Drew T 2391ee628d src(phase-37): T4 batch t4_D47 — restruct: batch t4_D47 rung D — 285 files (TUs; 285 drawable) · 279 declaration units: 250 canonical / 0 promoted / 1 K&R marked / 0 aliases typed / 0 __builtin_abs → abs / 29 kept · compiles 553 in 46 s wall · final 285/285 identical · written 251 files | R22 check-all: 218 passed, 0 failed of 218 | (census skipped) (restruct_cycle)

2026-09-12 20:06:58 -06:00

..

agents

phase-36: S105 — f11 landed 4/4 + 4 copies, R22 218/218, census 4,010 / 0 unmarked (exit 0); T7 ☑ ENDED (Drew) | 🛑 T8 NEXT

2026-09-11 20:48:00 -06:00

baseline

phase-36: the tail pass g6 + g6b (47 + 63 of 400 small classes; 186/191 siblings propagated), R22 218/218; 30,806 → 30,358 sites; --try (a candidate scored without a tree write), delever_pack.py (the 57 T7 packs + PROMPT.md), --restore refuses an empty snapshot; the checkpoint: T7 as one agent at a time, approved, starts in the fresh session

2026-09-09 23:32:57 -06:00

census

phase-37: T1 ☑ — tools/type_census.py + the struct map: 7,261 struct definitions (525 layouts, 206 duplicate classes / 2,789 names, 39 VARIANT), 503,016 raw pointer-cast dereferences in four forms (coverage OK; + 18,912 address-of, 36,681 typed cast-member), 98,648 lying declarations / 1,609 callees, 18,760 evidence-clustered types explaining 99.2 % of sites, 24 parked P36 classes in parked.tsv, controls 4/4; lever_census controls re-keyed and --strict made literal (6,717 direct GTE statements in bodies); progress.py types block + README sentence + 2 corrections; the readability series read by column name with 16 census columns; docs/struct-map.md; SETUP §P37 S106; dictionary row; kit corpus | 🛑 T2 NEXT (the probe, Max)

2026-09-11 23:54:15 -06:00

delever

src(phase-37): T4 batch t4_D47 — restruct: batch t4_D47 rung D — 285 files (TUs; 285 drawable) · 279 declaration units: 250 canonical / 0 promoted / 1 K&R marked / 0 aliases typed / 0 __builtin_abs → abs / 29 kept · compiles 553 in 46 s wall · final 285/285 identical · written 251 files | R22 check-all: 218 passed, 0 failed of 218 | (census skipped) (restruct_cycle)

2026-09-12 20:06:58 -06:00

engine

phase-36: T7 wave b — four agent closes plus the engine's own two (766 bodies; 26,074 → 24,915 sites, R22 218/218)

2026-09-10 13:12:08 -06:00

gte

src(phase-36): T5 sweep3 — 274 dead asm-bearing macro definitions deleted in 133 files (uses counted per governing definition; SHB 416 → 142) | R22 check-all: 218 passed, 0 failed of 218 | census: 34,087 sites all marked · per-TU asm macro definitions 9,540 → 314 · GTE levers 462 · the selftest no longer writes the real canonical table

2026-09-09 07:00:40 -06:00

permuter

src(phase-36): T6 rung D banked — the four exemplars the permuter closed, tidied and byte-judged (R22 218/218)

2026-09-09 12:21:42 -06:00

probe

phase-36: T2 — the oracle and the probe: tools/delever_oracle.py (4,284 recipes via make -n -W; a candidate compiled in place to a scratch object and compared with the fleet run's; calibrated 177/177 untouched + twin 32/32 + positive control DIFFERS) and tools/delever.py (positional rewrites per lever class, rung A strip-all + rung B greedy, headers serial after the TU phase); --probe on 283 distinct bodies: A identical 53/283, sites removed 561/1,309 (42.9 %), needed 748 (launder 77 %, barrier 82 %, pins 45 % — v0/v1 + a0–a3 dominant), 0.225 s per compile — T4 priced at ~94k compiles ≈ 0.5 h, the residue ≈ 81 % of bodies ≈ 1,800 distinct exemplars; dictionary + SETUP rows; kit corpus; the 🛑 block for T3

2026-09-09 02:03:04 -06:00