MS Life Science — the one page for James (12 October 2026, late evening)
Where the build stands (13 October, midday). Your Topic 2 review found two root causes, and both run through all three topics, so all three are being fixed at once. First, the closing line. Every lesson ended "So
Your five taps this morning are applied: the Topic 1 batch goes the moment the render server is back (every MS Chemistry video failed yesterday; a control render decides); two atoms stay on one lesson; the three assumed atoms get recall items before lesson 1; recall items live in the recall bank; the PP100 is a bank of six items per atom drawn from, so Topic 1's bank grows from 31 to 78 and Topics 2 and 3 get theirs. A plain tap now sends itself; only Other and Change this need Send.
On the elephant (lesson 10): "hundreds of thousands" is the accurate figure. A 6,000 kg African elephant against a 20 g house mouse is about 300,000 times heavier, so "millions" would overstate it. Left as written unless you want a different pair.
Your 13 October rulings on Topic 3 are being applied now (the intro cut, your lesson 2 script word for word, "the arrow means", two select-all items before the typed equation, the tomato item as a positive prediction with no chart); it then goes to the video pipeline as you asked. PP100s are one per topic, not clusters: banks of six items per atom, harder items first, are being written for Topics 1 to 3 now; the clusters page is closed. Two later-course samples are being authored for your read: Topic 11 (dominant, recessive and the Punnett square — a short, procedural topic) and Topic 7 (ecosystems and food webs — a long, systems topic). Topic 1 is at the video pipeline: eleven videos feeding since 13:14, first landings expected mid-afternoon.
Both later-course samples are on the scripts page now, and both are closed on keys (two cold readers each, an examiner, an adjudicator, a second read; one key changed in Topic 7 — a millipede eating dead leaves was a consumer by the lesson's own test, so it became a fungus). All three PP100 banks are closed on keys at six items per atom: Topic 1 78 items, Topic 2 66, Topic 3 48, each with two cold reads, an examiner, an adjudicator and a second read, and a draw-one list so a shorter test never pairs two items that hand each other an answer. Topic 1's bank · Topic 2's · Topic 3's. Topic 7 — Ecosystems and food webs: thirteen lessons in one meadow, introduced once and pointed at after that; water, air and a suitable place; the meadow coined an ecosystem; producer, consumer and decomposer sorted on one dead deer with your Topic 3 sorting shape; the chain drawn arrow by arrow with the wrong arrow called wrong; the web read both ways; the intro's rabbit disease answered with bars. Every lesson 169 to 178 words. Its cold reads start now. Topic 11 below is the Punnett-square sample.
Two Topic 7 calls for you (the rest of its blueprint decisions are listed under "Applied without asking"):
5. Biotic factor, abiotic factor and population are named by the Texas standard but have no atom. Add them?
The course says: Lesson 5 coins biotic and abiotic factor; lesson 13 coins population; the register has no row for any of the three.
Why it is a question: - TEKS 6.12A names all three, so a test can ask for them. - Autotroph and heterotroph were added the same way on 12 October. - Without a row they cannot be tagged in the PP100 bank.
Cost of changing it: three register rows, logged and validated.
My recommendation: add the three atoms.
A: Add the three atoms / B: Coin them, no atoms
6. Water, air and a suitable place: three short lessons, or one?
The course says: Lessons 1 to 3 teach the three needs as three videos of about 170 words each, one atom per lesson.
Why it is a question: - Each is one idea on one case: the dry July, the trout, the shade. - Your 13 October ruling lets one lesson carry two atoms, and the three together would still fit 180 words only by cutting the cases. - The grade 3 to 5 courses taught the needs; here they are retrieved, then sharpened into "suits it".
Cost of changing it: two scripts cut and one rewritten; the quizzes merged.
My recommendation: keep three short lessons.
A: Keep three short lessons / B: Fold into one lesson
7. Two videos in three come back with a reworded coining. Keep editing each one, or print the exact lines on screen?
The course says: Every line marked "spoken word for word" is checked; a paraphrase is an audio edit by id. Fourteen edits have run today and every one restored every line, at 20 to 40 minutes each.
Why it is a question: - Tonight's trial settled one thing: putting each exact sentence alone in its own short step changes nothing. Your Topic 3 lesson 2 script came back with one of ten lines intact in both shapes. - What the renderer does keep, every time, is the ending and whatever is printed on screen. Topic 1's closes held in all ten lessons, and each had its rule sentence printed on the closing screen as it was spoken. - MS Chemistry's Topic 1 fits the same pattern: the two lessons whose closing screen printed the rule sentence as one line held it; the five with bullet closes had it reworded. - Across all 20 of our first renders, 85 protected lines: lines printed on the closing screen held 73 percent of the time, unprinted lines 29 percent. Short lines survive on their own ("The arrow means react to make."); long compound sentences and the "Something that … is called …" coining form are the ones that go. - So two levers, neither a guarantee: write each exact line short, and print it on the closing screen. Printing runs against your 9 October ruling that a close is one short line OR bullets, with bullets for a multi-point close.
Cost of changing it: the screen style for coinings and rule lines; nothing in the scripts.
My recommendation: keep the audio edit as the standing route (22 of 22 edits across both courses restored every line today), and in the next topic's videos make every closing screen the one printed rule sentence, with bullets only where the rule genuinely has several parts, to see how many edits that removes.
A: Keep editing; next topic closes print the rule line / B: Keep editing every reworded line, no change
8. Topics 1 and 3's twenty videos are finished. Put them on the review site?
The course says: All eleven Topic 1 videos and all nine Topic 3 videos are delivered to the course's video folder, every protected line word for word, 54 to 95 seconds each.
Why it is a question: - Video never goes into the review site's own deploy (a 25 MB ceiling), so playing them there means an upload to the course's Cloudflare storage. - Nothing is uploaded without your say-so. - Until then they play from the course folder on this machine only.
Cost of changing it: one upload of twenty files, about 90 MB; the topic pages then link to them.
My recommendation: upload them, so you can watch Topics 1 and 3 as a student would, on their topic pages.
A: Upload the videos to the review site / B: Keep them on this machine for now
What to read next: the Topic 11 sample — Dominant, recessive and the Punnett square, on the scripts page, below Topic 3. Seven lessons on Mendel's pea plants on one growing drawing; every check moves to a fresh animal or plant; Punnett squares drawn, never narrated; every Punnett question closed (a count, a chance typed as "3 in 4", or a lettered box). Two cold readers and an examiner are on its questions now. Topic 3's fixes are in and its nine videos are at the pipeline. The Topic 7 sample (ecosystems and food webs) is being authored.
Three Topic 11 calls for you (the other blueprint decisions were taken at their recommendations and are listed under "Applied without asking"):
2. Pea plants carry two copies of each gene — stated as a given, and a new atom?
The course says: "Every pea plant carries two copies of the gene for flower color, one from each parent." (lesson 1, told as a fact)
Why it is a question: - Dominant and recessive mean nothing without the two-copies fact, and no atom in Topics 9 to 11 states it. - Topic 10's gene atoms say a feature travels as a gene and a gene comes in versions, but not that each body cell holds two. - A new fact atom before LIF-131 would give the sentence a home and a test.
Cost of changing it: one register row, logged and validated; no lesson changes.
My recommendation: state it as a given in lesson 1 now, and add the atom to the register.
A: State it and add the atom / B: State it, no new atom
3. One Punnett lesson, or two?
The course says: Lesson 7 teaches the Punnett square once: a worked cross in the video, then a scaffolded cross and two bare crosses in its questions.
Why it is a question: - One lesson keeps the topic at seven and the square in one place. - Two lessons would split "fill the square" from "read the chances off it", each with its own video. - The high-school course gives a whole lesson to "3 in 4 is a chance, not a promise"; here it is one spoken line.
Cost of changing it: one more script and quiz.
My recommendation: one lesson, with the chance-not-a-promise line spoken once.
A: One Punnett lesson / B: Two Punnett lessons
4. How is a chance written?
The course says: Lesson 7 says "3 in 4"; the chip shows "3 in 4 (75%)" once; every form is accepted when typed: "3 in 4", "3/4", "75%", "three quarters".
Why it is a question: - "3 in 4" reads as a count of offspring, which is how the square is built. - A fraction or percentage is what the state tests print. - Accepting every form costs nothing; the spoken form is the choice.
Cost of changing it: a handful of spoken lines and chips.
My recommendation: "3 in 4" spoken, the percentage once on the chip, every form accepted.
A: "3 in 4" spoken / B: Fraction spoken
What to read after that: Topic 3 (again, if you want to see the fixes), with the two root-cause fixes, on the scripts page. Topic 2 carries your 13 October notes and is closed on keys again; no re-read needed. Topic 1's batch is ready but not yet posted: its eleven briefs are in the Video Brief Guide's ten-field form, the one the renderer is proven on, with the recap closes and every lesson under 180 words, and the checker passes them. My message handing it to the video pipeline session was blocked by this session's permission layer, so it needs you: either allow the send in the chat, or paste the hand-off message yourself (it is written out in the batch's READY file). Three recall items now sit before Topic 1's first lesson for the assumed grade 3 to 5 ideas.
| Lessons | Essential questions | Attached questions | Mixed quiz | |
|---|---|---|---|---|
| Topic 1 | 10 | 76 | 10 | 16 |
| Topic 2 | 10 | 51 | 10 | 14 |
| Topic 3 | 8 | 46 | 8 | 11 |
Decisions
1. Does "vacuole" alone count as the third plant-cell mark?
The course says: The typed item asks for the three parts a plant cell has that an animal cell never has; the accept list requires "large vacuole", so "vacuole" alone is marked wrong.
Why it is a question: - You ruled the mark is ONE LARGE vacuole, because animal cells have small vacuoles too. - A student who types "cell wall, chloroplasts, vacuole" knows the three parts but not the point. - You accepted "membrane" alone and refused "wall" alone on the same page.
Cost of changing it: one accept list.
My recommendation: refuse "vacuole" alone — the word "large" is the whole point of the mark.
A: Refuse "vacuole" alone / B: Accept "vacuole" alone
Applied without asking
Bank calls decided and logged: a chart may sit on a read-the-data bank item (never on a prediction); "CO2" is an accepted typed form of carbon dioxide; sweat, not urine, is Topic 1's excretion case; the coal item keeps its energy chain as the one hard case beyond the plant's body.
Topic 7 decisions taken: the spoken energy sentence returns to lesson 10 (the atom says "trace the path of food energy"); the dead-leaf case stays unsaid (the high-school course counts it biotic, so nothing here contradicts it); the two direction rule lines stay as a plain pair; a sixteenth practice item covers the air atom; "population" is "all the living things of one species in a place", since Topic 6 has coined species by then.
Topic 11 blueprint decisions taken at their recommendations: pea plants in the lessons, animals in the questions; the letter convention set when each allele kind is coined (capital at dominant, lower case at recessive); seven lessons, not five, for the sample; typed share seven of sixteen in the practice set.
Topic 2 lesson 10 keeps "hundreds of thousands" (the arithmetic above).
Summaries parked in Topics 2 and 3 as in Topic 1 (nothing deleted). Topic 2 lesson 8 renamed "Plant cell or animal cell?" because it now tests both ways. Topic 3 lesson 2 keeps six questions (the four on-sight sorts stay in reserve as grade 4's); lesson 3 keeps four, its why carried by the attached question. The willow lesson stays taught, not assumed, as your exact-masses ruling treats it. The strand intro video stays (it teaches nothing and is not a summary). Topic 1 lesson 10's typed order closes a run of which-level questions on purpose: that is its quiz, where you put the which-level skill. Option order is not fixed across a topic; the page shuffles every item.
Your rulings, applied
- Topic 1 review (all 12 points), your seven page rulings (arrow "react to make"; "a green plant"; chloroplasts without "only"; three marks with the large vacuole; autotroph and heterotroph atoms; "hotter thing heats" for every band; the inbox), the willow exact with a difference question, reserves left reserved, and the 33 clusters on a simple page (held while Becky reworks the tests).