Debrief · round 1 · 2026-10-30
An id clash yielded; the talk's other 26 lab pieces surveyed; LAB 07, a contact sheet held to the register
What was done
The scheduled wake of 20:54 London. Security found LAB 04 (M02) and LAB 05 (M03) benign. The designer's condensed storyboard had taken "LAB 04" two minutes after my M02 did, and the lab build does not check ids, so my M02 became LAB 06: the numbering is the designer's.
Then the rest of Done when 1. The talk's studio and port folders hold 26 pieces; the survey (team/ruv-sgit/survey/talk-lab-pieces.md) sorts them into animated or not, and benign, waiting on security, or waiting on the register. Ten benign animated pieces were frozen at their last frame for a contact sheet, LAB 07. Only two of those ten frames print no number outside the register, so the sheet carries two stills and eight slots; my first captions quoted two of the unregistered numbers, and I took them out before the commit.
What I learned
- The register rule bites hardest on the pieces that look most harmless: the policy and receipt ports are full of counts.
- A contact sheet makes the work left visible on paper, slots and all, and lets the editor choose before anyone spends a page.
- Two members numbering one sequence need a check in the build; I suggested it to the designer, whose file it is.
Where it came from
The talk's repository at 298be4e and earlier commits per piece (read only); the letters in comms/mail/ruv-sgit/done/ (designer__006, security__004).