Skip to content

Commit 6e867d5

Browse files
docs(signals): a joined pass's children are the hold's (D-27, from A29) — reader-seat rulings 2026-10-28
Records the maintainer's 2026-10-28 rulings on the reader-seat rebuild's §7 two-readings items (today's behaviour over the brief's reading): - F-leaf: promoted A29 ruling A into D-27 as "A joined pass's children are the hold's" — a pass (a render effect included) that read a node born into the hold joins it, and a node it creates during that pass is born held; a dated italic note under A29 points at the promotion. Pins: lane-uninitialized-landing-3648, boundary-output-frame-reader, loading-fallback-in-flush-3540, fuzz-findings-l2. - F-ahead-posture: no rule change; D-27 as clarified today already rules it. Recorded in the brief's §1 table. - F-vlane: ruled today's behaviour (the fallback stays beside the held frame, #3540/#4337). The drafted D-25 paragraph is NOT written: as drafted it listed the #3884 pending re-pass among re-runs whose plain read of the held node is the screen's, which D-25's standing #3884 clause ("reads plainly — following the change it is fetching for") and superseded-transition-verdicts-3884.test.ts contradict. The brief's row records the ruling and the gap. RULES-INDEX.md regenerated (definition line numbers after A29 shift by two; no counts change). Docs only; no changeset. Co-authored-by: Cursor <[email protected]>
1 parent 7ebd4a3 commit 6e867d5

4 files changed

Lines changed: 58 additions & 51 deletions

File tree

‎documentation/plans/reader-seat-brief.md‎

Lines changed: 14 additions & 11 deletions
Original file line numberDiff line numberDiff line change
@@ -22,17 +22,20 @@ The pattern is the finding: the verdict path has drifted from the rules in
2222
several places at once, and each drift was found by a generator case. That is
2323
the case for the carve. It also makes P0 larger than the design priced.
2424

25-
| ruling | outcome | rule | where it is written |
26-
| ------ | --------------------------------------------------------------------------------- | ------------------------- | ------------------------------------- |
27-
| R1 | both orders show `[committed, true]`; not for the hold's own work | A10, D-25, D-27, #3884 | D-25 (clarified 2026-10-28) |
28-
| R2 | the parent waits for a lane's flights; the lane never waits for the parent | F-6, F-1, F-3, 2026-10-01 | F-6 (amended 2026-10-28, `d1b938d3c`) |
29-
| R3 | an untracked read of a live guess sees the guess, lane held or not | F-11, A17 | F-1 gloss (clarified, `40cde2cd6`) |
30-
| R4 | a pass reading two holds' display-ahead values is one reveal unit | A15 #3335 | D-25 |
31-
| R5 | `latest()` opens no hold; the pass re-derives once at a park | F-10, D-2, A10 | D-22 |
32-
| R6 | E8 is an equivalence for render effects only | D-2, F-9 | no change |
33-
| R7 | `v:0 l:1` through a memo as inline; verdict work shows the screen for plain reads | F-6, D-25, D-26 | D-26, A15 note |
34-
| R8 | `latest()` of a held first answer reveals; the plain read still throws | D-22, F-9, S-40 | D-22, S-45 |
35-
| D-27 | stands: a verdict mount is mainline; a plain read is never display-ahead | F-6, F-11 | D-27, Lane vocabulary |
25+
| ruling | outcome | rule | where it is written |
26+
| --------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------------------------- | --------------------------------------------------------------------------------------------------------------------- |
27+
| R1 | both orders show `[committed, true]`; not for the hold's own work | A10, D-25, D-27, #3884 | D-25 (clarified 2026-10-28) |
28+
| R2 | the parent waits for a lane's flights; the lane never waits for the parent | F-6, F-1, F-3, 2026-10-01 | F-6 (amended 2026-10-28, `d1b938d3c`) |
29+
| R3 | an untracked read of a live guess sees the guess, lane held or not | F-11, A17 | F-1 gloss (clarified, `40cde2cd6`) |
30+
| R4 | a pass reading two holds' display-ahead values is one reveal unit | A15 #3335 | D-25 |
31+
| R5 | `latest()` opens no hold; the pass re-derives once at a park | F-10, D-2, A10 | D-22 |
32+
| R6 | E8 is an equivalence for render effects only | D-2, F-9 | no change |
33+
| R7 | `v:0 l:1` through a memo as inline; verdict work shows the screen for plain reads | F-6, D-25, D-26 | D-26, A15 note |
34+
| R8 | `latest()` of a held first answer reveals; the plain read still throws | D-22, F-9, S-40 | D-22, S-45 |
35+
| D-27 | stands: a verdict mount is mainline; a plain read is never display-ahead | F-6, F-11 | D-27, Lane vocabulary |
36+
| F-vlane | today (rebuild §7 reading A): a fallback that went up beside the held frame stays there until the landing (#3540, #4337); the posture, not verdict-lane membership, is the rule; the #3884 clause stands | F-6, D-17, D-25 (#3884) | D-25 — the general paragraph is not yet written: as drafted it contradicted D-25's standing #3884 clause (rebuild §7) |
37+
| F-leaf | today (rebuild §7 reading A): a pass that read a born node joins, a render effect included, and a node it creates is born held | F-7, A29 ruling A, D-2, D-4 | D-27 (promoted from A29, 2026-10-28), A29 note |
38+
| F-ahead-posture | today: a mount gated by a display-ahead read is mainline; it is not "ahead" by its creator's verdict posture — only lane work's children live in the lane | F-6, D-27 | no change — D-27 (clarified 2026-10-28) already rules it; rebuild §7 |
3639

3740
Every thread on this area reads, before anything else: the precedence rule
3841
at the top of `RULES-FUNDAMENTALS.md` (a fundamental beats a derived rule

‎packages/signals/docs/RULES-DERIVED.md‎

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -224,6 +224,8 @@ A `<Show when={latest(x) > 0}>` opens at once (display-ahead). The `<Loading>` i
224224

225225
**Why the two differ, and what the mount's content may show (clarified 2026-10-28).** A plain read is never display-ahead; only a read that asks (`latest`, `isPending`) or a slot that is guessed (an optimistic value, F-11) is. A guess is _slot-level_: it is the value of its slot for the whole screen, so every plain read of it agrees and an optimistic mount's plain reads show it without tearing. `latest(x)` is _read-level_: only the reads that say `latest` see the held value. A mount's children are _gated_ by the display-ahead read, not _derived_ from it (F-6's "what is rendered from a display-ahead read" is the derivation — `fetch(latest(x))` — not the branch it opens); their plain `x()` is an independent read, and showing it the held value would make a plain read's answer depend on which `Show` wraps it — two plain reads of one signal disagreeing on one screen. So the content reads the screen, and the fallback it shows is `Loading` declining to tear over content that is, by the author's choice of a plain read, not yet the committed value. To show the held value inside the branch, read it there as `latest(x)`: `content {latest(x)}` shows `content 1` now, with no fallback.
226226

227+
**A joined pass's children are the hold's (promoted from A29, 2026-10-28):** a pass — a render effect included — that read a node with no screen to fall back to (a node born into the hold, never committed: D-2; D-4's tick-join) joins the hold: it has no committed value to show, so it is not a leaf reading the screen (D-3) but a reader that waits (A29 ruling A, 2026-10-01: "every reader derives from it — a render effect included — and joins"). A node it creates during that pass is born held: its value commits with the frame, and a mainline reader of it waits rather than showing it ahead of the frame (F-7) — the subtree-disagreement tear this rule guards against, a child's value on screen beside a parent frame still held. Shape: a lazy outlet's render effect reads a router query memo that has no committed value and creates a memo during that pass; the memo is the navigation's, even where its own first pass reads nothing held. This is the third mount beside the two above — a verdict reader's is mainline, an optimistic lane's is the lane's, a joined pass's is the hold's — and A29's "a held pass's children compute but do not effect and die with the commit that replaces their frame" is its archive form. Pins: `tests/lane-uninitialized-landing-3648.test.ts`, `tests/boundary-output-frame-reader.test.ts`, `tests/loading-fallback-in-flush-3540.test.ts`, `tests/fuzz-findings-l2.test.ts`.
228+
227229
**Collapses:** #3851/#3869 (2026-10-07), [A29](./SPEC-ASYNC-SEMANTICS.md#a29-a-tracked-read-served-a-live-transactions-staged-value-enters-that-transaction)'s 2026-10-07 amendment, L2 statement 3 as amended, #3835. **Status:** standing (`tests/verdict-mount-first-pass-3851.test.ts`, `tests/optimistic-mount-binding-3835.test.ts`).
228230

229231
### D-28. A read throws on its own; an earlier read does not disarm a later throw

0 commit comments

Comments
 (0)