Repository navigation
Merged
Conversation
… under __TEST__, differential, then swap Co-authored-by: Cursor <[email protected]>
…r __TEST__ beside the existing deciders core/seat.ts computes the reader seat (world/lane/ahead/posture) and the slot table (U/F/L/D/V/Vl/Vp/A/B rows) beside today's deciders; under __TEST__ every read and every recompute compares the two and records disagreements (INV-seat-1..3) into a registry (SEAT_SHADOW_REPORT / __SEAT_SHADOW_STRICT__). Nothing reaches a build: every statement is behind __TEST__ (false in every rollup define); SIGNALS_TEST_DEFINE=true builds a dev dist with it on for the solid/web/generator differential. Design doc §3/§6/§7 updated with the harness as built, the table rows found wrong against a rule (fixed in seat.ts), and the two-readings items the differential leaves open. Co-authored-by: Claude via Cursor <[email protected]>
…d published-inputs table rows, results in the design doc The 50,000-case generator runs (seeds 3289, 91501, 7, every pair) found two table rows wrong against a rule, fixed in seat.ts (test-only): - L4 / L2 / V1' (D-41): a held node's staged error throws for a leaf or a born-into frame reader; the rows said P. - V3 (F-9, #3305): a verdict reader's plain read of a node staged and in flight with INPUTS_PUBLISHED is the flight's (V4', NR), not the staging. Design doc §4.2 records the differential results (suites, 2k, 50k) and §6 the two rows; RULES-INDEX regenerated for the new #3305 citation. Build output unchanged (size harness: 15/15 bundling scenarios byte-identical). Co-authored-by: Claude via Cursor <[email protected]> Co-authored-by: Cursor <[email protected]>
…ost-fix re-run, plain-build abort indices, untraced residue The 50k row counted seed 3289's classes (133) against the three-seed reads/passes/instances; the aggregate is 220 classes / 8,821 instances before the two 50k table fixes and 209 / 8,706 re-run with the table as committed (the 11 classes the fixes cover go, none appear). The infinite loop abort on each seed is verified against a dist built without SIGNALS_TEST_DEFINE, at the same case index on both builds (7: 1158, 3289: 3963, 91501: 5807). Classes no §6 finding or §7 item names literally (623 instances / 43 classes) are listed and left untraced. Co-authored-by: Claude via Cursor <[email protected]>
…d / chosen 2026-10-28 Co-authored-by: Cursor <[email protected]>
…is the hold's own work, not a persisted posture; swap pins corrected Co-authored-by: Cursor <[email protected]>
…t @ 5e68925 Table fixes (design doc §6.1): V1'' staged error → E (D-41); the verdict seat is the hold the pass's own node's verdict lane names (vhold); the branch world (creatorsWorld) walks the owner chain of running passes (#4341); ahead only from a lane or verdict seat; verdict reads leave marks and do not raise; Vl4' (latest of a node staged and in flight serves the staging); D8 needs a value (u f is NR); lane rows for a boundary's structure are a derivation's (A15/#4283 — today is right, renderLeaf is EFFECT_RENDER); fallback's read(b._fb) brackets the seat (shadowSeatSave/Restore, #4337); INV-seat-1 amended per D-27 (a leaf that read a born node or the hold's own staging is in the hold's world). shadowKnown allow-list of next bugs the swap fixes by construction, each filed with a generator replay: #4426 (V1' P→C), #4427 (V6 Lv→P), #4428 (B2 born held via the owner-context creator), #4429 (probing leaks into a pulled pass; Vl4 world). §7 F-transitive reopened: today is a third reading; one-level and transitive each differ from it in disjoint born-held shapes. Differential: signals 393 files / solid / web / visibility oracle / generator 2k × {7, 3289, 91501} — zero classes besides the known entries; zero screen differences against next recordings. Build output byte-identical to next (scripts/size: every scenario equal in brotli and minified). Co-authored-by: Claude via Cursor <[email protected]>
🦋 Changeset detectedLatest commit: a4794c4 The changes in this PR will be included in the next version bump. This PR includes changesets to release 15 packages
Not sure what this means? Click here to learn what changesets are. Click here if you're a maintainer who wants to add another changeset to this PR |
…-29): the swap is stopped, S0+M1 land alone as the spec Appends §8 with the measured record from #4430: the four-column size table and the per-module floor breakdown after the levers; the deletions proven, not subsumed and undecidable; the 50k differential inventory with milestone attribution (20 of one D-37 class at M3b, two next bugs closed, two regressions — 3289 #39609 at M3a, 7 #45702 at M3b); the decision (the table replaces the deciders, ≈ −540 B in core.js, but S1 is additive — +294 B brotli on the core floor, +1,061 B on + isPending/latest — because the growth lives in the seam/reveal/hold layer the table does not touch); method lessons (2k is not evidence for a ruling; size is measured before design); what is salvaged and what the next decision is. §5 marked stopped; the brief's §4 is superseded by §8. Co-authored-by: Claude via Cursor <[email protected]> Co-authored-by: Cursor <[email protected]>
Size (brotli, eager entry chunk)
|
Coverage Report for CI Build 41019788916Coverage remained the same at 78.259%Details
Uncovered ChangesNo uncovered changes found. Coverage RegressionsNo coverage regressions found. Coverage Stats
💛 - Coveralls |
Merging this PR will regress 2 benchmarks2 warnings
Warning Please fix the performance issues or acknowledge them on CodSpeed. Performance Changes
Tip Investigate this regression by commenting Comparing 
Footnotes
|
…adoption, lanes split)
… __TEST__ shadow puts it past 5 s under coverage
ryansolid
added a commit
that referenced
this pull request
Oct 30, 2026
… at its stop (#4448) Follow-ups from #4438's review. - verify-release-artifacts.mjs (run by release.mjs before anything is published) fails when a fixed package's tarball carries `__SEAT_SHADOW__`, the mark a SIGNALS_TEST_DEFINE=true build leaves in @solidjs/signals' dist/dev-shared.js. A shadow build is a test build (rollup.config.js: "never for measuring or publishing"). - reader-seat-brief.md §4 says the swap stopped and points at reader-seat-rebuild.md §8, which supersedes it. Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 Co-authored-by: Claude <[email protected]>
This was referenced Oct 30, 2026
Merged
ryansolid
pushed a commit
that referenced
this pull request
Oct 30, 2026
Resolves verdict.ts beside #4438's shadow hook (the window's state is restored before the shadow's decision point) and the generated RULES-INDEX.md (regenerated). Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2
ryansolid
pushed a commit
that referenced
this pull request
Oct 30, 2026
…ask's flush The reader-seat table (#4438) predicted NotReady (row D3) for a memo with no value reading a first answer staged in the flush that re-asks it; under D-22 (amended 2026-10-30) the memo derives from it once. Row D3' predicts the staging (P), and `read` records the slot it serves. __TEST__ only: dist is unchanged. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2
ryansolid
pushed a commit
that referenced
this pull request
Oct 30, 2026
Resolves the generated RULES-INDEX.md (regenerated) after #4438. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2
ryansolid
added a commit
that referenced
this pull request
Oct 30, 2026
…s answered, then re-asked (D-22, F-9) (#4459) * fix(signals): a first answer landing in the tick a write re-asks it is answered, then re-asked (D-22, F-9) An async memo's first answer resolves, and in the same synchronous turn a plain write re-asks it; the promise reaction runs after the write, so the landing and the re-ask are processed in one flush. The runtime treated the answer as never produced: - `asyncWrite` dropped it outright when the write's notification had marked the node dirty (a stale-answer guard meant for re-asks). - Otherwise it landed staged, the re-ask left the node pending and uninitialized, and `isPending` gave it no verdict (`false` from the flush's probe, never re-run), while an identity memo or a derived store over it never derived from it (`latest` of either threw: [ROOT-L]). Now a first answer always lands; `isPending` of a node answered and re-asked in one flush is the re-ask's verdict (true); and a memo with no value of its own that reads it derives from the answer once, then re-runs on the missed-wake latch and waits on the re-ask holding it (a verdict engine hook, so the hello-world floor pays only the call). Amends D-22's 2026-10-28 clause per the 2026-10-30 ruling: a plain write re-asking the memo carries the answer like an action's. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * fix(signals): a first answer landing in the tick a write re-asks it is answered, then re-asked (D-22, F-9) An async memo's first answer resolves, and in the same synchronous turn a plain write re-asks it; the promise reaction runs after the write, so the landing and the re-ask are processed in one flush. The runtime treated the answer as never produced: - `asyncWrite` dropped it outright when the write's notification had marked the node dirty (a stale-answer guard meant for re-asks). - Otherwise it landed staged, the re-ask left the node pending and uninitialized, and `isPending` gave it no verdict (`false` from the flush's probe, never re-run), while an identity memo or a derived store over it never derived from it (`latest` of either threw: [ROOT-L]). Now a first answer always lands; `isPending` of a node answered and re-asked in one flush is the re-ask's verdict (true); and a memo with no value of its own that reads it derives from the answer once, then re-runs on the missed-wake latch and waits on the re-ask holding it (in `read`'s pending path, behind the verdict engine's hook so the floor folds it). Amends D-22's 2026-10-28 clause per the 2026-10-30 ruling: a plain write re-asking the memo carries the answer like an action's. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * test(signals): the seat shadow seats a first answer served in its re-ask's flush The reader-seat table (#4438) predicted NotReady (row D3) for a memo with no value reading a first answer staged in the flush that re-asks it; under D-22 (amended 2026-10-30) the memo derives from it once. Row D3' predicts the staging (P), and `read` records the slot it serves. __TEST__ only: dist is unchanged. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * chore(size): raise caps for the same-tick first answer isPending/latest +37 B brotli (+151 B minified) and the compiled live page: the first-answer serve in `read`'s pending path and the `isPending` verdict for it. The observe-attribution and compiled scenarios carry the core half (+10 B minified on every floor; the rest is brotli layout). Caps at measured + 10 B rounded up to 0.01 KB; recorded minified = measured. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * chore(size): raise the live page's frozen floor for the same-tick first answer 38.76 KB -> 38.79 KB, recorded minified 117,227 -> 117,379 B, measured at 38,778 B brotli. Size-Exception in the PR body. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * fix(signals): serve an answered-then-reasked first answer only to a reader waiting on it, off lanes, at any depth read()'s D-22 branch served the staged first answer to a reader that did not wait on it (a first pass: the seat table's D3 row predicts NR) and to a verdict lane's pass (V6: D-37, a lane sees the screen plus its own guesses). It now requires passLane === null and that the reader wait on the source a read of the node parks on (_error.source): the node itself one level down, the async it derives from at any depth, so identity(identity(m)) reads as m does (E2 at depth 2). The seat's D3' row says the same (!inSeat, the parked-on source, no staging of its own). handleAsync's answered-then-reasked guard is the verdict engine's (holdWorkHook): plain-only bundles keep the plain return. D-23 exception 1: an answered-then-reasked node is not a first load. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * fix(signals): type the parked-on source read in read() and the D3' row Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * chore(size): re-record #4459's caps over next @ 131ceb8 isPending/latest, compiled live SC and live + router grow with the verdict-engine branch (+159..+161 B minified); no frozen floor moves: plain-only bundles are unchanged in minified bytes. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * chore(size): re-record the live page floor's minified for #4459 Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * fix(signals): serve an answered-then-reasked first answer one pass a flush, out of line read()'s D-22 arm served a reader whose pass then threw anyway (a read beside the served one parked on the same source): the throw re-parked it on that source, the missed-wake latch re-ran it, and the re-run qualified again — a synchronous loop within one flush (async-gen 91501 #749, minimized: a boundary's children memo over two Shows that both park on the re-asked memo). The arm now stamps the served pass's clock on the reader (`_x._served`) and serves a reader once a flush; the re-run throws and waits. The seat's D3' row says the same. The arm moves out of read() into `servesStaged`, called behind `observeFlightHook` on the pending-throw path, so read() itself grows by a call (bytecode 1318 -> 1170; next 1139). Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * fix(signals): an answered-then-reasked first answer reaches a reader its landing released servesStaged served a reader with no value only while it still waited on the source. When the memo's input is read through latest() or a sync memo, the write's re-ask reaches the memo after its landing ran, and the landing had already released the identity memo's wait: it ran with no value, waiting on nothing, and threw, so latest(identity(m)) and a derived store stayed on [ROOT-L] while latest(m) showed the answer (async-gen E2/E3 3289 #310, 91501 #1993, #2609). D-22 as amended says a derivation with no value of its own reads like the memo at any depth; which the flush ran first is not the reader's. A reader that is not pending is served too (one served pass a flush, as before); one pending on another source is not. The seat's D3' row says the same. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * chore(size): re-measure the answered-then-reasked serve over next @ 230c670 Live server components 38.85 -> 38.96 KB (117,747; Size-Exception in the PR body), isPending/latest 11.17 -> 11.26 KB (31,002), compiled live 42.79 -> 42.85 KB (125,887); live + router re-records minified (143,150). Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * chore(size): re-measure the answered-then-reasked serve over next @ d0fc285 Takes next's ledger (recorded after #4471/#4479) and re-records only the scenarios this PR moves: isPending/latest 11.19 -> 11.28 KB (31,053), live server components 38.91 -> 38.96 KB (117,793; Size-Exception in the PR body), compiled live 42.83 -> 42.87 KB (125,938), live + router 48.51 -> 48.58 KB (143,201). Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 --------- Co-authored-by: Claude <[email protected]>
ryansolid
added a commit
that referenced
this pull request
Oct 30, 2026
…des the write's flush (#4495) (#4499) * fix(signals): a first answer landing in the tick a write re-asks it is answered, then re-asked (D-22, F-9) An async memo's first answer resolves, and in the same synchronous turn a plain write re-asks it; the promise reaction runs after the write, so the landing and the re-ask are processed in one flush. The runtime treated the answer as never produced: - `asyncWrite` dropped it outright when the write's notification had marked the node dirty (a stale-answer guard meant for re-asks). - Otherwise it landed staged, the re-ask left the node pending and uninitialized, and `isPending` gave it no verdict (`false` from the flush's probe, never re-run), while an identity memo or a derived store over it never derived from it (`latest` of either threw: [ROOT-L]). Now a first answer always lands; `isPending` of a node answered and re-asked in one flush is the re-ask's verdict (true); and a memo with no value of its own that reads it derives from the answer once, then re-runs on the missed-wake latch and waits on the re-ask holding it (a verdict engine hook, so the hello-world floor pays only the call). Amends D-22's 2026-10-28 clause per the 2026-10-30 ruling: a plain write re-asking the memo carries the answer like an action's. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * fix(signals): a first answer landing in the tick a write re-asks it is answered, then re-asked (D-22, F-9) An async memo's first answer resolves, and in the same synchronous turn a plain write re-asks it; the promise reaction runs after the write, so the landing and the re-ask are processed in one flush. The runtime treated the answer as never produced: - `asyncWrite` dropped it outright when the write's notification had marked the node dirty (a stale-answer guard meant for re-asks). - Otherwise it landed staged, the re-ask left the node pending and uninitialized, and `isPending` gave it no verdict (`false` from the flush's probe, never re-run), while an identity memo or a derived store over it never derived from it (`latest` of either threw: [ROOT-L]). Now a first answer always lands; `isPending` of a node answered and re-asked in one flush is the re-ask's verdict (true); and a memo with no value of its own that reads it derives from the answer once, then re-runs on the missed-wake latch and waits on the re-ask holding it (in `read`'s pending path, behind the verdict engine's hook so the floor folds it). Amends D-22's 2026-10-28 clause per the 2026-10-30 ruling: a plain write re-asking the memo carries the answer like an action's. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * test(signals): the seat shadow seats a first answer served in its re-ask's flush The reader-seat table (#4438) predicted NotReady (row D3) for a memo with no value reading a first answer staged in the flush that re-asks it; under D-22 (amended 2026-10-30) the memo derives from it once. Row D3' predicts the staging (P), and `read` records the slot it serves. __TEST__ only: dist is unchanged. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * chore(size): raise caps for the same-tick first answer isPending/latest +37 B brotli (+151 B minified) and the compiled live page: the first-answer serve in `read`'s pending path and the `isPending` verdict for it. The observe-attribution and compiled scenarios carry the core half (+10 B minified on every floor; the rest is brotli layout). Caps at measured + 10 B rounded up to 0.01 KB; recorded minified = measured. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * chore(size): raise the live page's frozen floor for the same-tick first answer 38.76 KB -> 38.79 KB, recorded minified 117,227 -> 117,379 B, measured at 38,778 B brotli. Size-Exception in the PR body. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * fix(signals): serve an answered-then-reasked first answer only to a reader waiting on it, off lanes, at any depth read()'s D-22 branch served the staged first answer to a reader that did not wait on it (a first pass: the seat table's D3 row predicts NR) and to a verdict lane's pass (V6: D-37, a lane sees the screen plus its own guesses). It now requires passLane === null and that the reader wait on the source a read of the node parks on (_error.source): the node itself one level down, the async it derives from at any depth, so identity(identity(m)) reads as m does (E2 at depth 2). The seat's D3' row says the same (!inSeat, the parked-on source, no staging of its own). handleAsync's answered-then-reasked guard is the verdict engine's (holdWorkHook): plain-only bundles keep the plain return. D-23 exception 1: an answered-then-reasked node is not a first load. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * fix(signals): type the parked-on source read in read() and the D3' row Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * chore(size): re-record #4459's caps over next @ 131ceb8 isPending/latest, compiled live SC and live + router grow with the verdict-engine branch (+159..+161 B minified); no frozen floor moves: plain-only bundles are unchanged in minified bytes. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * chore(size): re-record the live page floor's minified for #4459 Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * fix(signals): serve an answered-then-reasked first answer one pass a flush, out of line read()'s D-22 arm served a reader whose pass then threw anyway (a read beside the served one parked on the same source): the throw re-parked it on that source, the missed-wake latch re-ran it, and the re-run qualified again — a synchronous loop within one flush (async-gen 91501 #749, minimized: a boundary's children memo over two Shows that both park on the re-asked memo). The arm now stamps the served pass's clock on the reader (`_x._served`) and serves a reader once a flush; the re-run throws and waits. The seat's D3' row says the same. The arm moves out of read() into `servesStaged`, called behind `observeFlightHook` on the pending-throw path, so read() itself grows by a call (bytecode 1318 -> 1170; next 1139). Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * fix(signals): an answered-then-reasked first answer reaches a reader its landing released servesStaged served a reader with no value only while it still waited on the source. When the memo's input is read through latest() or a sync memo, the write's re-ask reaches the memo after its landing ran, and the landing had already released the identity memo's wait: it ran with no value, waiting on nothing, and threw, so latest(identity(m)) and a derived store stayed on [ROOT-L] while latest(m) showed the answer (async-gen E2/E3 3289 #310, 91501 #1993, #2609). D-22 as amended says a derivation with no value of its own reads like the memo at any depth; which the flush ran first is not the reader's. A reader that is not pending is served too (one served pass a flush, as before); one pending on another source is not. The seat's D3' row says the same. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * chore(size): re-measure the answered-then-reasked serve over next @ 230c670 Live server components 38.85 -> 38.96 KB (117,747; Size-Exception in the PR body), isPending/latest 11.17 -> 11.26 KB (31,002), compiled live 42.79 -> 42.85 KB (125,887); live + router re-records minified (143,150). Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * chore(size): re-measure the answered-then-reasked serve over next @ d0fc285 Takes next's ledger (recorded after #4471/#4479) and re-records only the scenarios this PR moves: isPending/latest 11.19 -> 11.28 KB (31,053), live server components 38.91 -> 38.96 KB (117,793; Size-Exception in the PR body), compiled live 42.83 -> 42.87 KB (125,938), live + router 48.51 -> 48.58 KB (143,201). Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 * fix(signals): every answer resolved in the tick a write re-asks it rides the write's flush (#4495) Two async memos m1, m2 answer, then a write re-asks both, in one tick. m1's landing flushed synchronously (`asyncWrite`), and that flush carried the queued write: it re-asked m2 before m2's answer's reaction ran, so the answer arrived superseded (`_inFlight !== result`) and was dropped. A memo over both then had no value and `latest` of it threw (D-22 says 13). A landing that arrives while a write's flush is queued (a staged value in `pendingNodes`, captured before the landing stages its own) no longer flushes: the write's flush carries it with every answer whose reaction is queued ahead of it (F-10), each answered, then re-asked (D-22, F-9). Only `latest` and `isPending` tell the difference, so it is gated on the verdict engine (`holdWorkHook`); a stream's next pull still waits on its landing's flush. Size: isPending/latest 11.28 -> 11.30 KB (31,087 B), compiled live server components 42.87 -> 42.90 KB (125,978 B); hello world unchanged. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_015hPxMkC3jg3Uivh65gvaf2 --------- Co-authored-by: Claude <[email protected]>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
S0 + M1 of
documentation/plans/reader-seat-rebuild.md, landed on their own as a CI spec. The S1 swap that was to follow (#4430) is stopped — the record and the decision are the design doc's new §8 "S1 result and decision (2026-10-29)", which ships in this PR. Supersedes #4395 (the original S0 draft, before the M1 re-baseline).What it is
packages/signals/src/core/seat.ts(test-only,__TEST__) computes the reader seat —{world, lane, ahead, posture}per pass — and the slot table from the brief: for every read, which world the reader sees (screen C, staging P, lane Lv/Dv, truth T, NR, E) and which marks it raises, each row carrying the rule id it is derived from (design doc §2.3, §6).__TEST__everyread,verdictValue,laneReadand everyrecomputehead/tail runs the table beside today's deciders (frameRead,readerSeesCommitted,serve,laneRead,verdictRead,route, the born-held arm, …) and compares slot, marks, seating and born-held decision; a disagreement is recorded as a mismatch class (kind · row · expected → actual · detail). Nothing in the production path changes behaviour: every call isif (__TEST__)-guarded.INV-seat(invariants.ts,INTERNALS-ASYNC-STATE.md): INV-seat-1..3 — a leaf never raises its world beyond its reads (amended per D-27 for the born-node read), posture is never inherited, the shadow agrees with the deciders.tests/setup/seat-shadow-report.ts(vitest setup, on whenSEAT_SHADOW_REPORT=<file>is set;SEAT_SHADOW_STRICT=1throws at the first disagreement),scripts/seat-shadow-summary.mjs(aggregates to a mismatch table),shadowKnown(the filednextbugs carried by name), andSIGNALS_TEST_DEFINE=true pnpm --filter @solidjs/signals buildto emit a dev dist with__TEST__on for the solid/web/generator runs (design doc §3.4). The default build is unchanged.How to run
cd packages/signals SEAT_SHADOW_REPORT=/tmp/seat-signals.jsonl pnpm vitest run node scripts/seat-shadow-summary.mjs /tmp/seat-signals.jsonl [--rows]For solid / web / the generator, build the signals dist with
SIGNALS_TEST_DEFINE=truefirst and setSEAT_SHADOW_REPORTthe same way (§3.4); rebuild plain (pnpm build) before measuring size.Results
Re-run on this branch (head e300252's tree, verified identical to
seat/s1-swap@ 4c413e1):M1's full table (design doc §4.3, measured at 4c413e1): signals 0 · visibility oracle 0 · solid 0 · web 0 · generator 2,000 × seeds 7 / 3289 / 91501 (every pair) 0 — zero mismatch classes everywhere besides the four
shadowKnownentries, each a filed bug onnextwith a generator replay in its issue: #4426 (V1' P→C), #4427 (V6 Lv→P), #4428 (B2 born held via the owner's stale flag), #4429 (Vl4probingleak). The 50k × 3 S0 run and its residue are in §4.2.Byte-identity (
pnpm build, default__TEST__false;scripts/sizeof this branch against a cleanorigin/next@ 5e68925 build, same native compiler): 22 / 22 scenarios identical in both brotli and minified bytes (core floor 7,415 / 19,908;+ isPending/latest11,026 / 30,408; compiled floor 9,961 / 27,271; server floor 1,275 / 3,255; …).packages/{solid,web}/distdiff -rqempty;packages/signals/dist/{prod,observe}differ only in statement braces and import order left where__TEST__blocks were removed (the minifier erases both — the scenario bundles are the proof). Addeddist/types/core/seat.d.tsis types only.Why it lands alone
S1 (the swap, #4430) was built through M5 and stopped on its gate: the table replaces the read/seating deciders (≈ −540 B minified in
core.js) but S1 is additive overall — +294 B brotli on the core floor, +1,061 B on+ isPending/latestover rc.20 — because the growth lives in the seam/reveal/hold layer the table does not touch, and the 50k differential found two regressions there (3289 #39609 at M3a, 7 #45702 at M3b) plus one class of 20 awaiting a ruling. Design doc §8 has the size table, the deletions, the inventory with milestone attribution, the method lessons (2k generator cases are not evidence for a ruling; size is measured before design) and the salvage list. §5 is marked stopped; the brief's §4 is superseded by §8.What this PR keeps is the part that proved itself: a spec of today's read decision that CI can run against the deciders, with 0 classes, at 0 bytes.
Commits
Cherry-picked from
seat/s1-swap— S0 (3609040, 40208a0, b71aafe, 4c6a2ee, 7718526, 697217f) and M1 (4c413e1) — ontonext; the resulting tree is identical to 4c413e1 (git diffempty). Plus the §8 doc commit. Changesets:.changeset/reader-seat-shadow.md,.changeset/reader-seat-s1-rebaseline.md(both@solidjs/signals: patch, test-only).Public API changes
None.