sots-re/findings
alex 1893751ffa PAR: roll parity -- an AI client's per-turn RNG word count is not fixed at any scope
The hypothesis under test was the lockstep discipline: that each run of an AI client consumes a
fixed number of draws regardless of the path it takes, so a reimplementation could keep the
generator aligned with the right COUNT and order of draws while getting the decisions wrong.

It is false, and it fails at four scopes. Measured with a new bracket on
StrategyClient::OnResumePlaying over the per-client generator at +0x134, six runs across VM140 and
VM145, five fresh processes; every unpinned run's autosaves are byte-identical to the published
oracle, so the instrument is behaviour-neutral (rules 19 and 26 both discharged).

  across clients   turn2->turn3: 3 / 0 / 0 words for AI players 32 / 496 / 512; human 0
  across turns     client 32: 3 words on turn 2, 7 on turn 1
  across processes client 512 makes ONE cl_RandRange call on turn 1 -- the research-target
                   tie-break at 0x006a8495, phase 18 -- and it cost 1 word in one process and
                   3 in another, because RNG_NextInt is an unbounded rejection loop
  per site         RNG_Chance costs ZERO words at p<=0 and p>=1

Twenty-one live draw sites in an AI turn, in twelve functions (plus two provably dead ones); two
fired on the reference turn, three on turn 1. Only three are unconditional, and all three only
given that their enclosing function was called. Six of client 32's seven turn-1 words come from
the ship-design composer 0x006ad700, which is also where the only loop-carried draw lives.

Also: cl_RandFloat 0x00579c70, a third cl_* RNG facade, found twice independently. It reaches
RNG_NextFloat by a TAIL JUMP, so no rel32 sweep for the entry points can see it -- which is why
ai-turn-logic.md 5's 'zero NextFloat calls from the AI module' reads as true and is not. All 29
call sites of the three facades are inside the AI band: the cl_* RNG facade is AI-only surface.

Positives for the engine: the AI draws from nothing but its own client's generator (foreign_words
0 on every bracket), the human client draws nothing at all, and the per-turn cost is single digits.
Rung B is unaffected. Rung C needs the decisions.
2026-09-08 19:33:00 -04:00
..
control-flow L1: CDiff crosses at frame 50, and the AI client seed is fresh in every process 2026-09-08 18:23:19 -04:00
objects board: lane L5 - reference pair non-deterministic; interest literals verified with a failing control; verified column held at 0 on principle 2026-09-08 17:55:47 -04:00
resolutions resolution: byte-match survives as C-exact/C-set; tie set is a draw support not equal cost; rule 26 on self-consistent controls 2026-09-08 18:38:37 -04:00
subsystems PAR: roll parity -- an AI client's per-turn RNG word count is not fixed at any scope 2026-09-08 19:33:00 -04:00
.keep.md init commit 2026-09-07 12:31:56 -04:00
00-strategy.md strategy: phase 2 decisions (two repos, c++ shim, accrual order) 2026-09-07 17:10:21 -04:00
01-fingerprint.md map: RTTI class inventory (1924 types, engine=Mars); reseed board with real targets 2026-09-07 14:48:43 -04:00
_template.md scaffold RE campaign: board, backlog, findings template, verify tree 2026-09-07 14:40:44 -04:00