From e7f811be3b4d5ed2c5bb441fe5282dff34808e1d Mon Sep 17 00:00:00 2001 From: Levi Neuwirth Date: Tue, 28 Jul 2026 17:24:02 -0400 Subject: [PATCH] docs: integrate main @ ad41cf1 and record what the ledger now needs MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit PR #185 merged during the re-scout, so `docs/active-work.md`, `docs/agent-handoff.md` and `COHERENCE.md` are no longer held by an open PR. #185 changed docs only — no source — so nothing in rev 5's census moves, and the two facts rev 5 cites from the durable records survive intact: the handoff §4 inventory still names `builtin/runtime/dired.lua:371` as a non-adopter of `set_generated_contents`, and COHERENCE §14 still names dired and still classifies it as the cheap half. This branch continues to touch only `docs/dired-stage2-framing.md`. A standalone docs-refresh PR is exactly the ledger contention treadmill the ops lessons warn about, and the ledger's own protocol puts a lane refresh with the work rather than ahead of it. What §16 now records instead: the landed `docs/active-work.md` lane still has this branch at head `ab42a79`, calls the document 1,570 lines, says the re-scout is under way, and states the rename census as FIVE path owners. Rev 5 makes all four stale — and the census is SIX (§5, W5). Same lesson in the other direction: a census is a reading, not a constant. The refresh should lift §5's table rather than re-derive it, and should note that Q#DR25 added scope the lane does not mention. Co-Authored-By: Claude Opus 5 (1M context) Claude-Session: https://claude.ai/code/session_01Lv428Fth9LRtffwJSsqH7T --- docs/dired-stage2-framing.md | 26 +++++++++++++++++++------- 1 file changed, 19 insertions(+), 7 deletions(-) diff --git a/docs/dired-stage2-framing.md b/docs/dired-stage2-framing.md index f7c3d37..10c84ad 100644 --- a/docs/dired-stage2-framing.md +++ b/docs/dired-stage2-framing.md @@ -2261,13 +2261,25 @@ to be worth fixing. **Ledger note (corrected in rev 5, W6):** this framing branch deliberately touches **only** this file. Rev 4 said the durable records -were held by open PR #169; **#169 merged** as `74301d1`. The current -holder is **PR #185** (`docs-landed-state-184`), which owns -`docs/active-work.md`, `docs/agent-handoff.md`, and `COHERENCE.md`. A -second edit to any of the three from here would conflict for no benefit, -so the Stage 2 lane goes in once #185 has landed. *(This is the ledger -contention treadmill the ops lessons name: with several PRs open, every -merge re-conflicts the rest in `docs/active-work.md`. Integrate late.)* +were held by open PR #169; **#169 merged** as `74301d1`, and its +successor **#185 merged as `ad41cf1` during this re-scout**, so +`docs/active-work.md`, `docs/agent-handoff.md`, and `COHERENCE.md` are +now unheld. This branch still does not touch them — the ops lesson is +that a standalone docs-refresh PR restarts the contention treadmill, and +the ledger's own protocol puts the lane refresh with the work. + +**Two things the landed ledger now needs, and this PR is not the place +for either.** `docs/active-work.md`'s dired Stage 2 lane still records +this branch at head `ab42a79`, describes the document as *1,570 lines*, +says *"the re-scout is under way"*, and states the rename census as +**five** path owners. Rev 5 makes all four stale: the head has moved, +the document is longer, the re-scout is done, and **the census is six** +(§5, W5). That is the standing lesson restated — *a census is a reading, +not a constant* — and it applies to the ledger's copy of this +framing's numbers exactly as it applied to the framing's copy of the +tree's. Whoever refreshes the lane should take the six-owner table from +§5 rather than re-deriving it, and should note that Q#DR25 added scope +the lane does not yet mention. ### Ownership warning: 2a must not run concurrently with Journey Stage 1b