Skip to content

bench: eighth official report, TeaCSS 0.10.5 under spec revision 10 - #53

Merged
billgo merged 1 commit into
masterfrom
bench-0.10.5-report
Nov 3, 2026
Merged

billgo merged 1 commit into
masterfrom
bench-0.10.5-report

Conversation

@billgo

@billgo billgo commented Nov 3, 2026

Copy link
Copy Markdown
Member

The first official run of TeaCSS 0.10: teacss-rs 0.10.5 and teacss-ts 0.10.5 against Tailwind CSS 4.3.3, in six groups (main, default configuration, PostCSS, Rsbuild, Bun, CLI), on Node 26.12.0.

Results at 20 000 utilities in 300 files

teacss-rs Tailwind v4 teacss-ts
Warm generation 42 ms 226 ms 466 ms
+10 classes to a running engine 6.2 ms 116 ms 116 ms
Dev save, +10 classes 74 ms 311 ms 271 ms
Vite build 388 ms 779 ms 1 043 ms
  • teacss-rs is faster than Tailwind in 116 of 117 comparisons, with 1 showing no measurable difference. It now also leads in all 15 small-project build tables.
  • 11 070 samples; load share 21 % to 42 %; verification passed on 78 000 cases.

Report generator fix

The engines and the Vite builds are byte-identical. The dev server sheets differ only in the four bundled fonts: @teacss/vite inlines them as data URIs and @teacss/vite-ts links them. Reducing each font to its digest makes the two sheets identical. The generator had blamed revision 9's Lightning CSS behaviour. It now counts the Vite build and the dev server separately and names the fonts. A test covers this.

Known caveat

Homebrew's Node 26.12.0 is dynamically linked, so an empty process takes 102 ms against 32 ms for the official 26.7.0 binary. Every Node process pays about 70 ms extra, all tools alike. The deltas over each baseline are unaffected, and the build ratios are understated (about 2.2x rather than 2.0x for Vite). The audit records the A/B.

Files

  • notes/reports/2026-11-03T13-22-12-mac.{md,summary.json,audit.md}, results/2026-11-03T13-22-12-mac.json
  • notes/articles/teacss-0.10.5.md and its index row
  • README status and summary block, plus a spec note on the dev fonts

🤖 Generated with Claude Code

The first run of TeaCSS 0.10: teacss-rs 0.10.5 and teacss-ts 0.10.5,
Tailwind CSS 4.3.3, six groups, Node 26.12.0. 11 070 samples, load
share 21 % to 42 %.

The report generator counted the Vite build and dev server together and
explained any difference with revision 9's Lightning CSS note. The builds
are byte-identical again; the dev servers differ only in the bundled
fonts, which @teacss/vite inlines and @teacss/vite-ts links. The report
now counts the two apart and names the fonts.

The audit records that Homebrew's Node 26.12.0 adds about 70 ms to every
Node process. It favours no tool and understates TeaCSS's build ratios.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Nov 3, 2026 •

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-11-03T14:33:54.687628Z 5c046c0 PR opened
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@billgo
billgo merged commit a13c1f8 into master Nov 3, 2026
1 check passed
@billgo
billgo deleted the bench-0.10.5-report branch November 3, 2026 14:33

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 5c046c01e5

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".


| 场景 | TeaCSS(Rust 引擎) | Tailwind CSS v4 | TeaCSS(TypeScript 引擎) |
| --- | --- | --- | --- |
| 全新进程,从启动到生成完 | 55 ms | 314 ms | 640 ms |

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Describe cold-engine timing without including process startup

The row describes elapsed time from process startup to completed generation, but runEngineScenario() starts its cold timer inside the already-running subject process, immediately before importing the subject and calling its engine (packages/benchmark/src/features/measure/engine-scenario.ts:43–46). Node startup and harness initialization are excluded. Consequently, the 55 ms figure is not a spawn-to-output measurement, and the audit's statement that cold-engine timings include the extra approximately 70 ms of Homebrew startup overhead is also incorrect. Label this row as subject loading plus first generation, and clarify that the startup-overhead caveat applies to externally timed builds and CLIs rather than cold-engine samples.

AGENTS.md reference: AGENTS.md:L11-L12

Useful? React with 👍 / 👎.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant