🧼 optimize install size by natemoo-re · Pull Request #35 · bombshell-dev/tty

natemoo-re · 2026-05-23T05:05:50Z

Reduces npm install size by 63% (723kB → 268kB unpacked) via:

optimizing clang build flags with explicit exports and dead code elimination
stripped source maps and src/ duplication from the npm package

Also bumps minimum node version to v22 (current LTS).

Future optimization opportunities identified (with unknown performance impact) include:

switching from -02 to -0z, adding wasm-opt step
compress wasm text (testing showed brotli + z85 encoding to be largest byte reduction)
strip data that can be extrapolated from wcwidth

pkg-pr-new · 2026-05-23T05:16:02Z

npm i https://pkg.pr.new/clayterm@35

commit: 79052b0

dreyfus92

overall looks really good. since you're optimizing install size, are you planning on the near future to add a "no size regression test", cause rn there's nothing in the CI that asserts the package stays under a certain size. without that the size could quietly creep back up over time.

Co-Authored-By: ghostdevv <ghostdevbusiness@gmail.com>

- replace sourceMap/declarationMap/target compiler options with skipSourceOutput: true - add type: "module" to package config - remove redundant files: ["esm"] (npm excludes lock files automatically, .npmignore handles the rest) Co-Authored-By: ghostdevv <ghostdevbusiness@gmail.com>

Co-Authored-By: ghostdevv <47755378+ghostdevv@users.noreply.github.com>

Co-Authored-By: ghostdevv <git@willow.sh>

cowboyd · 2026-05-24T03:19:49Z

There are a couple of optimizations here; some of which I'm 100% cool with, the others I think we need to make sure we do intentionally.

Unalloyed Wins:

The linker changes to only export the functions we need and to gc/strip the rest. I don't see a downside to this.
same with the npm bundle changes

Needs consideration

switching from -02 to -0z is going to result in a smaller bundle size, but there will be a degradation of runtime performance. How much? I've no idea. We should add some benchmarks so that we can understand the tradeoffs
the compression of the wasm text is going to add a tiny bit of overhead to the startup time. Do we know how much it will be?
as for the wcwidth. That's a piece of the codebase we took from termbox2, so while I'm not opposed to the changes, but I don think we need to make sure we understand it.

cowboyd · 2026-05-24T15:19:39Z

@natemoo-re

switching from -02 to -0z is going to result in a smaller bundle size, but there will be a degradation of runtime performance. How much? I've no idea. We should add some benchmarks so that we can understand the tradeoffs

Just so we don't block this work, can we merge all the other size optimizations and then consider these C options separately in the context of knowing how runtime perf is impacted?

natemoo-re · 2026-05-24T17:05:19Z

switching from -02 to -0z is going to result in a smaller bundle size, but there will be a degradation of runtime performance. How much? I've no idea. We should add some benchmarks so that we can understand the tradeoffs

Just so we don't block this work, can we merge all the other size optimizations and then consider these C options separately in the context of knowing how runtime perf is impacted?

@cowboyd yep that sounds like the right path forward! happy to split those out with some benchmarks

Reverts three changes that need benchmarks before landing: - -Oz / wasm-opt - brotli+Z85 wasm compression - wcwidth.c rewrite

cowboyd

Love it!

@ghostdevv

* ✨ add snapshot() for pre-packing directive subtrees Introduces a new `snapshot(ops)` constructor that pre-packs a directive array into its transfer encoding. The returned opaque `Op` can be spliced into any directive array, and during packing its bytes are copied directly into the command buffer without re-encoding. This enables higher-level frameworks to implement dirty tracking: unchanged component subtrees can reuse a cached snapshot, skipping the per-frame packing cost entirely. * docs: add maintainer build guide * 💄 format build docs * 🐛 improve pack string overflow errors * os matrix test in ci (#36) * 🧼 optimize build * 🧼 compress bundled wasm * 🧼 optimize wcwidth.c size * ⚙️ update npm settings * 🐛 install wasm-opt in ci * ⚡ use brotli-11 + z85 wasm encoding * 📌 pin @types/node to v22 * 🧼 apply @ghostdevv review suggestions from PR #35 Co-Authored-By: ghostdevv <git@willow.sh> * 🔨 add type to bundle-wasm * 💌 signed, sealed, delivered * chore: use hashes for versions * chore: don't save git credentials * chore: use array syntax for some reason the schema for the actions wants it to be an array * perf: set concurrency limits to reduce cost and improve dx Without this it means that, for example, if I push a change to a PR then shortly push again this workflow will be running twice. This change will cancel the old run before starting the new one, which reduces the overall actions cost and DX as you don't have extra runs * chore: use hashes for versions * chore: update node version * perf: set concurrency limits to reduce cost and improve dx Without this it means that, for example, if I push a change to a PR then shortly push again this workflow will be running twice. This change will cancel the old run before starting the new one, which reduces the overall actions cost and DX as you don't have extra runs * chore: don't save git credentials * chore: mitigate potential template injection See https://docs.zizmor.sh/audits/#template-injection * chore: update ::set-output command to new syntax https://github.blog/changelog/2022-10-11-github-actions-deprecating-save-state-and-set-output-commands/ * chore: use hashes for versions * chore: don't save git credentials * chore: limit id-token permission to the publishing steps * chore: explicitly disable npm cache to mitigate cache poisoning attacks * chore: mitigate potential template injection See https://docs.zizmor.sh/audits/#template-injection * chore: use oidc This should be using OIDC for publishing * chore: update ::set-output command to new syntax https://github.blog/changelog/2022-10-11-github-actions-deprecating-save-state-and-set-output-commands/ * 🙅 revert aggressive optimization experiments pending benchmark Reverts three changes that need benchmarks before landing: - -Oz / wasm-opt - brotli+Z85 wasm compression - wcwidth.c rewrite * Add CodSpeed performance benchmarks * 🔨 use deno * 🧼 deno fmt * ⚙️ vitest -> tinybench * 🧼 remove codspeed assets * 🔨 fix ci * add type module * fmt * chore: update github url (#38) * downgrade to tinybench@5 * move to examples folder with readme * 🔧 export animating from wasm build * 🧪 cover transitions in snapshots and validation * ✅ enforce nonnegative transition duration --------- Co-authored-by: Charles Lowell <cowboyd@frontside.com> Co-authored-by: Jacob Bolda <me@jacobbolda.com> Co-authored-by: Nate Moore <nate@natemoo.re> Co-authored-by: ghostdevv <git@willow.sh> Co-authored-by: Nate Moore <git@natemoo.re> Co-authored-by: codspeed-hq[bot] <117304815+codspeed-hq[bot]@users.noreply.github.com> Co-authored-by: Nate Moore <natemoo-re@users.noreply.github.com>

@ghostdevv

* 📝 add transitions design specification Design spec for clayterm transitions: frame-snapshot-compatible interpolation of element position, size, and color properties. Defines the deltaTime convention, the animating signal on RenderResult, declarative enter/exit semantics that replace Clay's function-pointer callbacks, and cancellation as a structural consequence of re-describing state. Implementation is gated on bumping the Clay submodule past the upstream transition commit. * ⬆️ bump Clay submodule to latest main (transitions API) * 🔧 adapt clayterm to new Clay signatures (OpenTextElement, EndLayout) * ⬆️ pin Clay to 938967a (work around upstream CLAY_WASM_EXPORT typo) * ✨ add deltaTime parameter to reduce() * 🔧 add deltaTime to Native.reduce signature * ✨ track deltaTime on Term, accept deltaTime override * ✨ add animating_count to Clayterm context * 🔧 expose animating() via Native binding * ✨ surface animating: boolean on RenderResult * 📝 rewrite transitions spec for v1 (Clay-supported subset) Scope v1 to what Clay currently supports without userData on transition callbacks: one duration and one easing per element, applied to all listed properties. Drop per-property longhand, enter/exit deltas, cubicBezier, and corner radius — each with an explicit "Deferred Until Upstream Clay" entry in §13 referencing nicbarker/clay#603 and the forthcoming exit-flag work. Easings are plain string literals ("linear" | "easeIn" | "easeOut" | "easeInOut") since v1 has no parametric easings. * ✨ add transition property names, bitmask helpers, and Easing * ✨ add transition field type to OpenElement * ✨ encode transition block in pack() * ✨ register Clay handlers, interpolate on property change Co-Authored-By: * ✨ reset deltaTime to 0 after idle (preserve transitions across long gaps) * ✅ verify color transitions work in line mode * 🎨 apply deno fmt and clang-format * ✨ add transitions demo (collapsing sidebar) * 📝 reference transitions-spec from renderer-spec * ✨ rewrite transitions demo as interactive full-screen menu * ✨ add clay-transitions demo port (v1-compatible subset) Ports the spirit of the raylib-transitions demo to clayterm: a 4×4 grid of colored boxes that animate position, size, and bg color. Shuffle (s) animates positions via Clay's transition system; recolor (c) toggles between two palettes with animated bg interpolation; hover tints each box by blending its bg toward white (overlay-color field is not yet in the v1 command buffer, so lighten-on-hover substitutes). Full mouse tracking is wired via mouseTracking() + pointer state from input events. * 🎨 let clay-transitions demo rows fill available height * 🎨 remove modeline from clay-transitions demo * 📝 note ct_active_context is a workaround for Clay userData PR * 🎨 use border-only boxes in clay-transitions demo * 🎨 prevent menu text from wrapping during sidebar transition * 🔥 drop unused grow import in transitions-run test * Update transitions branch with upstream changes (#53) * ✨ add snapshot() for pre-packing directive subtrees Introduces a new `snapshot(ops)` constructor that pre-packs a directive array into its transfer encoding. The returned opaque `Op` can be spliced into any directive array, and during packing its bytes are copied directly into the command buffer without re-encoding. This enables higher-level frameworks to implement dirty tracking: unchanged component subtrees can reuse a cached snapshot, skipping the per-frame packing cost entirely. * docs: add maintainer build guide * 💄 format build docs * 🐛 improve pack string overflow errors * os matrix test in ci (#36) * 🧼 optimize build * 🧼 compress bundled wasm * 🧼 optimize wcwidth.c size * ⚙️ update npm settings * 🐛 install wasm-opt in ci * ⚡ use brotli-11 + z85 wasm encoding * 📌 pin @types/node to v22 * 🧼 apply @ghostdevv review suggestions from PR #35 Co-Authored-By: ghostdevv <git@willow.sh> * 🔨 add type to bundle-wasm * 💌 signed, sealed, delivered * chore: use hashes for versions * chore: don't save git credentials * chore: use array syntax for some reason the schema for the actions wants it to be an array * perf: set concurrency limits to reduce cost and improve dx Without this it means that, for example, if I push a change to a PR then shortly push again this workflow will be running twice. This change will cancel the old run before starting the new one, which reduces the overall actions cost and DX as you don't have extra runs * chore: use hashes for versions * chore: update node version * perf: set concurrency limits to reduce cost and improve dx Without this it means that, for example, if I push a change to a PR then shortly push again this workflow will be running twice. This change will cancel the old run before starting the new one, which reduces the overall actions cost and DX as you don't have extra runs * chore: don't save git credentials * chore: mitigate potential template injection See https://docs.zizmor.sh/audits/#template-injection * chore: update ::set-output command to new syntax https://github.blog/changelog/2022-10-11-github-actions-deprecating-save-state-and-set-output-commands/ * chore: use hashes for versions * chore: don't save git credentials * chore: limit id-token permission to the publishing steps * chore: explicitly disable npm cache to mitigate cache poisoning attacks * chore: mitigate potential template injection See https://docs.zizmor.sh/audits/#template-injection * chore: use oidc This should be using OIDC for publishing * chore: update ::set-output command to new syntax https://github.blog/changelog/2022-10-11-github-actions-deprecating-save-state-and-set-output-commands/ * 🙅 revert aggressive optimization experiments pending benchmark Reverts three changes that need benchmarks before landing: - -Oz / wasm-opt - brotli+Z85 wasm compression - wcwidth.c rewrite * Add CodSpeed performance benchmarks * 🔨 use deno * 🧼 deno fmt * ⚙️ vitest -> tinybench * 🧼 remove codspeed assets * 🔨 fix ci * add type module * fmt * chore: update github url (#38) * downgrade to tinybench@5 * move to examples folder with readme * 🔧 export animating from wasm build * 🧪 cover transitions in snapshots and validation * ✅ enforce nonnegative transition duration --------- Co-authored-by: Charles Lowell <cowboyd@frontside.com> Co-authored-by: Jacob Bolda <me@jacobbolda.com> Co-authored-by: Nate Moore <nate@natemoo.re> Co-authored-by: ghostdevv <git@willow.sh> Co-authored-by: Nate Moore <git@natemoo.re> Co-authored-by: codspeed-hq[bot] <117304815+codspeed-hq[bot]@users.noreply.github.com> Co-authored-by: Nate Moore <natemoo-re@users.noreply.github.com> * remove uncecessary note * clean up and consolidate * re-organize transitions examples * 🔥remove redundant cast --------- Co-authored-by: Ryan Rauh <rauhryan@users.noreply.github.com> Co-authored-by: Jacob Bolda <me@jacobbolda.com> Co-authored-by: Nate Moore <nate@natemoo.re> Co-authored-by: ghostdevv <git@willow.sh> Co-authored-by: Nate Moore <git@natemoo.re> Co-authored-by: codspeed-hq[bot] <117304815+codspeed-hq[bot]@users.noreply.github.com> Co-authored-by: Nate Moore <natemoo-re@users.noreply.github.com> Co-authored-by: Ryan Rauh <rauh.ryan@gmail.com>

natemoo-re force-pushed the ref/opt branch from 6f0c618 to e67760b Compare May 23, 2026 16:52

natemoo-re mentioned this pull request May 23, 2026

🧼 split layout and input entrypoints #37

Open

dreyfus92 approved these changes May 23, 2026

View reviewed changes

ghostdevv requested changes May 23, 2026

View reviewed changes

Comment thread tasks/build-npm.ts Outdated

Comment thread tasks/build-npm.ts

Comment thread tasks/build-npm.ts Outdated

natemoo-re added a commit that referenced this pull request May 24, 2026

apply @ghostdevv review suggestions from PR #35

51cc77c

Co-Authored-By: ghostdevv <ghostdevbusiness@gmail.com>

natemoo-re force-pushed the ref/opt branch from 51cc77c to 820cde5 Compare May 24, 2026 00:08

natemoo-re added a commit that referenced this pull request May 24, 2026

apply @ghostdevv review suggestions from PR #35

bd406b2

Co-Authored-By: ghostdevv <47755378+ghostdevv@users.noreply.github.com>

natemoo-re force-pushed the ref/opt branch from 820cde5 to bd406b2 Compare May 24, 2026 00:26

natemoo-re and others added 10 commits May 23, 2026 22:35

🧼 optimize build

982ced5

🧼 compress bundled wasm

c14f591

🧼 optimize wcwidth.c size

82c9050

⚙️ update npm settings

f8c5c1b

🐛 install wasm-opt in ci

1ec215b

⚡ use brotli-11 + z85 wasm encoding

c6527e2

📌 pin @types/node to v22

f470b1f

🧼 apply @ghostdevv review suggestions from PR #35

db90740

Co-Authored-By: ghostdevv <git@willow.sh>

🔨 add type to bundle-wasm

1ea98a5

💌 signed, sealed, delivered

3437804

natemoo-re force-pushed the ref/opt branch from 122d57c to 3437804 Compare May 24, 2026 02:38

ghostdevv reviewed May 24, 2026

View reviewed changes

Comment thread deno.json Outdated

🙅 revert aggressive optimization experiments pending benchmark

79052b0

Reverts three changes that need benchmarks before landing: - -Oz / wasm-opt - brotli+Z85 wasm compression - wcwidth.c rewrite

natemoo-re requested a review from cowboyd May 24, 2026 17:29

cowboyd approved these changes May 26, 2026

View reviewed changes

natemoo-re merged commit ef9a3ed into main May 26, 2026
7 checks passed

natemoo-re deleted the ref/opt branch May 26, 2026 22:48

This was referenced May 28, 2026

🧼 build input.wasm for size (layout stays -O2) #47

Closed

🧼 generate compact wcwidth tables #49

Open

🧼 compress bundled wasm with brotli + z85 encoding #50

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

🧼 optimize install size#35

🧼 optimize install size#35
natemoo-re merged 11 commits into
mainfrom
ref/opt

natemoo-re commented May 23, 2026 •

edited

Loading

Uh oh!

pkg-pr-new Bot commented May 23, 2026 •

edited

Loading

Uh oh!

dreyfus92 left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

cowboyd commented May 24, 2026 •

edited

Loading

Uh oh!

Uh oh!

cowboyd commented May 24, 2026

Uh oh!

natemoo-re commented May 24, 2026 •

edited

Loading

Uh oh!

cowboyd left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Uh oh!

Uh oh!

Conversation

natemoo-re commented May 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pkg-pr-new Bot commented May 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dreyfus92 left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

cowboyd commented May 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Unalloyed Wins:

Needs consideration

Uh oh!

Uh oh!

cowboyd commented May 24, 2026

Uh oh!

natemoo-re commented May 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

cowboyd left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

natemoo-re commented May 23, 2026 •

edited

Loading

pkg-pr-new Bot commented May 23, 2026 •

edited

Loading

cowboyd commented May 24, 2026 •

edited

Loading

natemoo-re commented May 24, 2026 •

edited

Loading