The suite spent 88% of its step time asleep: 564 s of steps, 496 s of
them in sleep(n) - fixed spans estimated by hand and then copied from
step to step. That is why it was slow, and why it flickered.
The signal is afterFrames(count, timeout) in ui/render.ts, exposed as
__gm.afterFrames(). It wakes the loop itself, because an editor that
draws nothing would never redeem a promise of a frame, and it gives up
after a timeout instead of hanging the run. settle() in the harness
waits two drawn frames (~32 ms) instead of 150-300 ms and falls back to
sleep(150) when the hook is missing (an old build): silently not
waiting would make every later check a coin toss.
Every action goes through it now - click, doubleClick, wheel, drag,
dragAsset, press, typeText, select, mutate, undo, redo - and with that
about 550 sleep() calls behind those actions became redundant, helpers
that are nothing but a mutate in a coat included. What a frame cannot
answer got its own condition instead: waitUntil(expr), play()/stop()
waiting on __gm.state.mode rather than on 300 ms, saved() for the
autosave (the two hand-written polling loops sampled at 300 ms for
something that arrives after a second), and showTab() as a real tool -
47 places called evaluate("__gm.showTab(...)") raw, and 24 of them
waited for nothing at all and read the old panel.
564 s -> 243 s of step time, the waiting share 88% -> 76%. Three kinds
of wait were kept, and each one bit when it was taken away: loading
bytes (textures, fonts), game time that a repaint cannot stand in for
(time scale, shader and particle clocks, the live values a running
graph carries), and the inertia the scroll step measures on purpose.
Two things the new speed uncovered, both real: 25-animation read the
frame menu unfiltered and had frozen an incomplete list (the CI failure
on c990dbc), and 33-render-on-change measured idleness while the
autosave of the step before was still painting - it waits for
saveStatus to leave dirty/saving now, so it measures that nothing
happens any more instead of that it waited long enough.
Game Maker
A browser-based tool "one step before a game engine": create assets (scenes, cameras, sprites), place them into scenes, tweak their properties in an inspector and press Play – all in one program, all drawn inside a single WebGL canvas.
Live: https://nockal.com/game-maker/ · Widget playground: https://nockal.com/game-maker/?playground
The documentation is the wiki: https://git.nockal.com/meleki/game-maker/wiki – every concept, node, tab and shortcut, plus the architecture and the development workflow. This file is the short way in.
In one screen
- An asset is a named, typed thing (scene, camera, sprite, text, behavior, collider, container, texture, animation, state machine) with properties and no position.
- A scene object places an asset: position, sparse overrides of its properties, an optional parent. Only what differs from the asset is stored; the inspector paints it gold and offers ↺ / ↑.
- A container is a reusable group edited once and placed many times; its content is resolved from the template on every read. A screen is a container switched on and off.
- A behavior is logic as a node graph, placed like anything else and parented to the object it acts on. Play has no built-in behaviour – everything that happens is a graph.
- Play runs a throwaway copy of the project; every panel stays usable and Stop brings back exactly what was saved.
Quick start
Node 22 is required.
npm install
npm run dev # http://localhost:5175/game-maker/dist/ – API calls are proxied to production
npm run check # typecheck + lint + format:check + unit tests – the gate CI runs too
npm run build # writes dist/ and regenerates index.html; the folder is what nginx serves
The browser suite and the three example games need a machine with a GPU:
node tools/seed-e2e.mjs && node tools/e2e.mjs # the editor, 862 checks (GM_STEPS=20 for one step)
node tools/seed-jumprun.mjs && node tools/smoke-jumprun.mjs
node tools/seed-tetris.mjs && node tools/smoke-tetris.mjs
node tools/seed-actor.mjs && node tools/smoke-actor.mjs
Every push runs all of it on the Gitea runner, so locally take the gate plus the step your change touches and leave the full run to CI. Details, the GPU requirement and the detached runner are in Development.
Documentation
| Page | What it covers |
|---|---|
| Concepts | Assets, scene objects, overrides, play mode, the camera, the console |
| Scene editing | Placing, z-order, hierarchy, the editor offset |
| Behaviors | Node graphs: variables, the node set, structs and lists, loops, nesting, editing |
| Colliders | Collision events and blocking movement |
| Textures | Images, frames, the filter, upload limits |
| Animations | Animation assets and the Animation tab |
| State machines | The asset, named transitions, State branch, Set state and State transition |
| Containers | Templates, instances, screens |
| Keyboard | Every shortcut |
| Architecture | Stack, source layout, the rules that keep it organised |
| Development | Commands, the e2e suite, the example games, CI |
| API | The PHP JSON API and the project document |
Ideas, gaps and bugs
They live in the issue tracker: https://git.nockal.com/meleki/game-maker/issues. What is not there yet is listed on the wiki's Home page.