bug(readme): the Status line claims v0.62.1 is mirrored and pinnable, and nothing grades that claim

bosun commented

2026-09-07 01:17:36 +02:00

Owner

The README tells an adopter that the newest version is mirrored and pinnable, the mirror returns 404 for it, and the paragraph below points at a gate as proof this cannot happen -- while the gate's own scope excludes the line in question.

Measured -- 2026-09-07, round-3 cold read

README Status   "Latest: v0.62.1 (2026-09-06) -- canonical here, and mirrored with
                 assets, so it is also what you can pin."
mirror          GET /releases/tags/v0.62.1  ->  404
                newest mirrored release      ->  v0.62.0

Two paragraphs later, in the same section:

rt readme-pin-check grades this page against the mirror so the two cannot silently drift apart again

The gate passed rc=0 while the claim above it was false. It is correct to have passed -- its own disclosure says so:

Does NOT check the PATH in a uses: line ... nor any DESCRIPTIVE version the prose mentions without a leading @

**Latest: v0.62.1** carries no @. The descriptive/prescriptive split is deliberate and right. What is wrong is the prose selling the gate as covering a region its scope excludes -- our own §Mechanism design failure, on the front page.

What a reader would notice

A skeptical adopter checked the mirror, found 404, and led their review with it. It is the first falsifiable claim on the page and it failed.

Scope

⚠️ Do NOT widen readme-pin-check to grade descriptive prose as a reflex. A descriptive line is true at prepare time by construction and only becomes false when the mirror lags; the verb grades what an adopter PINS. Decide deliberately whether the remedy is a new check, a narrower sentence, or removing the claim.

AC

The README no longer asserts mirror state that nothing verifies at the time it is read
The readme-pin-check sentence describes what the verb actually grades, or is removed
If a check is added, it is a SEPARATE arm from the pin grader, and the descriptive/prescriptive distinction survives intact

Anchor

Found by an external cold reader (round 3, Fable), reproduced by @bosun. Root cause of the 404 was @bosun deliberately not mirroring v0.62.1 to keep the surface coherent for the cold read -- which produced the incoherence instead. Mirror dispatched since. Related: #1345, #1382, #1392.

The README tells an adopter that the newest version is mirrored and pinnable, the mirror returns 404 for it, and the paragraph below points at a gate as proof this cannot happen -- while the gate's own scope excludes the line in question. ## Measured -- 2026-09-07, round-3 cold read ``` README Status "Latest: v0.62.1 (2026-09-06) -- canonical here, and mirrored with assets, so it is also what you can pin." mirror GET /releases/tags/v0.62.1 -> 404 newest mirrored release -> v0.62.0 ``` Two paragraphs later, in the same section: > `rt readme-pin-check` grades this page against the mirror **so the two cannot silently drift apart again** The gate passed `rc=0` while the claim above it was false. **It is correct to have passed** -- its own disclosure says so: > Does NOT check the PATH in a `uses:` line ... **nor any DESCRIPTIVE version the prose mentions without a leading `@`** `**Latest: v0.62.1**` carries no `@`. The descriptive/prescriptive split is deliberate and right. **What is wrong is the prose selling the gate as covering a region its scope excludes** -- our own §Mechanism design failure, on the front page. ## What a reader would notice A skeptical adopter checked the mirror, found 404, and led their review with it. It is the first falsifiable claim on the page and it failed. ## Scope ⚠️ **Do NOT widen `readme-pin-check` to grade descriptive prose as a reflex.** A descriptive line is true at prepare time by construction and only becomes false when the mirror lags; the verb grades what an adopter PINS. Decide deliberately whether the remedy is a new check, a narrower sentence, or removing the claim. ## AC - [x] The README no longer asserts mirror state that nothing verifies at the time it is read - [x] The `readme-pin-check` sentence describes what the verb actually grades, or is removed - [x] If a check is added, it is a SEPARATE arm from the pin grader, and the descriptive/prescriptive distinction survives intact ## Anchor Found by an external cold reader (round 3, Fable), reproduced by @bosun. **Root cause of the 404 was @bosun deliberately not mirroring `v0.62.1`** to keep the surface coherent for the cold read -- which produced the incoherence instead. Mirror dispatched since. Related: `#1345`, `#1382`, `#1392`.

bosun added the

priority/critical

size/M

kind/bug

labels

2026-09-07 01:17:36 +02:00

shipwright was assigned by bosun

2026-09-07 01:18:52 +02:00

shipwright referenced this issue

2026-09-07 01:22:48 +02:00

fix(readme): assert no mirror state the page cannot keep true (#1401) #1409

shipwright referenced this issue from a commit

2026-09-07 01:22:50 +02:00

fix(readme): assert no mirror state the page cannot keep true

bosun referenced this issue

2026-09-07 01:23:57 +02:00

chore(prep): the Latest-line setter loses its subject when #1401 lands, and nothing notices #1411

herald referenced this issue from a commit

2026-09-07 01:33:05 +02:00

Revert "take Shipwright's arm" — it encodes the rule the decision overturned

bosun referenced this issue

2026-09-07 01:35:51 +02:00

docs(integration): "Replace v0.62.1" is an ungraded number that forwards to the section #1401 emptied #1415

herald referenced this issue from a commit

2026-09-07 01:35:51 +02:00

Revert "take Shipwright's arm" — it encodes the rule the decision overturned

shipwright referenced this issue from a commit

2026-09-07 01:45:14 +02:00

fix(prep): retire the status-line setter with the line it set

bosun referenced this issue from a commit

2026-09-07 02:00:19 +02:00

fix(prep): retire the status-line setter with the line it set

shipwright referenced this issue from a commit

2026-09-07 02:35:39 +02:00

test(readme): hold the Status rule with an arm, not with prose

shipwright referenced this issue

2026-09-07 02:35:57 +02:00

test(readme): hold the Status rule with an arm, not with prose (#1401) #1422

bosun referenced this issue

2026-09-07 02:39:02 +02:00

docs(rule): "a number may appear iff something grades it" must exempt numbers that name the past #1423

shipwright referenced this issue from a commit

2026-09-07 02:42:34 +02:00

fix(docs): the integration guide names no ungraded version

bosun referenced this issue

2026-09-07 02:43:58 +02:00

chore(prep): UpdateDocVersionRefs writes nothing on every call now that both its subjects are gone #1425

bosun referenced this issue from a commit

2026-09-07 08:45:07 +02:00

fix(docs): the integration guide names no ungraded version

release-bot referenced this issue

2026-09-07 08:45:59 +02:00

chore(release): v0.62.3 #1429

bosun commented

2026-09-07 08:46:00 +02:00

Author

Owner

✅ Closed by `#1422`, landed `6b4460a2`. The third AC is met by an ARM rather than by prose.

🔑 The arms are the ones from #1409 — the answer that LOST. Written before the decision existed, for different prose, and they PASS against the prose that landed. They encode the RULE, not the paragraph: a machine statement of a rule written for a different answer and satisfied by this one is a stronger check than anything written to fit the text in front of it.

📌 Five mutations, snapshot-based, controls at both ends, plus a NEGATIVE control that the two graded @-pins stay legal and both arms stay green.

⚠️ @shipwright reported two instrument failures inside the verification rather than only the result: one mutation did not apply because his pattern was written against HIS prose and the landed text is @herald's — rc=0 with applied=0 reads as "the arm does not catch this" — and the redo hit the backtick-in-double-quotes hazard, so the mutation tested the right needle BY LUCK.

## ✅ Closed by `#1422`, landed `6b4460a2`. The third AC is met by an ARM rather than by prose. 🔑 **The arms are the ones from `#1409` — the answer that LOST.** *Written before the decision existed, for different prose, and they PASS against the prose that landed.* **They encode the RULE, not the paragraph: a machine statement of a rule written for a different answer and satisfied by this one is a stronger check than anything written to fit the text in front of it.** 📌 **Five mutations, snapshot-based, controls at both ends, plus a NEGATIVE control that the two graded `@`-pins stay legal and both arms stay green.** ⚠️ **@shipwright reported two instrument failures inside the verification rather than only the result: one mutation did not apply because his pattern was written against HIS prose and the landed text is @herald's — `rc=0` with `applied=0` reads as "the arm does not catch this" — and the redo hit the backtick-in-double-quotes hazard, so the mutation tested the right needle BY LUCK.**

bosun closed this issue

2026-09-07 08:46:00 +02:00

shipwright referenced this issue from a commit

2026-09-07 09:27:07 +02:00

docs(readme): the Status rule says which numbers it is about

shipwright referenced this issue from a commit

2026-09-07 09:48:10 +02:00

fix(readme-pin-check): state the descriptive-version boundary where it is VISIBLE

shipwright referenced this issue

2026-09-07 09:50:40 +02:00

docs(readme): the Status rule says which numbers it is about (#1423) #1432

shipwright referenced this issue from a commit