Skip to content

Commit cb872d8

Browse files
ryansolidClaude via Cursor
andcommitted
chore(size): compiled-scenario caps at CI-measured + 10 B
The check job measured the merge commit onto next @ a8c98bd, where #3806 had added 4 B minified to the signals scheduler since the local measurement at b07fed5; the compiled floor landed 3 B over. The caps follow the repo's rule — CI head + 10 B, rounded up to 0.01 KB: floor 10,033 -> 10.05 KB, CSR 25,070 -> 25.08 KB, hydrating 30,899 -> 30.91 KB (unchanged). The compiled app's own bytes are identical on CI and locally (app=229 / 2,575 / 3,058); the delta is the runtime's. Co-authored-by: Claude via Cursor <[email protected]>
1 parent 6c3ef2b commit cb872d8

1 file changed

Lines changed: 25 additions & 19 deletions

File tree

‎scripts/size/scenarios.js‎

Lines changed: 25 additions & 19 deletions
Original file line numberDiff line numberDiff line change
@@ -2956,15 +2956,18 @@ module.exports = [
29562956
// `_$$click` and `delegateEvents`. Its delta against that floor is what
29572957
// the first real template costs over a hand-written app.
29582958
//
2959-
// Added 2026-10-05: first compiled-template scenarios; measured
2960-
// 10,019 B on next @ b07fed550 (28,184 B minified; signals=21,598
2961-
// web=6,098 solid=260 app=229) — +171 B brotli / +531 B minified over
2962-
// the hand-written floor's 9,848: `template`, `delegateEvents` and the
2959+
// Added 2026-10-05: first compiled-template scenarios; CI-measured on
2960+
// next @ a8c98bd2e (Linux): 10,033 B (28,188 B minified; signals=21,602
2961+
// web=6,099 solid=260 app=229) — +187 B brotli / +531 B minified over
2962+
// the hand-written floor's 9,846: `template`, `delegateEvents` and the
29632963
// delegated-handler branch of the dispatcher, and the compiled
2964-
// component itself. Cap at measured + 10 B rounded up to 0.01 KB.
2964+
// component itself. (Measured locally at 10,019 / 28,184 on next @
2965+
// b07fed550 before #3806's +4 B minified in the signals scheduler
2966+
// landed; the compiled app's own bytes are identical on both.) Cap at
2967+
// CI-measured + 10 B rounded up to 0.01 KB.
29652968
path: "fixtures/compiled/floor.jsx",
29662969
compile: { hydratable: false },
2967-
limit: "10.03 KB",
2970+
limit: "10.05 KB",
29682971
alias
29692972
},
29702973
{
@@ -2976,9 +2979,11 @@ module.exports = [
29762979
// `spread` reads a source through, whether or not the app calls
29772980
// `merge` itself (audit §1.3 finding 2).
29782981
//
2979-
// Added 2026-10-05: first compiled-template scenarios; measured
2980-
// 25,128 B on next @ b07fed550 (78,803 B minified; signals=62,628
2981-
// web=12,166 app=2,575 solid=1,431; lazy stats.js 182 B brotli). The
2982+
// Added 2026-10-05: first compiled-template scenarios; CI-measured on
2983+
// next @ a8c98bd2e (Linux): 25,070 B (78,807 B minified; signals=62,632
2984+
// web=12,167 app=2,575 solid=1,431; lazy stats.js 182 B brotli). (Locally
2985+
// 25,128 / 78,803 on next @ b07fed550, before #3806's +4 B minified in
2986+
// the signals scheduler; the brotli layout moved -58 B over it.) The
29822987
// store engine is ~24 KB minified of the signals figure (the app uses
29832988
// createStore, as the `+ createStore` scenario measures); the same app
29842989
// on signals alone measures 54,467 / 18,097 — the audit's §1.3 fixture,
@@ -2989,10 +2994,10 @@ module.exports = [
29892994
// merge still used; merge/omit/mergeProps alone 5,139 / 1,171; the
29902995
// spread alone retains 9.6 KB minified when nothing else uses merge,
29912996
// because it reads sources through the merge/omit view records). Cap at
2992-
// measured + 10 B rounded up to 0.01 KB.
2997+
// CI-measured + 10 B rounded up to 0.01 KB.
29932998
path: "fixtures/compiled/csr.jsx",
29942999
compile: { hydratable: false },
2995-
limit: "25.14 KB",
3000+
limit: "25.08 KB",
29963001
alias
29973002
},
29983003
{
@@ -3006,15 +3011,16 @@ module.exports = [
30063011
// `runHydrationEvents()` calls, `addEvent` in place of the `_$$click`
30073012
// property form).
30083013
//
3009-
// Added 2026-10-05: first compiled-template scenarios; measured
3010-
// 30,898 B on next @ b07fed550 (99,064 B minified; signals=62,812
3011-
// solid=16,599 web=16,594 app=3,058; lazy stats.js 196 B brotli):
3012-
// +5,770 B brotli / +20,261 B minified over the compiled CSR scenario —
3013-
// solid +15,168 (client/hydration.ts, including the store adapters the
3014-
// store import reaches), web +4,428 (`hydrate`, the claim walk,
3014+
// Added 2026-10-05: first compiled-template scenarios; CI-measured on
3015+
// next @ a8c98bd2e (Linux): 30,899 B (99,068 B minified; signals=62,815
3016+
// solid=16,600 web=16,594 app=3,058; lazy stats.js 195 B brotli):
3017+
// +5,829 B brotli / +20,261 B minified over the compiled CSR scenario —
3018+
// solid +15,169 (client/hydration.ts, including the store adapters the
3019+
// store import reaches), web +4,427 (`hydrate`, the claim walk,
30153020
// `getNextElement`/`getNextMarker`/`runHydrationEvents`/`setProperty`),
3016-
// app +483 (the compiled output's own growth). Cap at measured + 10 B
3017-
// rounded up to 0.01 KB.
3021+
// app +483 (the compiled output's own growth). (Locally 30,898 / 99,064
3022+
// on next @ b07fed550, before #3806.) Cap at CI-measured + 10 B rounded
3023+
// up to 0.01 KB.
30183024
path: "fixtures/compiled/hydrating.jsx",
30193025
compile: { hydratable: true },
30203026
limit: "30.91 KB",

0 commit comments

Comments
 (0)