# Author desk run-through · 29 September 2026

This is **simulation by the author**, not a real teacher, student or supporter interview, a timed classroom trial, QCAA approval or an accessibility study. The lesson clocks are planned allocations only.

## Teacher desk pass

I started with [README](README.md), opened [Day 1](LESSONS.md), chose Card A and the ratio-order print, and checked that a teacher can identify the exact prompt, five-part clock, three same-target response routes, an alternate setting and a misconception response without opening the key first. The sequence advances from ordered parts and part-to-whole, through simplification and matching units, to fixed-total division, simple scale and an error audit. On Days 5 and 10 I treated the [public checks](STUDENT-CHECKS.md) as first-work prompts and only afterward consulted the separate [public staff key](teacher/KEY-AND-NEXT.md). I confirmed the two check cases use different main numbers from Cards A–G. Public access means a teacher must use a fresh local parallel case if needed.

The [official-source crosswalk](SOURCE-AND-CROSSWALK.md) maps all six Ratios bullet groups at QCAA printed p. 15 and flags QCAA's **[complex]** markers on fixed-total division and simple scales. These sessions scaffold those two ideas in simple, fully specified settings. The ten periods are **4 h 10 min** against a **7-hour** notional ratio subtopic. Rates, Percentages, Units of energy, Unit 1 Topics 2–3 and formal school assessment remain outside the pack. No claim of whole-unit or whole-year completion is defensible.

## Learner desk pass

I used Card C without a worked answer, counted 8+12=20 badges, wrote story:information `8:12=2:3`, then story of **all** as `8/20=2/5`; the part-and-whole print keeps the different denominators visible. I followed Card E's five equal boxes to split 35 labels as 14 and 21, checked both sum and ratio, then followed Card F's unit path `4×200=800 cm=8 m`. These calculations were independently recomputed in the verifier. Actual comprehension by a learner has **not** been measured.

I opened the [offline ratio lab](ratio-lab.html) in local headless Chrome at desktop and mobile widths. The default text showed 2:3, equivalent 4:6, and 35 divided into 14 and 21. A query-state test with 4:7, factor 5 and total 55 rendered 20:35 and a correct 20/35 split. A non-divisible 2:3, total 36 state explicitly refused to round whole objects. I visually inspected desktop and mobile screenshots; input labels, bars and text remained legible, and bar meaning was restated in text. This is a browser desk check, **not** testing with an individual screen-reader, switch, keyboard or tactile user. The print and token routes remain available.

## Supporter/reuse desk pass

A parent, support worker or tutor can use the [learner cards](LEARNER-CARDS.md), choose an imagined [practice swap](PRACTICE-SWAPS.md), draw tokens on paper and compare against the [staff results](teacher/KEY-AND-NEXT.md) after a first attempt. No purchases, family finances, live survey, account or internet connection are needed. The [rights ledger](SOURCE-AND-RIGHTS.md) makes original content reuse explicit while keeping QCAA source and DejaVu font ownership separate. The pack does not suggest that every Year 11 learner is in Essential Mathematics, that paper ratio examples certify real allocation decisions, or that a public check is a formal grade.

## Print and source observations

I rendered all five original A4 PDFs to a contact sheet and visually inspected titles, number/order labels, answer spaces, condition notes, contrast, page edges and footer. Each SVG includes a title/description and each PDF has extractable text; [exact text/tactile equivalents](print/TEXT-ALTERNATIVES.md) preserve the learning prompts. Raised print/Braille production and real-user fit remain untested. The official QCAA v1.3 PDF was readable through the official web renderer; direct command-line download returned HTTP 403, so the pack does **not** claim a QCAA PDF byte hash. Local source identity and original file hashes are checked, with future live-source refresh still required.