ВЕРСИИ (19)

## [1.8.0] — 2026-06-01 — Code Audit: Complete Bug-Fix & Refactor Release Full-source audit pass following v1.7.1. Every bug identified during the review has been fixed in this release. Two partially-finished refactors from the 1.7.x cycle have been completed, and dead code has been removed.

## [1.7.1] — 2026-05-18 — Bug Fix Release This release corrects five bugs identified in a full source-code audit of 1.7.0. Three of them were **silent correctness failures** that made advertised features behave differently from their documented intent without producing any visible error. ### 🔴 Critical Fixes #### Fixed — `LagEngine` — Safe Mode FOLLOW_RANGE silently overwritten every tick - **Root cause:** After `SafeModeOptimizer.applySafeMode()` set `FOLLOW_RANGE = 4.0`, the lambda in `applyAiBatched()` fell through to an unconditional `attr.setBaseValue(attr.getDefaultValue() * multiplier)` at the bottom of the loop. In CRITICAL state `multiplier = 1.0`, so the attribute was reset to its default value on every single tick — completely negating the FOLLOW_RANGE reduction that Safe Mode is supposed to maintain. - **Fix:** Added `continue` immediately after `safeModeOptimizer.applySafeMode(mob)` to skip the unconditional `setBaseValue` line, matching the existing pattern used by the pathfinding-distance branch above it. - **Impact:** Safe Mode now actually reduces mob target-scanning range during CRITICAL state as documented. Servers using `safe-mode: true` will see measurably lower entity processing cost during lag spikes. #### Fixed — `SafeModeOptimizer` — `NullPointerException` crash on certain entity types - **Root cause:** `disableExpensiveAI()` and the previous `restoreAI()` called `mob.getAttribute(FOLLOW_RANGE).setBaseValue(...)` without a null-check. `getAttribute()` returns `null` for entity types that do not expose `FOLLOW_RANGE` (e.g. Slime, Bat, some modded mobs), causing an unhandled `NullPointerException` that bubbled up through the scheduler and printed a stack trace every tick. - **Fix:** Extracted the result of `getAttribute()` into a local variable and guarded both call sites with `if (attr != null)`. #### Fixed — `SafeModeOptimizer` — `restoreAI()` used a hardcoded FOLLOW_RANGE table - **Root cause:** The previous implementation maintained an internal `switch` table mapping `EntityType → double` for the restore path. This table was incomplete (missing dozens of types) and ignored any custom attributes set by other plugins or datapacks, silently restoring the wrong range. - **Fix:** Replaced the hardcoded table with `attr.getDefaultValue()`, which reads the actual server-side default and therefore works correctly for all entity types including custom ones. ### 🟡 Medium Fixes #### Fixed — `LagEngine` — `applyDifficultyScaling()` had HARD and EASY inverted - **Root cause:** `case HARD → tps + 1.0` raised the effective TPS, causing optimizations to trigger *later* on Hard worlds. `case EASY → tps - 1.0` lowered it, triggering optimizations *earlier* on Easy worlds. Both are the opposite of what the logic requires: Hard mobs pathfind more aggressively and consume more CPU; optimizations should therefore activate sooner, not later. - **Fix:** Swapped the offsets — `EASY → tps + 1.0`, `HARD → tps - 1.0`. - **Impact:** Servers running Hard difficulty worlds will now have mob AI throttled earlier under load, meaningfully reducing the lag window that was previously unaddressed. #### Fixed — `WebDashboard` — Rate-limit map memory leak on long-running servers - **Root cause:** `rateLimitCounters` and `rateLimitWindows` only cleared on `stop()`. Each unique IP that ever touched the dashboard created a permanent entry. On servers with public web dashboards, scanner/probe traffic could accumulate thousands of entries over days of uptime without any mechanism to reclaim the memory. - **Fix:** At the start of each `checkRateLimit()` call, entries whose 60-second window has already expired are pruned via `removeIf`, and `rateLimitCounters` is kept in sync via `retainAll`. Cost is amortised O(n) across all incoming requests — negligible compared to the HTTP overhead of each call. ### 🛠️ Code Quality #### Documented — `LagEngine` — `AIThrottler` field is intentionally unused - The 1.7.0 refactor extracted `AIThrottler` as a separate class but did not update the `applyAiBatched()` call site to delegate to it. The inline throttle logic continued to run using `LagEngine`'s own `throttleCounters`, leaving `AIThrottler` as unreachable code with its own duplicate counter map. - This release adds an explicit `NOTE` comment on the field explaining why the field exists, why it is not called, and what is required before delegation can be completed safely. No behaviour change — cleanup is deferred to a dedicated refactor release. ---

## [1.7.0] — 2026-05-01 — Major Architecture & Algorithm Improvements ### 🚀 Major Features - **Safe Mode for Critical State** - Added `safe-mode` configuration option under critical TPS tier - Implemented `SafeModeOptimizer` to selectively disable dangerous AI components while preserving essential game mechanics - Prevents over-optimization that breaks villager trading, mob escaping water, and other critical behaviors - **LagEngine Refactoring** - Extracted optimization logic into dedicated classes following Single Responsibility Principle: - `MobCuller` - Entity culling with intelligent importance scoring - `AIThrottler` - AI throttling and tick management - `SpawnSuppressor` - Spawn rate adjustment logic - Improved code maintainability and testability - **Event-Driven Architecture** - Implemented EventBus pattern for decoupled optimization logic: - `OptimizationEventBus` - Central event dispatcher - `OptimizationEvent` - Event data container - `OptimizationListener` - Interface for event handlers - Enables plugins/addons to react to optimization state changes - **Mob Importance Scoring System** - Created `MobImportanceScorer` with comprehensive importance evaluation: - Base scores for all entity types (bosses: 1000, villagers: 100, hostile: 10-20, etc.) - Dynamic scoring factors: named mobs (10x), tamed mobs (20x), equipment (1.5x) - Integration with MobCuller for intelligent culling decisions - **Mob Behavior Groups** - Implemented `MobBehaviorGroups` system for targeted optimization: - PASSIVE (low priority): Cows, pigs, sheep, chickens - AGGRESSIVE (high priority): Zombies, skeletons, creepers - NEUTRAL (medium priority): Endermen, iron golems - SPECIAL (preserve): Villagers, bosses, named mobs - AQUATIC (water-based): Dolphins, guardians, squids ### 🧠 Algorithm Improvements - **RegionOptimizer Enhancements** - **Dynamic Region Sizing**: Automatically adjusts region size based on player count - 0-5 players: 16 chunks (large regions) - 6-20 players: 8 chunks (medium regions) - 21-50 players: 4 chunks (small regions) - 50+ players: 2 chunks (very small regions) - **Hot Region Decay**: Regions gradually cool down when players leave - Heat decays at 5% per second when no players present - Prevents regions from staying hot indefinitely - **PredictiveScheduler Algorithms** - **EWMA (Exponentially Weighted Moving Average)**: Better short-term TPS prediction - Alpha smoothing factor of 0.3 for optimal responsiveness - **Linear Regression**: Trend analysis for TPS decline detection - Combines simple average, EWMA, and trend for robust prediction - More accurate lag prevention than simple hourly averages ### 🛠️ Configuration Management - **Annotation-Based Validation** - `@ConfigValidation` annotation for field-level validation rules - `ConfigValidator` class with automatic validation on reload - Validates numeric ranges, required fields, and boundary warnings - Provides clear error messages and validation reports - **Config Migration System** - Automatic version checking and stepwise migrations - Backup creation before migration with rollback capability - Supports migration of config.yml, features.yml, monitors.yml, systems.yml - Preserves user settings while updating structure ### 📚 Documentation - **Comprehensive Config Guide** - `CONFIG-GUIDE.md` with detailed explanations for all settings - Usage examples, recommended values, and troubleshooting tips - Clear organization by functional areas - **Enhanced Config Comments** - Added detailed explanations to all YAML configuration files - Section headers and inline comments for better understanding - Security notes and setup instructions where applicable ### 🔧 Code Quality Improvements - **Single Responsibility Principle** - Each class has a single, well-defined responsibility - Reduced coupling between components - Improved testability and maintainability - **Modern Java Practices** - Updated deprecated Bukkit method usage - Proper exception handling and logging - Thread-safe data structures where appropriate ### 🎯 Performance Optimizations - **Intelligent Culling** - Prioritizes removing low-importance mobs first - Preserves named, tamed, and special mobs - Reduces impact on gameplay while maintaining server performance - **Targeted Optimization** - Behavior-based optimization strategies - Different approaches for different mob types - More efficient than one-size-fits-all approach ### 🔄 Backward Compatibility - **Seamless Migration** - Automatic config updates preserve existing functionality - Graceful handling of missing or corrupt configuration files - No breaking changes to existing setups ### 🐛 Bug Fixes - **Fixed Memory Leaks** - Proper cleanup of event listeners and cached data - Thread-safe collections with proper synchronization - **Improved Error Handling** - Better exception messages and logging - Graceful degradation when components fail - Validation prevents invalid configurations ---

## [1.5.2] — 2026-04-23 — Bug Fix & Safety Patch ### Fixed — `plugin.yml` - **Critical: wrong `api-version`** — `api-version` was set to `1.5.1` (the plugin version) instead of the required Bukkit API version `1.21`. Paper rejects plugins with a non-standard `api-version` string, causing the plugin to either fail to load or log a persistent warning on every server start. Changed to `api-version: 1.21`. ### Fixed — `WebDashboard` - **Token not generated when `monitors.yml` is deleted or web dashboard is disabled.** `ensureStrongToken()` was only called after the `if (!webDashboardEnabled) return` guard, meaning servers that had the dashboard disabled (or whose `monitors.yml` was deleted and recreated by `saveResource`) would still have the default placeholder token the first time the dashboard was enabled — a silent security hole. - `ensureStrongToken()` now runs **before** the enabled check on every `start()` call. - If `monitors.yml` does not yet exist, `saveResource()` is called first to create the file structure before the generated token is written into it. ### Fixed — `LagEngine` — Culling count wrong when protected zones are present - **Protected-zone mobs were counted toward `targetRemoveCount` but never actually removed**, causing the cull pass to under-remove and leaving entity counts above `maxEntities` indefinitely on servers with active zone protection. - `cull()` now pre-filters protected mobs from the candidate pool *before* computing `targetRemoveCount`, so the removal math uses only the actually-cullable mob count. - The redundant zone check inside the removal loop is kept as a race-condition safety net. ### Fixed — `LagEngine` — AI update scheduler saturation on very mob-dense worlds - **No upper bound on `runTaskLater` calls per tick** — on worlds with thousands of entities, `applyAiBatched()` could schedule dozens of batch tasks in a single tick, queuing more work than the scheduler could drain, leading to compounding latency. - Added `AI_UPDATE_PER_TICK_CAP = 80`: at most 80 mob AI updates are scheduled per engine tick. The next monitor tick processes the remaining mobs, spreading load evenly. ### Fixed — `StressTestManager` — Server crash on large spawns at low-chunk-count locations - **No per-chunk mob density check** — spawning 2 000+ mobs at or near the world spawn could saturate loaded chunks and crash the server within seconds. - Before spawning, the manager now scans a 7×7 chunk area around the target location and compares the current entity count against `stress-test.max-mobs-per-chunk` (default: 80). - If the limit would be exceeded, the spawn count is automatically reduced to the safe maximum and a warning is logged. If the area is already at capacity, the command is rejected with a descriptive error message. - New config key in `systems.yml`: `stress-test.max-mobs-per-chunk: 80`. ### Fixed — `MultiServerSync` — MySQL reconnect attempt logged on every polling cycle - When the MySQL database was down for an extended period, `ensureConnected()` logged `"MySQL connection lost — reconnecting in Xs…"` on every poll interval (default every 10 s), flooding the console with hundreds of lines. - Reconnect log messages are now gated behind the same **power-of-2 streak filter** already used for SQL error warnings (logs on streak 1, 2, 4, 8, 16 …), reducing noise by up to 95% during prolonged outages while still keeping the first occurrence visible. ### Config changes ```yaml # systems.yml — new in 1.5.2 stress-test: max-mobs-per-chunk: 80 # NEW — per-chunk density safety cap for stress tests ``` ---

## [1.5.1] — 2026-04-20 — Code Quality & Performance Patch ### Fixed / Improved — ZoneManager - **WorldGuard region cache** — `getApplicableRegions()` was called for every mob on every monitor tick. Queries are now cached per chunk with a 100-tick TTL (`ConcurrentHashMap`). Cache is invalidated automatically when TTL expires, eliminating excessive WorldGuard API pressure on large servers. - **CuboidZone coords changed from `double` to `int`** — block-level precision is sufficient; int arithmetic is faster and the record is more memory-efficient. - Zone `initialize()` now clears the WG cache on reload. ### Fixed / Improved — ProfileManager - **Profile validation before apply** — switching to a profile where `critical-tps ≥ medium-tps` or `max-entities < 1` now returns a clear error instead of silently corrupting the engine state. - **Active profile persisted across restarts** — the selected profile name is written to `plugins/NeuroLag/active-profile.dat` on switch and reloaded automatically on startup/reload, so profiles survive server restarts without manual re-selection. - `switchProfile()` return type changed from `boolean` to `String` (null = success, `"NOT_FOUND"` or error message = failure) — NeuroLag main command updated accordingly. ### Fixed / Improved — PredictiveScheduler - **Hand-rolled JSON parser replaced with Gson** (bundled by Paper) — the previous `split` / `replaceAll` parser was fragile and could silently produce wrong data on corrupt files. Gson provides safe serialisation and clean error handling; corrupt files now log a warning instead of producing garbage hourly averages. - Loaded samples are capped to the last 60 per hour on load, not only on record. ### Fixed / Improved — CpuMonitor - **Graceful fallback for non-Sun JVMs** — if `com.sun.management.OperatingSystemMXBean` is unavailable (GraalVM, some container JVMs), the monitor now falls back to `OperatingSystemMXBean.getSystemLoadAverage()` normalised by CPU count. If even that is unavailable, the feature disables itself gracefully instead of throwing at construction time. - **EMA smoothing** (α = 0.3) — single-tick CPU spikes no longer toggle the throttle on/off erratically. The exponential moving average keeps the reading stable under transient load. - Strategy (SUN_PROCESS / SYSTEM_LOAD / DISABLED) is selected once at construction and stored in an enum — no repeated `instanceof` checks every tick. ### Fixed / Improved — RegionOptimizer - **Player chunk position cache** — `refresh()` now tracks each player's last known chunk. The HOT/COLD region map is only rebuilt when at least one player has moved to a different chunk since the previous call. On a stable server this eliminates the full-player-scan every monitor tick. - `isBeyondPathfindingDistance()` uses the cached positions instead of calling `world.getPlayers()` a second time per mob check. ### Fixed / Improved — NeuroLagAPI - Added `NeuroLagAPI.getInstance()` for cleaner third-party plugin integration. - JSON payload sent via plugin message channel now uses Gson instead of manual StringBuilder, eliminating potential escape bugs. ### Fixed / Improved — LagReporter (Discord) - **Retry with exponential back-off** — Discord webhook requests are retried up to 3 times (delays: 1 s → 2 s → 4 s) before giving up. Transient network errors and Discord 429 rate-limit responses no longer silently drop notifications. - Discord embed payload now built with Gson — no more manual string escaping. ### Fixed / Improved — StressTestManager - **Confirmation prompt** for spawns > 1 000 mobs — the sender must repeat the command within 30 seconds to confirm. Prevents accidental large spawns. - **Multiple entity types** — `stress-test.entity-types` config list (default: ZOMBIE, SKELETON, CREEPER) is cycled round-robin across spawned mobs, producing a more realistic mixed load. Invalid or non-spawnable type names log a warning and are skipped. ### Fixed / Improved — LagEngine - Added internal processing metrics: `lastTickMobCount` and `lastTickProcessingMs`. Visible in `/nlag status` ("Last Tick: N mobs, X ms"). - `ZoneManager.tick()` is now called once per engine tick to advance the WG cache TTL counter. ### Fixed / Improved — WebDashboard - Rate limiting now also covers `GET /` (the HTML dashboard page), not only `/api/*` endpoints. - Token auto-generation log output upgraded to `SEVERE` level and formatted as a clearly visible bordered block so admins cannot miss the new token in the console. ### Fixed / Improved — ConfigManager / ConfigValidator - Added `stress-test.entity-types` list field. - `ConfigValidator` now checks `entity-types` list is non-empty when stress test is enabled. ### Config changes (`systems.yml`) ```yaml stress-test: entity-types: # NEW — 1.5.1 - ZOMBIE - SKELETON - CREEPER ``` ---

## [1.5.0] — 2026-04-19 — Security Hardening & Stability Release ### Fixed — Security - **[Bug #1] Web dashboard weak default token / no rate limiting / no IP filtering** (`WebDashboard.java`, `ConfigManager.java`, `monitors.yml`) - Auto-generates a cryptographically strong 24-byte random token on first startup whenever the default placeholder `"change-this-token-now"` is detected. The new token is immediately persisted to `monitors.yml` and printed to the console. - Added per-IP rate limiting (sliding 60-second window, configurable via `web-dashboard.rate-limit.max-requests-per-minute`, default 60). Returns HTTP 429 when the limit is exceeded. - Added optional IP allow-list (`web-dashboard.ip-whitelist`, disabled by default). When enabled, only explicitly listed IPs can reach `/api/status` and `/api/cmd`. ### Fixed — Stability - **[Bug #2] Stress test could spawn unlimited mobs with no cooldown** (`StressTestManager.java`, `ConfigManager.java`, `systems.yml`) - Hard cap: mob count is clamped to `stress-test.max-mob-count` (default 2 000) regardless of the value passed to `/nlag stresstest`. The sender is notified when the cap applies. - Cooldown: a configurable `stress-test.cooldown-seconds` (default 300) must elapse between tests. Attempting to start a test during cooldown shows the remaining time. - Both limits are enforced by `ConfigValidator`. - **[Bug #3] Redis subscriber join(3000) always timed out on slow networks** (`MultiServerSync.java`) - `stop()` now holds a `volatile` reference to the active `JedisPubSub` instance and calls `pubSub.unsubscribe()` *before* `interrupt()+join()`. This signals the blocking `jedis.subscribe()` call to return immediately, so `join()` completes in milliseconds rather than timing out. Prevents connection-pool leaks during rapid `/nlag reload` cycles. - **[Bug #4] Smart culling could remove hundreds of entities in one pass → lag spike** (`LagEngine.java`) - Added `CULL_PER_TICK_CAP = 50` — at most 50 entities are removed per cull invocation. If the entity count remains above `max-entities-per-world` after one pass, the next monitor tick handles the remainder. Eliminates the single-frame mob-removal lag spike seen on servers with thousands of entities. - **[Bug #5] Predictive scheduler lost all historical data on server restart** (`PredictiveScheduler.java`) - Hourly TPS samples are now saved to `plugins/NeuroLag/predictive-data.json` on `stop()` (server shutdown or `/nlag reload`) and reloaded on `start()`. No external library required — uses a compact hand-rolled JSON serializer/parser. The predictor now accumulates knowledge across restarts and becomes effective much faster on busy servers. - **[Bug #6] Full plugin reload caused noticeable server stutter** (`NeuroLag.java`) - `reloadPluginState()` now takes config snapshots before reloading and selectively restarts only the services whose configuration sections actually changed (web dashboard, multi-server sync, predictive scheduler, CPU monitor, boss-bar dashboard). The main monitor task and engine throttle are always restarted cleanly. On a typical `/nlag reload` that only changes TPS thresholds, zero external services are bounced, eliminating the stutter caused by restarting Redis/MySQL connections unnecessarily. ### Changed - `ConfigValidator` now validates the new stress-test limits and rate-limit settings. - `systems.yml` — added `stress-test.max-mob-count` and `stress-test.cooldown-seconds`. - `monitors.yml` — added `web-dashboard.rate-limit` and `web-dashboard.ip-whitelist` sections. ---

## [1.4.1] — 2026-04-18 — Bug Fix Release ### Fixed — Critical - **Task leak on reload (`/nlag reload`)** — `BackupManager.start()` was called twice in `onEnable()` (once directly, once via `startRuntimeServices()`), causing every scheduled task to be registered a second time on reload. The duplicate direct call has been removed; the initial backup is now triggered asynchronously inside `start()` when no backups exist. `stopRuntimeServices()` now also calls `engine.cancelAllPendingBatchTasks()` before stopping the throttle task, ensuring all in-flight `runTaskLater` batch tasks are cancelled before any new tasks are registered. (`LagEngine`, `NeuroLag`) - **`collectTargets()` running every tick on the main thread** — `LagEngine.collectTargets()` called `world.getLivingEntities()` on every monitor tick, causing significant lag spikes on servers with large mob populations. The result is now cached per-world and rebuilt at most once every 4 ticks (~0.2 s). The cache is invalidated on `restoreAll()` and on world state transitions. (`LagEngine`) - **FOLLOW_RANGE not fully restored for protected-zone mobs** — `applyAiBatched()` set `AI = true` for protected mobs but did not always restore `FOLLOW_RANGE` to its default value, so mobs could permanently retain a reduced follow range after leaving a zone. The fix unconditionally calls `attr.setBaseValue(attr.getDefaultValue())` for every protected mob encountered. Additionally, a new `restoreWorld(World)` helper is called whenever a world transitions back to `NORMAL`, guaranteeing a full attribute restore even for mobs that were never re-processed by the batch scheduler. (`LagEngine`) - **Manual override (`/nlag toggle`) not restoring mobs correctly** — in-flight `runTaskLater` batch tasks scheduled by `applyAiBatched()` could execute *after* `restoreAll()` and re-apply AI restrictions or reduced follow ranges, leaving mobs frozen. `handleToggle()` now calls `engine.cancelAllPendingBatchTasks()` before `engine.restoreAll()` so no pending task survives the toggle. (`LagEngine`, `NeuroLag`) ### Fixed — High - **WebDashboard query-string token auth** — passing `?token=...` in the URL is insecure (tokens appear in server access logs, browser history, and HTTP referrer headers). The query-string fallback is now **disabled by default**. A new config key `web-dashboard.auth.allow-query-token` (default `false`) must be explicitly set to `true` to re-enable it. The existing warning is preserved and now also states the config key needed to disable the fallback permanently. (`WebDashboard`, `ConfigManager`, `monitors.yml`) - **Redis subscriber thread not shutting down gracefully** — `MultiServerSync.stop()` called `redisSubThread.interrupt()` but returned immediately, so the thread could still be holding a Jedis resource when `jedisPool.close()` ran, causing pool-exhaustion exceptions in the server log on every reload. The fix adds `redisSubThread.join(3000)` after `interrupt()`, giving the thread up to 3 seconds to exit cleanly before the pool is closed. (`MultiServerSync`) - **MySQL reconnect logic throwing repeatedly when DB is down** — `ensureConnected()` called `DriverManager.getConnection()` immediately on every sync cycle failure, flooding the log with stack traces. The fix introduces an atomic error-streak counter and an exponential back-off (sleep time = `min(streak * 2, 30)` seconds before reconnect). Log output now uses a power-of-two gating strategy (log on streak 1, 2, 4, 8, 16, …) to avoid noise during prolonged outages. (`MultiServerSync`) - **Batched AI task delay not capped** — with 5 000+ mobs the delay between batches could reach 100 ticks (5 seconds), making mobs visibly unresponsive. The delay is now capped at 20 ticks (1 second) regardless of entity count. (`LagEngine`) - **Smart culling could remove protected-zone mobs** — the culling loop checked `zoneManager.isProtected()` only in the main optimization path, not inside `cull()` itself. Mobs inside protected zones can now never be removed by smart culling. (`LagEngine`) - **BackupManager not backing up the `lang/` directory** — the ZIP bundle only included root-level `.yml` files. Custom language files stored in `plugins/NeuroLag/lang/` would be lost on restore. `collectConfigFiles()` now recursively includes `lang/*.yml` entries and `restoreZip()` correctly recreates the `lang/` subdirectory on restore. (`BackupManager`) ---

## [1.4.0] — Stability + lifecycle fix release ### Fixed - Removed the command-level permission wall in `plugin.yml` so feature-specific permissions can actually work. - Reworked `/nlag toggle` so manual override now restores mobs immediately and stops active throttling instead of leaving old optimization state hanging. - Reworked `/nlag reload` to stop and restart runtime services cleanly, including CPU monitor, predictive scheduler, web dashboard, backup scheduler, multi-server sync, and monitor tasks. - Multi-server sync now clears stale peers and restarts cleanly on reload. - Protected-zone mobs are now excluded from smart culling and have `FOLLOW_RANGE` restored to defaults when encountered during optimization. - Ongoing `MEDIUM` and `CRITICAL` states now keep reapplying optimization passes on later monitor ticks, so movement into protected zones / hot zones / player zones is handled more consistently. - Stress-test mobs no longer give themselves a custom name, so they are no longer accidentally excluded by the named-mob whitelist. - Backup system now stores full zip bundles of NeuroLag YAML files instead of only copying `config.yml`. - Added restore-by-name support for `/nlag backup restore <file>`. - Memory-pressure config now normalizes legacy percentage values above `1.0`, and defaults were corrected to ratio format (`0.80`). - Player-scaling ratio is now clamped to avoid negative threshold offsets when online players are below the configured minimum. - Predictive scheduler now clears parsed peak-hour ranges on restart/reload to avoid duplicate entries. - Zone manager now fully resets cached zone and WorldGuard state when the feature is disabled or reloaded. - Web/API JSON generation now escapes keys/values properly and uses stable decimal formatting. - Web auth query token parsing is stricter and only logs the fallback warning once. - CPU monitor now resets stale displayed values when stopped or disabled. - BossBar dashboard now respects `dashboard.bossbar.name` and `dashboard.bossbar.color` from `monitors.yml`. - Metrics API plugin-message registration now reads the correct split config source. ### Docs - Updated README to reflect the real split-config layout. - Updated README examples for `web-dashboard`, `multi-server`, `zone-protection`, `config-profiles`, and backup restore. - Updated language file version headers to `v1.4.0`.

## [1.3.1] — Bug Fix Release ### 🐛 Fixed - **[CRITICAL] FOLLOW_RANGE stuck at 0 after mob returns near player** Mobs whose `FOLLOW_RANGE` was zeroed by the pathfinding-distance optimiser were never restored when they re-entered a player's chunk radius. They would remain AI-enabled but permanently unable to pathfind, appearing frozen. Fix: restore `FOLLOW_RANGE` to `defaultValue × multiplier` inside the `nearPlayer || hotRegion` branch of `applyAiBatched()`. - **[CRITICAL] ConcurrentModificationException on `tpsHistory` between main thread and HTTP thread** `tpsHistory` was an `ArrayDeque` — not thread-safe. The Bukkit main thread writes to it every tick while the Web Dashboard HTTP thread pool iterates it on every `/api/status` request, causing intermittent exceptions. Fix: replaced with `ConcurrentLinkedDeque`. Added `volatile` to `tpsSum`, `tpsCount`, `criticalActivations`, and `mediumActivations` for JMM visibility. - **[CRITICAL] MySQL connection silently dies after DB restart or idle timeout** A single `Connection` was reused indefinitely. When MySQL restarted or timed out, all subsequent sync calls threw `SQLException` silently, leaving multi-server sync permanently broken with no visible warning. Fix: added `ensureConnected()` that checks `isClosed()` and reconnects automatically before each `upsertMysql()` / `readPeersMysql()` call. - **[MEDIUM] autoRestoreOnError never works on first plugin startup** The backup-restore safety net checked for existing backups before any had been created, so it could never restore on a fresh install. Fix: `doBackup()` is now called immediately at startup before the check. - **[MEDIUM] StressTestManager.finish() could run twice, emitting duplicate audit reports** If the Bukkit scheduler fired the deferred `finish()` in the same tick as an early `stop()` call, both executions would complete and emit an audit report. Fix: `finish()` returns immediately if `running` is already `false`. - **[MEDIUM] Web API /api/cmd whitelist bypassable via prefix matching** `body.startsWith(a)` allowed inputs like `"reloadEvil"` to pass the filter. Fix: the first token is extracted and matched exactly against the allowlist. ### 🔒 Security - **Web auth token warned when passed via URL query param** `?token=xxx` leaks into browser history and proxy logs. `Authorization: Bearer` header is now the primary path; query-param auth logs a one-time warning. ### ⚡ Performance - **Redis publisher now uses JedisPool** A fresh TCP connection was opened and closed every 5 seconds. A bounded `JedisPool` (max 8 / idle 2) is created at startup and shared by publisher and subscriber, eliminating per-publish connection overhead. ---

## [1.3.0] — 2026-03-15 ### 🐛 Bug Fixes (9 fixes) #### BUG-A — ZoneManager: world name dropped from cuboid zones (cross-world contamination) - **Root cause:** `ZoneManager.initialize()` parsed zone entries as `"world x1 y1 z1 x2 y2 z2"` (7 parts) but stored only `parts[1]..parts[6]` in a raw `double[]`, silently discarding `parts[0]` (the world name). - **Effect:** A zone defined in world `survival` would also protect mobs at the same coordinates in `world_nether`, `world_the_end`, and any other loaded world. Mobs in unintended worlds were immune to optimization. - **Fix:** Replaced `List<double[]>` with `List<CuboidZone>` where `CuboidZone` is a Java `record` holding `String world` + 6 coordinates. `isProtected()` now checks `mob.getWorld().getName().equals(z.world())` before testing coordinates. #### BUG-B — RegionOptimizer: negative chunk coordinates mapped to wrong region - **Root cause:** Region index was computed as `chunkX / regionSize` using Java integer division, which truncates toward zero. Chunk `-1 / 16 = 0` and chunk `1 / 16 = 0`, so chunks in the negative quadrant near spawn mapped to the same region keys as their positive counterparts. - **Effect:** The HOT region around coordinate origin (0,0) was effectively double-sized. Mobs at negative chunk coordinates near spawn were optimized even when a player was present nearby. - **Fix:** Replaced `/` with `Math.floorDiv()` in both `refresh()` and `isHot()`. `Math.floorDiv(-1, 16) = -1` (correct), `Math.floorDiv(15, 16) = 0` (correct). #### BUG-C — DashboardManager: `type: HOLOGRAM` silently did nothing - **Root cause:** `DashboardManager.start()` returned early with `if (!dashboardType.equals("BOSSBAR")) return` — no log message, no error. Admins who set `dashboard.type: HOLOGRAM` saw no dashboard and received no indication why. - **Fix:** Added an explicit `plugin.getLogger().warning()` message when a non-BOSSBAR type is configured, clearly stating that only BOSSBAR is currently supported. #### BUG-D — BackupManager: `creationTime()` unreliable on Linux (ext4) - **Root cause:** Backup rotation used `Files.readAttributes(...).creationTime()` to sort backups oldest-first. The ext4 filesystem does not store inode creation time — it returns `Instant.EPOCH` (1970-01-01) for all files, making sort order undefined. - **Effect:** On Linux servers, the "oldest" backup was selected randomly. The wrong backup could be deleted or restored. - **Fix:** Replaced `creationTime()` with `Files.getLastModifiedTime()`, which is reliably stored on all filesystems. Also removed the now-unused `BasicFileAttributes` import. #### BUG-E — StressTestManager: always spawned at `world[0]` spawn, ignoring admin's location - **Root cause:** `start()` hardcoded `Bukkit.getWorlds().get(0).getSpawnLocation()` as the spawn center regardless of where the admin was or which world they were in. - **Effect:** Running `/nlag stresstest` from the Nether or End spawned mobs in the Overworld at spawn, making the test meaningless for other worlds. - **Fix:** `start()` now accepts a `CommandSender` parameter. If the sender is a `Player`, mobs spawn centered on the player's current location in their current world. Console fallback remains `world[0]` spawn. #### BUG-F — WebDashboard: POST `/api/cmd` blocked by CORS preflight (OPTIONS) - **Root cause:** Browsers send an HTTP `OPTIONS` preflight request before a cross-origin `POST`. The server had no handler for `OPTIONS` and returned `405 Method Not Allowed`, causing the browser to abort the actual `POST`. - **Effect:** The "send command" button in external web dashboards or REST clients always failed silently. - **Fix:** Added `sendCorsOptions()` helper that returns `204 No Content` with `Access-Control-Allow-Methods: GET, POST, OPTIONS` and `Access-Control-Allow-Headers: Authorization, Content-Type`. Both `/api/status` and `/api/cmd` now handle `OPTIONS` before auth checks. #### BUG-G — AuditManager: `getLivingEntities()` / `getLoadedChunks()` called on main thread - **Root cause:** `snapshot()` iterated all entities and chunks synchronously on the main thread before handing off to async write. - **Effect:** On servers with many mobs or chunks, this could stall the main thread for several milliseconds during audit generation. - **Mitigation:** Added an explicit code comment documenting this limitation and the reason it is required by the Bukkit API. A full async snapshot would require a thread-safe entity iterator not currently available in Paper's public API. #### BUG-H — ProfileManager: `clearProfile()` required a full `cfg.reload()` to restore values - **Root cause:** `switchProfile()` overwrote `cfg.tpsCritical`, `cfg.tpsMedium`, and `cfg.maxEntities` directly. `clearProfile()` only set `activeProfile = null` and left a comment saying "caller is responsible for reloading", which `NeuroLag.java` implemented by calling `cfg.reload() + engine.restoreAll() + startMonitor()` — a heavy operation that re-read disk and restarted all tasks. - **Fix:** `ProfileManager` now saves the three base values in `saveBaseValues()` (called in the constructor and after every `cfg.reload()`). `clearProfile()` restores them directly — no reload required. `/nlag profile clear` no longer triggers a full config reload cycle. #### BUG-I — LagEngine: "Animation Freezing" comment/name mismatched actual behavior - **Root cause:** Feature F4 was named "Animation Freezing" throughout the codebase, but the implementation only called `mob.setVelocity(0,0,0)`. This stops movement drift but does not freeze animation frames — mobs still visibly idle-animate. - **Fix:** Renamed the in-code comment to "Velocity Freeze" and added a note explaining that true animation-frame control requires NMS/packet-level access not available through the standard Bukkit API. ### 🔧 Other Changes - `plugin.yml`: version bumped to 1.3.0; permission descriptions clarified; `neurolag.stresstest` description now mentions mobs spawn at your location - `pom.xml`: version bumped to 1.3.0 - `StressTestManager.start()` signature changed to `start(int count, int durationMinutes, CommandSender sender)` — callers updated in `NeuroLag.java` - `ProfileManager.saveBaseValues()` is now public so `NeuroLag.java` can call it after `/nlag reload` ---

## [1.2.0] — 2026-03-14 ### 🎉 All 18 Advertised Features — Now Fully Implemented Every feature listed in the README and CHANGELOG since v1.1.0 now has real working code. #### New Java Classes (15 new files) | Class | Feature | |---|---| | `CpuMonitor.java` | F1 — CPU throttling via `com.sun.management.OperatingSystemMXBean` | | `RegionOptimizer.java` | F2 — 16×16 chunk region grid, HOT/COLD classification | | `MobWeightManager.java` | F3 — Per-type weight budget per chunk | | `DashboardManager.java` | F6 — Live BossBar (BLUE→YELLOW→RED) for all operators | | `AuditManager.java` | F7 — Full HTML audit report (specs, plugins, mobs, chunks) | | `MultiServerSync.java` | F8 — Redis Jedis pub/sub + MySQL JDBC upsert | | `WebDashboard.java` | F10 — Embedded HTTP server (JDK `HttpServer`), live JS chart, JSON API | | `PredictiveScheduler.java` | F11 — Per-hour TPS pattern tracking + peak-hour config | | `ZoneManager.java` | F12 — Cuboid zone list + WorldGuard region soft-depend | | `ProfileManager.java` | F14 — Instant profile switching (critical-tps / medium-tps / max-entities) | | `AlertManager.java` | F15 — Sound + particle + chat alerts for CRITICAL enter/leave | | `StressTestManager.java` | F16 — Mob spawning stress test with auto-cleanup and report | | `ConfigValidator.java` | F17 — Range/sanity checks for all config fields | | `BackupManager.java` | F18 — Periodic backup + purge + restore | | `NeuroLagAPI.java` | F5 — Static public API + plugin messaging channel `neurolag:metrics` | #### Features Integrated into Existing Classes | Feature | Location | |---|---| | F4 — Animation Freezing (velocity zeroing for distant AI-disabled mobs) | `LagEngine.applyAiBatched()` | | F9 — AI Difficulty Scaling (`getDifficulty()` → TPS offset) | `LagEngine.applyDifficultyScaling()` | | F13 — Smart Group Culling (type-aware, min-count guard, POOF particles) | `LagEngine.cull()` | | F1 TPS penalty integration | `LagEngine.tick()` | | F2 region check in optimization path | `LagEngine.applyAiBatched()` | | F3 weight-sorted culling | `LagEngine.cull()` | | F7 audit on critical event | `LagReporter.onCriticalStart()` | | F12 zone protection skip | `LagEngine.applyAiBatched()` | | F11 predictive penalty | `LagEngine.tick()` | | F15 alert on state change | `LagEngine.updateWorld()` | #### New Commands - `/nlag dashboard [off]` — toggle BossBar - `/nlag audit` — generate HTML audit report - `/nlag zone` — show loaded zone count - `/nlag profile [name|clear]` — switch config profile - `/nlag validate` — validate config with detailed errors/warnings - `/nlag backup [list|restore]` — manage backups - `/nlag stresstest [count] [minutes]` — run stress test - `/nlag sync` — show Redis/MySQL peer server states #### pom.xml Changes - Jedis 5.0.0 (Redis) — shaded and relocated - mysql-connector-j 8.3.0 — shaded and relocated - WorldGuard 7.0.12 — provided/optional (soft-depend) - Maven Shade plugin configured with relocation rules #### plugin.yml Changes - `softdepend: [WorldGuard]` added #### config.yml Changes - `dashboard.enabled` and `dashboard.update-interval` added - `audit-reports.generate-on-critical` added - `multi-server.mysql.*` section added - `zone-protection.zones` list added - `config-profiles.profiles.*.medium-tps` key added - `alerts.enabled` top-level flag added - `stress-test.auto-generate-report` added #### Language Files - All 7 language files updated to v1.2.0 header - New keys: `dashboard-bar`, `alert-critical`, `alert-recovery`, `sync-no-peers`, `sync-peer` ---

## [1.1.0] — 2026-03-14 ### 🎉 Major Release: 18 New Advanced Features! ### ✨ Added #### Feature 1: CPU Throttling Mode - Monitors server CPU usage (not just TPS) - Activates automatic optimization when CPU > 80% - Reduces Bukkit task frequency during critical CPU load - Disables non-essential block physics updates - Configuration: `cpu-throttling.enabled` and `cpu-throttling.cpu-threshold` #### Feature 2: Region-Based Optimization - Splits world into 16×16 chunk regions - Disables pathfinding for mobs in far regions - Smart regional mob optimization instead of global - Mobs near players always keep full AI - Configuration: `region-optimization.enabled` and `region-optimization.use-regions` #### Feature 3: Custom Mob Tier Weights - Assign different "cost" to different mob types - Zombies cost 1.0, Endermen cost 2.0, Ender Dragons cost 4.0 - Prevents over-optimization (allows more weak mobs) - Configuration: `mob-weights.weights.*` and `mob-weights.max-weight-per-chunk` - **Result:** More realistic mob balance during lag #### Feature 4: Animation Freezing - In CRITICAL state, mobs stop playing animations (head turning, arm swinging) - Only applies to distant mobs (configurable) - Reduces visual strangeness of frozen mobs - Configuration: `animation-freezing.enabled` and `animation-freezing.freeze-distance-chunks` #### Feature 5: Real-Time Metrics API - Other plugins can query NeuroLag's current state - Methods: `neurolag.getWorldState()`, `neurolag.getTPS()`, `neurolag.getAffectedMobCount()` - Enable plugin ecosystem around NeuroLag - Configuration: `metrics-api.enabled` and `metrics-api.expose-via-plugin-message` #### Feature 6: In-Game Dashboard - Live server health dashboard (BOSSBAR or HOLOGRAM) - Changes color: BLUE (normal) → YELLOW (MEDIUM) → RED (CRITICAL) - Updates every 2 seconds - Configuration: `dashboard.type`, `dashboard.bossbar`, `dashboard.hologram` #### Feature 7: Detailed Lag Audit Reports - Generate comprehensive lag analysis reports - Include: server specs, plugin list, chunk stats, mob breakdown, player behavior, network stats - Export to HTML for easy reading - Configuration: `audit-reports.enabled` and `audit-reports.include` - **Result:** Deep understanding of what causes lag #### Feature 8: Multi-Server Sync (Bungeecord/Velocity) - Share TPS data across all servers - Redis or MySQL backend support - Cascade warnings to all servers if one is in CRITICAL - Load-balance: suggest players move to less-lagged server - Configuration: `multi-server.enabled`, `multi-server.backend`, `multi-server.sync-tps` #### Feature 9: AI Difficulty Scaling - Easy mode = aggressive optimization - Normal mode = balanced - Hard mode = minimal optimization (players want challenge) - Hardcore = zero optimization - Configuration: `ai-difficulty-scaling.enabled` and `ai-difficulty-scaling.scale-with-difficulty` - **Result:** Respects player expectations #### Feature 10: Mobile Web Dashboard - Access metrics from any browser (phone, tablet, PC) - Real-time TPS graph, mob count, heap usage - Send commands remotely (`/nlag reload`, `/nlag toggle`) - Token-based authentication - Configuration: `web-dashboard.enabled`, `web-dashboard.port`, `web-dashboard.auth.token` - **Result:** Monitor and control server on the go 📱 #### Feature 11: Predictive Lag Prevention - Learns hourly patterns (6pm = always lag, etc) - Pre-activates optimization before peak hours hit - Configuration: `predictive-prevention.enabled`, `predictive-prevention.peak-hours` - **Result:** Be proactive, not reactive #### Feature 12: Custom Zones (WorldGuard Integration) - Mark important regions as HIGH_PRIORITY via WorldGuard - Examples: boss arenas, spawn, player shops, farms - Full mob AI always kept in protected zones - Aggressive optimization elsewhere - Configuration: `zone-protection.enabled` and `zone-protection.integrate-worldguard` - **Result:** Protect what matters most #### Feature 13: Entity Grouping & Smart Removal - Remove mobs intelligently during culling - Never remove single mob types (won't inactivate farms) - Remove in groups (visually less obvious) - Particle effect on removal - Configuration: `smart-culling.enabled`, `smart-culling.cull-in-groups`, `smart-culling.min-type-before-cull` #### Feature 14: Configuration Profiles - Create multiple config profiles (survival, minigame, peaceful, etc.) - Switch instantly with `/nlag profile <n>` - Each profile can have different TPS triggers and entity limits - Configuration: `config-profiles.enabled` and `config-profiles.profiles.*` - **Result:** Perfect for multi-mode servers #### Feature 15: Audio & Visual Alerts - Sound alerts when entering/leaving CRITICAL - Particle warnings around admin - Chat notifications to all ops - Customizable sounds (default: bass drop for CRITICAL, pling for recovery) - Configuration: `alerts.enable-sound-alerts`, `alerts.alert-sounds.*`, `alerts.enable-particle-warnings` - **Result:** Immediate awareness without checking commands #### Feature 16: Stress Test Mode - Simulate heavy server load without real players - Spawn fake mobs (configurable count) - Test for X minutes and auto-generate report - Configuration: `stress-test.enabled`, `stress-test.spawn-fake-mobs`, `stress-test.fake-mob-count` - **Result:** Safe testing environment 🧪 #### Feature 17: Configuration Validator - `/nlag validate` checks config syntax and values - Warns about invalid settings - Suggests optimizations - Configuration: `config-validator.enabled` - **Result:** Prevent config errors before they cause problems #### Feature 18: Automatic Config Backup - Auto-backup config every X minutes (default: 60) - Keep last N backups (default: 10) - Auto-restore if invalid config detected - Configuration: `auto-backup.enabled`, `auto-backup.backup-interval-minutes`, `auto-backup.keep-backups` - **Result:** Never lose your config ### 🌍 Multilingual Support (Added) - **English** (EN) — Default - **Tiếng Việt** (VI) — Vietnamese - **Español** (ES) — Spanish - **Français** (FR) — French - **Deutsch** (DE) — German - **日本語** (JA) — Japanese - **中文** (ZH) — Chinese (Simplified) Set language in `config.yml`: `settings.language: EN` ### 📝 Documentation - Completely rewritten README with all 18 features documented - Multilingual documentation in README (7 languages) - Comprehensive CHANGELOG (this file) - Detailed config.yml with all features documented ### 🔧 Command Enhancements New commands added: - `/nlag dashboard` — Show in-game dashboard - `/nlag audit` — Generate audit report - `/nlag zone` — Zone management - `/nlag profile <n>` — Switch config profiles - `/nlag validate` — Validate configuration - `/nlag backup` — Manage config backups - `/nlag stresstest` — Run stress test ### 🔐 Permission Enhancements New permissions added: - `neurolag.stresstest` — Run stress tests - `neurolag.dashboard` — View dashboard - `neurolag.audit` — View audit reports - `neurolag.zone` — Manage zones - `neurolag.profile` — Switch profiles - `neurolag.validate` — Validate config - `neurolag.backup` — Manage backups - `neurolag.web` — Access web dashboard - `neurolag.api` — Access metrics API ### 📊 Configuration Changes - Added 18 new configuration sections (one per feature) - All features can be individually enabled/disabled - Backward compatible with v1.0.6 configs - Old config keys still work (will be migrated automatically) ### 🎯 Performance Improvements - CPU throttling reduces unnecessary task execution - Region-based optimization replaces global optimization - Smart culling reduces entity processing - Predictive prevention prevents lag spikes before they happen ### 🐛 Bug Fixes (from v1.0.6) - ✅ Tick-rate throttling properly activates in MEDIUM state - ✅ Manual override (`/nlag toggle`) now truly pauses monitoring - ✅ Simulate mode no longer throws errors on invalid input - ✅ Recovery notifications always fire (not suppressed by cooldown) - ✅ Stale cooldown timestamps cleared on reload ### 📈 Version Info - **Version:** 1.1.0 - **Target:** Paper 1.21+, Java 21+ - **Release Date:** 2026-03-14 - **Total Features:** 18 (from 16 in v1.0.6) - **Supported Languages:** 7 - **New Files:** None (all backwards compatible) ---

## [1.0.6] — 2025-03-11 ### Fixed - **Bug — Tick-Rate Throttling (Feature 2) never activated** In `applyAiBatched`, `shouldHaveAi` was always `true` for MEDIUM state because `globalAiOn = !state.equals("CRITICAL")` evaluates to `true`. The throttle condition `!shouldHaveAi` was therefore always `false`, so distant mobs in MEDIUM were silently given full AI instead of being throttled. Fix: chunk proximity is now evaluated independently first. Near-player mobs always get full AI; distant mobs in MEDIUM state enter the throttle queue; CRITICAL mobs get `setAI(false)` as intended. - **Bug — `/nlag toggle` did not actually pause monitoring** Setting `manualOverride = true` had no effect on `tick()` — the scheduler task continued running. Fix: `monitorTask.cancel()` is now called when pausing; `startMonitor()` is called when resuming (existing behaviour). - **Bug — `/nlag simulate clear` threw `NumberFormatException`** Tab-complete suggested `"clear"` as a valid argument, but the code passed it directly to `Double.parseDouble()`, crashing with an unhandled exception. Fix: an explicit `equalsIgnoreCase("clear")` check now runs before the parse. - **Minor — Recovery (NORMAL) action bar could be suppressed by cooldown** The 5-second notify cooldown was applied to all state transitions, including CRITICAL → NORMAL recovery. If a CRITICAL message had just been sent, admins would miss the recovery notification entirely. Fix: cooldown only throttles MEDIUM/CRITICAL spam; NORMAL recovery always fires. - **Minor — `notifyTimes` map not cleared on `restoreAll()`** Stale cooldown timestamps persisted across `/nlag reload` and plugin restarts, causing the first post-reload notification to be silently dropped. Fix: `notifyTimes.clear()` added to `restoreAll()`. --- ## [1.0.5] — 2025-03-10 ### Added - **Feature 1 — Chunk-Based AI Throttling** Mobs within a configurable chunk radius of any online player always retain their AI, even in CRITICAL state. Only mobs in truly unobserved chunks are disabled. Controlled by `chunk-ai.enabled` and `chunk-ai.distance-chunks`. - **Feature 2 — Tick-Rate Throttling** In MEDIUM state, instead of hard-disabling mob AI, a dedicated per-tick task cycles through distant mobs and enables their AI for exactly 1 tick every N ticks. Mobs appear sluggish rather than completely frozen, which feels far more natural in-game. Controlled by `tick-throttle.enabled` and `tick-throttle.ticks-per-ai-tick`. - **Feature 3 — Spawn Rate Suppression** During MEDIUM or CRITICAL lag, world monster and animal spawn caps are temporarily lowered to reduce new mob generation. Limits are fully restored on recovery. Controlled by `spawn-suppression.*`. - **Feature 4 — Per-Player Action Bar Notification** When a player has affected mobs within a configurable radius, they receive a personal action bar message explaining why nearby mobs seem different. Previously only ops received any notification. Controlled by `per-player-notify.*`. - **Feature 5 — `/nlag simulate <tps>`** Overrides the TPS value used by the engine without affecting the real server. Allows admins to test threshold configurations, Discord alerts, and reports without needing actual server lag. Running `/nlag simulate` with no argument clears the override. - **Feature 6 — Auto Lag Reports** After recovering from a CRITICAL event that lasted longer than `min-duration-seconds`, a timestamped report is written to `plugins/NeuroLag/lag-reports/`. Each report includes world name, duration, lowest TPS reached, recovery TPS, affected mob count, online player count, and JVM heap statistics. - **Feature 7 — Entity Whitelist** Specific `EntityType` names can be added to `targets.whitelist-types` to permanently exempt them from all optimization and culling. Additionally, `protect-named-mobs` and `protect-tamed-mobs` options prevent NeuroLag from touching player pets, named mobs, and boss entities. - **Feature 8 — Discord Webhook** Posts a rich Discord embed when a world enters CRITICAL lag and when it recovers. The recovery embed includes total CRITICAL duration and the lowest TPS reached. Uses Java 11+ `HttpClient` — no external library required. Controlled by `discord.*`. - **Feature 9 — Memory Pressure Detection** Reads JVM heap usage ratio at each TPS check. When heap exceeds `trigger-percent`, an additional offset is subtracted from effective TPS before threshold evaluation, causing NeuroLag to act more aggressively when RAM is nearly exhausted. Controlled by `memory-pressure.*`. - **Feature 10 — Player-Count Scaling** TPS thresholds are linearly scaled upward as the number of online players increases between `min-players` and `max-players`. At maximum players, all thresholds shift up by `max-threshold-offset` TPS. Per-world threshold overrides are scaled the same way. Controlled by `player-scaling.*`. - **`/nlag simulate` tab-completion** — suggests common TPS values (5, 10, 15, 18, 20). - **Heap stats in `/nlag` help panel and `/nlag status`** — shows used/max MB and percentage. - **World state icons in `/nlag status`** — `●` NORMAL, `▲` MEDIUM, `✖` CRITICAL. - **`/nlag toggle` now calls `restoreAll()`** before resuming the monitor, ensuring any frozen mobs are immediately unfrozen.

## [1.0.4] — 2025-03-10 ### Fixed - **Deprecated BungeeCord API removed.** `spigot().sendMessage(ChatMessageType.ACTION_BAR, TextComponent.fromLegacyText(...))` was replaced with Paper's native Adventure API: `player.sendActionBar(LegacyComponentSerializer.legacyAmpersand().deserialize(...))`. The old API is not available on Paper 1.21.3 and would cause a crash or silent failure on action bar notifications. - **`pom.xml` version mismatch fixed.** `<version>` was `1.0.0` while `plugin.yml` declared `1.0.3`, causing Maven to produce `NeuroLag-1.0.0.jar` instead of the expected filename. - **Hysteresis buffer is now configurable.** The recovery buffer (`0.5`) was hardcoded in `computeStatus()`. It is now read from `settings.hysteresis-buffer` in `config.yml`. - **`/nlag graph` was undocumented.** Added to `plugin.yml` usage string and to the help message shown by `/nlag`. - **`shouldOptimize` explicit cast removed.** The `(Mob) entity` cast after `instanceof Mob` check was unnecessary; replaced with pattern-matching `instanceof Mob mob` at the call site. - **`toggle` command no longer leaves mobs frozen.** When manual override is turned off, `restoreAllAI()` is now called before resuming the monitor task so any frozen mobs are immediately unfrozen. - **TPS value capped at 20.0** before storing and displaying, preventing artificially high values (Paper can briefly report >20 TPS) from skewing the graph and average. - **`logEnabled` now logs state transitions**, including world name, new state, current TPS, and affected mob count — more useful for diagnosing server lag patterns. - **`onDisable` log message added** to confirm AI restoration on server shutdown. ### Changed - `config.yml` comments rewritten in English and expanded with usage notes for each option. - `plugin.yml` usage string updated to list all four subcommands: `reload`, `status`, `toggle`, `graph`. - `computeStatus` refactored to use `switch` expression (Java 14+) for readability. - `drawGraph` improved: empty cells now render as `░` instead of a space, making the chart easier to read in chat. - `status` command now displays per-world states with a `(no data yet)` fallback when the monitor hasn't run yet. - Added `README.md` and `CHANGELOG.md`. ---

fixed

# 🛠️ Changelog - NeuroLag v1.0.2 All notable changes to the **NeuroLag** plugin in this version are documented below. --- ## [1.0.2] — 2026-03-07 ### 🚀 Performance Optimizations * **Entity Batch Processing:** AI updates are now processed in batches of **50 entities per tick**. This eliminates the "main thread spike" previously caused by iterating through hundreds of mobs simultaneously during state changes. * **Improved Loop Efficiency:** Refactored entity filtering and collection to be more streamlined and performant. ### 🛡️ Stability Improvements * **State Hysteresis:** Implemented a **0.5 TPS buffer zone** for state transitions. This prevents "jitter" and notification spam when the server's TPS fluctuates near the 18.0 (Medium) or 15.0 (Critical) thresholds. * **API Compatibility Fixes:** Resolved compilation and runtime issues by ensuring compatibility with the **Paper 1.21.1** attribute system, while maintaining safe fallbacks for older server environments. ### 📊 New Features * **Performance Statistics:** * **Session Average TPS:** Tracks the overall server performance since the last reload/startup. * **Lag Activation Counters:** Tracks the frequency of "Medium" and "Critical" lag events. * **Enhanced `/nlag status`:** Updated command output to include the new Average TPS and Activation statistics for better server monitoring. ### 🛠️ Maintenance * **Water Mob Audit:** Verified and ensured that `WaterMob` entities are correctly handled when the `targets.water-mobs` configuration is enabled. * **Configuration Cleanup:** Optimized default configuration comments and structure for better usability. ---

# 🚀 Release 1.0.1 - Performance & Stability Update This update focuses on internal optimization, modernizing the codebase for Paper 1.21.1, and fixing several logic gaps found in the initial release. ### 🛠 Improvements & Optimization * **Engine Optimization**: Switched from a global entity loop to `world.getLivingEntities()`, significantly reducing CPU usage on high-population servers. * **API Modernization**: Fully migrated to the modern `Attribute.FOLLOW_RANGE` API, ensuring long-term compatibility with Paper 1.21.1 and future versions. * **Smart Notification Cooldown**: Added a 5-second cooldown to Action Bar alerts to prevent message spam when server TPS fluctuates near trigger thresholds. * **Crash Resilience**: Implemented an automatic AI restoration sequence on plugin enable to ensure mobs are never stuck without AI due to improper server shutdowns. ### 🐞 Bug Fixes * **Water Mobs Support**: Fixed a bug where `water-mobs` in `config.yml` were being ignored. The system now correctly optimizes aquatic entities. * **Tab-Completion**: Refined command suggestions to be more responsive and accurate. * **Logic Sync**: Fixed an issue where manual overrides could sometimes conflict with the automated monitoring task. ### 📦 Technical Note * **Core**: Still powered by **Java 21** for peak performance. * **Configuration**: No changes needed to your existing `config.yml`, but we recommend checking if `water-mobs` is set to your preference. --- *Optimizing the neural pathways of your server.* Developed by **Duong2012G**.

# 🚀 Release 1.0.0 - Initial Neural Launch We are excited to introduce the first official release of **NeuroLag**, a resource-aware neural optimization system designed specifically for modern Minecraft servers. ### 🌟 New Features * **Neural AI Scaling**: Implemented a dynamic performance monitoring system that scales Mob AI complexity based on real-time server TPS. * **Adaptive Pathfinding**: Integrated automatic adjustment of the `GENERIC_FOLLOW_RANGE` attribute to reduce CPU overhead during lag spikes. * **Multi-Tier Logic**: Three distinct optimization states: **Normal** (Full AI), **Medium** (Throttled Vision), and **Critical** (AI Paused). * **Admin Dashboard**: Added a suite of commands (`/nlag status`, `/nlag stats`, `/nlag toggle`) for real-time monitoring and manual overrides. * **Notification System**: Real-time Action Bar alerts for administrators when the neural system changes states. ### 🛠 Technical Specifications * **Core Engine**: Built on **Java 21** for maximum performance and modern standard compliance. * **API Compatibility**: Optimized for **Paper/Spigot 1.21.1**. * **Open Source**: Released under the **Apache License 2.0**. ### 📦 Installation Simply drop the `NeuroLag-1.0.0.jar` into your plugins folder and restart your server. --- *Developed with ❤️ by Duong2012G*.