Three people on the beach, and the cubes come down
The first phase judged from where a visitor stands, and four rules the judge got wrong
Phase 10 was the first phase judged from the ground. The four aerial cameras were held, and every number on them came back where phase 09 left it. The work was at cameras E and F, where a visitor will stand.
Three things were asked for at the skyline and the beach. The floating cubes were to leave the skyline of E and F. The one boatman who walked behind a flag in phase 09 was to be fixed at his two known faults and placed in the census with a shadow. Two more people were to be made through the route that had worked, a woman carrying a pot along the lane past F and a man at the boats at E. All three were built. The cubes are gone, replaced below thirty metres by flat cards of the distant houses and a lighter version of the palms between sixty and four hundred metres; the frame rate at E went from thirty-five to sixty-eight frames a second with them on. The boatman now stands within four degrees of vertical, walks at his own clip's speed of seven tenths of a metre a second with his feet sliding at most nine centimetres, and casts a shadow. The woman and the man both rigged first time, both have faces that read at three metres, and both entered the census as declared records at c.1700 and nowhere else. Eight paid jobs, one attempt each.
The judgment held the aerial score at 8.0 and gave the first eye-level score: 6.2 at the beach, 6.0 in the lane. Then it found four faults, and all four were the judge's own rules. The skyline gate said no pixel above the palms may differ between the normal render and the full-geometry render; that counted every lighter palm the same brief had asked for, while the thing the eye objects to, something drawn where the geometry shows sky, had fallen from over four thousand pixels in twenty blocks to a few hundred hairlines. The boatman's path was the test path from phase 09, written into two briefs without checking that a raft lay across its first four metres. The smoke was asked to change three thousand pixels and did, but every wisp rose in front of a palm crown and never reached the sky, so nothing reads. And the rule that made the figures' skin matte named every figure asset, while the men whose shoulders shine at E are not figures at all: they are modelled into the raft they sit on, texels of a boat.
The implementer's report had all four in it, plainly, with the numbers that proved them. It also volunteered that its count of the baseline skyline disagreed with the coordinator's by a factor of two, and it was right; the coordinator's figure did not reproduce. Every measurement in the report matched the coordinator's own re-measurement to the digit, for the seventh phase running.
Phase 11 moves the boatman's path two and a half metres along the beach to open sand, re-sets the four gates on the statistics the eye actually uses, gives the raft's seated men the same matte skin, matches the three new figures' colours to the sheets they were drawn from, and adds people only where a written dossier and a matrix of evidenced daily life put them.
- D-2026-09-15-01, phase 10 judged
- D-2026-09-15-02, phase 11 controls



