In a dead-drawn endgame — the two most recent journal moves are both
passes, so the board and the robot's rack are frozen and the robot is
bound to pass again — the robot still waited out its long late-game think
time (up to 90 min) before passing, needlessly dragging out a decided game.
Shorten that delay to a [0.8, 1.5]x band around the human's last-move think
time (the gap between the last two journal entries), clamped to [30s, 8min]
and taken as a min with the normal schedule, so the robot never moves
slower. A slow human collapses to the 8-min cap; a fast human is tracked,
with the floor keeping the robot from passing suspiciously instantly. The
anchor reads the move journal only (no schema change), stays deterministic
from the seed, and still defers to the sleep window.
RobotTurns now carries EndgamePass + OppLastMove, filled by one batched
journal query on the scan; the honest-AI single-game trigger keeps the
normal path (it moves at once). NextMoveAt (admin ETA) is left as the
normal-schedule upper bound.
The robot followed its per-game playToWin/lose intent on every move, which made
the outcome too predictable. It now flips that intent for a single move on ~20%
of opening/midgame turns (a winning robot eases off, a losing one surges ahead),
so the chosen strategy may not pan out — which favours the human. The chance
tapers linearly to 0 over the last 14 tiles in the bag and is 0 once the bag is
empty, so the endgame follows the chosen strategy strictly.
The decision is deterministic from the seed (mix(seed,"deviate",moveCount)) and
applies to both robot paths via the shared selectMove; the per-game play-to-win
intent the admin card shows is unchanged. Adds deviateProb/deviates helpers and
unit tests (taper bounds + monotonicity, never-in-endgame, determinism, ~20%
distribution); bakes the behaviour into ARCHITECTURE §7, FUNCTIONAL (+_ru),
backend/README, PRERELEASE and PLAN Stage 5.
New Game's quick game gains an explicit opponent selector — 🤖 AI (default)
or 👤 Random player. AI starts a game seated with a pooled robot that joins
and moves at once: no per-move timeout (a 7-day inactivity loss reusing the
turn-timeout sweeper), chat/nudge disabled, no statistics, the opponent shown
as 🤖 everywhere. The random path (disguised robot) is unchanged.
Driven by one game flag (games.vs_ai), set only on AI-started games so the
disguised path is never revealed; Matchmaker.StartVsAI seats the robot
directly (no open pool); the robot replies event-driven via the game service's
after-commit/after-create hook. Wire: vs_ai on EnqueueRequest + GameView.
A single tile that only extended a word perpendicular to the client-declared
direction was rejected: the UI always sent dir=H for one-tile plays (the
dirOverride/Controls toggle was orphaned in the Stage 7 game rework), so placing
"А" above "БАК" to form "АБАК" failed the solver's main-word-length check even
though the word is in the dictionary.
Make the backend infer a play's orientation from the placed tiles and the board
(internal/engine.resolveDirection): two or more tiles by the line they share, a
lone tile by the axis it abuts (longer word wins, horizontal on a tie). Direction
becomes an output, not an input: drop dir from the SubmitPlay/Eval wire requests
and add it to EvalResult. Journal replay keeps trusting the stored "H"/"V"
(SubmitPlayDir) so a rebuilt game matches the one committed.
UI: stop computing/sending direction; the preview now shows the words a move
forms with its total score (game.previewWords); the make-move control is disabled
until the play is confirmed legal; the "your turn" label hides while tiles are
pending. Delete the orphaned Controls.svelte.
Regenerate the FlatBuffers bindings (Go + TS) and update the gateway transcode
and the loadtest edge client to the new contract. Bake the decision into
ARCHITECTURE.md (§5/§9.1), FUNCTIONAL.md (+ _ru) and the backend README.
Two owner-reported defects from a live contour game.
A. Frequency: the robot's proactive nudge fired hourly for 12h+ (a 12h idle threshold
then the 1h cooldown, uncapped). Replaced with a lengthening, randomized schedule
(proactiveNudgeGap): the first nudge ~60-90 min into the human's turn, each later gap
growing toward 1-6h (uniform sample in [60min, ceil], ceil ramping 90min->6h over 12h
of idle, measured from the previous nudge), so a long wait gets a handful of
increasingly-spaced reminders instead of a stream.
B. Language: out-of-app push routed by the recipient's GLOBAL service_language
(last-login-wins), so after re-logging via the RU bot an English game's nudges came
from the RU bot. Now a game push (your_turn, game_over, nudge, match_found) carries
the game's own language (engine.Variant.Language) on push.Event, and the gateway
routes by it (falling back to service_language for non-game pushes). The New-Game
variant-gating guarantees the game's bot is one the player has started, so delivery is
never blocked.
Tests: proactiveNudgeGap unit + retimed TestRobotProactiveNudge; TestVariantLanguage;
emit your_turn/game_over language; TestNudgeRoutedByGameLanguage integration. Docs:
ARCHITECTURE (§7 nudge, §10/§13 routing), FUNCTIONAL (+ _ru), PLAN tracker.
Ilia Denisov