Ci benchmark workflows

[brnrdog]

No description provided.

[github-actions]

ReScript Signals benchmark: PR vs main

Compared implementations:

• ReScript Signals (main)
• ReScript Signals (PR)

Overall:

Version	Total ms	Avg ms/test
ReScript Signals (main)	17575.19	878.76
ReScript Signals (PR)	18012.50	900.62
Delta (ReScript Signals (PR) - ReScript Signals (main))	437.31	2.49%

Per-test delta (lower is better):

Test	Main ms	PR ms	Diff ms	Diff %
2-10x5 - lazy80%	708.02	769.48	61.46	8.68%
25-1000x5	4811.98	4973.60	161.63	3.36%
3-5x500	358.14	387.26	29.12	8.13%
4-1000x12 - dyn5%	7275.85	7364.27	88.42	1.22%
6-100x15 - dyn50%	515.37	539.13	23.76	4.61%
6-10x10 - dyn25% - lazy80%	434.64	452.80	18.15	4.18%
avoidablePropagation	386.97	386.37	-0.59	-0.15%
broadPropagation	332.05	342.18	10.13	3.05%
cellx1000	22.69	21.95	-0.74	-3.27%
cellx2500	58.43	59.16	0.73	1.25%
createComputations	371.51	307.89	-63.61	-17.12%
createSignals	5.17	4.95	-0.22	-4.22%
deepPropagation	126.93	131.16	4.23	3.34%
diamond	232.27	246.75	14.48	6.23%
molBench	39.67	39.81	0.15	0.37%
mux	416.67	432.81	16.13	3.87%
repeatedObservers	36.90	37.85	0.95	2.57%
triangle	82.87	78.47	-4.40	-5.31%
unstable	48.73	52.34	3.61	7.41%
updateSignals	1310.32	1384.24	73.91	5.64%

Note: single-machine run in CI. Numbers can vary with runner load and Node/V8 version.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 4886daa71c

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".

[chatgpt-codex-connector]

Guard PR comment upsert against fork pull requests

This step runs for every pull_request event, but forked PRs get a read-only GITHUB_TOKEN; in that context, the github.rest.issues.*Comment calls fail with 403 and the benchmark job is marked failed even though benchmarking succeeded. Add a fork-aware condition (for example, skip when github.event.pull_request.head.repo.fork is true) or move comment writing to a trusted workflow context; the same pattern is also present in .github/workflows/benchmark-pr-vs-main.yml.

Useful? React with 👍 / 👎.

[chatgpt-codex-connector]

Pin benchmark suite revision before cloning

The workflow clones milomg/js-reactivity-benchmark from its moving default branch, so upstream changes can break this repo’s PR checks without any change here. That risk is concrete in this commit because scripts/ci/benchmark-pr-vs-frameworks.mjs expects specific framework names and throws if they change. Cloning a fixed tag or commit SHA will make CI results reproducible and prevent flaky external breakages.

Useful? React with 👍 / 👎.

[github-actions]

Reactivity benchmark: PR vs top frameworks

Compared implementations:

• Alien Signals
• Preact Signals
• SolidJS
• Svelte v5
• Vue
• ReScript Signals (PR)

Overall ranking (lower total ms is better):

Rank	Framework	Total ms	Avg ms/test
1	Alien Signals	3602.86	180.14
2	Preact Signals	4060.88	203.04
3	Vue	5608.95	280.45
4	Svelte v5	8098.11	404.91
5	SolidJS	10007.48	500.37
6	ReScript Signals (PR)	17293.06	864.65

Per-test runtime (ms):

Framework	2-10x5 - lazy80%	25-1000x5	3-5x500	4-1000x12 - dyn5%	6-100x15 - dyn50%	6-10x10 - dyn25% - lazy80%	avoidablePropagation	broadPropagation	cellx1000	cellx2500	createComputations	createSignals	deepPropagation	diamond	molBench	mux	repeatedObservers	triangle	unstable	updateSignals
Alien Signals	335.52	656.32	185.93	503.61	282.56	241.10	95.64	200.85	13.51	34.85	97.64	4.13	62.08	127.90	28.38	167.70	32.48	39.31	47.04	446.33
Preact Signals	426.11	894.63	203.57	573.27	296.43	240.51	112.34	211.34	11.57	31.43	94.81	3.72	86.74	141.17	28.29	163.18	22.10	46.53	36.75	436.38
Vue	620.78	985.86	292.99	714.88	396.66	344.56	244.75	254.19	29.30	84.53	147.26	4.97	114.93	211.25	61.78	243.71	28.81	65.42	51.65	710.68
Svelte v5	1135.07	1015.27	338.85	900.57	405.46	373.95	570.40	420.31	18.37	51.07	169.96	3.50	175.72	434.81	30.17	249.66	76.91	124.02	110.25	1493.78
SolidJS	2099.56	1387.72	488.42	1106.77	652.93	682.88	332.67	554.44	27.16	75.31	204.59	5.21	205.18	400.76	35.37	298.09	94.02	137.08	134.08	1085.24
ReScript Signals (PR)	666.95	4824.42	319.91	7265.32	499.54	417.48	331.24	317.33	24.10	64.02	253.79	5.51	94.37	215.95	40.16	396.10	31.00	64.29	41.21	1420.38

Note: single-machine run in CI. Numbers can vary with runner load and Node/V8 version.