulary,
check-docs, docs-gen + docs-check, bootstrap-cfree (seed is a fixpoint).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Ludic's ECS is already pool-based — the allocator recycles freed entity slots
through a freelist (L_alloc pops @L_freen before growing @L_entc), and component
storage is fixed per-entity arrays, so spawn/despawn churn (bullet-hell/horde)
does no per-spawn heap allocation and cannot fragment. Expose that with a Pool.*
namespace so a game can watch reuse: Pool.live (alive now), Pool.free (recycled
slots waiting), Pool.reserved (high-water — stays flat across a steady
spawn/despawn loop, proving reuse not reallocation), Pool.capacity (the fixed
cap). Zero-cost inline reads of the existing counters.
Example pool.ludic proves the key property: after despawn+respawn,
Pool.reserved() stays 3 (freed slot reused) — prints 0 0 3 3 0 2 1 3 0 3 1.
4 docs pages. Full suite 118/0, goldens byte-identical, fixpoint holds.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Assets loaded synchronously in Boot stalled the first frame(s). Adds an
Assets.* preload queue over the #81 atlas: Assets.enqueue(name, path) queues a
named image without loading it, Assets.pump(max) loads up to max per frame
(returns how many), and Assets.total/loaded/ready/progress (0..100) drive a
progress bar. A loading scene pumps a few per frame, draws Assets.progress(),
and becomes the play scene once Assets.ready() — the deterministic, no-threads
form of async preloading (work spread across frames; same enqueue+pump order
loads identically every run). Loaded assets are reachable by name via
Assets.get / Sprite.named.
Example preload (enqueue 3, pump incrementally 0->33->66->100, ready flips, get
by name) prints 3 0 0 0 1 33 66 1 100 1. 6 docs pages. Full suite 117/0,
fixpoint holds.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
`namespace Name { export function foo(...) ... internal function bar(...) ... }`
declares a Name.* namespace once and controls its public surface declaratively,
instead of annotating every function with @Namespace(Name). Inside the block
each `function short(...)` is emitted as `namelower_short`; export (the default)
makes it callable as Name.short(...), internal keeps it a private helper
(emitted, callable by short name from siblings — calls are rewritten — but
Name.internalOne() is a compile error). Block sugar for the per-function
@Namespace annotation; a package's public API reads at a glance. Namespaces
declared the old way are unchanged (gameplay_foundation still passes).
namespace added to LUDIC_KW_DECL + JetBrains/TextMate + docs page + inventory
(vocab/impl/docs checks green). Example namespace_block (export + internal +
sibling calls + internal-visibility compile error verified). Full suite 116/0,
fixpoint holds.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A windowed action game can hide the OS cursor and lock/confine the mouse to the
window. Input.cursor_mode(mode): 0 normal, 1 hidden (draw your own reticle),
2 locked (hidden + dissociated — the mouse feeds relative motion via
Input.mouse_dx/dy and Input.mouse_x/y is a clamped virtual cursor, FPS/twin-stick
aim), 3 confined (dissociated but visible; the mouse can't leave the window).
The platform auto-releases (shows + reconnects) while the window is not key
(Cmd-Tab) and on close, so the cursor is never left captured. Headless it is a
no-op (DCE'd).
Native macOS impl in cocoa.ll: [NSCursor hide]/[unhide] (ref-counted, toggled
only on change so the count stays balanced across focus changes),
CGAssociateMouseAndMouseCursorPosition, and CGGetLastMouseDelta for the relative
virtual cursor, behind a new win_cursor_mode intrinsic. Windowed-only behaviour
(not in the headless golden suite); example compiles headless and links
windowed. Full suite 115/0, fixpoint holds.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Sprite loading was a bare png_load — one file per 16x16 sprite, no way to load
one sheet and address a cell by grid coords or name. Adds a Sprite.*/Assets.*
runtime (atlas.ludic) over the variable-size image loader, so a cell is a
sub-rect of the kept image and is NOT restricted to the 16x16 sprite table:
Sprite.sheet(path,cw,ch), Sprite.cell(sheet,col,row),
Sprite.cell_span(sheet,col,row,cols,rows) (a sprite may span >1 cell),
Sprite.define/named (name + lookup), Sprite.draw/draw_scaled (through
camera/zoom/clip like Screen.sprite), Sprite.width/height, and
Assets.image/load/get. Spliced on demand (Sprite.sheet/… or Assets.*), so a
program using neither is byte-identical.
Example examples/library/atlas.ludic (verified against a real 12x11 Kenney
sheet, incl. a 2x3 multi-cell span and named lookup). 14 docs pages. Full
suite 114/0, goldens byte-identical, fixpoint holds.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The compiler auto-splices runtime/native/* for any ECS game, but resolved it
relative to the build CWD, then fell back to the package module root
($LUDIC_MODULES) — forcing every external project to copy/symlink the engine
runtime into ludic_modules/. The runtime is part of the toolchain, not the
project: do_import now resolves a runtime/... import that isn't found locally
from $LUDIC_HOME (default: the compiler binary's dir — where cocoa.ll/audio.ll
already come from), before the module root. So ludic_modules/ holds only
third-party packages.
In-repo builds are byte-identical (the runtime resolves locally there, so the
$LUDIC_HOME fallback never fires; fixpoint holds). Verified by a hermetic test
that builds an ECS game from an external CWD with no runtime/ or ludic_modules/
under it, resolving the runtime from LUDIC_HOME. Full suite 113/0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
#79 — Input.axis_i(neg,pos) -> int returns a -1/0/1 movement intent from the
multi-key device set, so WASD-to-movement needs no bool->int glue and feeds an
int mover directly (dx = Input.axis_i('a','d')).
#84 — a Bounds config entity (rect + policy, from ludic.core) drives the
engine-owned world-bounds system (LateUpdate): clamp / wrap / bounce (clamp +
flip Body velocity) / kill (despawn a body fully outside). Reads the Collider
size so the box stays inside; off by default, spliced only when Bounds is
declared (byte-identical otherwise). Adds World.despawn(e) — the by-id
reflective despawn (runs @OnDespawn + frees) via a new world_despawn intrinsic
whose @fn_world_despawn helper is emitted in emit_program's tail once a use is
seen (the g_uses_* prelude pattern), used by the kill policy and callable from
any system.
Examples input_movement + world_bounds. Full suite 112/0, goldens
byte-identical, fixpoint holds.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Since #83 the loop commits the device layer once per frame (input_drive), but a
game that ALSO called Input.poll by hand committed a second time in the same
frame; input_device_commit copies in_held into in_prev at the top of every
commit, so the second commit left in_prev == in_held and the edges (held &&
!prev) could never see a transition.
Fix: an in_have_frame_driver flag. The loop's input_drive sets it; a manual
Input.poll under the loop then becomes a no-op returning the frame's key
instead of re-committing. An entry-driven harness has no loop, so the flag
stays false and each Input.poll commits a frame as before (the #7/#50
record/replay + device tests are unchanged). Frame loop now calls input_drive.
Example input_edge (press edge on the down frame, release edge on the up
frame). Full suite 110/0, goldens byte-identical, fixpoint holds.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The engine already auto-ticks SpriteAnim/Motion; it now auto-DRAWS too.
Declare a Sprite component (id/offx/offy/scale/flip/tint/hidden, shipped from
ludic.core) on an entity with a Position and esys_sprite draws it each Render
frame — no hand-written Render handler, no hand animation (adds the SpriteAnim
frame when present). Registered on the engine-system registry (Render) and
spliced only when the game declares Sprite, so a game that never declares it
compiles byte-identically; opt out by omitting Sprite or `disable system
esys_sprite`.
Deprecates the bare draw_sprite/draw_sprite_scaled globals in favour of
Screen.sprite/Screen.sprite_scaled: a direct bare call emits a one-time
compile-time deprecation note (the bare form still lowers, since Screen.sprite
uses it); migrates the chronorift demo to the namespaced calls (golden render
byte-identical).
Example sprite_render (pixel-readback: engine draws the sprite, respects
hidden). Full suite 109/0, goldens byte-identical, fixpoint holds.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
@ClearColor(0xRRGGBB) declares the framebuffer clear colour, so the engine
owns the per-frame clear and flip: the Render phase clears to the colour at
the top and presents after the handlers run. Games drop the repeated
Screen.clear(color)/Screen.show() boilerplate, and the colour is configured
declaratively (an annotation) rather than in the handler body. Opt-in and
backward-compatible: a program with no @ClearColor is byte-for-byte identical
(it clears/presents itself, or the light system owns the present).
Parser reads @ClearColor(int) into g_clear_color/g_has_clear_color;
emit_game_main emits rt_clear before and rt_present after the Render phase,
gated on the flag. Example clear_color (pixel-readback verified — an undrawn
pixel holds the clear colour, proving the engine cleared), docs page +
inventory entry. Full suite 108/0, goldens byte-identical, fixpoint holds.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Fixes the papercut: the generated frame loop called rt_poll() (feeding only
Input.key) but never input_poll(), so Input.active/key_down/mouse/pad read
empty unless the game called Input.poll() by hand. The loop now calls
input_poll() when the game uses any Input runtime method — committing the
held-key/mouse/gamepad state, and record/replay — and stores its return as the
frame key so Input.key still works. A game using no Input runtime keeps the
plain rt_poll path, byte-identical.
Adds the Input-Manager API: Input.action(name,key) ships a default binding
(kept if already bound, so a rebind/loaded map isn't clobbered),
Input.bind_pad(name,button) makes an action device-agnostic (keyboard OR pad),
and Input.active/just_pressed/just_released read the multi-key device layer
with clean on-press/on-release edges (deterministic, dispatch-free — a handler
polls the edge; a replay fires identically).
Examples input_manager + input_auto, 5 docs pages. Full suite 107/0, goldens
byte-identical, fixpoint holds.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The #78 investigation rejected hardware f32/f64 for the coordinate types
(they would desync lockstep/replay/save) and identified camera zoom as the
one genuinely-missing render feature. Ship it: Camera.zoom(scale) scales the
whole view about the screen centre by a Q16.16 factor, threaded through the
same two framebuffer chokepoints (rt_put_px/rt_fill_rect) that carry the
camera offset, so it composes with Camera.set/follow/shake. Gated by an
internal rt_cam_zoomed flag so a game that never zooms renders byte-for-byte
identically (golden renders unchanged); Camera.zoom(1.0) turns it back off.
The world coordinate types stay integer px + Q16.16 velocity, so it's a pure
render-time transform and itself deterministic.
Example examples/library/camera_zoom.ludic (pixel-readback verified),
docs page, RFC updated (docs/RFC-POSITION-TYPES.md). Full suite 105/0,
fixpoint holds.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The reference six-lever controller: Platformer component (jump feel as data),
decomposed input/move/gravity/jump/anim engine sub-systems (each disable-able),
JumpRequested/Landed/StateChanged events, gravity policy enum, and the opt-in
scaffolding (moving/crumble platforms w/ rider carry, pickups+score, springs,
hazards+Life, checkpoints/goal). Reuses the shared esys_move collision.
Also fixes a latent SSA-name collision in ludic_sweep_entity that triggered
once a program declared >=11 events. Reseeded; C-free fixpoint holds.
3 new regression cases (100 passed, 0 failed).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Shared building blocks for the builtin gameplay controllers (#57): the
ludic.gameplay source package (Cooldown timer, Stats + timed modifier stack,
Faction table, Combat damage pipeline with cancel/mutable hooks), plus the
compiler `disable system <esys_fn>` extensibility lever. Deterministic,
integer-only; C-free bootstrap fixpoint holds.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Infinite/chunked maps: <chunk x y width height> (TMX) and JSON chunks[]
(default 16x16) decode and flatten into the dense layer array, sized to the
chunk union; the map's 0/0 header dimensions fall back to the flattened bounds.
- .world stitching: Tiled.world / Tiled.world_count / Tiled.world_map read a
.world (JSON, reusing Json.parse) and list its member maps at their offsets.
- base64+zstd: a self-contained pure-Ludic Zstandard decompressor
(runtime/native/zstd.ludic, RFC 8878) — the design's "largest single item,
explicitly last". Frame header + raw/RLE/compressed blocks; raw/RLE and
direct-weight Huffman literals; the full FSE sequence path (predefined,
transcribed exactly from zstd's hardcoded tables since they're not rebuildable
from the default distributions; RLE; FSE-described) with repeat offsets and
execution. Decodes the low-entropy GID streams a tilemap produces; a high-
entropy FSE-compressed-Huffman-weights block fails cleanly with -1 rather than
emitting wrong bytes (documented scope).
Proven by library/tiled_p6.ludic (12 assertions): TMX + TMJ chunk flattening,
.world offsets, and a real zstd-compressed tile layer decoding byte-exactly to
its CSV baseline. x test: 97 passed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Image layers: <imagelayer> renders with parallax + optional repeatx/repeaty
tiling across the view; the image loads relative to the map file. Group layers
flatten at build (children render in file order); layer offset/opacity/tint are
queryable (Tiled.layer_kind / layer_offsetx / layer_offsety / layer_opacity /
layer_tint), a group's tint applying recursively.
- Orientation transforms: Tiled.cell_x / Tiled.cell_y compute a tile cell's
screen position for orthogonal, isometric ((x-y)*tw/2, (x+y)*th/2), staggered,
and hexagonal (rows step by (tileh+hexsidelength)/2, alternate rows shoved per
staggerindex) — the tile draw now places cells through them.
- Wang: exported Wang-set GIDs are ordinary GIDs and resolve through the standard
GID resolver; the terrain-corner authoring concept is editor-side (design cut).
Proven by library/tiled_p5.ludic (14 assertions) over the real
isometric_grass_and_water.tmx (iso + Wang) and hexagonal-mini.tmx, plus a
hand-authored image+group fixture. x test: 96 passed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Object layers: all shapes (rectangle/ellipse/point/polygon/polyline/text) and
custom properties parse and are queryable — Tiled.object_count / Tiled.object /
Tiled.object_shape, and Tiled.prop / prop_int / prop_type over any object /
tile / layer / map.
- Custom types: Tiled.load_types reads an objecttypes.xml project custom-type
table so a class property resolves its default; enum values resolve as strings.
- Templates: Tiled.template reads a .tx/.tj, and Tiled.load merges each object's
`template` reference under the instance's overrides (field inheritance).
- Opt-in spawning: Tiled.spawn / Tiled.spawn_layer map an object's class +
properties onto Ludic components through the reflection ABI (Position from x/y,
each property to a like-named field). Off by default — no gameplay coupling in
the core load. Split into tiled_spawn.ludic, spliced only for a Tiled *game*
(the world table exists only under has_ecs), so a plain map-reading tool never
links against the reflection ABI.
Proven by library/tiled_p4.ludic (18 assertions) incl. the design's named
orthogonal-outside.tmx object shapes. x test: 95 passed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Animated tiles: a tileset <animation>'s {tileid, duration} frames advance as a
pure function of the fixed 60/s engine frame counter (the same clock SpriteAnim
rides), so an animated GID resolves at draw to the current frame's GID with its
flip flags preserved — deterministic, frame-identical across runs, ticks for
free. Tiled.frame_gid(map, gid, frame) / Tiled.animated(map, gid) expose it;
Tiled.draw_anim(map, camx, camy, frame) draws a map with animations advanced.
- Tile objects: object-layer entries with a `gid` draw the tile image (with their
own flip flags), bottom-anchored at the object position, as placeable sprites;
tmap_draw_full now renders object layers alongside tile layers.
Proven by library/tiled_p3.ludic (14 assertions): frame advance at authored
durations + wrap, flip flags riding an animation swap, tile-object parse + draw +
flip mirroring, and the design's named rpg/beach_tileset.tsx (33 <animation>
blocks / 131 frames). x test: 94 passed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Normalise three collision sources into the byte tilemap esys_move / Grid.* /
Path.* read, in the design's priority order (§3.5):
- per-tile <objectgroup> hitboxes (Tiled.tile_shapes) — the precise source;
- the solid/oneway/trigger bool property convention (Tiled.collision_kind);
- the designated collision layer — any non-zero GID solid (Tiled.project),
the fallback source, applied automatically on load.
Tiled.collide drives collision from a visual layer's per-tile metadata alone
(a tile with no collision metadata stays passable). tmap_collision_kind
classifies a GID (0 none / 1 solid / 2 one-way / 3 trigger) from its metadata;
the projection adds the collision-layer fallback.
Proven by library/tiled_p2.ludic (14 assertions): kind classification for each
source, the property convention and collision-layer fallback producing an
identical Solids feed, a per-tile-<objectgroup> floor blocking an esys_move
mover, and A* over a loaded map matching the hand-authored baseline. x test:
93 passed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The heart of Tiled support: load a map and draw it.
- rt_tmap model (Tmap/TmLayer/TmTileset in tiled.ludic): map header + ordered
layers (dense int32 GID arrays, heap-allocated to w*h, lifting the 96x64 cap)
+ tilesets. Built from the intermediate Value tree, so the TMX and TMJ paths
both feed it.
- GID resolver (Tiled.resolve): gid -> (tileset, localId, flipH/V/D); the three
flip flags masked off before the local-id lookup, returned alongside. gid==0
is empty.
- Image-backed render (Tiled.draw): every visible tile layer in file order,
blitting each tile from its tileset image with flips applied at draw.
- Compatibility projection (Tiled.project / auto on load): a designated
collision layer projects to the legacy byte tilemap ('#' solid, '=' one-way
via the oneway property, ' ' empty) so Grid.*/Path.*/esys_move are unchanged.
- Tiled.load resolves external tilesets + images relative to the map file and
auto-projects a collision/solids/walls layer.
- The grid and physics_tiles demos now run off a loaded map (grid_maze.tmx /
physics_map.tmx) instead of hand-authored Map.row strings, byte-identically.
Two compiler fixes fell out of this (see the changeset):
- emit_index_addr set g_addr_ty before evaluating the index, so slice[obj.field]
came back mis-typed; set it last, like the raw-pointer branches.
- @strcmp was declared by both the world table and the fs prelude; centralise
it in the head prelude so a game that uses Fs/Path links.
Proven by library/tiled_p1.ludic (14 assertions: loads+renders Kenney map
identically from .tmx and .tmj, flip mirroring) + the converted grid/
physics_tiles demos. x test: 92 passed; self-host bootstrap fixpoint intact.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Tiled.read / Tiled.read_tsx (runtime/native/tiled.ludic): map the native
XML formats (TMX/TSX/TX) onto the SAME intermediate the JSON path produces
— a Value tree in Tiled's JSON schema — so one format-independent core
(P1) consumes either reader.
- tmx_to_value walks a <map> into {orientation, width/height, tilewidth/
height, tilesets[], layers[]}; every tile layer's <data> is decoded to a
dense GID int list (CSV split, or base64 -> zlib/gzip inflate ->
little-endian u32s), so a CSV .tmx and a base64 .tmj of the same map read
structurally identically.
- External tilesets keep the {firstgid, source} reference (as TMJ does);
embedded tilesets and standalone .tsx (tsx_to_value) inline full geometry
+ per-tile metadata (animation frames, collision objectgroup, class,
properties) for later phases.
- Object layers, shapes (rect/ellipse/point/polygon/polyline/text),
image/group layers and custom properties are parsed into the tree now so
P2/P4/P5 just read it.
- tmj_normalize decodes base64 `data` strings in a parsed .tmj so the JSON
path matches the XML path.
Proven by library/tiled_p05.ludic (30 assertions) incl. the in-repo Kenney
sampleMap.tmx + external sampleSheet.tsx and TMX-vs-TMJ structural
identity. Docs + inventory added; x test: 91 passed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The three parsing primitives the TMX path needs that were not already in
the tree (zlib inflate + the JSON reader already shipped):
- Xml.* — a minimal, deterministic pure-Ludic XML reader for the
element/attribute/CDATA subset TMX/TSX/TX use, spliced on demand. Handles
nested elements, single/double-quoted attributes, text + <![CDATA[…]]>,
comments, the <?xml?> prolog and <!DOCTYPE>, the five predefined entities
and numeric character references.
- Base64.* — standard base64 (RFC 4648) decode + a matching pure-Ludic
encoder; the decoder ignores the whitespace Tiled wraps into <data>.
Splicing Base64.* also pulls in inflate.ludic so a plain tool can run the
full base64 -> zlib/gzip decode chain.
- z_gunzip — gzip framing (RFC 1952) over the existing DEFLATE inflater:
skip the 10-byte header + optional fields, inflate the body, ignore the
CRC32/ISIZE trailer.
Golden corpus vendored under assets/tiled-fixtures/ (mapeditor/tiled
examples, attributed, + hand-authored multi-encoding fixtures). New
example library/tiled_p0.ludic proves all three (21 assertions); docs
pages + inventory added so check-impl stays green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Phase 0 of the gameplay-controller work (#57): turn the static Collision.*
overlap tests into a real physics-lite moving-body-with-collision layer that
the platformer / shooter / NPC-AI / RPG families build on.
New engine-owned system esys_move (runtime/native/systems_move.ludic), spliced
and Update-phase-registered when a game declares `Body` (parse.ludic), standing
entirely on the reflection ABI like the SpriteAnim / Motion / Light2D systems:
- Body { vx, vy, gravity, max_fall, rx, ry, policy, on_ground, hit_wall,
hit_ceiling } — Q16.16 velocity + engine-owned sub-pixel accumulators.
- Collider { w, h, offx, offy, is_trigger, one_way, layer, mask, hit, entered,
exited } — AABB shape off Position, layer/mask filtering, one-way + triggers.
- Solids { tile, wall, oneway } — optional config entity enabling the tile-grid
broadphase over the Map.* tilemap.
esys_move integrates velocity + gravity, then resolves per-axis swept AABB
against both solid Collider entities and the tile grid (no tunneling), handles
one-way platforms (block only a downward landing), reports trigger/sensor
overlaps without resolving, and sets on_ground / hit_wall / hit_ceiling for a
controller to poll. No float, no hidden singletons, no runtime dispatch —
integer + deterministic, so replay / lockstep / world_save hold. Contact is
surfaced as polled flags (the SpriteAnim.event_fired shape), so a game raises
its own CollisionResolved / TriggerEntered events with no engine coupling.
Two self-asserting examples (entity + tile broadphase) wired into `x test`;
docs added to the collision section. A build with no Body compiles byte-for-byte
the same (bootstrap fixpoint + all golden renders unchanged).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Registry-izes the two hooks that made stdlib namespaces and engine systems
compiler-hardcoded, so a package registers them with no compiler edit — the
Phase-1 prerequisite for shipping the controller libraries (#58–#61) as real
packages rather than in-repo stdlib.
- Engine systems are a data-driven registry (component, esys-fn, phase). The
core three (SpriteAnim/Motion — Update, Light2D — Render) are seeded in that
exact order, so uses_engine_systems / emit_engine_systems_for_phase are now
registry-driven with byte-identical output (verified: anim_ecs, light_ecs,
snake IR unchanged; 87/0 golden renders; C-free fixpoint holds). A package
appends with `@EngineSystem(Component, Phase)` on its esys function.
- Namespaces are a registry too: a package marks a provider with
`@Namespace(Foo)`, and emit_ns_call aliases an otherwise-unknown Foo.method to
the bare foo_method (the same generic path the core namespaces use) — after
every hardcoded core block, so core dispatch is untouched.
- Both annotations are keyword-free (like #64's @System), so no vocabulary /
grammar churn.
Proven end-to-end (hermetic, source path, runs everywhere): a package registers
Score + esys_score via @EngineSystem and coach_bonus via @Namespace; a consumer
game imports it and prints "4 99" — the engine system ran each Update and
Coach.bonus() dispatched, with no compiler edit for the package. Package suite
18/0.
Core stdlib namespaces stay on their optimized hardcoded blocks by design
(byte-identity + determinism); the generic path is proven to carry a namespace
and packages ride it.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The runtime + compiler foundation for prebuilt binary packages, over the
existing reflection C-ABI (EV0–EV8 already gives dynamic components via
ludic_register_prop).
- ludic_register_system(fn, phase) + a registry the frame loop dispatches after
each phase's own handlers — the systems analogue of ludic_register_prop,
mirroring the EV6 foreign event-listener array. Emitted for every ECS program;
a zero-length registry means a non-hosting game is output-identical.
- --emit-module: compile a package to a position-independent module — no main,
no world table — that declares the host reflection ABI it calls and carries a
load-time constructor which registers its @System(Phase) functions and runs
module_init (where it registers its dynamic components). Built as a dylib with
-undefined dynamic_lookup, it binds ludic_* back to the host image at load.
- @System(Phase) annotation marks a module function as a runtime-registered
system; the compiler supplies its address (Ludic source cannot take one).
- The reflection ABI (world table) is now emitted for every ECS program, so any
game can host binary modules with no flag; unused defs dead-strip at -O2, so
golden renders stay byte-identical and the C-free bootstrap fixpoint holds.
Proven end-to-end: a module dylib registers a component + an Update system; a
host game that never saw its source links it and the system mutates the shared
world each frame. Full suite 87/0, test-tools 30/0, fixpoint intact.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Implements the v1 direction decided in the RFC as a set of `x` subcommands
plus a small, contained compiler change.
* URL-as-identity, no registry — a dependency is named by its git import
path and a `git tag vX.Y.Z` publishes a version.
* Minimum Version Selection — a `require` is a minimum; the resolver picks
the greatest required minimum per module, then the reachable closure at
those versions. Deterministic, no SAT solver (tools/x/pkg.ludic).
* Content-addressed global store + per-project links — packages live once in
~/.ludic/store keyed by a content hash; each project links them under
ludic_modules/. package.ludic (manifest) + package.lock.ludic (lock).
* Namespace registration for source packages via a module-root import
fallback in the compiler: do_import resolves a non-local, non-absolute
import under $LUDIC_MODULES (default ludic_modules/), so a fetched
package's Ludic compiles into the consumer the way the built-in stdlib
does. Collisions and missing prebuilt targets are hard errors.
Commands: x add / x get / x update / x verify / x vendor. New hermetic suite
`x test-pkg` (stands up throwaway git repos, offline) is gated inside `x test`.
Existing programs compile byte-for-byte identically (the import fallback only
fires when the local path is absent); the C-free bootstrap fixpoint holds and
the seed is regenerated. Full suite: 87 passed, package suite: 12 passed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A layered concurrency library, safe by default. The recommended tier is
Job.* / Promise.*: a Job is a future — Job.run(kind, arg) starts a
cooperative background compute that advances each Job.pump(budget) and
finishes after enough frames (heavy work spreads out instead of hitching),
or Job.defer + Job.fulfill/fail/cancel drives one by hand. Poll with
done/ok/failed/cancelled, read result/error, count outstanding work with
Job.pending. Promise.all/race combine handle lists into a group job resolved
on the main thread; Promise.count_done/all_done power a loading bar.
The advanced, opt-in Sync.* tier (mutex/atomic/channel + cpu_count) is the
"here be dragons" surface for engine-level message passing.
The whole thing is a deterministic cooperative scheduler: results are
collected on the main thread and a Job never touches the ECS world, so
lockstep and replays stay bit-exact — same jobs + same budget reproduce
byte-for-byte on every target, and a preemptive OS-thread backend can slot
behind this same API later. Ludic has no closures, so a Job carries a
compute kind + int arg (or a hand-driven defer) rather than fn()->…, and
Promise progress is polled rather than chained through then.
Written in Ludic and spliced on demand (like Regex/Dict/Numeric): a program
that never mentions Job.*/Promise.*/Sync.* compiles byte-identically and the
C-free bootstrap fixpoint is untouched. New: runtime/native/jobs.ludic,
emit_ns_call dispatch, parse-time splice, examples/library/jobs.ludic (31
self-asserting checks), 33 docs pages + inventory, changeset.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Extend `enum` from named int constants to a tagged union: a variant may
carry a payload (`enum Tile { Empty, Wall, Door(int), Portal(int, int) }`).
Such enums box to a heap record (an i32 tag at offset 0, then one 8-byte
slot per payload position); an all-bare enum keeps its zero-cost compile-
time-ordinal representation, byte-for-byte unchanged (every golden render
and the bootstrap fixpoint still hold).
- Parser: variant payload declarations, stored as N_PARAM kids on the
variant node.
- Construction: by name — `Door(3)`, `Portal(x, y)`, bare `Empty` — resolved
ahead of the function-call fallback and boxed with the payloads coerced to
their declared types.
- match: destructures a tagged scrutinee, switching on the tag and binding
each arm's payload names in a scoped local frame.
- Checking pass: a tagged `match` must be exhaustive (cover every variant or
end in `_`), and constructor/pattern arities and binding forms are checked
— all reported where the scrutinee's type is known.
Adds selfhost/tests/enums.ludic to the regression suite and documents the
feature in LANGUAGE.md and the enum/match pages.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A const reference lowered to `val(itoa(g.a.ival), "int")` in emit_call.ludic —
the initializer's raw integer bits, hardcoded as `int`. For a fixed const like
`const X: fixed = 10.0` that yielded the Q16.16 bits (655360) typed as int, so
every fixed comparison/arithmetic against it silently broke (it caused an
infinite loop in runtime/native/numeric.ludic, previously worked around with
inline literals).
Fix: a const reference now emits its initializer expression via emit_expr(g.a),
which carries the initializer's real type (E_FLOAT->fixed, E_BOOL->bool,
E_STR->string) and even handles computed initializers. Every existing const is
an int literal, for which this is byte-identical to the old immediate — the
C-free bootstrap fixpoint and all golden renders are unchanged.
- selfhost/tests/const.ludic + sh_case pin fixed/int/bool const behaviour.
- runtime/native/numeric.ludic restored to named fixed consts (HUGE_TEN etc.),
which the workaround had inlined; the numeric example (20 assertions) still
passes, validating the fix under runtime splice.
All suites green: x selfhost-test 31/31 (fixpoint holds, goldens byte-identical),
x test 85/85, x test-tools 30/30.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
