136 lines
8.6 KiB
Markdown
136 lines
8.6 KiB
Markdown
# Postboot contextual integration delta — 2026-09-11
|
|
|
|
Codex-ytr83r /root/integration_resume. Read-only delta review against root
|
|
`fb77b857d26b64b80d99eabb4c65cf6785f3f451`; read AGENTS, issues 1/2 through
|
|
276/269, no open PRs, Authoring20260911, BehaviorAuthoring20260911 and Home.
|
|
Coordination: issue 1 comment 274. This supplements, rather than repeats,
|
|
`release-staged-source-integration-ytr83r-20260911.md`.
|
|
|
|
Root will finish its frozen B4 correction/gates while the source owner resumes
|
|
validation. Do not import the author's unvalidated reboot WIP as an accepted
|
|
milestone. The smallest updated contextual source, once those gates mature, is
|
|
the coherent df6 implementation plus the two-file ca53 correction described
|
|
below, or its explicitly frozen successor. No chooser/canvas/core WIP is needed.
|
|
|
|
## Ordered import recommendation
|
|
|
|
1. Preserve root's B4 clip, lifecycle, CLI and current ledgers. Recheck owner
|
|
heads and comments immediately before integration; the recorded 9e head is
|
|
`ca53f574e57381a523e946d5277b4081cff74355` on codex/release1-9e8r6t.
|
|
2. Import contextual implementation `df6f59562643b7771ecf0c972f8655f3e3f1c4da`
|
|
and its associated oracle/fixtures/review/archive as the earlier preflight
|
|
specifies. Resolve only the behavior.rs module and guide.rs walkthrough
|
|
unions, retaining B4. Its scenario workspace overlap remains comment-only
|
|
on this host. Do not replace those complete files with the peer tree.
|
|
3. Apply only the ca53-versus-df6 production delta: creation.rs `picker_view`
|
|
caption (six added/one removed lines), and source_draft.rs essentials-first
|
|
view plus its three prepared mounted tests. These are the only two production
|
|
paths changed after df6. ca53's parent c4bad4b and grandparent d95b1ab contain
|
|
scenario and CLI publication evidence respectively, not source prerequisites.
|
|
The two-file correction does not touch App, scenario/topology workspace,
|
|
FormShown, core/server/CLI, schema, Cargo.toml or Cargo.lock.
|
|
4. Bring the ca53 corrected staged-source fixture README and appended
|
|
release-staged-source-authoring-oracle report. Save after child Use can adopt
|
|
a valid parent: inspect before Save or Cancel parent then Save for an
|
|
unchanged staged-only check. The frozen oracle and base source do not change.
|
|
Preserve the observed missing-runtime loader failure and scoped DLL remedy.
|
|
5. Preserve ca53's contextual-source-native-20260911 and
|
|
contextual-source-layout-reboot-20260911 archives, and reports
|
|
release-contextual-source-native, release-contextual-source-layout-review,
|
|
release-contextual-source-layout-independent and release-staged-owner-label
|
|
(all suffixed -20260911.md). Keep the peer RESTART_HANDOFF as linked,
|
|
source-specific history; do not overwrite root's RESUME or shared ledgers.
|
|
Earlier CLI/scenario evidence can be deliberately imported separately;
|
|
neither intervening complete commit is required for this source correction.
|
|
6. Union the raw-byte attributes before staging. Verify imported index and later
|
|
checkout bytes against original inventories, then capture fresh combined
|
|
source/binary provenance and run root's affected gates/native acceptance.
|
|
Any later peer correction must be reviewed as a new exact delta.
|
|
|
|
## Overlap and still-open acceptance
|
|
|
|
The ca53 source_draft delta starts at view line 327; the rest is a new test
|
|
module after Draft::command. All pre-view state, tokens, serialization, update,
|
|
combination, final command, existing parent controls and pinned footer remain.
|
|
Kind choices precede unbounded parent detail; a single-kind button disappears
|
|
only when the draft already holds that permissible kind. The caption changes
|
|
only the selected exact staged Owner's presentation. Neither path intersects
|
|
root's B4 render.rs/render_tests.rs or its draw_markers scale argument.
|
|
|
|
The author's independent review at 89B08EF6 closed a long-parent-before-kind
|
|
finding. ca53 source C8D36BD6 also includes mechanical fixes for two actual test
|
|
helper compile errors (column macro ambiguity and Id::new(String)). Corrected
|
|
focused2 was interrupted before tests. Expected contextual26, strict Clippy,
|
|
formatting and a matching new GUI remain pending. The three tests compose a
|
|
caption plus child card at 340/500 by 215; they do not mount full scenario or
|
|
topology workspaces. Real 1100x700 ordinary-panel Name/kind/root-parent and
|
|
owner-caption acceptance, plus long-parent/recovery/wide controls, is still due.
|
|
|
|
Peer 956's published restart main is
|
|
`8d55d2d2eb05aed826b96a917cf776e1d8c3b690`; isolated core WIP is
|
|
`b96ff3935ab3f36b7144566e46559a05a29292a5`, incomplete canvas WIP is
|
|
`e511c4153fb5bb484652434340e0e4c7000c93d0`. Do not import these to obtain
|
|
the card/caption correction. Issue 1/275 keeps them paused pending that owner's
|
|
post-reboot return. Its source-derived, unexecuted P2 focus findings remain:
|
|
normal Scenario child Back/Use/Escape remounts a parent whose initial focus flag
|
|
was consumed; an already accepted FormShown task later operates using only a
|
|
global NAME_INPUT and can disturb a replacement session. Their report at exact
|
|
8d55d2d2 is `harness/release-chooser-contextual-integration-review-20260911.md`.
|
|
The remedy belongs to that owner's exact current-mount/session focus work;
|
|
ca53 does not resolve it. Keep delayed-task and ordinary child-return tests,
|
|
modal ownership, parent scroll retention and live native validation pending.
|
|
This does not assert the global task reaches Popup's separate search input.
|
|
|
|
Historical GUI52 native acceptance is distinct: Lifeline Save-adoption and SA
|
|
Chain direct Create each passed complete model/source/files and one Undo/Redo
|
|
with original IDs. SA Machine Create/Save and Undo/Save passed; Machine Redo and
|
|
reopen did not. The live trial was closed at saved Machine Undo; reboot does not
|
|
preserve in-memory history. Use its original pinned first-commit carrier for
|
|
detached-file expectations, and a fresh trial for missing history. Direct
|
|
Lifeline Apply/Create, cancellation/recovery and corrected-layout native gates
|
|
remain separate. None of these form/B4 slices closes issue 2.
|
|
|
|
## Exact Git-byte audit
|
|
|
|
Authorized fetch of exact ca53 used --no-tags --no-write-fetch-head, populated
|
|
objects only, and completed successfully. No ref/index/source/commit/Cargo or
|
|
native mutation was performed. Only this report was written by this reviewer.
|
|
|
|
All 13 layout manifest payloads match exact Git blobs in size and SHA-256.
|
|
Manifest SHA-256: 7639dbd81f1a766f656e563013c16910950eec359c92c5765c39c4c0981eef70.
|
|
All 77 native original records / 51 unique payloads match, and all 51 checksum
|
|
entries match. Native manifest SHA-256:
|
|
9a2ad6fb186606ef2a4fd1101d68c8e45bc96953093670fe9451ed9fb90a2ea6;
|
|
SHA256SUMS.txt: f69019970c42c52e5a0506b5f0369f804c216a3f0280e37d0e9b1eedf4a9523a.
|
|
The 69 total files in the two archives include README/manifests/preservation
|
|
method files outside those payload lists. Preserve all, including zero-byte
|
|
stdout/stderr, initial helper/loader failures and interruption records.
|
|
|
|
The unchanged oracle and both base source blobs remain byte-identical to df6:
|
|
9eb9dd3868b6cb772f907d16ae02995ad10a16f85c087012c24948e6e6fe775d,
|
|
aa844950ec4fb86f262a257e1fc370a26557ea7978d8a4cab78265037660234c,
|
|
ad9de02963c20d1f43856455a797deb05ca674659e45f2c4b96b52783024f8bd respectively.
|
|
|
|
ca53 source_draft.rs is 38,949 bytes and exactly matches frozen C8D36BD6BCD0818AAD2FBBC404B29085F4ADCAF8191B95A04208A7C2346C36E0.
|
|
creation.rs is 29,301 LF bytes, SHA-256
|
|
438a245ead397060767c790c51a8f21cd5021342627999deb4695cf3bd22091a.
|
|
The author's inputs3 instead records 30,053 bytes / 0FB19BE3A5688C4BD3CA3643C47A97DA7E7F85ADD8A860CD046A0D9CB3CE9DFC.
|
|
Uniform CRLF conversion gives 30,067 bytes / 5d9d5e72036c54b1446eb207eb20c951e728832307c7090e1b22d289e54c5f79,
|
|
so uniform normalization alone has not established exact author-source equality.
|
|
A mixed-newline original is plausible, but is not proven by this review. The
|
|
published caption diff exactly matches the report; no extra semantic source
|
|
delta appears. Keep the original manifest unchanged and obtain fresh import
|
|
build provenance; do not claim creation's committed bytes equal its old freeze.
|
|
|
|
Additional raw rules for these two ca53 archives (retain the earlier union):
|
|
|
|
```gitattributes
|
|
harness/experiments/contextual-source-native-20260911/** -text whitespace=cr-at-eol
|
|
harness/experiments/contextual-source-native-20260911/**/*.arcadia whitespace=-blank-at-eol,-blank-at-eof,cr-at-eol
|
|
harness/experiments/contextual-source-layout-reboot-20260911/** -text whitespace=cr-at-eol
|
|
harness/experiments/contextual-source-layout-reboot-20260911/run-gates2.ps1 whitespace=-blank-at-eof,cr-at-eol
|
|
```
|
|
|
|
No additional dependency blocker was found. This is a delta/import handoff,
|
|
not a performed integration, new test/native pass or release milestone.
|