## Summary
`Mobile.DamageEntries` was a `List<DamageEntry>` allocated for every mobile, including the ~99% that never take damage. It is now an inline `ValueLinkList<DamageEntry>` (24 bytes in the `Mobile` object, no separate allocation) ordered least recent → most recent.
- `DamageEntry` implements `IValueLinkListNode<DamageEntry>`.
- `RegisterDamage` moves the entry to the tail in O(1) instead of `Remove` + `Add` on a list.
- Expired entries are always a head prefix, so pruning walks from the head and stops at the first live entry. The `DamageEntries` getter prunes on access.
- `DamageEntries` is exposed as `ref readonly`; enumerate with `foreach` or `.ByDescending()`. Mutation goes through `RegisterDamage` / `ClearDamageEntries`.
- `BaseCreature.GetLootingRights` and `BaseCreature.ComputeBonusDamage` take `in ValueLinkList<DamageEntry>`; all callers compile unchanged. Files that `foreach` over `DamageEntries` need `using Server.Collections;` for the enumerator extension.
- RunUO migration docs (`dev-docs/runuo-migration-docs/09`, `11`) and the `migrate-items-mobiles` skill document the change.
Saves one object and 16 bytes per mobile (~8 MB and 500k gen2 objects on a 500k world). Second of three PRs from the lazy per-mobile collections design (first: #2604). Branched from `main`; the two diffs touch disjoint hunks of `Mobile.cs`.
## Breaking change
- `Mobile.DamageEntries` is no longer a `List<DamageEntry>`. Indexing, `.Clear()`, `.Add()`, `.Remove()` no longer compile; use `foreach`, `.ByDescending()`, `.Count`, `ClearDamageEntries()`, and `RegisterDamage`. Calling a `ValueLinkList` mutator on the `ref readonly` property compiles but operates on a copy while still unlinking the real nodes; do not.
- `BaseCreature.GetLootingRights` and `BaseCreature.ComputeBonusDamage` signatures changed to `(in ValueLinkList<DamageEntry>, …)`.
Save format is untouched: damage entries are not serialized.
## Behavior
Recency order, `allowSelf`, tie-breaking in `FindMostTotal`/`FindLeastTotal` (most recent wins), `Responsible` accounting, and loot-rights ordering are unchanged and covered by the new `DamageEntryTests` and `LootingRightsTests`.
## Testing
- `dotnet build -c Release` clean.
- New `DamageEntryTests` and `LootingRightsTests` plus full `Server.Tests` and `UOContent.Tests`.
Fixes walk-up aggro latency (up to a full 10 s of obliviousness) and hardens the reacquire gate so no state can silence acquisition, while turning `AcquireOnApproach` into the reaction-time knob for future per-creature intelligence tuning.
### Why
`AcquireFocusMob` re-armed the 10 s `ReacquireDelay` **before** scanning, success or failure. A creature that scanned an empty room was blind for 10 s to a player walking up — walk-up aggro latency was uniform in 0..10 s. Waking from sector sleep stacked the AI timer's 0–3 s construction stagger on top. And `NextReacquireTime` is not serialized: on hosts whose tick counter starts negative (GCP pass-through), the 0 default blocked **all** acquisition shard-wide after a restart until the counter crossed zero.
### What
**Event-driven reaction — `AcquireOnApproachDelay` (the intelligence gradient)**
- The paragon `AcquireOnApproach` bool becomes a `TimeSpan` on every creature: an enemy moving inside `AcquireOnApproachRange` (10 for all creatures — on-screen reactive aggro; the periodic scan keeps the wide `RangePerception` sweep) *clamps* the next scan to at most the delay. Repeated steps cannot shorten it further — one scan per delay period, not per step or think.
- `Zero` (paragons) also prods the AI timer: the ranked scan engages within a wheel turn — the old snap, minus the special-cased engage path. The target now comes from the normal FightMode ranking instead of whichever mobile happened to move, and the `Combatant == null` guard stops re-engage spam.
- The 2 s default reads as "took a beat to notice you"; larger values are dumber; `ReacquireDelay` alone is the oblivious floor. Mover checks are the approach logic's `IsEnemy` + `CanBeHarmful` (so pets count and hidden movers are excluded via `CanSee`), with `IsEnemy` first to cheaply reject same-team wild creatures wandering past. The check rides the `OnMovement` callback every step already pays for — no polling added.
**Gate correctness**
- Every scan re-arms the full `ReacquireDelay`, success or failure (classic semantics; reaction time is the approach path, not the poll).
- Self-healing by construction: a deadline further out than `ReacquireDelay` is an illegal state and reads as open — no wedged or wrapped value can silence acquisition beyond one delay period.
- `NextReacquireTime` is seeded from a live tick on deserialize (the GCP negative-tick blackout).
**AI timer wake**
- Activation (sector wake, spawn, resurrection) starts within a 0–256 ms spread instead of the 0–3 s construction stagger, which read as lag.
- The stagger's real job — keeping same-speed cohorts out of lock-step (the RunUO town artifact) — is now a zero-mean ±period/8 jitter on each **idle** think, so phases random-walk apart within seconds and can never re-lock. Instrumentation showed why a one-shot spread can't do this job: the timer wheel fires within ±1 ms, so with 10 creatures on a 500 ms period some pair collides on nearly the same phase ~75% of the time (birthday paradox) and then steps in the same loop iteration *forever*. Jitter is scoped to passive speed: engaged cadence stays exact, since pursuit timing anchors to real step times.
**Debug**
- The `AcquireFocusMob` scan message no longer re-arms the shared 5 s debug cooldown, which swallowed every AI's "I have detected X" transition line.
**API change** for custom scripts: `AcquireOnApproach` (bool) → `AcquireOnApproachDelay` (TimeSpan). Documented in `content-patterns.md` § Target Acquisition, `runuo-migration-docs/09` + `11`, and the migration skill checklist.
### Tests
`AcquisitionTests`: both scan outcomes honor `ReacquireDelay`; a 60 s-wedged gate still acquires; enemy movement clamps the deadline (same-team wild movers and out-of-range movers ignored); repeated movement cannot shorten below the delay; `Zero` opens the gate and prods without a direct engage. Full suite: 755 UOContent green.
Stacked on #2594. Fixes jerky creature movement (lich / Fast-bucket melee chases) by choosing the client animation flag from the actual step pace instead of a caller-supplied `run` argument, and fixes three step-pacing defects in the move budget found while verifying it with paired server/client traces.
### Why
The `Direction.Running` bit does nothing for creatures server-side (`Mobile.OnMove` reads it only for the player throttle and stealth reveal). Its whole effect is on the client, which animates each step over a fixed time selected by that bit: walk 400 ms / run 200 ms on foot, 200 / 100 ms mounted. ClassicUO queues up to 5 steps and *drops* the sixth, so a creature stepping every 300 ms while flagged as walking backs the queue up until it snaps forward — the observed jerk.
The `run` argument never carried the one fact that matters (the step interval). RunUO passed `true` in combat / `false` for pets and gated it on `dist > 5`; #2271 flipped every combat site to `false`; pets passed `currentDistance > 2`. None of that is a coherent signal.
### What
**Pace-derived run flag**
- `BaseAI.ShouldRun()`: run iff the effective step delay (move clock + badly-hurt inflation) is shorter than `Movement.WalkFootDelay` / `WalkMountDelay` (mounted or flying) — with a continuity rule: an *isolated* step (taken after standing at least a walk interval) goes out as a walk, because the client renders each step alone and a lone run-flagged step is a 200 ms dart. Only a continuing cadence flags run; a true sprinter (pace under the run interpolation) always runs, since a walk-rendered first step would flood the client's 5-step queue. This reproduces RunUO's close-in feel (its `dist > 5` gate) from first principles.
- `DoMoveImpl` stamps the bit; it is the single place the flag is set.
- `run` removed from `MoveTo`, `WalkMobileRange`, `ApproachTarget`, `MoveToPoint`, `MoveToWithGroup`, `MoveToWithCollisionAvoidance`, the move intent, and `PathFollower.Follow`. All 35 call sites updated. **API change** for custom scripts — documented in the RunUO migration docs (`09-items-mobiles-creatures.md`, `11-api-reference.md`) and `content-patterns.md` § Creature Speeds.
**Move-budget pacing fixes** (each confirmed by UTC-aligned server/client step traces)
- A stall no longer banks catch-up steps: the budget's snap-to-now released up to three steps in ~300 ms when a creature resumed chasing after standing beside its target — rendered as a teleport.
- Debt accrual removed entirely: a step landing sub-period late (think-grid vs budget misalignment during reactive mirroring) kept the remainder and fired a follow-up ~100 ms later — a dart pair. `ConsumeMoveBudget` now paces every step from when it was actually taken; in continuous pursuit the move-wake lands within wheel resolution of the deadline, so the cost is single-digit-ms drift.
- Net effect: a creature can never step faster than its pace, verified across a full chase session (zero sub-pace steps; metronomic 350 ms cadence for a 0.3 s lich).
- Test fixture now runs `Movement.Configure()` (the walk delays were 0 in tests).
### Accepted trade-off
Animal (LOW group) bodies without a run animation slide on their stand frames when flagged as running. Most are slow enough to stay flagged as walking; the client-side fallback is in ClassicUO/ClassicUO#1930.
### Tests
`RunFlagTests`: foot thresholds (0.3 / 0.125 run; 0.4 / 0.45 / 1.05 walk), flying uses the mount threshold, badly-hurt inflation flips a 0.35 s creature back to walk, a real `DoMove` stamps the bit, isolated steps drop to walk (sprinters keep running), a stall restarts the cadence with no banked steps, and a late step earns no quicker follow-up. Full suite: 837 Server + 747 UOContent green.
Closes#2593. Closes#2595.
Two related pet-AI fixes: the post-#2591 pacing/wake regression (#2593), and the guard order silently converting to Attack during combat (#2595). Root-cause analyses are in the issues.
## #2593 — pets follow slowly; stale AITimer wakes
**Why pets slowed:**
- The per-step budget grew from **half a think interval** (`CurrentSpeed * 500`) to the full RunUO-parity move table (`CurrentMoveSpeed * 1000`). Medium-bucket pets (Horse, Dog, most tamables): passiveMove **1.05s/step**.
- Pet order speed depended on stale `Warmode`: `HandleGuardOrder` set it once, but `OnCombatantChange` clears it whenever the combatant drops, so obedience ran active or passive **by combat history** — usually passive. Net: Guard/Come at ~1.05s/step (~2.1x slower than pre-#2591), vs a player running at 0.1–0.2s/step.
- The AITimer never rescheduled its pending wheel entry: the wheel reads `Interval` only after the next fire, so a speed-up or a fresh order (`Activate()` no-ops while running) waited out the stale wake — up to a full passive think, stacked on the residual move budget on Guard → Follow.
**What changed:**
- **Order handlers own obedience speed** (RunUO `OnCurrentOrderChanged`/`DoOrder*` parity, re-derived continuously): issuing a movement order (Come/Follow/Guard/Attack) sets the **active** think clock, resting orders (Stay/None/Transfer) set passive, and the guard/follow peaceful branches write **RunUO's AOS `CurrentSpeed = 0.1` sprint** — RunUO's guard else-branch had the identical write as follow. The bespoke 0.1 fuses to both clocks through #2591's existing classification, so `CurrentMoveSpeed` stays **pure herding + classification** with no obedience special case, and `DoMoveImpl`'s per-step flip skips obeying pets (their handler owns the pace) and loses its old follow-only 0.1 write. Combat still re-derives organically via warmode/combatant.
- **`AITimer`**: tracks the pending wake and reschedules (`Stop`, `Delay` = remaining, `Start`) when a speed-up or fresh order moves the earliest deadline up; changes inside a tick still flow through `ScheduleNext`. New `Prod()` wakes the AI immediately on player commands — including from a stopped timer, so stable claims no longer wait out the random construction stagger. Sector/spawn wakes keep the stagger. Spam-safe: a prodded think grants reaction, never action — steps/swings/casts/abilities are gated by their own budgets and timers.
The residual move budget is deliberately **not** cleared on order change — that would let order-spam macros grant free steps. Deadline changes reschedule the timer; rate changes take effect at the next deadline computation.
## #2595 — Guard order converts to Attack during combat
**Why:** `FindCombatant()` set `ControlOrder = OrderType.Attack` when engaging, so a guarding pet left the Guard order for the whole fight: OPL tags wiped (pet `1080078` + master `501129`), no retargeting (`DoOrderAttack` locks its target), `TeleportPets` left the pet behind on recall/gate, and every engage→kill→resume cycle replayed the guard flourish.
**What changed:**
- **`FindGuardTarget()`** (was `FindCombatant`): a pure selector — prefers the aggressor **closest to the master** (RunUO guard parity, dynamic retargeting to protect the owner), keeps the current combatant unless a strictly closer one exists, and never mutates order state. `DoOrderGuard` engages through it while **staying in Guard** the whole fight.
- **Persistent-order semantics** (the ModernUO improvement over RunUO): an explicit `all attack` completes → `ResumePersistentOrder()` returns to Guard → the guard scan engages remaining threats in-order. The Attack-chaining fallback (`FightMode.Closest/Aggressor`) now applies only to non-guard persistent orders. Resuming Guard no longer replays the sound/"is now guarding you" message.
- **Peaceful guard stands down deterministically** (`Warmode`/`Combatant`/`FocusMob` cleared) and returns to the master at the RunUO sprint (see above); at the master's side it stays organically active.
- **`WalkMobileRange` honors the caller's run flag** (the internal hardcoded `dist > 5` gate silently overrode it). Run is animation-only server-side; the only callers passing anything but `false` — follow, guard, clone — gate on their own thresholds.
## Resulting behavior (Medium-bucket pet)
| Scenario | Broken | This PR |
|---|---|---|
| Guard trailing master (AOS) | ~1.05s/step, think-grid quantized | 0.1s/step sprint (RunUO parity), smooth move wakes |
| Guard during combat | order flips to Attack; tags lost; no retarget; left behind on recall | stays Guard; retargets to master's closest aggressor; teleports with master |
| `all attack` while guarding | resume spams guard flourish per kill; chains into Attack | resumes Guard silently; guard scan takes over |
| Come / friend-follow | 1.05s/step | activeMove 0.45s/step (≈ pre-#2591 feel) |
| Guard → Follow reaction | up to ~1.5s dead time | think within one wheel turn |
| Follow master (AOS sprint) | 0.1s/step | 0.1s/step (unchanged) |
| Wild creature chase | RunUO-parity move table | unchanged |
Also documents two contracts this work leaned on: the `ControlOrder` setter deliberately fires on every assignment (a reissued order is a command — retarget/break-off/re-anchor), and `OnThink`/`MonsterAbility` must be excess-call tolerant (`dev-docs/content-patterns.md` § OnThink: the excess-call contract).
## Testing
- Full suite passes (1570: 837 Server + 733 UOContent).
- `PetPacingTests`: order-issue think-clock parity, follow-master sprint via Obey, guard organically active at the master's side, combat-chase and herding boundaries, plus two deterministic timer-wheel tests (8ms-lockstep slicing) proving a fresh order and a mid-wait speed-up wake the AI promptly.
- `GuardOrderTests`: engage keeps the Guard order; retargets to the aggressor closest to the master; explicit attack resumes Guard without chaining into Attack; peaceful guard stands down. Setup self-validates LOS/terrain.
- `GuardFollowTests`: guard-following registers a move intent, steps toward the master, sprints at 0.1 under AOS (per-step flip must not undo it), and runs active pre-AOS.
- All behavioral tests were written first and failed for the documented reasons.
Splits creature speed into two clocks so movement pace can be tuned without touching reaction time:
- **Think clock** — `ActiveSpeed`/`PassiveSpeed`/`CurrentSpeed`: seconds per AI decision. Unchanged in meaning, storage, and cadence.
- **Move clock** — `ActiveMoveSpeed`/`PassiveMoveSpeed` (+ resolved `CurrentMoveSpeed`): seconds per step. `0` = inherit the matching think value.
### How
- Move speeds come from optional `activeMove`/`passiveMove` in `npc-speeds.json`, are `[props`-tunable per instance (set `0` to re-inherit), and serialize (BaseCreature v22).
- `SetSpeed()` keeps its legacy one-clock semantics — sets the think clock **and clears move overrides** — so existing callers cannot half-configure a creature. `SetMoveSpeed()`/`ClearMoveSpeed()` configure movement explicitly; `ScaleMoveSpeed()` scales overrides for buffs.
- `CurrentMoveSpeed` is derived by classifying `CurrentSpeed`: a verbatim active/passive think value maps to the matching move value; a bespoke pace written directly (mount boosts, follow sprint) stays fused to both clocks. External `CurrentSpeed` writers need no changes.
- `AITimer` schedules the earlier of the two deadlines. Decisions run at the think cadence exactly as before; while a pursuit/investigation is live, the timer also wakes when the movement budget elapses and advances one step with no decisions. Steps no longer snap to the think grid, so any step delay paces smoothly on the 8ms wheel. A blocked creature schedules no move wakes.
- The movement budget is RunUO's `m_NextMove` accumulate-and-clamp at a full step, so long-run pacing averages `CurrentMoveSpeed` exactly.
### Behavior changes
- **`npc-speeds.json` buckets get RunUO `TransformMoveDelay`-parity move values**: creatures step at RunUO pace while thinking/reacting at current speed. The situational +0.1/+0.2 offsets are deliberately omitted.
- **Existing saves migrate on load**: a pre-v22 creature whose think speeds still match its npc-speeds entry (never hand-tuned) adopts the table's move values — worlds and pets pick up the new pacing without a respawn. Tuned creatures keep movement inheriting their think clock.
- **Paragons scale movement by `SpeedBuff` (1.2x)**: RunUO had no deliberate policy here — dividing by 1.2 knocked most speeds off `TransformMoveDelay`'s exact-equality table (raw pass-through, 2x+ faster), while 0.3/0.6 creatures landed back on it for ~1.33x. This applies the uniform 1.2x the buff always claimed. UnConvert snaps speeds back to exact table values within 1e-4 — /1.2 then ×1.2 drifts 0.45 and 0.9 by an ulp, which would read as hand-tuned (and defeat a future skip-table-conformant-values serialization pass); tuned speeds keep.
- **Herding paces the movement clock**: the old `CurrentSpeed` getter hack is gone. A herded creature walks at a fixed 0.3s/step — RunUO's forced pace, without its `TransformMoveDelay` inflation to 0.6 — so herding is never penalized by a slow creature. Thinking is untouched, and `CheckHerding` walks through `MoveToPoint`, so herded creatures path around obstacles.
- **Badly-hurt slowdown now inflates the step delay only** (RunUO parity), computed from the base each step. Previously it wrote `CurrentSpeed = CurrentSpeed + 0.05..0.15` back on every successful step — compounding unboundedly while hurt and slowing decisions too.
- Removes the vestigial `MoveSpeedMod` (never read, written, or serialized).
- With no bucket or per-instance move values, both clocks carry identical values and creatures pace as before.
### Testing
- Full suite passes (1557, including 12 new `MoveSpeedTests`: resolution classes, `SetSpeed` clearing, `0`-re-inherit, v22 round-trip with exact-consumption check, save migration adopt/skip, buff scale/snap, herding).
- In-game verified via local diagnostics build (per-step budget tracing): steady 700ms step cadence on a 0.3s think grid with one-step catch-up after idle, think grid unperturbed by move wakes.
### Summary
* Upgrades Serialization Generator to v3. This contains numerous bug fixes and a significant performance improvement.
* Bumps other dependencies.
Two features ran on every shard out of the box, each polling on its own 60s timer for files most shards never generate, neither ever asked for. Fixing that turned into untangling why they shared a config file — and then into the on-loop cost of the three lists behind them.
## Before / after
Measured on the shipped defaults. On-loop numbers are what freezes the world; the tick budget is 8 ms.
| | before | after |
|---|---:|---:|
| Blocklist poll on a shard with no list | every 60s, forever | **none** (opt-in) |
| Manual allowlist poll on a shard with no carve-outs | every 60s, forever | **none** (opt-in) |
| Promote-guard sweep timer | leaked on `Stop()` | stopped, and only started when hits are reported |
| Login allowlist flush, on-loop | O(n) walk + 2 arrays **every 60s**, LOH past ~5,300 entries | reused buffers, **hourly**, zero steady-state allocation |
| Auto-denylist, accept path | 9.1 ns/call | **6.1 ns/call** |
| Auto-denylist, sustained flood at cap (60k rejected) | 26.7 ms | **9.3 ms** |
| Auto-denylist, flood end — **worst single call** | 9.49 ms | **0.05 ms** |
| Auto-denylist cap | 65,536 (stranding 9,895 slots) | **324,449** (exact `HashSet` capacity, ~19 MB) |
The auto-denylist row that matters is the third: the on-loop stall at flood end drops **190×**, because retiring lapsed holds is now the number expiring rather than the number held.
## Why this design
It is built for the shape of attack these shards actually see: **hundreds to a few thousand connections per second**, occasionally tens of thousands, sustained over minutes rather than delivered instantly. Against that shape the cap now covers the whole observed range (50k–250k distinct sources) in memory, and the work of expiring them spreads across the accept calls that were already happening.
There is one case this design is *worse* at than the old one: if every held entry lapses within the same millisecond, retiring them costs ~10.7 ms against the old ~8.9 ms, because the ring's random-access set removals lose to a sequential dictionary scan. Reaching it requires an entire flood to arrive inside one millisecond. **A shard absorbing 324,449 connections in a millisecond is finished at the accept path no matter what this list does** — that is the point where the answer is upstream security and scrubbing (an L4 proxy, edge filtering, a bouncer at the kernel), not a data structure in the game loop. We chose the design that fits the attacks we see and degrades honestly past them, rather than over-engineering for one we do not.
## Blocklist — now opt-in
`BlocklistFilter.Start` only bailed when `_path == null`, which needs `file` to be empty. The default is `"Configuration/ip-blocklist.txt"`, so on any default install both `Task.Run(PollLoop)` and a recurring `SweepGuard` timer started unconditionally, logging *"Blocklist inert: no list at …; polling every 60s"* and then doing exactly that forever.
Adds `"enabled"`, default `false`, using the `_enabled = s.Enabled && <preconditions>` idiom already in `LoginAllowlist` and `AutoDenylist`. **Upgrade is deliberately loud**: a missing key binds to the default, so `LogWhyDisabled()` splits three cases and a shard with a list on disk but no `enabled` key gets a **Warning**, not silence.
## `FileAllowlist` → `ManualAllowlist`, with its own config
Moves to `Configuration/ip-allowlist.json` (`enabled` default `false`, `files`, `reloadInterval`) and into `Network/ManualAllowlist/`, mirroring `Network/LoginAllowlist/`.
It was never a sub-feature of the blocklist. `ManualAllowlist.Contains` has two callers:
| Caller | Could anything else do it? |
|---|---|
| `BlocklistFilter.Evaluate` | **Yes** — the generator already subtracts these files at generation time |
| `BanExemptions.IsExempt` | **No** — sole mechanism for suppressing behavioural ban contributions |
The second reaches `BanChannel.IsExempt` with no blocklist in the path. A shard running **no blocklist** still needs this so the admin's own IP isn't auto-banned by rate-limit detection, so a shared flag couldn't express it — the implication is asymmetric. They still work together via a startup warning when the blocklist is on and the allowlist is not.
On the name: "File" described the storage. The distinction from `LoginAllowlist` is **provenance** — declared by an operator versus earned by authenticating — and "Manual" matches `BanReasons.Manual`. `allowlistFiles` is removed from `BlocklistSettings` outright; blocklists have not shipped long enough for anyone to have set it.
## Login allowlist flush
`Flush()` allocated two arrays sized to the live entry count and copied the whole dictionary into them **on the game loop**, every 60s. `UInt128` is 16 bytes, so past ~5,300 entries that first array was an LOH allocation once a minute, forever. The file write was already off-loop; the walk was not.
Static buffers grown geometrically; the writer owns them until it posts completion back through `Core.LoopContext`, so `_writing`/`_dirty` stay loop state (rule #10). Interval → 1 hour against a 90-day TTL. Clean shutdown writes synchronously via `EventSink.Shutdown`; `HandleClosed` skips `InvokeShutdown` when crashed, so the crash path subscribes separately and only writes when it is actually on the loop thread. Also fixes a pre-existing hole where `_dirty` was cleared *before* the write, so a failed write dropped entries despite the comment promising a retry.
## Auto-denylist: expiry ring
Reclaiming lapsed holds was O(entries held) — every cap-triggered reclaim during a flood walked the whole dictionary to find the few that expired, and `_warnedFull` suppressed the log, not the work.
A hold is **never refreshed** now: the first detection sets the expiry, later ones leave it. That makes insertion order equal to expiry order, so a ring of the same keys is sorted by construction and retiring stops at the first live record. Nothing is lost — the rate limiter runs *ahead* of the connection filters (`NetState.Network.cs`) and reports to the ban channel, so a flooder whose hold lapses is re-held on its next attempt.
Because the ring carries the expiry, the membership side only answers "present?", so it is a `HashSet` — measured at **36 B/slot against the dictionary's 52**. `HashSet` and `Dictionary` share `HashHelpers`, so the from-empty capacity progression is identical (36,353 → 75,431 → 156,437 → 324,449 → 672,827) and the cap still lands on one exactly. The ring is parallel `UInt128[]`/`long[]` rather than an array of structs — `UInt128` forces 16-byte alignment, so a packed pair costs 32 bytes where these cost 24, and the drain reads only the `long[]`.
Rejected after measuring: splitting the drain into a scan loop plus a removal loop (inside noise — both issue N hash removes, and the pointer math was never the bottleneck), and `Dictionary<UInt128,bool>` with tombstoning instead of removal (10% slower *and* unbounded, which breaks the cap).
## Testing
Build clean, 0 warnings. **1,530 tests pass** — 708 UOContent, 822 Server.
Tests were reworked rather than patched: the refresh test inverts to `Repeat_detection_does_not_extend_the_hold`, the obsolete sweep-throttle test is deleted along with the throttle, and four were added for the ring — set/ring parity, release-then-re-hold not being retired by the stale record, exact fill of a non-power-of-two cap, and the moved allowlist config's casing contract. The throttle test added mid-PR was verified to fail without its fix before being deleted.
One commit is comments only (verified: a diff filtered of `//` lines is empty), removing development narration — a `"(Task 2)"` plan reference, `"matching the per-feature JSON config pattern used by X"` across four loaders, a duplicated threading note — and repointing `Firewall` at `dev-docs/ip-bans-and-allowlists.md` instead of a "ban-channel design doc" that does not exist.
Note `Distribution/Configuration/blocklist.json` is gitignored (`.gitignore:14`) and generated from the record defaults on first boot, so the record default *is* the shipped default.
### Summary
Players reported an exploit: decipher a treasure map, then run a ClassicUO/Razor organizer agent that pulls the gold out of the chest in small amounts. Each pull spawned more monsters, turning one chest into an unbounded farmable spawn generator.
### Root cause
`TreasureMapChest.OnItemLifted` grants a 10% guardian spawn roll per first-time-lifted item, deduplicated by the instance-keyed `_lifted` set. But a partial lift goes through `Mobile.LiftItemDupe`, which re-adds the stack remainder to the chest as a **brand-new item instance** (engine-side `AddItem`, bypassing the `CheckHold` block on refilling). Every subsequent pull lifts an instance the `_lifted` set has never seen, so each one re-rolls the 10% spawn chance:
- A level 4 chest holds 4,000 gold → pulled coin by coin, ~400 spawned creatures (plus more from reagent stacks), hands-free, per chest.
- Spawns use `guardian: false`, so nothing tracks or caps them.
- Legit full-stack looting yields roughly 5–8 bonus spawns per chest for comparison.
The code is inherited from RunUO, so descendant shards likely share the hole.
### Fix
Mark every item that enters the chest **after the initial fill** as already lifted, via an `OnItemAdded` override gated by a non-serialized `_filled` flag (set at the end of the constructor and in `[AfterDeserialization]`). Ordering makes this exact: `LiftItemDupe` re-adds the remainder *before* the chest's `OnItemLifted` runs, so the lifted original still gets its one legitimate roll while the remainder is pre-marked.
This also covers packing items *into* the chest (e.g., merging gold back in to lift it out again) and bounce-backs — anything not part of the original loot can never grant a spawn roll.
### Tests
- `PartialLift_MarksSplitRemainderAsLifted` — drives the real `Mobile.Lift` path with a 1-coin pull and asserts the split remainder is marked (failed before the fix).
- `ItemAddedAfterFill_IsMarkedLifted` — post-fill additions are marked (failed before the fix).
- `OriginalFillLoot_IsNotMarkedLifted` — original loot keeps spawn-roll eligibility.
Full `UOContent.Tests` suite: 701 passed.
## Why
An Argon2 verify is **~8.9 ms of frozen world per login attempt** — more than half a 16 ms frame. Failed attempts cost exactly the same as successful ones, by design, so a credential-stuffing flood is a full-cost stall per packet without needing valid credentials. `SetPassword` derives a hash too, so `[password`, the admin gump and account creation each pay the same.
## What the measurement says
Off-loading does not delete the cost, it relocates it. Three things stay on the loop:
| Component | Measured |
|---|---:|
| Inline verify (today) | **8.92 ms** |
| Dispatch to the worker | 210 ns |
| Drain the continuation off `LoopContext` | 13 ns |
| Loop's own work slowed by shared-L3 eviction | **0.05 – 5.44 ms** |
Net gain **3.5 – 8.9 ms** of on-loop time per login. Harness in `ModernUO-Benchmarks` (`Benchmarks/Argon2OffLoop/`): it models the loop as a dependent-load pointer chase swept across working-set sizes, which is an upper bound on cache-latency sensitivity, and copies `EventLoopContext` so the hand-off cost is the real one.
Two results shaped the design:
- **The contention tax peaks in the middle of the working-set range**, not at the top — 5.44 ms at 8 MiB (a quarter of this chip's L3), but 0.76 ms at 30 MiB and 0.10 ms at 256 KiB. A tiny hot set has nothing in L3 to lose; a huge one is already DRAM-bound.
- **Per-login tax falls as concurrency rises** (5.44 → 2.56 → 1.60 ms at 1/2/4 hashers) while *total* loop damage rises. Contention is shared, not additive, so a login rush is not the disaster case — a single login is.
## Why exactly one worker
It is load-bearing three times over, which is also why it must not quietly become a pool:
- **Cost bound.** Off-loop loses to inline only if a hash steals ~82% of the loop's throughput. One hasher contending for one core leaves the loop ~50%. **A single background hasher cannot cost the loop more than the inline verify under any scheduling regime**, which is what lets the measurement hold on hardware we cannot inspect — AMD, VPS, oversubscribed VM. Four hashers drop the loop to ~20% and break it.
- **Memory.** Exactly one hashing arena is live at a time whatever the login volume.
- **Ordering.** Writes apply in dispatch order *only* because a single thread drains FIFO. A second worker would need ordering reintroduced; `WritesApplyInDispatchOrder` fails if that happens.
Throughput is ~110 verifies/sec. Only loop time matters, not login latency, so head-of-line blocking during a rush costs nothing.
## Making every protection safe off-thread
The worker was initially Argon2-only. That was the right call for the wrong reason — it was blamed on Argon2's salt RNG, which is a stateless syscall wrapper and was never a problem. The real blockers were elsewhere, and both are fixed at the source:
| Protection | Was | Now |
|---|---|---|
| MD5/SHA1/SHA2 | shared `HashAlgorithm.ComputeHash`, which carries the running digest across `HashCore`/`HashFinal` through process-wide singletons | static `HashData` into a `stackalloc` span — no state, no allocation, identical bytes |
| PBKDF2 | `Utility.RandomMinMax` → shared `System.Random`, thread-unsafe *and* game state | `RandomNumberGenerator.GetInt32`, matching the salt beside it |
| Argon2 | already safe (`Verify` is static + stackalloc) | unchanged, singleton reused |
Literal digests are pinned in a test **before** the change and still pass after it. These are compared as strings against every account database, so any casing or encoding drift would lock out every SHA and MD5 account at once.
With all three safe, the worker no longer knows which algorithm it runs and the dispatch conditions collapse to "is off-loop available".
## Correctness
- **Phrase derivation** moves to `AccountSecurity.DerivePhrase`, so verification (stored algorithm's rule) and rehash (target algorithm's rule) cannot disagree. Deriving with the wrong one is the shape of the lockout fixed in #2562.
- **Liveness** is checked at dequeue *and* at apply — a connection can drop while queued or while the result sits in the loop queue. A job with no connection attached, such as an admin password change, runs regardless.
- **Queue overflow rejects** a login rather than verifying inline; steering work back onto the loop is what a flood wants. A password change instead falls back to hashing inline, because unlike a login it must not be dropped.
- **Shutdown and crash** both just stop the thread, and pending jobs are dropped. No save is initiated once shutdown begins — saving is the operator's choice up front, via the admin gump's save/no-save variants, and `WaitForWriteCompletion` honours one already in flight — so a write applied during teardown would reach no disk. The crash path needs its own subscription because `HandleClosed` skips `InvokeShutdown` when crashed.
## Bounding
`MaxPending` is 4096 — a backstop, not a flood defense. `SentFirstPacket` holds a connection to one pending verify and the engine caps connections at 4096, so the queue is already bounded by construction and this can only trip if that invariant breaks. A cap low enough to blunt an attack would reject real players first; during a mass reconnect they *are* the queue. Flood defense belongs at the connection layer.
The real DoS improvement is elsewhere: today every attempt stalls the world, and after this a flood occupies one core while the loop keeps ticking.
## Gate
Release builds on 4+ cores. Below that there is no spare core to move work to, so off-loading buys nothing by construction; `DEBUG` is excluded because dev boxes and test shards have few logins. Both modes call the same code — the gate only chooses where it runs.
## Engine change
One property, `AccountLoginEventArgs.Deferred`, so a subscriber can say "no verdict yet". `EventSink.AccountLogin` is `Action<...>` with no continuation, and the packet handler replies in the same call. Approved separately since it touches `Projects/Server/`.
## Docs
`dev-docs/threading-model.md` and the threading skill gain a vetted-workers section. The forbidden-patterns table bans `new Thread`, `ConcurrentQueue<T>`, `Interlocked` and `volatile` in `UOContent`, and its exceptions covered only `Projects/Server/` — the existing Advanced Search fan-out already sat outside it. The new section leads with proving the need (measure on-loop time, not wall-clock; gate on core count; record the measurement), keeps game logic on the loop via chunking, and documents the hand-off protocol in both directions.
## Testing
698 UOContent tests, 810 Server tests, Release build clean.
Covered: verify and rehash outcomes, phrase rules for SHA1/SHA2 vs Argon2, stored-format stability for MD5/SHA1/SHA2, jobs with no connection attached, and dispatch ordering through the real queue. The liveness and ordering guards are mutation-verified.
## What
- Bind the login auth id to the account **and** origin address that earned it, make it a CSPRNG draw, expire it after two minutes, and spend it only once its owner presents it.
- Skip the password verify on `GameLogin` (0x91) when the presented id vouches for the submitted username and address.
## Why
A full client login hashes the password twice — `AccountLogin` (0x80) and then `GameLogin` (0x91). At the current Argon2 parameters that is **most of a 16 ms frame each, on the single-threaded game loop**, for every login attempt.
The second verify is redundant. `GameLogin` already requires an id from `_authIDWindow`, and that window is only populated by `GenerateAuthID`, called from `PlayServer` — reachable only after 0x80 has already authenticated the account **in this same process**. ModernUO Gateway has its own auth-id passing mechanism and is out of scope here.
## Why the id needed hardening first
Skipping the verify promotes the id from a correlation token to a bearer token, and it was not one:
- drawn from `Utility.Random` → `BuiltInRng`, a non-cryptographic PRNG
- bound to nothing — `AuthIDPersistence` carried only `Age` and `Version`
- never expiring; `Age` was only read to pick an eviction victim
A guessed id got you nothing while the password was still checked. Without that check it would have been an account takeover, so the id is now a CSPRNG draw, single-use, two-minute TTL, and bound to both the account and the origin address.
What remains is observing a live id on the client's network or machine — which the server cannot defend against under any design, and which already yields the password itself, since the client transmits it in the same handshake.
Network switching mid-login is deliberately unsupported.
## Behaviour
A full verify was always required before this change, and ids never expired, so every "before" is a password check.
| Case | Before | After |
|---|---|---|
| Id absent | Disconnect | Disconnect |
| Address mismatch | Verify | **Disconnect** |
| Account mismatch | Verify | **Disconnect** |
| Expired | Verify | **Verify** |
| Id vouches | Verify | **Skip** |
No case grants access the previous code would have denied. Expiry deliberately falls back to the verify rather than disconnecting — a player can idle, and turning that into a lockout would be a regression for no gain.
## Look, then take
An id is not consumed until the presenter has shown it is theirs. Removing it first would let anyone who lands on a live id burn it, and its owner would arrive to `"Unable to find auth id."` and have to log in again over a packet they had no part in.
The **address is compared before the account**, so a guesser from anywhere else is rejected before a username is ever looked at. That is what makes it safe to leave the id in place on a mismatch: there is no username-enumeration risk to trade against, and the only presenter who could enumerate is already on the victim's own address.
## The window is not a cap
It was 128 entries with the oldest evicted to make room. That is a cap on *concurrent logins*, not a resource bound: 800 people picking a server at once would have live ids discarded and those clients would arrive to `"Unable to find auth id."` — a failed login caused by nothing except other people logging in.
Issuing now sweeps expired entries and lets the window grow if everything in it is still live. Unbounded is safe here: an entry costs a **successful** password verify to create and dies after two minutes, so its size tracks logins genuinely in flight.
Removing an id when its connection drops is not an option, and this was checked rather than assumed — `NetState.cs:787` disconnects the login connection *deliberately*, immediately after the id is issued, and that disconnect is never cancelled. Surviving it is the whole purpose of the id. Expiry is the only correct reclamation.
## Handshake hardening
Choosing a server queues a disconnect, but the queue drains on the *next* slice, so a client pipelining into the same recv buffer can reach the handshake handlers again. Two had no do-once guard:
- `LoginServerSeed` (0xEF) now rejects when `state.Seeded` is already set.
- `PlayServer` (0xA0) now rejects when `state.AuthId != 0` — otherwise a connection that had already spent its id would be handed the spent one back.
Issuing is also idempotent (`EnsureAuthId`), so a connection holds exactly one id by construction and an orphan is impossible rather than something to clean up. The login state machine itself is untouched.
Also fixes a fall-through: the "Unable to find auth id" branch disconnected without returning, then continued with a default entry and nulled `state.Version`.
## Testing
`ConsumeAuthId` is a seam with no `NetState` dependency, so the auth decision is tested directly: vouching, account mismatch, address mismatch, case-insensitive usernames, IPv4-mapped-IPv6, unknown ids, single-use by the owner, **a rejected attempt leaving the id redeemable**, expiry-into-verify, and an 800-id login rush that must evict nobody. Expiry is driven by moving `Core._now`, not by waiting. Every new clause was verified to discriminate by removing it and confirming only its own tests fail.
## Cost
Halves the per-login game-loop cost. This does not make hashing cheaper or move it off the loop — that is gated on a measurement described in `docs/handoffs/2026-08-07-off-loop-argon2-hashing.md`.
> ⚠️ **Rollback hazard — one-way door once logins are taken.** Serialization is unchanged, so a save
> written by this build still *loads* on the previous one. Its contents do not survive the trip: on
> its first successful login each account is rehashed to `$argon2id$`, and the previous build ships
> Argon2.Bindings 1.19.0, whose `Verify` is gated by the verifier's own configured type and answers
> `false` for an `$argon2id$` hash. **After a shard running this build has accepted logins, do not
> roll back past this commit** — every account that logged in is locked out on the older binary, and
> the only recovery is rolling forward again or resetting passwords by hand. Roll back only from a
> save taken before the first post-deploy login.
Requires [Argon2.Bindings 1.20.0](https://github.com/modernuo/Argon2.Bindings/pull/14), now published.
## What
- Consume `Argon2.Bindings` 1.20.0, which resolves the Argon2 type from the stored PHC string rather than from the verifier's own configuration.
- Default to **Argon2id, m=16384, t=1, p=1** — 8.51 ms against the old Argon2i 8 MiB t=3 at 10.11 ms. Cheaper *and* stronger.
- Rehash on a successful login whenever the stored parameters are stale, not only when the algorithm changes.
- Fix `SetPassword`, which derived the password phrase from the outgoing algorithm while storing it under the incoming one.
## Why
**Verification was gated by the verifier's configured type.** `Verify` passed the instance's own `ArgonType` to native `argon2_verify`, whose `decode_string` rejects a disagreeing `$argon2i$`/`$argon2id$` prefix and returns `DECODING_FAIL` — folded into `false`, the same answer as a wrong password. Switching the default type would have locked out every existing account, and `VerifyAndUpdate` could not have migrated them either: it delegates to the same type-fixed `Verify` and never compared `ArgonType`. Fixed upstream in 1.20.0. The pinned legacy-`$argon2i$` test here fails on 1.19.0 for exactly that reason, which is what makes the package bump load-bearing rather than incidental.
**Changing the defaults would otherwise have reached nobody.** Argon2's PHC string embeds `m`, `t` and `p`, so verification uses the parameters stored with each account, not the configured ones — and verification is the hot path. `CheckPassword` only rehashed when the *algorithm* changed, never when its cost parameters did, so on an established shard the new defaults would have applied to new accounts only. `IPasswordProtection.NeedsRehash` closes that: it defaults to `false`, so PBKDF2 and the `HashAlgorithm` protections are untouched — only Argon2 carries its cost inside the stored value.
**`SetPassword` picked the phrase rule from the wrong algorithm.** SHA1 and SHA2 salt the phrase with the username; Argon2 and PBKDF2 do not. It chose the rule from the *outgoing* algorithm while storing under the *incoming* one, so any algorithm change wrote a credential its own next verify could not reproduce. It now assigns `PasswordAlgorithm` first and derives the phrase from that. Note this ordering is load-bearing and invisible — `UpgradingAlgorithm_DoesNotLockTheAccountOut` is what pins it.
## Cost
Verification is re-derivation, so these are login numbers. A full login calls `CheckPassword` twice — `AccountLogin` (0x80) then `GameLogin` (0x91): **~20 ms before, ~17 ms after**, plus a one-time ~8.5 ms rehash on each account's migrating login.
That cost is still paid on the game loop. Moving hashing off-loop is deliberately **not** in this PR — it needs a pending-auth state in the login handlers, bounding of in-flight hashes, and login rate limiting.
## Summary
Players could not place **any door** while customizing a house, and placing other pieces could disconnect them outright. Staff saw neither problem: `HouseFoundation.Designer_Build` only enforces `ValidPiece` below `GameMaster`.
Original report and diagnosis by @SynPDX.
## Root cause 1 — no door is ever registered
The retail client's `doors.txt` separates its header rows with lines of **bare tabs** (it is the only sheet that does):
```
int<TAB>int<TAB>...<TAB>string
<TAB><TAB><TAB><TAB><TAB><TAB><TAB><TAB><TAB><TAB> <-- 10 tabs, not an empty line
Category<TAB>Piece1<TAB>...<TAB>FeatureMask<TAB>Comment
<TAB><TAB><TAB><TAB><TAB><TAB><TAB><TAB><TAB><TAB>
0<TAB>1657<TAB>1659<TAB>...
```
`Spreadsheet.ReadLine` skipped a line only when `line.Length > 0`. A 10-tab line has length 10, so it was returned as the **names row** — every column ended up named `""`, `GetColumnID("Piece1")` and friends returned `-1`, and not one of the 230 door graphics was registered. Unregistered item IDs keep the `-1` sentinel, and `CheckValidity` rejects those, so `ValidPiece` refused every door.
ClassicUO skips these lines (`string.IsNullOrWhiteSpace` in `HouseCustomizationManager.ParseFile`), which is why the client happily offers doors the server then rejects.
Measured against a retail 7.0.x `doors.txt` using the shipped `Spreadsheet`:
| | `FeatureMask` column | door graphics registered |
|---|---|---|
| before | `-1` | **0** |
| after | `9` | **230** |
## Root cause 2 — `IndexOutOfRangeException` out of the packet handler
Every sheet ends in a cosmetic `Comment` column that ModernUO never reads, and client sheets write an empty comment as a plain newline with no trailing tab. `Split('\t')` then returns one field fewer than the header declares, and the parser indexed past the end:
```
System.IndexOutOfRangeException: Index was outside the bounds of the array.
at Server.Multis.Spreadsheet..ctor(String path)
at Server.Multis.ComponentVerification.LoadSpreadsheet(...)
at Server.Multis.ComponentVerification.IsItemValid(Int32 itemID)
at Server.Multis.HouseFoundation.ValidPiece(Int32 itemID, Boolean roof)
at Server.Multis.HouseFoundation.Designer_Build(NetState state, ...)
```
The client's own parser only requires the columns up to `FeatureMask` — ClassicUO's `CustomHouseMisc.Parse` guards on `scanf.Length >= 12` for a 13-column `misc.txt` — so such a row is valid data listing real pieces. Missing trailing fields are now treated as empty rather than dropping the row, which would unregister every piece the row lists and reproduce the door symptom.
`EnsureLoaded` also set `_loaded` before loading, so once the throw escaped, an all `-1` table stayed cached and rejected everything for players from then on — the same player-visible symptom as #2500.
## Also made explicit rather than accidental
- **Named the table sentinels.** `NotAComponent` (-1) is the anti-cheat guard and the initial state; `NoFeatureRequired` (0) is a piece with no expansion gate — how `walls.txt` encodes pre-AOS base pieces and what `housing.bin` collapses to under `HousingTierMask` (#2500).
- **A sheet with no `FeatureMask` column is refused and logged.** `GetInt32` on a missing column returns 0 = `NoFeatureRequired`, which would have silently marked every piece in that sheet unconditionally placeable regardless of expansion. This was previously only harmless by accident.
- **A sheet matching none of its expected tile columns is refused and logged** — that is what `doors.txt` was doing silently. Individual missing columns stay tolerated, since older sheets predate columns such as `walls.txt`'s `SecondAltWindowS`/`E`.
- **Catch per sheet**, so one unreadable file no longer costs the other six.
- **Header guards**: an empty file or a types-only file raised a `NullReferenceException`; a names row shorter than the types row indexed past the end.
- **Fall back to the component sheets when `housing.bin` cannot be read**, instead of passing `null` into a `SpanReader`.
`_loaded` is still set before loading, deliberately: this runs from the design packet handler, and retrying would re-read every sheet on each subsequent placement attempt.
Sheet precedence is **unchanged** — the client's copies stay authoritative and `Data/Components` remains the fallback.
## Verification
- Retail 7.0.x client `doors.txt` through the shipped `Spreadsheet`: 0 door graphics before, 230 after.
- 5 new tests in `SpreadsheetTests` covering the tab separators, the omitted trailing field, per-row recovery, and both header guards. All 5 fail against `main` and pass here.
- `dotnet build` clean (0 warnings, 0 errors); `UOContent.Tests` 642/642.
## Why
The shard owner, on a Starlink CGNAT address, was blocked by the imported reputation blocklist.
The cause was not CrowdSec. The address was a literal line in `ip-blocklist.txt`, so `BlocklistFilter` denied it at accept and then promoted it — and clearing the CrowdSec decision could not fix it either, because the file entry re-reports within `promoteSuppression` of every reconnect attempt.
This is structural, not a one-off. Reputation feeds list shared consumer address space constantly: on CGNAT one public address fronts many subscribers **at the same time**, so a single abusive customer gets the address listed and everyone else behind it is blocked with them. Where leases rotate, a listing says little about whoever holds the address now. Around 1,000 Starlink addresses sit in the current list.
So exemptions go where they cost nothing, and escalation is driven by what a connection actually does.
## Generator — `tools/Export-IpBlocklist.ps1`
`-AllowlistFile` takes multiple paths, subtracted from the merged set before the output is written. Defaults to every `ip-allowlist*.txt` beside the output, merged into one allow set:
- `ip-allowlist.txt` — operator exemptions, created once and **never rewritten**
- `ip-allowlist-<name>.txt` — a carve-out you built, regenerable and copyable between shards
**Subtraction is range-correct.** An allowlisted address inside a blocked CIDR splits that CIDR around the hole rather than being silently ignored. This also fixes `-ExcludeAnonymizers`, which parsed CIDR entries into `$anonCidr` and then only ever subtracted singles.
**No carve-out ships.** A carve-out names a real network, and which ones a shard should exempt depends on where its players actually are — so publishing one would make that policy call for every shard and put a specific provider's address space in the repo. The script builds them on request instead:
```powershell
.\Export-IpBlocklist.ps1 -AddCarveout starlink -Asn 14593
```
Carve-outs are **discovered, not configured**: every `ip-allowlist*.txt` beside the output is subtracted, by the generator and by the shard, so a file an admin adds needs no config edit and no code change. Each carries an `asn=` marker in its header, which is how `-RefreshCarveouts` rebuilds it without the script keeping a list of anyone's networks; a hand-written allowlist has no marker and is never rewritten.
Prefixes come from **announcements, not ownership records**, because registry data disagrees with what is actually routed and silently caps result sets: ARIN whois returns at most 256 rows and gives per-customer /24s, and `206.83.96.0/19` reads as APNIC in RDAP even though `206.83.96/21` is announced by Starlink.
Editing an allowlist bypasses `-MinInterval`, so a just-added exemption isn't indistinguishable from the allowlist not working. A Starlink carve-out, if you build one, costs **~4,300 IPs + ~144 CIDRs of 4.2M (0.10%)**.
## Allowlists
**`FileAllowlist`** reads the same files the generator subtracts, so an operator entry means "leave this address alone" for real. Subtraction alone only covers being *blocked*; behavioural detections never consult the blocklist, so without this a carve-out was quietly routed around — one scanner behind a shared address was enough to get everyone behind it contributed and firewalled, with nothing in the shard's own config explaining why. Reading the files also means an entry applies on the next reload rather than the next regeneration, which is what matters when someone is complaining now.
**`LoginAllowlist`** is earned by authenticating, with a 90-day TTL because an address that logged in years ago is a stranger. Its own store rather than `Account.LoginIPs`, which has no timestamps and cannot be backfilled. An entry is evidence rather than a licence: 10 suppressed contributions in an hour revokes it, and a fresh login forgives the tally.
Both are consulted **only after the blocklist has already matched**, so a normal accept pays nothing for them and the accept gate stays allowlist-free. `BanExemptions` combines them behind `BanChannel.IsExempt` and suppresses escalation only — every local defence still applies.
Two limits, both deliberate and documented in the class: `LoginAllowlist` **cannot bootstrap** (an entry is only earned by getting in, so it never repairs an existing false positive), and it is weakest on rotating CGNAT. That is why `FileAllowlist` is the fix for those, and why it is manual.
## Behavioural detection
| Reason | Trigger |
|---|---|
| `silent-connect` | Reaped after 5s having sent **zero bytes** |
| `invalid-seed` | Opened with a zero seed |
| `foreign-protocol` | Positively identified as HTTP, TLS or SSH |
**`ForeignProtocol` inverts the test.** Asking "is this a good UO client?" cannot work: `LoginEncryption.ClientDecrypt` is a byte-for-byte stream XOR, so a legitimate client with encryption enabled when the shard expects none sends a structurally perfect connection whose payload is noise. "Speaks HTTP" is safe where "unreadable" is not — however misconfigured a UO client is, it never sends `GET / HTTP/1.1`.
Nothing assumes arrival framing. TCP has no message boundaries, so a rule of the form "these bytes must arrive together" is broken by construction and drops real players on poor links. A prefix match with too few bytes to confirm waits for more. A four-byte seed can legitimately spell `GET ` (the address 71.69.84.32) or `0x16 0x03 0x0?` (22.3.x.x), so confirmation requires the request line to continue in printable ASCII or an actual ClientHello inside a plausible record — a real client's fifth byte is a packet id (`0x80`, `0x91`, `0xEF`), none of them printable, so those collisions fall through.
Everything is keyed on **bytes-received rather than elapsed time**. A connection that sent something and ran out of time is far more likely a slow link than an attack, and banning those produces the worst failure mode available: the player retries, trips the rate limiter, and compounds a bad connection into hours of being firewalled off.
## `AutoDenylist`
A short-lived local hold (15m) on behavioural detections, as `IConnectionFilter` + `IBanReporter` over one store so the engine detection sites never reach into content.
This closes the gap where a flood pays for a socket, buffer and `NetState` slot per connection while waiting for the OS bouncer — the verdicts that matter most are reachable only *after* reading bytes — and it is the entire defence on a shard running no bouncer, which is the default config. Not persisted: a holding pen that survives restarts is a ban without a ban's review.
Cost: one dictionary lookup on a usually-empty dict per accept.
## `BanReasons`
Centralises the reason slugs. `IsBehavioral` is an **opt-in** set, not "everything except manual", so a future reason escalates normally instead of silently inheriting an exemption or entering a local denylist.
This caught a real bug during review: the first cut of the exemption swallowed `manual` admin bans (`Commands.cs`, three sites in `AdminGump`) for any allowlisted address.
## Fixes found in review
- **`BanConfiguration.Settings` was null until `Configure()` ran**, while the reap path dereferences it every `Slice()`. A harness driving `NetState.Slice()` directly hit an NRE that presented as flaky because it depended on whether an earlier test had already called `Configure()` — which is why it failed on some CI platforms and not others. Now starts at the record's defaults, with idempotency tracked by a flag; this also removes the same latent NRE from the pre-existing rate-limit path.
- **`-AllowlistFile` was typed `[string]`** while documented and used as a list, so passing two paths would have collapsed them into one string.
## Layout and docs
Content network code moves out of `Misc/` into `UOContent/Network/`, one concern per folder — `AutoDenylist/`, `Blocklist/`, `CrowdSec/`, `Firewall/`, `LoginAllowlist/`, `Packets/`. **Namespaces are untouched**, so these are pure file moves (git tracks all 16 as renames).
`dev-docs/ip-bans-and-allowlists.md` documents the subsystem, leading with the operator process for unblocking a player — including the three things that look sufficient and are not: deleting the CrowdSec decision alone, editing `ip-blocklist.txt` by hand, and `cscli allowlists` alone. `.gitignore` covers the new config files.
## Testing
Build clean. **Server.Tests 810 passed**, **UOContent.Tests 637 passed**, zero warnings. This branch adds 38 tests; the rest of the delta is main's, since this is rebased on current `main`.
New coverage: TTL boundary and renewal, private-address exclusion, manual-ban-never-exempt, unopted-reason-never-exempt, strike revocation, quiet-window reset, login forgiveness, file-allowlist CIDR coverage, file-allowlist not spending the earned list's strikes, denylist expiry-on-read, cap enforcement, lapsed-entry reclaim, HTTP/TLS/SSH identification, seed-collision fall-through, and encrypted-login-is-not-foreign.
Generator verified end-to-end against live feeds: a clean run ships no carve-out, `-AddCarveout starlink -Asn 14593` fetches and collapses 213 prefixes to 115 ranges in 0.1s over 4.2M entries, `-RefreshCarveouts` rediscovers it by its `asn=` marker, a hand-written allowlist is left untouched, and deleting a carve-out drops it rather than having it rewritten. CIDR splitting verified exhaustively: a single-IP hole in a /24 leaves exactly 255 of 256 addresses blocked.
## Operator note
Existing installs are unaffected until the generator next runs, which creates `ip-allowlist.txt` and nothing else. To unblock someone: add the address to that file and delete any live CrowdSec decision — the existing ban outlives the config change. The shard picks the entry up on its next reload, so re-running the generator is optional.
A shard whose players are on CGNAT (satellite, mobile, or an ISP short on IPv4) will likely also want `-AddCarveout`; see `dev-docs/ip-bans-and-allowlists.md`.
## Also included: a latent CI failure this PR surfaced
`fix(tests): serialize test classes that rent through STArrayPool` touches a property-list test file that has nothing to do with this feature. It is here because it was failing macOS CI, and it is trivially cherry-pickable out if you would rather it went to `main` on its own — **which may be the better call, since it is failing `main` today.**
CI has since gone green with it applied.
`STArrayPool` is single-threaded by design and its bucket cache is a plain `static`, not `[ThreadStatic]`, with a check-then-act initialize in `Return()`:
```csharp
var cacheBuckets = _cacheBuckets ?? InitializeBuckets();
```
Two threads both see null, both initialize, and the loser trips `Debug.Assert(_cacheBuckets is null)`. Anything renting from it has to stay off parallel test threads — which is what the `DisableParallelization` collections are for.
- `ObjectPropertyListReentrancyTests` and `ObjectPropertyListNestedBuildTests` (added in #2555) build property lists, which rent the interpolation buffer, but were not in the sequential collection — unlike `PropertyListInvalidationDuringBuildTests` in the same file. This is a **latent failure already on `main`**; it is timing-dependent, so it shows on some platforms and not others.
- `AutoDenylistTests` (added here) has the same exposure: its cap tests reach `AutoDenylist.Sweep`, which rents a `PooledRefList` without `mt`. The blocklist tests need no marking because `BlocklistSnapshot.Build` asks for the `mt` pool explicitly.
No production change — `STArrayPool` is the right pool on the game loop, where both `Sweep` and the property list actually run.
## Deliberately not included
Waiting for a fragmented four-byte seed at `AwaitingSeed`. It looked like a bug but the disconnect is a deliberate defence: only pre-0xEF clients reach it (0xEF goes through `HandlePacket`, which already waits for its 21 bytes), and waiting converts an instant drop into a full 5s slot hold for a client sending one or two bytes, or a loris dribbling a byte every few seconds. Against a fixed 4096-entry `MaxConnections` table that trades capacity that matters for a fragmentation case a reconnect already fixes.
## The bug
Any property getter reached from `GetProperties` that calls `InvalidateProperties` takes the tooltip build down with it:
```
System.ArgumentNullException: Value cannot be null. (Parameter 'array')
at Server.ObjectPropertyList.AppendStringDirect(String value)
at Server.Mobiles.PlayerMobile.GetProperties(IPropertyList list)
```
`InvalidateProperties` rebuilds **in place** — `Reset()`, then `GetProperties()` again on the same instance — and `Reset()` does two destructive things to a build already in flight:
1. **It returns the pooled interpolation buffer.** The compiler rents it in the handler ctor and returns it in the closing `Add`, so *every hole is evaluated while it is live*:
```csharp
var handler = new InterpolatedStringHandler(1, 2, list); // InitializeInterpolation() RENTS
handler.AppendFormatted(pl.Rank.Title); // <-- getter runs HERE
handler.AppendLiteral("\t");
handler.AppendFormatted(faction.Definition.PropName);
list.Add(1060776, ref handler); // consumes span, RETURNS
```
```
GetProperties(list)
├─ InitializeInterpolation() -> _arrayToReturnToPool = Rent(256) buffer LIVE
├─ « hole 1: pl.Rank.Title »
│ └─ PlayerState.Rank.get (lazy recompute)
│ └─ Invalidate() -> InvalidateProperties() -> m_PropertyList.Reset()
│ └─ Dispose(): Return(buf); _arrayToReturnToPool = null buffer GONE
└─ handler.AppendFormatted("Knight")
└─ _arrayToReturnToPool.AsSpan(_pos..)
└─ ArgumentNullException (Parameter 'array')
```
It surfaces as `ArgumentNullException` rather than `NullReferenceException` because the `Range` overload of `AsSpan` must read `array.Length`, so the BCL null-checks and names the parameter `array`.
2. **It rewinds the packet cursor**, so properties already written are overwritten by the nested pass — a silently corrupted tooltip even where the buffer survives.
## The fix: refuse, don't recover
There is no correct recovery, and retrying the build would only hide the defect. A nested invalidation now logs an error with a stack trace, **throws in `DEBUG`** so it gets found and fixed, and in `RELEASE` returns without touching the list — a possibly stale tooltip, but no crash, no corrupted packet, and nothing leaked back to the pool. Getters that genuinely must invalidate should defer:
```csharp
Timer.DelayCall(InvalidateProperties);
```
The guard flag lives on the `ObjectPropertyList`, not the entity: it is that list's own lifecycle, it costs nothing (both `Item` and `ObjectPropertyList` absorb it in existing padding, and the list is allocated lazily), and it stays correct when builds for different entities nest.
Base instance sizes are unchanged from `main`: Item 128 B, Mobile 792 B, ObjectPropertyList 72 B, PlayerMobile 1216 B.
`PropertyList` also publishes the list into `m_PropertyList` **before** building it rather than assigning through `??=` afterwards, so a nested `InvalidateProperties` sees the build in progress instead of recursing into a second throwaway list whose work is discarded.
`ObjectPropertyList` re-rents its scratch buffer instead of spanning a null array, so a stray `Reset()` from any other caller degrades rather than aborting `GetProperties`.
## Factions `PlayerState`: maintained, not lazily computed
The getter that surfaced this is now a plain field read — the whole `if (m_InvalidateRank)` block and the flag itself are gone:
```csharp
public RankDefinition Rank => m_Rank;
```
`UpdateRank()` recomputes at each point an input actually changes:
| Site | Why |
|---|---|
| `RankIndex` setter | this player's index changed |
| end of `KillPoints` setter | two paths write `m_RankIndex` directly, bypassing the setter; runs once the swap bookkeeping and `ZeroRankOffset` have settled |
| `Faction.AddMember` | *after* the insert — the member count is not settled during the ctor |
| `FactionState` load | once ordering and `ZeroRankOffset` are final |
Supporting fixes this forced out:
- **Both ctors seed the lowest rank.** Nothing recomputes on read any more, so `Rank` has to be usable immediately — including for members that never get a `RankIndex` assigned, which is *every member with no kill points*. Without this, `Rank.Title` NREs.
- **`Rank` always resolves.** Ranks are ordered by `Required` descending ending at `0`, so a *negative* percent (`RankIndex` out of sync with `ZeroRankOffset`) matched nothing and left `m_Rank` null. It no longer divides by a zero `ZeroRankOffset` either.
- **A pre-existing staleness bug.** The `KillPoints` setter writes `m_RankIndex` directly in two places, so the cached rank was never refreshed when a player crossed zero kill points.
All six readers of `Rank` were checked; none relied on the old side effect.
One behaviour change worth flagging: rank refreshes are now **eager** where they used to be lazy, so a `KillPoints` change invalidates each swapped player as it happens. The swap loops break as soon as ordering is satisfied — typically 0–2 swaps — but it is on the path that runs on every faction kill.
## Documentation
The rule is written down so it is enforceable rather than folklore:
- **CLAUDE.md** audit rule 19
- **`dev-docs/property-lists.md`** — new "Never Invalidate From Inside `GetProperties`" section with the failing/passing pattern
- **`dev-docs/claude-skills/modernuo-property-lists.md`** — key rule + anti-pattern
- **`dev-docs/claude-skills/modernuo-code-audit.md`** — rule 19, ERROR severity
## Tests
- `ObjectPropertyListReentrancyTests` — `Reset()` and `Dispose()` re-entered mid-hole (both red against `main` with the exact exception above), nesting behaviour, and the new contract: `DEBUG` throws, `RELEASE` survives, and the build is never retried into a loop.
- `FactionRankTests` — `Rank` is populated before anything reads it, tracks `RankIndex` without a read, is stable across reads, and still resolves when `RankIndex` is out of sync with `ZeroRankOffset`. Red-verified: removing the ctor seed fails the first one.
793/793 `Server.Tests` and 608/608 `UOContent.Tests` pass.
## Noted, not addressed here
`~ObjectPropertyList()` returns the rented array to `STArrayPool<char>.Shared` from the **finalizer thread**, and that pool is single-threaded by design. Left alone as a separate concern.
## Problem
Contributing a ban to CrowdSec failed against a real LAPI — `POST /v1/alerts` answered **500**, and depending on the shard's locale, auth answered **401**. Three independent defects, each sufficient on its own.
## Fixes
**`scenario_hash` / `scenario_version` were never serialized.** LAPI dereferences both unconditionally when persisting an alert, so omitting them is a nil deref and a 500 rather than a validation error. Both are now emitted with the values a watcher without a hub scenario is expected to send (`""` and `"1.0"`).
**`start_at`/`stop_at` were formatted without an `IFormatProvider`.** `:` is the time separator *specifier* in a custom .NET format string, not a literal — a shard running under a culture like `fi-FI` emitted `T15.04.05.123Z`, which Go's `time.RFC3339` rejects, producing another 500. Non-Gregorian cultures (`th-TH`, `ar-SA`) would also shift the year. Formatting is now pinned to `InvariantCulture` in `FormatTimestamp`, which additionally converts non-UTC input — the trailing `Z` is a literal and was previously an unchecked claim.
**The `User-Agent` was a plain product string.** LAPI's default watcher profile matches the `crowdsec/` prefix and answers 401 without it, so the header is a protocol constraint, not cosmetic. It is now an `internal const` carrying that reason.
Also fixed, same root cause as the timestamp bug: the login-expiry parse used a bare `DateTime.TryParse` on LAPI's RFC3339 `expire`. Under a mismatched culture that silently fails and falls back to a fabricated `UtcNow + 1h`, pushing re-auth past the real expiry and costing a 401-relogin round trip on every send.
`capacity` now defaults to `1` instead of `0`, matching the one-decision-per-alert shape actually being sent.
## Note on scope
The two 500 causes are independent. On an `en-US` shard only the missing scenario fields were biting; the date bug was latent and would have surfaced as an unexplained regression the first time someone ran a shard under a European locale.
## Verification
The emitted payload is field-for-field identical to a hand-verified request that a live LAPI accepts:
```json
[
{
"scenario": "modernuo/rate-limit",
"scenario_hash": "",
"scenario_version": "1.0",
"message": "ModernUO rate-limit ban for 192.0.2.123",
"events_count": 1,
"start_at": "2026-07-27T15:04:05.123Z",
"stop_at": "2026-07-27T15:04:05.123Z",
"capacity": 1,
"leakspeed": "0s",
"simulated": false,
"events": [],
"remediation": true,
"source": { "scope": "Ip", "value": "192.0.2.123" },
"decisions": [
{
"origin": "modernuo",
"type": "ban",
"scope": "Ip",
"value": "192.0.2.123",
"duration": "300s",
"scenario": "modernuo/rate-limit"
}
]
}
]
```
Regression tests assert the required scenario fields on the **serialized JSON** rather than the DTO — the DTO is not what goes on the wire — and cover the timestamp as a `[Theory]` across `fi-FI`/`th-TH`/`ar-SA`.
`dotnet test --filter "FullyQualifiedName~CrowdSec"` → **21/21 passed**, build clean with 0 warnings.
Reshapes IP banning around one idea: **core owns the question, content owns every answer.**
Core gains a single accept-path seam — `IConnectionFilter` — and loses everything that used to implement one. The firewall moves to UOContent, a new file-backed blocklist joins it there, and CrowdSec is repositioned from an in-app enforcer to a contribute-first reporter.
## The seam
```csharp
public interface IConnectionFilter
{
string Name { get; }
void Configure();
void Start(CancellationToken token);
void Stop();
bool ShouldDeny(IPAddress address);
}
```
The accept path went from hardcoded branches to one question:
```csharp
else if (ConnectionFilters.ShouldDeny(remoteIP, out var deniedBy))
{
logger.Debug("{Address} denied by connection filter '{Filter}'", remoteIP, deniedBy);
}
```
Filters register during the Configure sweep. The registry is a plain array walked by an indexed loop — no enumerator, no closure, no allocation — and the first denial short-circuits. An interface dispatch is noise next to the `accept()` syscall, so pluggability costs nothing measurable on the path that has to survive a DDoS.
Whatever a hit implies — persisting, promoting to an OS bouncer, contributing to the ban channel — is the filter's business, not the accept path's.
A filter that throws is **unregistered and the connection fails open**. A filter that faults once faults for every subsequent connection, so leaving it registered means an exception and a log line per accept — exactly the amplification an attacker wants — and a broken filter must not be able to deny everyone either.
This deliberately does **not** reuse `EventSink.InvokeSocketConnect`: that fires later and allocates a `SocketConnectEventArgs` per connection, which is what the accept path avoids for rejected traffic.
## What ships behind it
**`firewall`** (UOContent) — the existing admin-curated set. Collapsed from `Firewall` + `AdminFirewall` + a threaded enforcer into one single-threaded store with **zero concurrency primitives**: the accept path, admin gump, TTL expiry and boot load all run on the game loop. Persists to `Configuration/firewall.json` with automatic migration from the legacy `firewall.cfg`. No behavior change for operators — same namespace, same gump, same commands.
**`blocklist`** (UOContent) — new. Holds a millions-strong list in-app and **demand-pages** hits up to CrowdSec, which promotes them to the OS firewall.
The motivation is concrete: CrowdSec's Windows bouncer cannot load the ~3.9M IPs that 91 community feeds produce, but it handles ~100k fine. So the millions live in-process behind a binary search, and only addresses that *actually connect* get promoted. A `PromotedGuard` suppresses re-reporting an address until the bouncer picks it up.
The list is parsed straight from UTF-8 file bytes with no per-line string allocation, off the game loop, and published as an immutable snapshot swapped through a single `volatile` reference. Reloads yield to world saves.
**`tools/Export-IpBlocklist.ps1`** — the producer. Requires PowerShell 7 and runs on Windows, Linux and macOS; Windows PowerShell 5.1 is refused up front via `#requires`. Merges a thin, non-overlapping feed set into one de-duplicated, bogon-filtered file. Parsing runs in a compiled `Add-Type` hot loop (~1s for ~4M lines instead of minutes). Written to a `.tmp` sibling and swapped with `File.Replace`, so the shard never reads a half-written list, and a total feed outage refuses to overwrite a good list with an empty one. Re-running is idempotent — it exits without downloading anything while the list on disk is younger than `-MinInterval` (default 2h, the anchor feed's own refresh period), so a misconfigured scheduler can't hammer upstream.
## CrowdSec: contribute-first
`IBanReporter` + `BanChannel` fan locally-decided bans out to external systems. `CrowdSecReporter` (UOContent) posts to LAPI `POST /v1/alerts` and retracts via `DELETE /v1/decisions`.
Reporting is **enqueue-only** on the accept path: a bounded, coalescing channel drained off-loop with bounded retry, counted drops on overflow, and a flush on shutdown. Under a DDoS the accept path never does synchronous or lock-contending per-IP work.
### Why not pull decisions from CrowdSec?
The original design streamed decisions into an in-app snapshot and enforced them at the accept gate. That's the wrong layer: by the time the shard sees the connection, the TCP handshake and socket setup are already paid for. `cs-firewall-bouncer` drops the same traffic **at the kernel**, and it's what CrowdSec is built to do. So the shard now contributes what it uniquely knows (rate-limit trips, blocklist hits from real connection attempts) and lets the OS enforce.
The one thing the OS can't do — hold millions of entries on Windows — is exactly what the in-app blocklist covers, and it feeds the same pipeline.
## Threading policy
`CLAUDE.md` rule #3 is rewritten as an explicit three-part policy, with rule #10 restated in tandem:
- Anything touching game state runs **only** on the main loop.
- Heavy work that *needs* game state must be **chunked** across ticks, never threaded.
- Heavy work that does *not* need game state (large-file parse, external I/O) **must** run off-loop **and must yield to world saves**.
Results come back via an immutable snapshot swapped through a single `volatile` reference, or `Core.LoopContext.Post` — never by letting the scheduler decide where heavy work runs. Both new subsystems follow it.
## Shared primitives
`SortedRangeIndex<T> where T : IBinaryInteger<T>` — coalesced disjoint interval arrays plus a binary search. The firewall, the blocklist, and (as of this PR) core's reserved-network tables all use it.
Coalescing is a correctness requirement, not an optimization: multi-feed lists nest CIDRs (`/24` containing a `/32`), and a search that inspects only the rightmost run whose minimum is ≤ the value is sound **only** over disjoint runs. That bug was caught in review and is covered by regression tests.
`IPAddressUtility` collects the allocation-free `IPAddress` ↔ `UInt128` conversions and CIDR parsing that were previously scattered or duplicated.
## Config
| File | Owner | Keys |
|---|---|---|
| `Configuration/bans.json` | core | `reportRateLimitTrips`, `autoBanDuration` |
| `Configuration/blocklist.json` | content | `file`, `reloadInterval`, `reportHits`, `banDuration`, `promoteSuppression` |
| `Configuration/crowdsec.json` | content | `lapiUrl`, `machineId`, `password`, `origin`, `manualBanDuration`, `flushInterval`, `maxQueue` |
| `Configuration/firewall.json` | content | persisted firewall entries (migrated from `firewall.cfg`) |
Everything is inert by default. CrowdSec self-disables without credentials; the blocklist self-disables until its file exists. A shard that changes nothing sees no behavior change.
## Notes for review
- **Core no longer references `Firewall` or `IFirewallEntry` anywhere.** `NetworkUtilities` used to build its reserved-network tables out of `CidrFirewallEntry`, which coupled core to the firewall for something unrelated to banning; those are now a `SortedRangeIndex<UInt128>`, same semantics and public API.
- **`BanChannel.Stop()` no longer persists the firewall** — a contribution coordinator has no business saving an enforcement store. That's the firewall filter's `Stop()`.
- **A dead `whitelisted` parameter was dropped** from the blocklist gate: it was hardcoded `false` at its only call site, and no whitelist concept exists in core.
- **The blocklist filter is an instance, not a static.** The static version forced its tests onto the sequential collection with a reset hook; they now run in parallel.
- `dev-docs/networking-packets.md` documents the seam for content authors, plus a known wart in the `IPAddress` ↔ `UInt128` normalization flagged for a follow-up PR.
- The generator was verified on Linux, macOS and Windows under a temporary CI matrix (since removed). It caught two portability bugs — a Windows-only path separator, and a culture-sensitive duration parse that read `2.5` as `25` on comma-decimal locales and *silently* turned a 2.5h cooldown into 25h — plus a third that made the script unparseable on Windows PowerShell 5.1. The source is ASCII-only for that last reason: `#requires` is only honored once a file parses, so non-ASCII in a BOM-less script produces parse errors instead of the version message.
## Tests
**1344 pass** (782 `Server.Tests`, 562 `UOContent.Tests`). New coverage: filter registry (registration, short-circuit, fault-disable), blocklist parsing/CIDR/coalescing, snapshot reload markers, promote-guard TTL, ban-channel fan-out, CrowdSec alert building/dedup/flush-on-stop, and the generator's output-format contract pinned against the reader.
## Summary
Hardens the **Advanced Search** engine (`Projects/UOContent/Engines/Advanced Search/`) — the GM entity finder that fans searches across background worker threads. A code review surfaced 14 defects (A–N), including a shard-crasher reachable from a single admin typo and a path that silently disables autosave for the rest of the shard's uptime. Each behavioral fix ships with a test.
Full `UOContent.Tests` suite: **530/530 green** (21 new AdvancedSearch tests).
## Fixes
### Crash / data-loss
- **A — Shard crash on a malformed Property Test.** `AdvancedSearchThreadWorker.Execute` had no `try/catch` and the worker `Thread` is foreground, so a parse throw (`Hits>abc`, `Layer=onehanded` — `Enum.Parse` was case-sensitive, `Hits>1@` — empty sub-expression indexing) terminated the process. Now: `ParseValue`/`CompareValues` use `TryParse`/`Enum.TryParse(ignoreCase)` and return no-match instead of throwing; the per-entity filter is wrapped in `try/catch` (logs + skips); empty expressions are guarded.
- **C — Overlapping searches corrupt state + brick autosave.** `_threadWorkers`/`_threadId` were `static` but `DoSearch` is an instance method; a second search (double-click / two admins) stomped shared worker state and could leave a drain waiting forever on the shared `AutoResetEvent`, so `AutoSave.SavesEnabled` was never restored. Now: an `Interlocked` re-entrancy guard rejects concurrent searches.
- **G — Autosave restore not guaranteed.** The restore lived only in the success callback. Now it's in a `finally` (plus an outer `catch` covering the synchronous setup and a `catch` on the drain body), so autosave + the guard are always released.
### Wrong results
- **D — `@`/`|` operator precedence.** `a@b|c` evaluated as `a && (b || c)` instead of `(a && b) || c`. OR now binds looser than AND (`AdvancedSearchUtilities.EvaluateBoolean`, unit-tested).
- **E — Descending sort, partial last page rendered blank** (the index decreased in descending mode and the `break` early-out killed the loop). Now a bounded `VisibleCount`-driven loop renders the last page in both directions.
- **F — Deleted entities** were not skipped (ghost rows). Now `DoEntitySearch` skips `entity.Deleted`.
- **N — Reference-type comparisons** threw (`Comparer<T>.Default.Compare` on non-`IComparable`) and compared references to a string. Now equality is by value and ordering is guarded to `IComparable` (no throw).
### Worker perf / hardening
- **H** busy-spin → `Thread.Yield()` in the drain; **I** `GetProperties()` cached per `Type`; **J** `HandleValidInternal` moved behind the cheap map/range/region filters; **K** worker threads are `IsBackground` + `Exit()` tolerates an already-terminated worker; **L** `_filter == null` guard; **M** consistent `Volatile` access on `_pause`/`_exit`.
### Documented
- **B** — the residual worker/event-loop read race is documented on `AdvancedSearchThreadWorker`: workers read live entity state concurrently with the loop, so value-type reads may be stale-but-safe and getter exceptions are swallowed; fully eliminating it would require snapshotting entity fields on the main thread (deferred).
## Notes
- New test-only seams (`TryBeginSearch`/`EndSearch`/`IsSearchInProgress`/`VisibleCount`/`TryParseValue`/`EvaluateBoolean`) are `internal` via the existing `InternalsVisibleTo("UOContent.Tests")`.
- Dead `public ParseValue<T>` removed.
- `ConcurrentDictionary` for the reflection cache is intentional — these workers are genuinely parallel.
## Problem
Items dropped on the ground never decay. Corpses do, which makes the breakage look selective — but corpses are unaffected only because `Corpse.BeginDecay` runs its own `InternalTimer` and never touches `DecayScheduler`. Ordinary items are the only things that depend on the scheduler.
## Root cause
`Item.MoveToWorld` called `SetLastMoved()` — which triggers `UpdateDecayRegistration()` — at the *top* of the method, before detaching the item from its parent and before assigning the new map. `CanDecay()` reads `Decays`, `Parent`, **and** `Map`, so registration was evaluated against the item's *pre-move* state.
Because the `Item` constructor sets `m_Map = Map.Internal`, and `Mobile.Lift` calls `item.Internalize()` to put an item on the cursor, registration was consistently one step behind:
| State | Tracked for decay? | |
|---|---|---|
| Item on the ground | **No** | never decays |
| Item held on the cursor | **Yes** | backwards |
Nothing corrected it afterwards: the later `m_Map = map` assigns the field directly, bypassing the `Map` property setter, and that setter does not refresh registration either. With no parent, `RemoveItem` (which *does* re-register) never runs.
World load masked this — `ItemPersistence.PostDeserialize` re-registers every item against its final state, so decay appears to work for items that survive a restart. Only freshly dropped items are affected.
**Fix:** stamp `LastMoved` up front so decay math stays correct, then call `UpdateDecayRegistration()` once the parent, map, and location are final.
## Audit of the rest of the call sites
All 16 `SetLastMoved()` call sites were reviewed. `SetLastMoved()` must keep refreshing registration — `LastMoved` feeds `ScheduledDecayTime` and therefore which bucket an item belongs in — but it may only run once parent/map are final. The vendor, house, lift and drop sites already satisfy that. The rest of this PR fixes the ones that did not, plus what the audit turned up:
- **`Item.Deserialize`** stamped via `SetLastMoved()` before the version data was read, registering against an unread `Map`/`Parent`. Safe only by accident (the `Item(Serial)` ctor leaves flags at 0, so `Decays` is false), and it cost an unregister per item per world load. Now stamps only; `PostDeserialize` does the registration.
- **`Item` constructor** registered then immediately unregistered every item — the `Movable` setter saw `m_Map` still null, so `CanDecay()` was true. Also removes a `Configure`-order landmine: constructing an `Item` before `DecayScheduler.Configure()` would have thrown in `Shared.Start()`.
- **`Container.Destroy`** stamped `LastMoved` immediately before `MoveToWorld`, which now stamps it itself.
- **`Unregister` was documented O(1)** but scanned twelve buckets and did a linear `PriorityQueue.Remove`, on every construction, deserialize and move. Items now record where they are tracked in `Item.DecaySlot` (1 byte), so untracked items — the common case — leave in O(1).
- **A refused decay silently dropped the item.** `ProcessActiveQueue` deleted on `OnDecay() == true` but did nothing when a region refused, leaving the item dequeued, untracked and on the ground forever. It now restarts the decay clock; re-registering as-is would spin, since `ScheduledDecayTime` is already past.
## Two content bugs of the same class
The decay system replaced a polling sweep. Under polling, a `Decays`/`DecayTime` override could read live state every pass. Under a registration model it cannot — the scheduler drops items that stop being eligible, but nothing enrols one that becomes eligible while untracked.
- **`TreasureChestLevel1-4`** overrode `DecayTime` as `Utility.Random(15, 60)` — a fresh roll on every read. `ScheduledDecayTime` is read repeatedly (to bucket, to re-bucket on rotation, to test whether due), so those reads disagreed: the chest re-bucketed every tick and decayed early instead of after its intended interval. Now rolled once per chest. Distribution unchanged — `Utility.Random(from, count)` is RunUO's `from + Next(count)`, so this is 15–74 minutes, as before.
- **`StrongBox`** overrode `Decays` with a live check on `_house`, `_owner.Deleted` and `IsCoOwner`. Nothing notifies the box when any of those change, so it was never enrolled and the override never decayed anything — and it could not have: `HouseRegion.OnDecay` refuses a secured item inside a standing house, and the box is in `Secures`. Decay was never the mechanism here.
- A strongbox is only ever its owner's. Without a house, or without an owner still co-owning that house, it would be a free container anyone could loot, so `Validate()` destroys it. The old check missed exactly those two cases — it required a non-null owner and treated a null house as valid. A deleted owner deserializes back as null, which `IsCoOwner` rejects. The now meaningless `Decays`/`DecayTime` overrides and an unhelpful `Console.WriteLine` are gone.
`DecayScheduler` now documents both constraints.
## Tests
`DecayRegistrationTests` (Server) covers world placement, lift/drop, container round-trip, cursor-held (must *not* track), `Container.Destroy` spill, `DecaySlot`/structure agreement, refused decay, and a full decay lifecycle driven through the scheduler. `TreasureChestDecayTests` (UOContent) locks `DecayTime`/`ScheduledDecayTime` stability across all four chest levels.
To make the lifecycle testable deterministically, `DecayScheduler` gains `internal` members (visible only via existing `InternalsVisibleTo`): `IsRegistered()`, `ProcessTick(now)` — extracted from `OnTick()` with no behaviour change — and `ResetForTests()`.
Red/green verified. Without the `MoveToWorld` fix, 5 of 6 of the original tests fail, including `ItemOnGround_ActuallyDecaysAfterDecayTime`, which shows a ground item never decays even after a full simulated hour. Without the chest fix, 8 of 8 chest tests fail. Without the refused-decay fix, that test fails.
Server.Tests 737/737 and UOContent.Tests 509/509 pass.
Started as an allocation pass over `StepCache` and grew into a cleanup of the surrounding pathing engine. Four commits, each independently reviewable; net **−560 lines**.
Build clean (0 warnings). All 122 `Server.Tests.Pathfinding` tests pass.
---
## 1. `perf`: pool the strata buffer, cut a hot-path dictionary lookup
**The headline is that `TryGetMask` — the actual hot path — was already allocation-free.** `StepMask` is a readonly struct, `StaticTileEnumerable` is a `ref struct`, `ChunkMissState` is a struct in a `Dictionary`. So most of this is a bake-throughput and GC-churn win, with one exception noted below.
`BuildChunk` accumulated packed multi-Z strata into a `List<byte>` that grew by doubling (256 → 512 → 1024 → …) and then paid a final `ToArray()`. A full map bake runs it ~114k times. It now writes into a `byte[]` rented from `STArrayPool<byte>.Shared` through a span writer, and hands the chunk one exact-size copy.
**This required fixing a latent out-of-bounds guard.** The record-fit check reserved headroom for **8** strata (`StratumByteLength * 8`) while `ComputeStandableSurfaceZs` can return up to **16** — so a cell could write 305 bytes starting from a 65,383-byte offset. Against a `List` that was benign (it just grew past 64 KB, and emitted offsets stayed under the `NoStrata` sentinel). Against a fixed-size rented buffer it is an out-of-bounds write, so tightening it was a *prerequisite* for the pooling, not a drive-by. The guard is now exact, which additionally proves no emitted offset can collide with `NoStrata == ushort.MaxValue`.
**One genuine query-path win:** `ShouldPromoteAfterMiss` did *two* dictionary lookups per miss — a `TryGetValue`, then an indexer assignment that re-hashes and re-probes. It now mutates in place via `CollectionsMarshal.GetValueRefOrNullRef`. This runs on every uncached chunk touch during A* expansion. The window-expiry branch keeps its explicit early return, so `MissPromotionThreshold == 1` still resets rather than promoting.
Also dropped `StepProbe.ComputeStrataAt` / `ComputedStratum` (dead code, zero callers) and collapsed six 18-argument `new StepMask(0, 0, …, kind)` blocks into `Fallthrough(kind)`.
**Considered and rejected:** pooling the `Direction[]` that `Find` returns. It *escapes* the call — `MovementPath` holds it across ticks while `PathFollower` walks `m_Index` through it — so it cannot be rented-and-returned, and it cannot be borrowed from the shared `BitmapAStarAlgorithm.Instance` without one creature clobbering another's in-flight path. `CheckPath` rate-limits repaths to one per 2s per creature, putting this at roughly 60 KB/sec at 1,000 pathing creatures. Not worth a public API break plus a use-after-return footgun.
## 2. `docs`: rewrite the comments for publication
The comments had accumulated as development notes: internal phase jargon (`Tier 4`, `the Phase-2 synthesizer`), change narration aimed at a reviewer (`which the old ComputeStandingZ anchor missed`, `legacy behavior`), benchmark anecdotes (`benchmarked as near-optimal`, `a ~20 ns lookup`), and paragraphs restating the code.
Rewritten to keep the rationale you cannot recover by reading the code — why the source-Z guard cannot be widened, why multis fall through with a halo, why the promotion gate counts Finds rather than calls, why `ComputeFingerprint` must hash the *files* and not the live tile tables — and drop the history that got us there.
Three comments were **factually wrong**, not just wordy:
- `CacheEvictionTimer` and `CacheStats` documented a class called `StaticWalkabilityCache`. No such class exists — it is `StepCache`.
- `StepCacheFile` declared `File layout v8` while `FormatVersion` is 9, and called the current record layout "the v6 layout" in four places. The layout descriptions are now unversioned so they cannot drift again.
- `StepProbe.ComputeStandingZ` claimed `StepCache` uses it to bake `SourceZ`. It has not since the baker moved to the clearance-aware `ComputeStandableSurfaceZs`; only a parity test calls it.
## 3. `refactor`: simplify `StepCacheFile.Write`, consolidate the format tests
`SaveToFile` walked `_keysList` **twice** — once to count the map's chunks, then again through a `ChunkEnumerator` closure to emit them — because `Write` needed the count up front to size its index array. Both loops had the same root cause. Passing a **span** collapses them: the count is just `span.Length`.
That deletes the `ChunkEnumerator` delegate, the closure over the list enumerator, and **both `InvalidOperationException` throws**, which existed only to police the delegate's "yield exactly `chunkCount` chunks" contract — a contract a span makes unrepresentable.
`Write` now patches the header's `IndexOffset` by seeking back to it rather than reaching into the writer's live buffer with `BinaryPrimitives`. That also retires `IndexOffsetFieldPosition`, a hand-maintained byte offset that had to track the header layout, and sidesteps the stale-array hazard that motivated the manual patch (`BufferWriter` reallocates on growth).
**Tests:** `StepCacheFileV6/V7/V8Tests` were named for the format version that introduced each transform — and the format is now **v9**, so all three names described formats the loader rejects outright. Beyond triplicated builders and plumbing, two things were actually broken:
- The three near-identical rejection tests each cited a `MinSupportedVersion` that had since moved (`"version 5 < MinSupportedVersion 6"`, `"6 < 7"`, `"7 < 8"`). They passed for the wrong reason.
- `AssertBaseEqual` (used by V7 and V8) **silently skipped the swim and strata trailers**. A regression dropping either would not have failed those tests.
Now one `StepCacheFileFormatTests`, named for behavior — predictive-Z elision, compression, compact index — with a single `AssertIdentical` that does check both trailers, the three rejection tests folded into one theory that also covers a future version, and a zero-chunk case the delegate-based writer never had coverage for.
## 4. `test`: consolidate the parity and lifecycle tests
Three files tested "parity" and none of the names said *which*. They were three different layers, and the seams are the useful part, so they are now one `StepCacheParityTests` that names them:
| Test | Compares | Answers |
|---|---|---|
| `ProbeMatchesSlowPath` | StepProbe vs MovementImpl | Is the bake right? |
| `CacheMatchesProbe` | StepCache vs StepProbe | Is it stored and returned intact? |
| `CacheServesReachableWalkStates` | StepCache vs MovementImpl | End to end, over the states A* visits |
Merging removed a duplicated stub `Mobile`, duplicated region seeds, and a filename/class mismatch (`StepProbeParityTests.cs` declared `StaticWalkabilityParityTests`). `SwimBake_ProducesWetCells` moved with it — it lived in the cache parity file but never touched the cache.
Tests reached into `StepCache._chunks` via `GetField` in **9 places**, each rebuilding the key encoding and cell-index arithmetic by hand. `StepCache` now exposes `GetResidentChunk` and `ResidentIndexInSync` alongside the internal test hooks it already had (`LazyReaderHasChunk`, `CurrentFindGeneration`), and the shared arithmetic moved to `PathingTestSupport`. All 9 reflection blocks are gone.
`StepCacheLifecycleTests` is regrouped by what it covers — promotion gate, fallthrough routes, strata, swim layer, eviction — with the `Tier4*` names dropped. Removed `Singleton_IsAvailable`, which asserted an inline-initialized static property was not null; that is the entire 123 → 122 test-count delta.
---
## Verification
Tests were mutation-checked rather than just run, since round-trip and parity tests can pass while a transform silently no-ops:
- Injecting an off-by-one into the `IndexOffset` patch fails **15 of 123** — the format tests are load-bearing.
- Offsetting the cache's cell index by one fails **7 of 10** parity cases, and the 3 that stay green are exactly the ones that do not touch the cache. The layering localizes a fault rather than just reporting one.
## Problem
Three coupled issues, each hiding the next:
1. **CI passed despite failing tests, with no test logs.** ([example run](https://github.com/modernuo/ModernUO/actions/runs/28639143286/job/84931544255) — the `Test` step produced zero output and the job went green.)
2. **Two `EmitsLowerStatReqWhenPassed` tests** fail with `KeyNotFoundException: '1060435'`.
3. Once CI actually ran the tests, **~337 UOContent tests failed** with `FileNotFoundException: tiledata.mul was not found` — the test bootstrap force-loaded copyrighted client data that CI doesn't have.
## Root causes & fixes
### 1. CI ran zero tests (`fix(ci)`)
The `Test` step ran `dotnet test --no-restore`, but the `Build` step only restores/builds `Application` — never the test projects. Without a restore, the test projects have no `project.assets.json`, so `Microsoft.NET.Test.Sdk`'s targets aren't imported, they aren't recognized as test projects, and `dotnet test` runs the `VSTest` target against **zero** projects → no output, **exit 0**.
- Both jobs now run `dotnet test --logger trx --results-directory ./TestResults` (test projects restore and run) **plus a guard** that fails the job if no `.trx` is produced — a permanent backstop against silent zero-test passes.
### 2. Impossible OPL tests (`fix(ci)` + `test(opl)`)
#2501 deliberately emits `LowerStatReq` (`1060435`) **inline in each item**, not in `GetProperties`. A follow-up "fix" dropped the `lowerStatReq:` argument to make the tests compile but left the assertions expecting `1060435`.
- Removed the two impossible tests, then removed the **entire `Tests/PropertyList/` OPL attribute set** from #2501: these assert exact cliloc/value/order of OPL emission per item base — a one-time proof of the #2501 rewire, now a permanent tax on modding (any admin reorder/value change/added line reddens the build). The one non-trivial case (LowerStatReq) is what just broke, because the test was wrong. Inline emission stays covered by the `BaseArmor`/`BaseClothing` tests.
### 3. Tile-data-dependent tests crashed CI (`test(uocontent)`)
UOContent.Tests' collection-fixture constructor force-loaded `tiledata.mul` unconditionally. On CI (no client files) it threw, and xUnit failed **every test in the collection** with the same error — mostly collateral (packet/scheduler/spawner tests that don't need tile data).
- Mirror Server.Tests' graceful pattern: `TestServerInitializer` probes for `tiledata.mul` and only loads tile/multi data (and runs the tile-dependent configure steps) when present, exposing `TileDataLoaded` so the fixture no longer throws.
- Add a shared `TileDataRequirement.SkipIfMissing()` guard and apply it to exactly the **31** pathfinding/multi/AI tests that genuinely need real tile data (`[SkippableFact]`/`[SkippableTheory]`).
## Verification (all local)
| Scenario | Server.Tests | UOContent.Tests |
|---|---|---|
| **Client data absent (CI)** | 726 pass, 17 skip, **0 fail** | 469 pass, 32 skip, **0 fail** |
| **Client data present (dev)** | 726 pass, 0 skip, **0 fail** | 501 pass, 0 skip, **0 fail** |
- Full `dotnet test` exits **0**; TRX files produced; the no-test guard trips (exit 1) only when zero `.trx` are produced.
Supersedes #2376 (@jwvalentine). This is a reviewed, corrected, and scoped-down **Phase 1** of Joe Valentine's throwing implementation — his original commits are cherry-picked here with authorship preserved, plus a fix/scoping pass. Phase 1 lands only the **core gargoyle Throwing skill**; the incomplete content is excised for follow-up PRs (see below).
## What's included (core skill)
- `BaseThrown` combat mechanics on top of the existing skeleton: close-quarters penalty, below-min-range penalty, shield penalty, STR-scaled range, overthrow damage penalty.
- Base weapons: **Boomerang, Cyclone, SoulGlaive** (gargoyle-only), + blacksmith crafting (SA-gated).
- Two symmetric hooks on `BaseWeapon` (`ModifyHitChance`, new `ModifyDamage`) that are inert no-ops for every other weapon.
## Fixes over the original
- **Overthrow damage**: was dead code (the swing gate already guarantees you're within `MaxRange`, so the old `ComputeDamage` check never fired). Reimplemented as `finalDamage × 0.53` applied *after* all offensive bonuses via a new `ModifyDamage` hook, firing at the outer range ring.
- **`DefMaxRange`**: clamped to `[MinThrowRange, MaxThrowRange]` (uncapped before → e.g. range 13 at 200 Str) and guarded against a latent divide-by-zero.
- **Close-quarters mitigation** now uses `RawDex` (matches ServUO/OSI; deterministic under stat mods).
- **Return-throw timer** guarded against a deleted/unmapped thrower/target.
- Reverted the `MovingShot` change (it rebalanced archery — belongs in a separate PR).
- Kept only complex-logic tests (hit-chance/range/damage math); dropped property-value assertions.
## Excised for follow-up PRs (Phase 2/3)
7 named artifacts, the Into-the-Void quest + Agralem, GargishOutcast, the Bladeweaver vendor, and the SA loot tables — these were unwired/non-functional (loot never triggered, quest/creature never spawned) and will return properly wired. `StormCaller` also needs its missing Battle Lust, and the quest its correct void-creature target.
## Verification
- Build: 0 warnings / 0 errors.
- `UOContent.Tests`: **501/501** passing (the `ModifyDamage` hook causes zero regressions across all weapons).
- Full whole-branch review completed: no must-fix defects.
## Summary
Consolidates the duplicated inline AOS attribute → `ObjectPropertyList` emission that each item base copy-pastes into per-family `GetProperties(IPropertyList)` methods, mirroring the existing `AosSkillBonuses.GetProperties` precedent.
### Per-family `GetProperties(IPropertyList)`
- **`AosAttributes`** — the 24 common attributes in canonical cliloc-ascending order, with optional `damageBonus` / `hitChanceBonus` / `luckBonus` params so item-computed bonuses (e.g. `GetDamageBonus()`) stay out of the family type.
- **`AosWeaponAttributes`** — `UseBestSkill`, the `Hit*` block (1060416–1060430), `MageWeapon` (`30 - prop`), `SelfRepair`.
- **`AosArmorAttributes`** — `MageArmor`, `SelfRepair` (the always-direct members; `LowerStatReq`/`DurabilityBonus` stay inline since they're item-computed in armor but container-direct in clothing).
### Rewired all 6 `AosAttributes`-emitting item bases
`BaseJewel`, `BaseArmor`, `BaseClothing`, `BaseWeapon`, `BaseTalisman`, `Spellbook` now call the family methods instead of inlining the chain. Net: large dedup in `BaseWeapon`/`BaseArmor`/`BaseClothing`/`BaseTalisman`/`Spellbook`.
## Behavior change: tooltip line **order** (set preserved)
This is **not** a pure no-op refactor, and that's unavoidable. Today the families are emitted **interleaved in cliloc order**, and the relative order differs per item class — e.g. `BonusDex` (1060409) is emitted early in `BaseArmor` but **after** the `Hit*` block in `BaseWeapon`. No single emission order reproduces every class byte-for-byte, so consolidating into contiguous per-family blocks necessarily **de-interleaves**: lines regroup **specific → general** (family-specific, then common `AosAttributes`).
- The **set** of emitted `(cliloc, argument)` lines per item is preserved **exactly** — nothing dropped, added, or value-changed.
- Only the **order** of lines within a tooltip changes for `BaseArmor` / `BaseWeapon` / `BaseClothing`. `BaseJewel` / `BaseTalisman` / `Spellbook` were already canonical, so those are byte-identical.
## Tests
- **Golden set-invariance tests** per item base (`BaseArmor/Clothing/Jewel/Weapon/Talisman/Spellbook PropertiesTests`) — each was written to pass against current `main` **before** the rewire (locking the emitted-line set), then confirmed still passing after, proving no line is lost/added/changed.
- Family-level unit tests for each `GetProperties` (canonical order, computed-bonus folding, the `AosArmorAttributes` exclusions).
- `dotnet build` clean; full `UOContent.Tests` green. (Pre-existing `AccountPacket`/`GumpPacket`/`MobilePacket`/`ClientEnumerator` golden-test failures reproduce on unmodified `main` and are unrelated to this change.)
## Summary
Removes the dated `DynamicJson` JSON helper and migrates spawner JSON (de)serialization to a polymorphic `record SpawnerDto` hierarchy. `DynamicJson` was the last remaining consumer (regions moved off it in #1400).
The key correctness improvement: **System.Text.Json deserializes plain DTO records, never a live `Item`.** Previously, deserializing directly into an `Item` meant STJ constructed world-registered objects *before* the data was validated — a malformed/hand-edited spawn file could leave orphaned spawner Items in the world save. Now a parse failure is GC-only; `dto.ToSpawner()` constructs the spawner only from a fully-validated DTO and self-cleans on failure.
## What changed
- **New:** `SpawnerDto` (abstract) + `SpawnerDataDto` / `RegionSpawnerDto` / `ProximitySpawnerDto`, each marked with a reusable `[JsonDiscoverableType]` opt-in attribute. Auto-discovered at the `Configure` phase — no manual registration list (avoids the regions `Register<T>()` footgun), open to custom spawner subtypes.
- **Symmetric mapping:** `BaseSpawner.ToDto()` (export) ⇄ `SpawnerDto.ToSpawner()` (import). `ToSpawner()` deletes-and-rethrows on any failure, so the importer can never orphan an Item.
- **`SpawnerJsonSerializer`** wires `$type` polymorphism on the `SpawnerDto` root with loud collision/constructibility validation.
- **Import/export commands** rewired to the typed DTO path (reflection `FindTypeByName`/`CreateInstance` removed).
- **Data migration:** the 109 `Distribution/Data/Spawns/**` files moved from `"type"`→`"$type"` and legacy `homeRange`→`spawnBounds`. The runtime still *reads* legacy `homeRange` for external files. The `homeRange→spawnBounds` formula is proven equivalent to the runtime conversion (`BoundsEquivalenceTests`, hr=0/1/3/7).
- **Deleted:** `Projects/Server/Json/DynamicJson.cs`.
Sparse export output matches the legacy `ToJson` (nullable DTO properties + `WhenWritingNull`). Binary world-save serialization is untouched.
## Tests
- DTO round-trip per spawner type; sparse-default omission; legacy `homeRange` read; export/import file round-trip.
- `Import_MalformedFile_LeaksNoWorldItems` — proves a mid-array parse failure constructs zero world Items.
- `AllSpawnFilesLoadTests` — every migrated spawn file deserializes and builds.
- Duplicate-discriminator validation.
- UOContent.Tests 485/485, Server.Tests 710/710, build clean.
## Follow-up (not in this PR)
`Rectangle3DConverter.Write` (in `Projects/Server/`) omits `z1/z2` when `Start.Z == -128`, so a future server-side export of a `homeRange`-style spawner round-trips depth 256→255. Pre-existing and out of scope here (Server change); the migrated data reads correctly. Worth a separate small converter PR.
## Problem
Houses and boats (multis) were pathed correctly only by **delegation to the slow path**: `StepCache.TryGetMask` returns `Fallthrough_Multi` for any multi-covered cell, and `GetSuccessors` ran `CheckMovement` **8× per cell** (each re-resolving the tile stack via `GetStaticAndMultiTiles`) — a sustained per-step cost near every house/boat. There was also no automated test pinning multi pathfinding.
This branch is the full multi-pathfinding effort in phases on one branch.
## Phase 1 — characterization tests (the oracle)
Implementation-agnostic invariants: a cache-on≡cache-off whole-path invariant, a per-cell sweep vs `CheckMovement` over footprint+halo (incl. destination Z), hand-verified routing (around walls, demolish-reopens, foundation-redesign-honored), classic-house / foundation / boat fixtures, non-vacuity guards. These gate every later phase byte-for-byte.
## Phase 2 — live single-pass synthesizer
`StepProbe.ComputeMultiMaskAt` synthesizes a covered cell's full 8-direction `StepMask` in one pass (the existing surface/step logic over `GetStaticAndMultiTiles` instead of 8× `CheckMovement`). `GetSuccessors` routes `Fallthrough_Multi` cells through it. No new cache, no `.swb` change. **~1.5×**, zero added allocations.
## Phase 3 / 3.1 — warm per-`multiID` interior cache (airtight)
`MultiMaskCache` caches each fixed multi's local-frame `StepMask` for **interior** cells (cell + all 8 neighbours covered → terrain-neighbour-free → position-invariant), keyed by `multiID & 0x3FFF`, built lazily from the MCL. Interior cells become ~20 ns lookups.
The cache is gated on a **per-instance footprint-clean flag** (`BaseMulti.PathInteriorCacheState`): an instance whose whole footprint terrain is below its floor (`maxTerrain < minFloor`) serves from the cache; a **dirty** instance (terrain intrudes — a contrived/GM placement) **degrades to live-synth, never a wrong mask**. This closes a cross-instance soundness gap (the cached mask depends on neighbour terrain too) found in a holistic review. The gate resets whenever the footprint's world-terrain relationship can change — **location, map, or ItemID** (a boat's heading swaps the MCL).
**Boats are cached too.** Their per-`multiID` deck masks are movement-invariant (built once per heading), so a sailing boat never rebuilds them; only the cheap clean-flag rescan repeats per move (and only when pathed near). Narrow existing boats have little interior; wide galleons (`multi.mul`) would gain Castle-class. `HouseFoundation` (per-instance runtime `DesignState`) is the one type that stays on the live path.
## Verification
- `UOContent.Tests` **454/454**, `Server.Tests` **708/708**, 0 failures.
- The Phase-1 oracle (`MultiPathInvariantTests`, cache-on ≡ cache-off) stays **byte-identical** with the synthesizer + interior cache active.
- Tests pin: footprint-cleanliness (clean vs sunk), dirty/cluttered placement degrades to live-synth while still pathing, clean placement serves, and the gate resets on move/ItemID change.
## Performance (modernuo/ModernUO-Benchmarks#8, full-fixture)
Houses at **Green Acres** (flat staff region → clean footprints, the legit-placement case):
| Route | Slow path | Phase 3.1 (interior cache) | Speedup |
|-------|----------:|---------------------------:|--------:|
| `around_a` (29 steps) | 238.3 µs | **49.1 µs** | **4.85×** |
| `around_b` (29 steps) | 224.3 µs | **49.5 µs** | **4.53×** |
~130 of ~167 multi cells/route serve from the cache (~20 ns) vs 37 live-synth. Per-cell, the slow path's 8× `CheckMovement` grows with multi complexity (GuildHouse ~857 ns → Castle ~1,194 ns), the synthesizer is a flat ~780 ns, and the cache serve is ~20 ns — so big/tall multis (and wide galleons) gain most. Identical allocations throughout.
Fixes#1690. Addresses Blood Oath holistically — three bugs found while researching the spell against RunUO, ServUO, the UODemise/uo.com guides, and the archived UOGuide page.
## Bugs fixed
### 1. Expiry timing (the filed issue)
The `ExpireTimer` polled every 1s, so expiry and death/delete cleanup lagged up to ~1s. Replaced with a **single-shot** timer plus centralized `[OnEvent]` handlers on `PlayerDeathEvent`/`PlayerDeletedEvent`/`CreatureDeathEvent`/`CreatureDeletedEvent` — the oath now breaks immediately on death/delete of either party.
### 2. Duration formula
Used `/80` (the bugged in-game tooltip value) instead of the real OSI formula `((SpiritSpeak - Resist) / 8) + 8`. Confirmed by RunUO, ServUO, the emulator guides, and the code's own fixed-point comment. At GM Spirit Speak this changes duration from ~9.5s to 23s and makes Spirit Speak actually affect duration.
### 3. Damage reflection (`BaseCreature.Damage` vs `PlayerMobile.Damage`)
`BaseCreature.Damage` diverged: it attributed the reflected hit to the attacker itself (`from.Damage(amount, from)`) instead of the caster, reflected the bonused (not original) amount, used `×1.1` vs `×1.2`, lacked the caster-survival guard, and had no Publish 48 resist mitigation.
Unified both paths: reflect the **original** damage attributed to the **caster** at `×1.2`. Publish 48 resist mitigation now applies only to creature casters and is gated behind `Core.SA`.
## Internals
- Collapsed the parallel `_oathTable` into a single `_table` keyed by both participants → shared timer, so `RemoveCurse` resolves from either side (required by the event handlers).
- Extracted `GetDurationSeconds` and `ComputeReflectedDamage` as testable statics.
## Tests
13 new tests (duration formula, reflection mitigation, oath lifecycle, end-to-end event-driven removal). Full suite: **436/436 pass**.
## Summary
Fixes the pathfinding step-cache (`.swb`) prebake so it bakes **once** and skips when a valid cache already exists, instead of re-baking on every boot. The root cause was the staleness fingerprint hashing mutable in-memory tile data rather than the on-disk files. This PR makes the fingerprint a pure function of the client data files and separates dynamic multis (houses/boats) from the static cache.
> This branch builds on the `ConfigurePrompts` first-boot-prompt unification (commit `5df8d0bd`, also included here) — that commit accounts for the `ServerConfiguration.cs` and `dev-docs/server-lifecycle.md` changes in the diff.
## The bug
With `pathfinding.prebakeMaps` set, the cache re-baked on **every** boot. The `.swb` staleness fingerprint hashed the live `TileData.LandTable`/`ItemTable` flags, which the server patches at runtime (`ItemFixes`, `LOSBlocker`, `PotionKeg`, `CTF`) at nondeterministic lifecycle points (Initialize-phase methods share a priority; static ctors fire lazily). So a fingerprint stamped at runtime (`[PathBake`) never matched the one recomputed during startup `Initialize()`, and the cache rebaked every time.
## Changes
**1. Fingerprint the files, not the in-memory tables** (`fix`)
Hash `tiledata.mul` (cached, computed once) plus the per-map `.mul`/`.uop` files — never the runtime-mutated `TileData` tables. The fingerprint is now lifecycle-stable. Existing `.swb` files rebake once after deploy, then stay stable.
**2. Compute the fingerprint once per boot** (`refactor`)
`Configure()`'s `AutoLoadAtStartup()` already opens and fingerprint-validates a reader for every up-to-date `.swb`. `Initialize()` now skips baking any map that already has an open reader (`StepCache.HasLazyReader`) instead of recomputing the fingerprint a second time.
**3. Bake static-only; route multis to the live path** (`refactor`)
Multis (houses/boats) are dynamic, so they're no longer baked into the static chunk cache — they were tagged with `BuiltMultisVersion`, a non-persisted session counter, which made persisting them unsafe (false matches / wasted re-bakes).
- Chunks bake land + `statics.mul` only.
- At query time, any cell whose sector (or its 1-cell halo) contains a multi routes to `Fallthrough_Multi` → the existing live, multi-aware `CheckMovement` path. The halo prevents a cell proposing a walkable edge into a neighbouring wall; interior (multi-free) cells pay one sector lookup.
- Adds `Sector.HasMultis` (one engine accessor); `.swb` format → v9 (rejects old multi-baked files); new `Fallthrough_Multi` telemetry.
- Behaviour-preserving: multis use the same live path the engine used before the cache existed.
**4. Comment polish** (`style`) — no behaviour change.
## Testing
All green: **92** pathfinding (incl. a new fingerprint-stability test and a multi-halo fallthrough test), **423** UOContent, **708** Server.
## Follow-ups (not in this PR)
- **Background bake worker** — make `[PathBake` and the boot prebake non-blocking (game thread serves tile reads to an off-thread worker).
- **Per-multi MCL cache** — cache walkability in each multi's own frame (keyed by multiID, movement-invariant) so houses/boats get a fast path instead of the live fallback.
- **House-pathfinding equivalence tests** — the one area not yet covered by a dedicated automated test; multi pathing is currently correct by delegation to the live path.
## Summary
Adds authentic **T2A-era (pre-UO:Third-Dawn) packet-based crafting menus**, enabled via the **`t2aCraftMenus` server setting** (read once at startup; default **`!Core.UOTD`**, so a pre-UO:TD shard gets them automatically). When enabled, double-clicking a crafting tool opens the classic `0x7C`/`0x7D` item-list menu — skill- and material-filtered — instead of the modern gump, covering all 8 tool/skill crafts (blacksmithy, tailoring, tinkering, carpentry, alchemy, bowcraft/fletching, inscription, cartography). It is **not** a runtime/admin-flippable feature flag.
This is the **definitive, reconciled** branch and **supersedes**:
- **#2181** (Delphi — `T2A_CraftingMenus`): the original effort.
- **#2381** (Jack/UOLL — `t2a_crafting_menus`): the research-grounded superset (Delphi's base + 12 corrections), rebased onto current `main`.
Original authorship is preserved across the cherry-picked history: foundation commit **@Delphi79**, mechanic fixes **@jackuoll (Jack Ward)**, reconciliation/fixes/docs mine.
## How it was built
1. Cherry-picked Jack's 13 commits onto current `main` (superset of Delphi's; only 2 trivial FeatureFlags conflicts).
2. Applied targeted fixes (below) with tests.
3. Full convention audit, build, and test pass.
Grounded in independent historical research plus Jack's deep dive. Maintainer reference: `dev-docs/t2a-crafting.md`.
## Mechanics (highlights)
- Double-click tool → target resource → skill/material-filtered menu → craft. Resource pre-selection per skill; make-last by targeting the tool.
- **Stacked-gem jewelry:** target a gem stack → the **full stack** is consumed and the piece is named by count ("a 1000 diamond ring"); count persists (`BaseJewel` serialization **v4 → v5**, new `_gemCount`).
- **Tool-less inscription & cartography** (skill-list invoked; no pen/sextant); inscription consumes reagents+scroll on success and failure, mana only on success.
- **Tailoring matching-hue consumption:** targeting hued cloth/leather consumes only that hue. Crafted items take color from their **`CraftResource`** (not the dyed hue), so dyed leather/cloth don't tint the product; in T2A only colored ingots/ore color items (metal armor/shields).
- **Half-resources on failed non-scroll crafts** (pre-UO:TD).
- **Maker's mark** always prompted for exceptional items, via the shared `QueryMakersMarkGump`.
- Server-side menu infra changes are additive (`ItemListEntry.CraftIndex`, `Entries` setter, `HasSent`).
## Notable changes on top of the cherry-pick
- **Toggle is a startup server setting, not a feature flag.** Removed `ContentFeatureFlags.T2ACraftMenus` (and its admin-flippable plumbing); the value is read once via `ServerConfiguration.GetSetting("t2aCraftMenus", !Core.UOTD)` into `T2ACraftSystem.Enabled`. Since the default tracks the era and it can't be flipped at runtime, there's no incoherent "menus-on / UO:TD-era" state.
- **Stacked-gem consumption (B3a/B3):** consume the full `PendingGemCount` (was deliberately consuming 1 while naming by the stack), null-safe gem type, plain-piece fallback + message. New `T2AJewelGemCraftTests`.
- **Convention audit:** `new List<Item>()` → `PooledRefList<Item>` on the hue-aware consume path; removed dead code.
## Decisions & deviations
- `make-last` kept as **QoL** (post-T2A gump-era feature).
- `half-on-failure` (non-scroll) kept as a **reconstruction** (not OSI-confirmed).
- **Stacked-gem** behavior set per shard authority (overrides the "single gem" reconstruction).
- **Cooking** out of scope (no T2A crafting menu existed for it).
- **No colored items from dyed materials:** crafted color comes from the `CraftResource` type. Pre-AOS leather has no colored variant, so leather is always uncolored; weapons retain resource color only in AOS+ (unchanged, intended).
## Test plan
- Automated: `dotnet build ModernUO.slnx -c Debug` clean; `dotnet test Projects/UOContent.Tests` → **421 passed** (incl. 3 new jewelry tests).
- Manual (needs a running T2A shard + client):
- [ ] Each of the 8 skills opens the correct menu; empty-menu guard fires.
- [ ] Make-last repeats the last craft (jewelry re-prompts gem).
- [ ] Jewelry consumes the full targeted gem stack and names by count.
- [ ] Cartography consumes blank maps only with T2A enabled / maps+scrolls when disabled.
- [ ] Tailoring consumes only the targeted-hue material; crafted items are not tinted by dyed cloth/leather.
- [ ] Maker's-mark prompt on exceptional.
- [ ] Failed non-scroll craft consumes half resources.
- [ ] Inscription: reagents+scroll on success/failure, mana only on success.
- [ ] T2A disabled: gump crafting unchanged.
## Credits
Co-authored-by: @Delphi79
Co-authored-by: @jackuoll
## Problem
Running the full `UOContent.Tests` suite, the test host **hangs ~2.5 minutes at shutdown and then crashes** (`Test host process crashed` / run aborted). The tests themselves are fine — they complete in ~1s — but the process can't exit.
Captured via `--blame-hang` dump. The blocking thread:
```
System.Threading.WaitHandle.WaitOne()
Server.SerializationThreadWorker.Sleep() SerializationThreadWorker.cs:54 (_stopEvent.WaitOne())
Server.SerializationThreadWorker.Exit() SerializationThreadWorker.cs:61
Server.World.ExitSerializationThreads() World.cs:429
Server.Tests.UOContentFixture..ctor()
```
### Root cause
Both collection fixtures (`UOContentFixture` and `PathfindingTestFixture`) each run the **full process-global ModernUO bootstrap**. `World.Load()` is guarded to run once per process, so the **second** fixture's `World.Load()` is a no-op and does **not** respawn the serialization workers — but `World.ExitSerializationThreads()` is **not** guarded, so the second fixture calls `Exit()` on workers whose threads have already terminated. `Exit()` → `Sleep()` → `_stopEvent.WaitOne()` then blocks forever (a dead thread never sets the event). The first collection's tests run; the second collection's fixture deadlocks in its constructor; the host eventually gets killed.
This is why single-collection (filtered) runs were fine — only one fixture ever bootstraps — but the full suite hangs. It's not a parallelization race: even strictly sequential, the second fixture deadlocks.
## Fix
**(a) Engine — idempotent `SerializationThreadWorker.Exit()`**
A second `Exit()` is now a safe no-op instead of a permanent block. Only the owning (main) thread calls `Exit()`, so no synchronization is needed, and the single-call production shutdown path is unchanged.
**(b) Tests — one shared bootstrap, strictly sequential collections**
- New `TestServerBootstrap.EnsureInitialized()` runs the superset global init **exactly once per process** (lock + once-flag).
- `UOContentFixture` / `PathfindingTestFixture` slim down to delegate to it and no longer tear down global state (which the single-bootstrap model owns for the host's lifetime).
- `[assembly: CollectionBehavior(DisableTestParallelization = true)]` so collections never overlap.
## Result
| | Before | After |
|---|---|---|
| Tests run (full suite) | 258 (UOContent collection deadlocked) | **418** |
| Outcome | 2.5-min hang → host crash | **418 passed, clean exit** |
| Wall time | killed | **~7s** |
## Summary
Fixes two related pet-behavior bugs and the underlying design flaw behind both:
1. **Post-combat erratic** — after a pet killed its `all kill` target it milled around erratically at the kill site (or failed to return to the master) until the player issued `all follow`/`all stop`.
2. **`all stop` returns home** — a pet with a non-zero `Home` walked back toward that location on `all stop` (on ML; on non-ML the old `DoOrderStop` was a no-op, so the sighting there came from a residual `Stay`).
### Root cause
`ControlOrder` and the wild-creature `Home` field were overloaded to express several distinct ideas, mutated/read inconsistently across order transitions:
- `Home` doubled as the controlled-pet "stay anchor" (`HandleStayOrder` set `Home = Location`) but nothing cleared it when the pet left the staying state; `HandleStopOrder` was the only handler that never touched it.
- `DoOrderStop` had dropped RunUO's `Home = Location` re-anchor, so on ML it walked to a stale anchor.
- The post-combat fallback was a fragile `_lastPetOrder` hack in `DoOrderNone` that re-anchored a resumed `Stay` at the corpse.
- Controlled idle-wander bypassed the `CheckIdle()` rest gate that every non-controlled creature uses, so idling pets jittered every AI tick.
## Approach
Separate three concepts that were tangled together:
- **`ControlOrder`** — the active order (may be transient: Come/Attack/Drop).
- **Persistent command** (`PersistentOrder` ∈ `{None, Stay, Follow, Guard}`) — the standing directive a pet falls back to when a transient order completes. Runtime-only (not serialized; reset to `None` on load) and **derived from master proximity on login** (near → Follow, far → Stay).
- **Anchor** (`Home`) — a pure function of the persistent command, set only when that command changes (never on transient transitions or fallback-resume), so it can't go stale.
### Behavior
- **Stop** is resolved immediately from what the pet was doing: Attack/Come → resume the persistent command; Follow/Guard → cancel to idle where it stands; Stay → stay put.
- **Stay** holds its post (returns only if displaced, e.g. after a fight) — no shuffle.
- **Idle** (`None`) is a gentle wander routed through `CheckMove/CanMoveNow/CheckIdle`, so idling pets take the same 15–25s rest periods as other creatures, on both ML and non-ML.
- **Post-combat** the pet resumes its persistent command (a staying pet returns to its original post, not the corpse).
- **Release** without a spawner anchors where the pet stands instead of pathing to a stale anchor.
This restores the RunUO-intended behavior (verified against the RunUO reference) while fixing the ModernUO regressions.
## Tests
New `PetOrderTests` (13 deterministic xUnit tests) cover: anchor lifecycle, the full Stop truth table, report 1 (post-combat return to post), report 2 (no stale-anchor walk-home), frozen-Stay/gated-idle wiring, release fix, derive-on-login, and a non-ML spot-check. The subjective wander *feel* is covered by a manual-QA checklist in the implementation plan.
## Notes
- Engine project (`Projects/Server`) untouched; the one `BaseCreature.cs` change is the `ControlOrder` setter passing the previous order to `OnCurrentOrderChanged`.
- `DoOrderCome` keeps auto-converting to `Stay` on arrival, which under the new model cleanly means "come and hold near me."
- Commits in this PR are temporarily **unsigned** (the signing agent's passphrase cache expired mid-session); happy to re-sign / amend on request.
## Summary
Phase #3b (final roadmap item), stacked on #2470. Compacts the index trailer from 20 to 8 bytes/chunk.
Trammel: 19.2 MB → 17.9 MB. Roadmap total: 565 MB → 17.9 MB (−96.8%).
## Details
- Trailer stores `{ u32 packedKey = (ChunkX << 16) | ChunkY, u32 recordLength }` per chunk, in record write order; the file offset is dropped and reconstructed by cumulative recordLength from HeaderSize.
- No record reordering, no varint; fixed-stride, TryReadChunk unchanged.
- Also simplifies the accumulated `.swb` code comments across the stack.
- Format v8; v7 files rejected and re-baked once.
## Tests
v8 multi-chunk round-trip (cumulative offset reconstruction) + the v6/v7 suite; full pathfinding suite green; Release build clean.
## Summary
Phase #3a, stacked on #2469. Compresses each chunk record independently with libdeflate (random access preserved).
Trammel: 124.7 MB → 19.2 MB (−85%).
## Details
- Whole-record framing `[u32 UncompressedLen][payload]`; records that don't shrink (tiny Uniform) are stored raw, detected as payload length == UncompressedLen.
- Codec chosen by full-Trammel spike: libdeflate VeryHigh (16.5 MB, 1.83 µs/chunk decompress) over zstd L19/22 (17.4 MB) and managed Brotli q11 (16.4 MB) — best native ratio, fastest decompress, already the repo's packet codec (no new dependency).
- Reuses cached thread-static bindings: `Deflate.Maximum` for bake, `Deflate.Standard` for reads.
- Compression is bake-time only; decompression is one-time per chunk (LRU-cached).
- Format v7; v6 files rejected and re-baked once.
## Tests
v7 unit tests + the v6 suite run through the compression path; full pathfinding suite green; Release build clean.
## Summary
Phase #2 of the `.swb` step-cache size-reduction roadmap (after #2465, v5 uniform elision). Stores the 16 base directional Z arrays as masked residuals against each cell's own SourceZ and omits any array that matches its prediction. Lossless, byte-identical reconstruction.
Trammel: 231.9 MB → 124.7 MB (−46%).
## Details
- Predictor: `predict = mask bit ? SourceZ : 0` (matches the baker's 0 on unwalkable directions); residual `Z − predict` via unchecked two's-complement (byte-exact for all inputs); reconstruct `Z = predict + residual`.
- A `u16 ZArrayMask` flags which of the 16 base arrays differ from prediction; matching arrays are omitted and synthesized from mask + SourceZ at read.
- Serializer-layer only: StepChunk, the cache, the algorithm, and the baker are unchanged.
- Format v6; v5 files rejected and re-baked once.
## Tests
21 v6 unit tests; full pathfinding suite green; Release build clean.
## Issue
Fixes#2452. A player with 30 Ninjitsu reported that the Animal Form menu showed **every** form; selecting one above their skill (e.g. Dog, req 40) **consumed mana** and returned "you need at least 40 skill", and afterwards the **gump never reopened** — every recast silently re-attempted the unusable form and drained more mana.
## Root cause
Three linked bugs, all reproduced from the code:
1. **Gump not gated by skill.** `AnimalFormGump.BuildLayout` compared `Skill.Fixed` (which is `Value * 10`, so 30 skill → `300`) against the raw 0–100 `ReqSkill` (Dog = `40`). `300 >= 40` is always true, so all forms were shown. `Morph` itself correctly uses `.Value`.
2. **Mana charged on a no-skill cast.** `Morph` returns `MorphResult.NoSkill` for an under-skilled form, but both call sites (`OnCast`, `OnResponse`) only special-cased `MorphResult.Fail`; `NoSkill` fell through to the branch that deducts mana.
3. **Menu never reopened.** Per OSI ([uo.com](https://uo.com/wiki/ultima-online-wiki/skills/ninjitsu/), [uoguide](https://www.uoguide.com/Animal_Form)), casting while **standing still always opens the selection menu**, and casting while **moving** quick-transforms into the last selected form. ModernUO only opened the menu when `lastAnimalForm == -1`, so once any form was selected a stationary recast skipped the menu.
## Fix
- Add `AnimalForm.CanSelectEntry` (compares `Skill.Value` to `ReqSkill`, plus the talisman check) and use it for the gump's per-entry enable check.
- `OnCast`: standing still always opens the menu; moving quick-transforms into the last form. `NoSkill` no longer costs mana.
- `OnResponse`: handle `Success` / `Fail` / `NoSkill` explicitly so `NoSkill` costs no mana.
## Tests
Adds `AnimalFormTests`:
- `CanSelectEntry` rejects forms above skill, accepts forms at/below skill, and requires a talisman for talisman-gated forms.
- `Morph` returns `NoSkill` (without transforming) when under-skilled, and `Success` when sufficiently skilled.
Verified the gating test catches the regression (reintroducing `.Fixed` fails it). Full solution build is clean; the 5 new tests plus 284 other UOContent tests pass (the pathfinding/AI sequential tests were excluded only because they deadlock under concurrent local runs — they are unrelated to this change).
## Summary
Sub-project #1 of the `.swb` step-cache size-reduction roadmap (`dev-docs/pathfinding.md` § Future work). Adds **uniform-chunk elision** to the `StepCacheFile` format, bumping it **v4 → v5**.
A fully-uniform 16×16 chunk — no strata, **no swim layer**, all 19 base arrays constant (open ocean, Green Acres, void) — serializes to a **~28-byte record** (`KindUniform`) instead of ~5,393, and reconstructs **byte-identically** via `Array.Fill`. Non-uniform chunks use the existing v4 body (`KindFull`) with the swim-layer and strata trailers **fully preserved** — the Kind byte is just prepended.
## Calibrated result (measured, not projected)
Baked Trammel via `SaveToFile`:
| | |
|---|---:|
| Chunks | 114,688 |
| Uniform (swim-aware) → elided | 62.7% |
| Swim-layer chunks (stay Full) | 8.9% |
| Strata chunks (stay Full) | 1.8% |
| Baseline (full records) | 592.2 MB |
| **Actual v5 `.swb`** | **231.9 MB (−61%)** |
The residual is ~150 MB of non-uniform land Z-blocks (targeted by #2 predictive-Z) + ~81 MB of swim-layer trailers (#2/#3). #2 and #3 are separate follow-up PRs.
## Implementation
- `StepChunk.IsUniform()` — false if it has strata **or a swim layer**, else true only when all 19 base arrays are constant (the "all-same" check uses the SIMD-accelerated `ContainsAnyExcept`).
- `StepCacheFile` v5 — `Kind` byte (`KindFull=0`/`KindUniform=2`, 1 reserved); uniform write/read; `FormatVersion`/`MinSupportedVersion` → 5 (v4 files rejected on open and re-baked). No `StepCache`/algorithm/index changes; fingerprint logic untouched.
## Tests
7 `StepCacheFileV5Tests` (uniform round-trip + `<200 B` compactness, varied-full, swim-layer-full, strata-full, swim+strata combined, v4 version-gate rejection) + the existing StepCache/pathfinding suite — **70 pass**, including the prior `SwimLayer_RoundTrips`. An independent review verified write/read symmetry, cast round-tripping, swim/strata preservation, and the version gate (READY TO MERGE).
Fixes#2462
## Summary
Removes the per-object `VirtualHairInfo` heap wrapper for mobile/corpse hair. Hair is now stored **inline** on `Mobile` and `Corpse` as `int _hairItemId` / `int _hairHue` plus a lazily-allocated, **non-serialized** ephemeral `Serial _hairSerial` (in the high virtual-serial range) — and likewise for facial hair. The `VirtualHairInfo` class is deleted, with a **lossless** save migration.
This delivers three things:
1. **Fixes a hair-removal bug.** `Delta(MobileDelta.Hair)` is deferred (it enqueues; `ProcessDeltaQueue` runs later in the tick). The old `HairItemID = 0` setter nulled `_hair` *immediately*, so by the time `ProcessDelta` built the remove packet the equipped virtual serial was already gone — the old `??=` code then re-materialized a **fresh** serial (≠ the equipped one), so clients never removed the right entity, and it left a phantom ItemId-0 object behind. The serial now lives on the entity and **persists across removal**, so remove packets carry the correct serial.
2. **Lightens the entity.** No heap hair object; bald mobiles allocate nothing (the serial is minted lazily only when hair is present). This was the original reason `HairItemID`/`HairHue` exist.
3. **Removes `VirtualHairInfo` entirely**, keeping the high-range virtual serial behavior.
## How
- **Mobile** (manual serialization): inline `_hairItemId/_hairHue/_hairSerial` (+facial); lazy `HairSerial`/`FacialHairSerial`; `ProcessDelta` reads those. Serialization **v36 → v37** — the v30-v37 deserialize is unified, reading the legacy per-hair `VirtualHairInfo` version int only when `version < 37`. Setting item id to 0 clears the hue (matching the old object-nulling) while retaining the serial.
- **Corpse** (codegen serialization): decomposed to `[SerializableField] int _hairItemId/_hairHue` (+facial) + ephemeral serial; **v16 → v17** with `MigrateFrom(V16Content)`.
- **Lossless migration:** the loader validates exact byte length, and the old corpse hair is a presence-bool-gated block, so a tiny **migration-only** `LegacyHairInfo` reader (no runtime role) consumes the legacy `[bool][int ver][int itemId][int hue]` bytes. Frozen `Corpse.v14/v15/v16.json` are retyped to it; `v17.json` describes the new int fields.
- All consumers updated to discrete accessors: `OutgoingMobilePackets`, `CorpsePackets`, corpse subclasses (`MilitiaFighterCorpse`, `SchmendrickApprenticeCorpse`), and the packet test mirrors.
- `VirtualHair.cs` renamed to `OutgoingVirtualHairPackets.cs` (the only type left in it after `VirtualHairInfo` was removed).
## Test Plan
- [x] Full solution build: **0 warnings, 0 errors** (`TreatWarningsAsErrors`).
- [x] `Server.Tests`: **708 passed** (incl. new `RemoveHairUsesEquippedSerial` / `RemoveFacialHairUsesEquippedSerial` proving the serial survives removal + hue clears).
- [x] `UOContent.Tests` corpse/hair: **6 passed** (incl. `CorpseHairMigrationTests` asserting the legacy hair bytes are consumed exactly — the loader's length invariant).
- [x] Generated migration code inspected: V14/V15/V16 readers consume the legacy block byte-for-byte; serial never written to disk.
## Upgrade notes
- Old Mobile (v30–v36) and Corpse (v13–v16) saves load losslessly.
- Minor cosmetic-only change: `SchmendrickApprenticeCorpse` hair/facial-hair RNG draws shift order within each pair (same draw count); irrelevant for a quest NPC corpse.
## Summary
Closes the Cold-cache regression flagged in PR #2450. `StepCache.TryGetMask` no longer eagerly runs `BuildChunk` on the first miss for a chunk that isn't in a `.swb` lazy reader. Instead it returns `Fallthrough_NotBuilt` and the caller (`BitmapAStarAlgorithm`) takes the per-cell slow path. The chunk is only promoted to the bitmap fast path after the **second** miss within a 30-second window, filtering single-touch pass-throughs.
This makes BitmapAStar's worst-case (cold cache + short hops) collapse from **12–47× slower** than FastAStar to **roughly the same**, which is the floor the slow path can deliver. Steady-state warm performance (the actual deliverable) is unchanged from PR-5 — it was always the cache fast path.
## The pet-follow scenario this fixes
A mounted player at ~4 tiles/sec with a pet/hireable following will trigger an NPC pathfind every 100–300 ms. Each pathfind is 1–6 tiles. As the player crosses chunk boundaries (~4 sec/chunk), the pet's first pathfind in the new chunk under the previous behavior triggered a full ~700 µs `BuildChunk` for a chunk the player would leave shortly after. At 50–100 mobiles per shard, this exceeded the 8 ms tick budget. PR-5 BDN data showed scenarios 6–9 (2–8 tile NPC perception) at 2,300–3,700 µs Cold vs FastAStar's 80–200 µs.
Under the new gate:
- First miss → `Fallthrough_NotBuilt` → caller uses slow path (~30–50 µs short path). No `BuildChunk`. No allocation.
- Player keeps moving → chunk never gets a second touch within window → never promoted, no rot.
- NPC patrolling a fixed territory → repeatedly hits the same chunks → second touch within window → promote → cache fast path on subsequent calls.
## What changed
- **`CacheHitKind.Fallthrough_NotBuilt = 6`** + **`CacheStats.FallthroughNotBuilt`** counter. `IsHit=false`, so the caller routes to slow path.
- **`StepCache._chunkMissTracker`** — `Dictionary<long, ChunkMissState>` capped at 4096 entries. State is `(byte missCount, uint lastMissTickStamp)` keyed by chunk key. Window-expired entries reset count to 1; capacity overflow prunes window-old entries first.
- **`StepCache.MissPromotionThreshold`** (default `2`) and **`StepCache.MissPromotionWindowMs`** (default `30_000`) — tunable, can be wired through `ServerConfiguration` if shards want different policy. Setting threshold to `1` restores legacy eager-build behavior (used by tests that prime chunks via single `TryGetMask` call).
- **`StepCache.TryGetMask` miss branch** — try lazy reader first (file-loaded chunks bypass the tracker entirely; an `.swb` represents an explicit prior decision to keep the chunk warm). Otherwise consult the tracker.
- **`BitmapAStarAlgorithm.GetSuccessorsSlowPath`** now layers `IsBlockedByDynamic` on top of `CalcMoves.CheckMovement`. Previously the slow path only ran for `CanFly` creatures and rare cache fallthroughs — `CheckMovement` doesn't iterate same-cell mobiles, so the bitmap fast path's `IsBlockedByDynamic` was the only mobile-blocking check. Now first-touch pathfinds run through the slow path, so the gap had to close.
## Tests
50 pathfinding tests pass (was 47). New / updated:
- **`TryGetMask_FirstTouchOnUnbuiltChunk_DefersBuildAndReturnsFallthrough`** — single TryGetMask call returns `Fallthrough_NotBuilt`, no chunk built, no allocation.
- **`TryGetMask_SecondTouchWithinWindow_PromotesAndBuilds`** — second call inside the 30s window builds + serves.
- **`TryGetMask_SecondTouchAfterWindow_RestartsCounterAndDefers`** — second call outside the window restarts the count, returns Fallthrough again.
- **`TryGetMask_DistinctChunks_TrackedIndependently`** — counters are per-chunk; one touch on each of two adjacent chunks both stay in fallthrough.
- **`LazyReaderHit_BypassesMissTrackerOnFirstTouch`** — open `.swb` + first touch hits without consulting the tracker. Production with `.swb` loaded skips the gate entirely.
- **`MultisVersion_Bump_TriggersDirtyRebuild`** — updated to reflect the new 3-step flow (Fallthrough → Miss_NotBuilt → Miss_DirtyRebuild).
- Tests that prime chunks via a single `TryGetMask` call (multi-Z, Tier4, lifecycle, parity, BitmapAStar uses-cache) set `MissPromotionThreshold = 1` to opt into eager behavior.
## Expected BDN impact
The Cold column from PR-5's BDN should change as follows once the bench's submodule pointer is updated to this branch:
| # | Scenario | Cold (PR-5) | Cold (PR-6 expected) | FastAStar Cold |
|--:|-----------------|-------------:|---------------------:|---------------:|
| 2 | sewer corridor | 1,627 µs | ~36 µs | 36 µs |
| 4 | causeway | 1,533 µs | ~39 µs | 39 µs |
| 6 | pet 2-tile | 2,364 µs | ~80 µs | 81 µs |
| 8 | npc 5-tile | 3,708 µs | ~140 µs | 141 µs |
| 9 | npc 8-tile | 2,386 µs | ~200 µs | 197 µs |
WarmNoFile and LazyWarm rows should be unchanged — they were always cache-warm. The miss tracker only fires when neither resident chunks nor the lazy reader can satisfy the request.
## Future work (not in this PR)
- **Background-thread bake**: builds outside the game thread so even promoted chunks don't pay the 700 µs build cost on the main thread. Rule 10 (no Task.Run) applies, so this needs careful design — the bake is a pure data transform but main-thread synchronization on chunk-state transitions has to be threaded through. Defer to a follow-up.
- **Long-traverse BDN scenario**: a multi-Find benchmark simulating 50 pet repaths across chunk transitions. Requires restructuring the bench harness; the existing 10-scenario corpus + Cold provider already exercises the gate.
- **Swim sourceZ bake**: scenario 5 (sea serpent) shows 56 B alloc on warm paths because the cache's SourceZ is computed under default-walker rules. Swim creatures fall through to slow path. Independent of this PR.
## Summary
Creatures (pets following, monsters chasing, NPCs approaching) would **oscillate — "pace back and forth really fast"** at concave obstacles (reported at the Britain Inn L-desk: a pet at `(1493,1614,20)` never reaching its master at `(1494,1605,21)`) instead of routing around them.
**Root cause** (the A* pathfinder itself was correct): all goal-seeking funnels through `MoveTo` and `WalkMobileRange` → `MoveTowardsOrAwayFrom`, which step greedily via `DoMove(dir, badStateOk:true)`. `DoMove` returns `true` even when the direct step was blocked and the creature merely **auto-turned and sidestepped** (`MoveResult.SuccessAutoTurn`), and the caller then set `Path = null`, discarding the `PathFollower`. So at a concave obstacle a non-progressing sidestep was mistaken for progress and the creature never committed to a route. (AOS pet-follow runs at `CurrentSpeed = 0.1`, hence the "really fast" shuffle.)
## What changed
- **New centralized `BaseAI.ApproachTarget(target, run, range)` primitive.** A greedy step is committed only when it **fully succeeds (`MoveResult.Success`) and actually gets closer**; otherwise the creature commits to a **persistent `PathFollower`** that routes around the obstacle and is never discarded by a greedy step. The open-terrain fast path (one greedy step, no pathfinding) is preserved. `MoveTo`, `MoveTowardsOrAwayFrom`, and `MoveToWithCollisionAvoidance` all delegate to it — public signatures unchanged, so no AI-class call site changes.
- **Best-distance give-up + idle.** A creature that cannot reach a **stationary** in-range goal stops shuffling and idles after `ApproachGiveUpTicks` (40) ticks without lowering its closest-ever distance; a **moving** goal (active chase) never gives up. It resumes the moment the goal moves.
- **Pathfinder fix (required):** `BitmapAStarAlgorithm.IsBlockedByDynamic` now skips the dynamic mobile-block check **at the goal cell only** (`MoveImpl.Goal`). Previously A* returned `null` whenever the target mobile stood on the goal cell, so creatures could never pathfind *toward* another mobile — only toward empty ground. The follower stops within `range` short of it. Static/item blocking and all non-goal mobile blocking are unchanged.
## Tests
New AI-loop integration tests in `ApproachTargetTests.cs` drive the real `BaseAI` primitives against live Britain Inn map statics: exact-repro pet follow, open-terrain (asserts zero pathfinding), `MoveTo` chase (static + walking-away target), route-around-a-dynamic-wall, and walled-off give-up-and-idle.
- Pathfinding + AI subset: **52/52** pass.
- Full `UOContent.Tests`: **301/301** pass. (Note: the test host lingers on shutdown — a pre-existing infra quirk unrelated to this change; all tests complete and pass.)
- Full solution build: clean (0 warnings / 0 errors).
## Notes
- Branched off `main`; independent of the in-flight step-cache work.
- Out of scope (future work): proactive "SmartAI" look-ahead pathfinding so clever creatures plan a route before walking into the obstacle, rather than reacting after they hit it.
## Test Plan
- [X] In-game: order a pet to `follow`/`come` across the Britain Inn L-desk; confirm it routes around and reaches you instead of pacing.
- [X] Aggro a monster and kite it around a building/treeline; confirm it chases around obstacles.
- [X] Confirm open-terrain following/chasing feels unchanged (no extra latency).
- [X] Confirm a creature with a genuinely unreachable target idles rather than shuffling forever.
## Summary
Multi-Z cells (bridges, stairs, paver-over-ground, multi-floor structures) now carry **per-stratum walkability data** in the cache instead of falling through to the slow path. The data is computed at chunk-build time, persisted in the `.swb` file, and selected at query time by matching the request's `sourceZ` against each stratum's `zCenter` (within `StepHeight` tolerance).
This is the Tier 4 strata feature, deferred from PR #2447 / PR #2448 / PR #2449. Builds on PR #2449's lazy backing store and public bake helpers.
## Wire format change (v1 → v2)
`StepCacheFile.FormatVersion = 2`. `MinSupportedVersion = 2`. v1 `.swb` files are silently rejected at open time (treated as missing) and overwritten on the next `SaveToFile` / `BakeMap`. **No migration** — older files just get re-baked.
The `MinSupportedVersion` sentinel is the model going forward: bump the constant when an incompatible change lands; admins re-bake on the next deploy. No matrix of v1↔v2↔v3 migration logic to maintain.
## What changed
- **`StepProbe.ComputeStrataAt(map, x, y)`** — enumerates walkable standing-Zs at the cell (one per land surface plus one per walkable static), collapses Zs within `2*StepHeight`, runs `ComputeMaskAt` at each surviving Z. Returns `null` for single-Z cells (caller uses the chunk's main mask).
- **`StepChunk`** — replaces the old `MultiZCells` bitmap with a **strata storage pair**:
- `ushort[256] StrataOffsetByCell` (sentinel `NoStrata = 0xFFFF` = "no strata for that cell")
- `byte[] StrataData` packed: `u8 stratumCount`, then `count × 19-byte stratum`
- `sbyte zCenter, byte walkMask, byte wetMask, sbyte walkZ_N..NW (8), sbyte swimZ_N..NW (8)`
- `IsCellMultiZ` derives from `StrataOffsetByCell[cell] != NoStrata` — same semantics, single source of truth.
- **`StepCache.BuildChunk`** — populates strata for cells flagged multi-Z via `SetStrata`. Chunks with zero multi-Z cells pay zero strata overhead (offset array + data array stay null).
- **`StepCache.TryGetMask`** — for multi-Z cells, scans strata with `TryStratumHit`; returns the matching one with `HitKind=Hit`. Falls through to slow path only when no stratum matches the query `sourceZ`.
- **`StepCacheFile`** — v2 serialization with strata trailer per chunk + `recordLength` in index entry. Lazy reader sizes scratch per-chunk-record using the recorded length, growing on demand for multi-Z-heavy chunks. Patches `IndexOffset` on `w.Buffer` (BufferWriter's current backing array) since it grows during variable-size chunk writes.
## File layout v2
```
Header (48 bytes):
u32 Magic = 0x42575300 ('SWB\0')
u32 Version = 2
u32 MapId
u64 Fingerprint XxHash3 over LandTable + ItemTable flags + map files (mapX.mul/.uop, staidxX.mul, staticsX.mul)
u64 BakeTimestamp informational
u32 ChunkCount
u64 IndexOffset position where chunk index begins
Per chunk (variable size):
u16 ChunkX, ChunkY
u32 BuiltMultisVersion
u8 HasStrata 0 = no strata trailer; 1 = strata trailer follows
byte WalkMask[256], WetMask[256]
sbyte SourceZ[256], WalkZN..WalkZNW[256], SwimZN..SwimZNW[256]
// Strata trailer (only when HasStrata == 1):
u16 StrataOffsetByCell[256] // NoStrata sentinel = 0xFFFF
u32 StrataDataLength
byte StrataData[StrataDataLength]
Per multi-Z cell: u8 count, then count × Stratum (19 bytes)
Index trailer (20 × ChunkCount bytes):
per chunk: { u64 chunkKey, u64 fileOffset, u32 recordLength }
```
## Summary
Adds two pieces of pathfinding tooling on top of PR #2448's lazy `.swb` infrastructure:
- **`PathfindRecorder`** — admin-toggled JSONL telemetry capture; one record per `BitmapAStarAlgorithm.Find` call. Output format matches the BDN harness corpus, so production traffic can be captured and replayed in benchmarks without an adapter.
- **Public bake helpers on `StepCache`** — `ComputeLiveTileDataHash`, `TryReadTileDataHashFromFile`, `BakeMap`, `ClearResidentChunks`. Lets the benchmark project (and any future bake utility) drive cache fill + persist without exposing internal types.
The companion BDN harness update lives in [ModernUO-Benchmarks#kb/pathfinding-pr4-bench](https://github.com/modernuo/ModernUO-Benchmarks/tree/kb/pathfinding-pr4-bench): porting `Benchmarks/PathfindInGame/` from the `kb/ai_pathfinding` branch to the API shipped in #2446–#2448.
## What's in this PR
### `PathfindRecorder` (`PathfindRecorder.cs`)
- Holds a single `StreamWriter` open while recording; its internal buffer absorbs per-record writes without per-call `File.AppendAllText`.
- Single `bool` check on the hot path; cheap when disabled.
- Disabling flushes + disposes; an IO failure during write also disables the recorder.
- Server config:
- `pathfinding.recorder.enable` — bool, default `false`. Read on boot via `GetOrUpdateSetting`.
- `pathfinding.recorder.path` — default `<basedir>/Data/Pathfinding/recordings/pathfinds.jsonl`.
- Hooked into `BitmapAStarAlgorithm.Find` — runs once per call, does nothing when disabled.
- Admin command: `[PathRecord [on|off|flush|status]` (default `status`).
### Public cache helpers
- `static ulong StepCache.ComputeLiveTileDataHash()` — wraps the file module's hash function for staleness checks.
- `static bool StepCache.TryReadTileDataHashFromFile(string, out ulong)` — peeks at a `.swb` file's hash field (20 bytes).
- `int StepCache.BakeMap(int, string)` — walks every chunk in the map, populates resident set, saves. Offline / fixture use; blocks for many seconds on a full-map walk.
- `void StepCache.ClearResidentChunks()` — drops chunks + zeros counters but keeps lazy readers open. Lets benchmark loops measure "first query after boot" cost across iterations without the lazy-reader reopen overhead.
## Summary
Adds a binary disk format + lazy reader so the step cache can warm-start from a precomputed file without paying chunk-build cost on the first pathfind through a region. **Resident memory stays bounded by `MaxResidentChunks` regardless of file size** — opening a `.swb` reads only the header + chunk-offset index (~16 bytes per indexed chunk), and individual chunks are seeked + deserialized only when `ResolveMissingChunk` asks for them.
The lazy design (vs. an eager bulk load): a 250 MB bake on a RAM-constrained shard never materializes more than the LRU cap (~40 MB at the default 8192-chunk cap), and unwanted regions never enter memory at all.
Builds on PR #2447.
## What changed
- **`StepCacheFile`** — binary reader/writer module. Writer emits header → chunks (offsets recorded) → index trailer, then patches the header's `IndexOffset` field. Reader is `OpenForLazy(path)` returning a `LazyReader` that holds an open `FileStream` + offset dictionary.
- **`StepCacheFile.LazyReader`** — `TryReadChunk(chunkX, chunkY)` does a single seek + bulk read for one record. `Dispose` releases the underlying stream. Files are opened with `FileShare.Read | FileShare.Delete` so admin tooling can replace them.
- **TileData fingerprint via XxHash3.** The `.swb` header carries a hash of `LandTable + ItemTable` flags. Load rejects any file whose hash doesn't match the running server. Computed via `HashUtility.ComputeHash64` (engine-blessed hasher) — adds a `ReadOnlySpan<byte>` overload alongside the existing `ReadOnlySpan<char>` one for parity.
- **`StepCache.SaveToFile(path, mapId)`** — writes resident chunks for the given map.
- **`StepCache.TryOpenLazyReader(path, mapId)`** — opens the file, validates header, holds the reader for the map's lifetime.
- **`StepCache.ResolveMissingChunk`** — now consults the lazy reader before invoking the runtime baker. A loaded chunk whose `BuiltMultisVersion` doesn't match the live sector falls through to the baker (snapshot was made before a multi was added/removed in that sector).
- **`StepCache.Clear` closes lazy readers.** Test cleanup can delete `.swb` files cleanly.
- **Auto-load at startup.** `PathCacheCommands.Configure()` opens `Data/Pathfinding/<mapId>.swb` as a lazy reader for every map.
- **`[PathCacheSave`** / **`[PathCacheLoad`** — admin commands for the same workflow.
- **`pathfinding.maxResidentChunks` shard-tunable.** Read from `server.cfg` at boot via `ServerConfiguration.GetOrUpdateSetting` (default 8192 ≈ 40 MB). Small shards can tune down; large shards with substantial bakes can tune up to reduce eviction churn. Default is written back to `server.cfg` on first boot, matching the engine pattern used by other settings.
## File layout (v1)
```
Header (48 bytes):
u32 Magic = 0x42575300 ('SWB\0')
u32 Version = 1
u32 MapId
u64 TileDataHash XxHash3 over LandTable + ItemTable flags (HashUtility)
u64 BakeTimestamp informational
u32 ChunkCount
u64 IndexOffset file position where the chunk index begins
Chunk records (fixed size, ~5,393 bytes each, +32 if multi-Z):
u16 ChunkX
u16 ChunkY
u32 BuiltMultisVersion
u8 HasMultiZ
byte WalkMask[256], WetMask[256]
sbyte SourceZ[256], WalkZN..WalkZNW[256], SwimZN..SwimZNW[256]
[byte MultiZCells[32] when HasMultiZ == 1]
Index trailer (16 × ChunkCount bytes):
(u64 chunkKey, u64 fileOffset)
```
## Memory math
| Scenario | Disk file | RAM at boot | Notes |
|---|---|---|---|
| Empty / no `.swb` files | — | 0 | Silent; cache builds on demand. |
| Admin-curated towns (5K chunks) | 25 MB | 0 + per-query | Index ≈ 80 KB. Resident grows to the configured cap under steady-state queries. |
| Full-map bake (50K chunks) | 250 MB | 0 + per-query | Index ≈ 800 KB. Same configured cap. Cold areas never load. |
| All 5 maps fully baked | 1.25 GB | 0 + per-query | Index ≈ 4 MB total. Same configured cap. |
## Hash choice (FNV-1a → XxHash3)
The original draft used inlined FNV-1a-64. Switched to XxHash3 via `HashUtility`:
- ~30× faster on this workload (~30 GB/s SIMD vs ~2 GB/s byte-by-byte). Boot-time only, so absolute saving is microseconds — the real wins are elsewhere.
- Stronger collision resistance and distribution.
- Drops ~25 lines of inlined hash code; matches the rest of the codebase's hashing pattern.
- Hash is stable as long as `HashUtility`'s `xxHash3Seed` constant doesn't change (already marked `// DO NOT CHANGE THIS NUMBER`).
## Summary
Builds on PR #2446's cache-direct A*. The previous PR conservatively routed players + creatures with capability flags entirely through the slow path. This PR pushes that line: most mobile classes now use the cache, with the right rule set layered on top per-mobile, and the cache fast-path now does the dynamic items / mobiles check that PR #2446 had silently skipped.
## What changed
- **Non-GM players** now use the cache. Diagonal corner-cut applies the strict AND-rule (BOTH cardinal partners walkable) by reading the same source-cell mask byte the creature OR-rule reads — both rules are evaluable from one byte.
- **Creatures with `CanOpenDoors` / `CanMoveOverObstacles`** now use the cache. Reading `MovementImpl` confirmed those flags only affect dynamic items, never static tiles, so they were over-conservatively excluded before.
- **Swim creatures** now use the cache via a capability overlay. `StepProbe` bakes a second rule set (`canSwim=true, cantWalk=true`) producing `WetMask` + `SwimZ_*`. The algorithm composes `effectiveMask = (walkMask & !cantWalk) | (wetMask & canSwim)` per direction; walk Z preferred when both apply.
- **Dynamic-obstacle pass.** Cache fast-path now mirrors `MovementImpl`'s per-cell items + mobiles collision check (`GetItemsAt` / `GetMobilesAt` at the target cell, with `CanOpenDoors` / `CanMoveOverObstacles` / spell-field overrides). This closes a correctness gap from PR #2446 — the cache fast-path was silently skipping dynamic obstacles entirely.
- **`StepCache.TryGetMask` returns `StepMask` struct** instead of 11 out parameters. `HitKind` rolls into the struct with an `IsHit` accessor. Sets up wet/swim without ballooning the call site.
- **`StepChunk.MultiZCells` is lazy-init.** Most chunks are entirely single-Z; allocating the 32-byte bitmap up-front wasted ~256KB at full cap.
- **Admin commands.** `[PathCacheStats` (resident chunks + hit/miss/eviction counters) and `[PathCacheClear` (drop everything, zero counters).
- **Feature flag.** `bitmap_pathfinding_cache` (default true) gates the cache fast-path. Flipped off, every cell expansion routes to `MovementImpl` — equivalent to PR #2446's slow-path-only behavior. Safety net for shipping the new behavior.
`RequiresSlowPath` shrinks to just `CanFly` — flying creatures Z-jump arbitrarily, which the cache's static-Z model can't accommodate.
## Summary
Replaces `FastAStarAlgorithm` with `BitmapAStarAlgorithm`: one cache lookup per cell expansion (8-direction mask + per-direction destination Z) instead of 8 separate `MovementImpl.CheckMovement` calls. Adds the supporting cache infrastructure to back it.
Public API unchanged — `MovementPath` / `Mobile.Move` / `CalcMoves.Find` return the same shapes; the algorithm swap is internal.
## What's in this PR
- **`BitmapAStarAlgorithm`** — A* that issues one `StepCache.TryGetMask` call per cell expansion. Inline fallthrough to the per-cell slow path for multi-Z, off-map, source-Z mismatch, and non-default walkers.
- **`StepCache`** — singleton chunk store keyed by `(mapId, chunkX, chunkY)`. Lazily built on first query, invalidated by `Sector.MultisVersion` mismatch, memory-bounded by sampled probabilistic LRU.
- **`StepProbe`** — computes static-only walkability for a single cell, mirroring `MovementImpl.Check` minus the item / mobile collision phases.
- **`StepMask` / `StepChunk`** — value / storage types for the per-cell results.
- **`CacheEvictionTimer`** — periodic cap backstop (60s interval; early-returns when not over cap).
- **`Map.Sector.MultisVersion`** promoted to `public` so the cache can detect dynamic-static invalidations cheaply.
## Eviction strategy
Sampled probabilistic LRU (Redis-style). Per eviction, sample 5 random keys from a parallel `List<long>` kept in lockstep with the chunk dictionary; evict the oldest of the sample via swap-and-pop. O(1) per eviction regardless of resident count, so sustained cap pressure has no perpetual perf hit.
## Capability handling (interim)
Non-default walkers (non-GM players, creatures with `CanSwim` / `CanFly` / `CanOpenDoors` / `CanMoveOverObstacles`) route entirely through the per-cell slow path via `BitmapAStarAlgorithm.GetSuccessorsSlowPath`. The 2-pass design (cache + capability overlay + dynamic-obstacle pass) lands in the follow-up PR.
Kamron Batman
