26 sections author `span: 'full'` and do not span the row at modal width — measured pixel-identical to writing nothing, and 24 of them were rewritten INTO that spelling by PR #1605

[os-steve]

Restart-when: a published @objectstack/console > 17.4.0 whose .objectui-sha contains objectstack-ai/objectui@bd09957380 (PR #9253) — probe the installed console bundle: span: 'full' in a 3-column section emits @md:col-span-2 as well as @2XL:col-span-3; not the issue state

Measured while discharging #1694 (the absolute-colspan-discouraged question), in Chromium against objectstack dev on pinned @objectstack/* 17.4.0. ⛔ Filed rather than fixed: the sites are this repo's, the cause is the renderer's.

The measurement

span: 'full' compiles to a single class, @2xl:col-span-3, gated at the top breakpoint only. Rendered geometry for one field in an authored 3-column section:

authored	390px (1 col)	720px (2 col)	1700px (3 col)
(no attribute)	358px	285px	495px
span: 'full'	358px	285px	1517px
colSpan: 2	358px	586px	1006px

⇒ at the modal width, span: 'full' is pixel-identical to writing nothing — one column of two — while the colSpan spelling the platform discourages spans correctly there. Overflow was 0px in every configuration at every width; the readings were taken twice with a settle and were stable both times.

Why this repo has 26 of them

grep -rn "span: 'full'" src/ → 26 today. grep -rn "colSpan" src/ → 0.

24 of the 26 are there because PR #1605 rewrote colSpan: 2 into span: 'full' — to clear the absolute-colspan-discouraged warning. #1694 has now measured that warning's claim ("a fixed span only aligns at one width") and found it false at all three widths it names.

⇒ the rule steered this app out of a spelling that aligned everywhere and into one that silently does nothing at the middle width. ⭐ That is the finding: not a noisy rule, a rule whose advice made the reference app measurably worse — and no gate could see it, because the app is conformant.

What is actually broken for a reader today

Every authored "this field takes the whole row" intent is honoured on a wide page and dropped in a modal, which is where a record is most often edited. Nobody has audited which of the 26 sites care; some may look fine at half width by accident.

⛔ What NOT to do

• ⛔ Do not revert to colSpan. It happens to align today, but it is discouraged by a live rule and the rule's fate is the upstream card's to decide. Reverting 24 sites into a warned spelling on the strength of one measurement swaps one silent problem for a loud one, and would have to be undone again if the platform fixes the breakpoint gating instead of the rule.
• ⛔ Do not silence, exempt, or comment around the rule locally (AGENTS.md Scope rule 4 — a bad platform default is fixed at the default).
• ⛔ Do not add a guard. Epic Epic: bring this repo's test farm back under the 2026-08-31 ruling — platform-first, os lint --strict first, then retire the local re-implementations by family #1579's fence stands, and a local gate cannot see a renderer breakpoint anyway.

What unblocks it

objectstack#17328 carries both halves: the false rule message and this breakpoint gating (the objectui renderer half is flagged there for platform triage). When it ships in a released line, this card is a re-measurement — ⭐ re-measure, ⛔ do not act on the table above; it is stale from the moment it was written — and then whichever spelling actually spans at all three widths.

⚠️ If the platform's answer is "span: 'full' is correct and the top-breakpoint gating is intended", then this card becomes a content question instead: 26 authored intents that the product deliberately does not honour in a modal, and the docs and the authoring guidance both say otherwise.

Refs: #1694 (the measurement, closed) · PR #1605 (the rewrite that created the 24) · #1524 (the noise question, closed) · objectstack#17328 (upstream) · #806 (why source reading is not acceptable here)

---

Generated by Claude Code

[claude]

Half-state repair — R61 board audit (repo:hotcrm seat, session_0132iDHq4FLW2zS9VPXqbf9f, 2026-09-11T01:4xZ)

The line was already in the body and in the right place — but written **Blocked-by: ...**. Markdown bold defeats the ^Blocked-by: predicate, so this card was invisible to the reverse index despite looking correct to a human reader. De-bolded only; the reference is unchanged:

Blocked-by: objectstack-ai/objectstack#17328

⭐ Worth recording as a lane fact: a correctly-authored blocker can still be index-invisible. The fix is convergence on the one spelling, ⛔ never loosening the predicate.

---

Generated by Claude Code

[objectstack-fleet]

repo:hotcrm seat, session_01X8U3asekbiC7yWoEPWR4Dg · stock re-triage group 6 (maintainer-confirmed ten-card group; maintainer reply verbatim: 「同意」) · 2026-09-25T04:35Z

pm:blocked → pm:on-hold — fixed upstream, not yet installable

#17328 closed via objectstack PR #17671, a lint-only withdrawal. The renderer half was fixed in objectui PR #9253 (2026-09-12). Both landed after 17.4.0 (2026-09-09). On restart, re-measure at 390/720/1700px as the card instructs. Expected: span: 'full' fills the row at modal width ⇒ close completed with zero diff. ⛔ Do not revert to colSpan.

---

Generated by Claude Code

[objectstack-fleet]

Unlock: pm:on-hold → pm:queue. The restart condition is met on the 17.6.0 pin. repo:hotcrm seat, session_01ER8ntXZhYebyQ66aXWdjfT, 2026-10-02T23:00Z.

• The line: Restart-when: a published @objectstack/console > 17.4.0 whose .objectui-sha contains objectstack-ai/objectui@bd09957380 (PR #9253) — probe the installed console bundle …
• Pin: main installs @objectstack/console@17.6.0 (PR chore(deps): upgrade @objectstack/* to 17.6.0 #1982). Its dist/.objectui-sha reads 31971ff1e28f89cfc45f0c19bc5b05e443f28b79.
• Containment: in a full, non-shallow objectui clone, git merge-base --is-ancestor bd09957380 31971ff1e2 exits 0. Control: the root 9947e9af exits 0.
• Bundle: the installed bundle carries both @md:col-span-2 (dist/assets/ui-components-*.js, vendor-objectstack-*.js) and @2xl:col-span-3. ⚠️ That is string presence only. ⛔ The seat did not render a 3-column span: 'full' section, so the dispatch measures that first.
• Double-check ①: the latest transition comment 5826810478 names this body's line. Double-check ②: no PR has cross-referenced this card since.

---

Generated by Claude Code

[objectstack-fleet]

Claim: PM loop round R70
Session: session_01ER8ntXZhYebyQ66aXWdjfT
Account: hotlong (the seat's linked user as GET /user answers it; always the card's assignee)
Branch: claude/issue-1870-span-full-remeasure
Worktree: hotcrm-issue-1870
Domain: repo:hotcrm (single-lane repo, no domain:* taxonomy)
Seat: repo:hotcrm#1
File surface: none expected (a geometry re-measurement on the 17.6.0 pin; the recorded expectation is to close completed with zero diff). If the measurement shows span: 'full' still fails at modal width, ⛔ no edit: report, and the seat routes it (stop on breach; explain in the report)
Container & model: S, mode:subagent, model: default (dispatch-gates --tier --repo objectstack-ai/hotcrm answers "no path-derived mandate … the tier stays the PM's per-card judgment call"; a browser measurement is a judgement surface)
Clause-②: no
Thread-read: 5962805271
Serial constraints cleared: there is no file surface. In flight: the #806 / #970 / #1452 fold (record:details sections of three *_detail.page.ts, claim 5963239470) is disjoint from the *.view.ts form sections this card measures. src/sales/views/opportunity.view.ts is held by draft PR #1950, which is read-only for this card.

---

Generated by Claude Code