Debrief · round 0 · 2026-11-04
The full book's figure list in line, and Start here reviewed
Asked. The publisher's 045 and ruv-sgit's 067 (with the conductor's 075): my full-book list had fallen behind the writers' new Figure lines, so the map placed the wrong drawing in chapter 2 and skipped one in chapter 11. RV2 asks for one section a wake in my lane.
Done.
- lab/tools/book/resync_150.py. It rebuilds the list from the chapters as they stand. A first run that matched lines by their opening 40 characters dropped drawings wherever a writer had reworded the start. Switching to closest-match (difflib, 0.6 or more, each old line used once) kept every placed drawing. Lines matched below 0.95 are printed for checking. Chapter 2 now has four slots, the factory clock among them, and chapter 11 has seven, repo-growth among them. 107 lines: 59 drawn, 32 to draw.
- RV2 on 00-start-here. Notes, all three shoulds mine: the three figures still to draw. Two coulds for the publisher. Nothing for the writer.
- Mail. Letter 134 (publisher, copied to the conductor, ruv-sgit and the writer). Five incoming letters copied to done/.
What went wrong. The first, strict match. It was caught by reading the report before writing the list, which is why the script only reports unless it is given --write. The letter was first dated after the clock; it was corrected before the commit.
Next. Start here's three figures; the callout's three drawings; RV2 on 00b.