EvaluatePlay (the hottest gameplay call, fired on every tile placement) now uses
the warm live-game cache directly: an active game stays cached (mutated in place
across moves, evicted only on finish), so the cached engine game and its immutable
seat list answer the membership check and the score with no DB read. The cold path
(eviction / first load) still loads and validates via the store. The seat list is
cached alongside the engine game for the membership fast path.
GetGame also folds its two round-trips (game, then seats) into one LEFT JOIN,
preserving the contract (same Game, a seatless game still returns empty seats, seat
order kept) — one round-trip for every remaining caller.
Measured at 500 players: evaluate p99 halves (200 -> 100 ms) and the per-op query
count drops. It does NOT cut postgres CPU — that is write-bound (per-move CommitMove
plus draft upserts and journal replays), the cheap indexed GetGame reads were never
its bottleneck, and postgres runs with headroom (~1.5 of 2 cores). So this is a
latency / query-volume optimization, not a DB-CPU one.
Regression cover: a non-player evaluate against a warm game asserts the cached-seat
membership path; the integration suite exercises GetGame's join across every game op.
Squash the 12 goose migrations into one 00001_baseline.sql (there is no prod
data; verified schema-identical to the chain via a pg_dump diff + the green
integration suite) and rename the game-variant labels
english/russian_scrabble/erudit -> scrabble_en/scrabble_ru/erudit_ru across the
backend, the FlatBuffers wire values and the UI.
dawg filenames and the Go enum identifiers are unchanged; the i18n display keys
are kept. Adds PRERELEASE.md (the R1-R7 pre-release tracker), linked from
CLAUDE.md. Contour DB wipe and the scrabble-dictionary tidy are follow-ups.
internal/game drives the engine over a single match and owns everything the
engine does not: event-sourced persistence (a games row + an append-only decoded
move journal, with the live engine.Game kept warm in a cache and rebuilt by
replay on a miss), the play/pass/exchange/resign transitions with
validate-at-submit scoring, an unlimited score/legality preview, the hint
(per-game allowance + profile wallet), the word-check tool with complaint
capture, per-player game state, history and GCG export (Poslfit dialect), and a
background turn-timeout sweeper that auto-resigns overdue turns honouring each
player's daily away window. Like Stages 1-2 it is a service/store layer with no
HTTP; the gateway surface lands in Stage 6.
Engine: additive decoded domain API (Direction, SubmitPlay/SubmitExchange/
EvaluatePlay/HintView/Hand, MoveRecord.{Dir,MainRow,MainCol}, Registry.Lookup,
ParseVariant) so internal/game never imports scrabble-solver; and a Resign fix
so the resigner keeps their score yet never wins (the other player wins a
two-player game). Timeout reuses Resign.
Persistence: migration 00002 adds games, game_players, game_moves, complaints,
account_stats and extends accounts with away_start/away_end/hint_balance; go-jet
regenerated. account gained SpendHint. Config adds BACKEND_DICT_DIR (required),
BACKEND_DICT_VERSION, BACKEND_GAME_TIMEOUT_SWEEP_INTERVAL, BACKEND_GAME_CACHE_TTL;
main loads the registry at boot (hard dependency) and starts the sweeper.
Tests: engine resign + decoded-API tests; game unit tests (GCG, away-window
boundaries, hint budget, cache, keyed mutex, payload); inttest integration
(lifecycle, replay equivalence, timeout sweep with away grace, resign stats,
hint policy, word-check/complaint, per-game-lock). Docs (PLAN, ARCHITECTURE,
FUNCTIONAL +_ru, TESTING, README) updated.
Ilia Denisov