Changelog 2.0.1

🚀 Dual-Platform Engine: Paper & Fabric First-Class Support

• Multi-Module Gradle Architecture:
  • Restructured the entire codebase into a clean, modern multi-module architecture:
    • :api: Platform-independent public interfaces, models, and events (DreamVoiceAPI, VoiceLocation, Speaker, VoiceWiretap, VoiceEvent, etc.).
    • :common: Platform-agnostic core acoustic engine (DreamVoiceCommon), DSP audio pipelines, raycasting logic, persistence, and service implementations.
    • :paper: Paper / Purpur 26.1.2+ server plugin (DreamVoice-paper-2.0.1.jar) utilizing Paperweight and shadow jar packaging.
    • :fabric: Fabric 26.1.2+ server mod (DreamVoice-fabric-2.0.1.jar) built with Fabric Loom 1.17, Fabric API, and Cloud Fabric commands.
• Platform Abstraction Layer (VoicePlatform):
  • Implemented PaperVoicePlatform and FabricVoicePlatform bridging server scheduling (sync/async/timer), entity tracking, block occlusion raycasts, particle dispatches, and loggers.
  • Native multiplatform coordinates via VoiceLocation, allowing services to operate identically regardless of whether Bukkit or Minecraft ServerLevel is running underneath.
• Dual Configuration Directories:
  • Automatically loads and saves data in standard platform locations:
    • Paper: plugins/DreamVoice/
    • Fabric: config/dreamvoice/

---

🔌 Unified DreamVoiceAPI Service Locator

• Universal API Entrypoint:
  • Unified entry point DreamVoiceAPI.get() accessible from both Paper plugins and Fabric mods without platform-specific boilerplate.
  • Direct accessor methods for all acoustic services:
    • api.speakerService(): 3D spatial speakers, dual-channel audio, entity tracking.
    • api.wallService(): VoiceWall raycast occlusion, material decibel loss, diffraction.
    • api.roomService(): Soundproof Cuboid zones and reverberation presets.
    • api.recordingService(): Zero-CPU Opus capture, timeline slicing, interactive cassettes.
    • api.radioService(): Frequency channels, analog band-pass filters, push-to-talk clicks.
    • api.wiretapService(): Covert listening bugs, mobile entity bugs, live eavesdropping.
    • api.projectionService(): Decoupled voice anchors, CCTV feeds, bilateral audio.
    • api.transmitterService(): Point-to-point directional voice transmitters.
    • api.broadcastService(): Server-wide and localized podium/intercom broadcasts.
    • api.filterService(): File-first DSP filters, hot-compilation, and stack management.
    • api.speechService(): Offline Vosk keyword spotting and transcription stations.
    • api.playerService(): Voice player wrappers (VPlayer) and connection metadata.
    • api.persistenceService(): Asynchronous disk flushes and auto-save coordination.
    • api.voiceService(): Direct Simple Voice Chat bridge and raw packet dispatch.

---

📡 Fabric Network Packet System & Inter-Mod Protocols (@DreamPacket)

• Annotation-Driven Networking Pipeline:
  • Introduced @DreamPacket and @DreamServerReceiver with automatic classpath scanning via PacketAnnotationProcessor and ServerPacketAnnotationProcessor.
  • Registered all packet codecs with Fabric's PayloadTypeRegistry under the dreamvoice:* namespace.
• Preshot Server-Bound & Client-Bound Packet Catalog (40+ Packets):
  • Configuration & Visualizer: ServerBoundRequestConfigSyncPacket, ClientBoundConfigSyncPacket, ServerBoundVisualizerTogglePacket, ClientBoundVisualizerSyncPacket.
  • 3D Spatial Speakers: ServerBoundSpeakerCreatePacket, ServerBoundSpeakerDeletePacket, ServerBoundSpeakerLinkPacket, ServerBoundSpeakerUnlinkPacket, ServerBoundSpeakerPlaySoundPacket, ServerBoundSpeakerStopSoundPacket, ServerBoundSpeakerUpdatePacket, ClientBoundSpeakerSyncPacket.
  • Wiretaps & Surveillance: ServerBoundWiretapCreatePacket, ServerBoundWiretapDeletePacket, ServerBoundWiretapSubscribePacket, ServerBoundWiretapUnsubscribePacket, ClientBoundWiretapSyncPacket.
  • Radios & Walkie-Talkies: ServerBoundRadioCreatePacket, ServerBoundRadioJoinPacket, ServerBoundRadioLeavePacket, ClientBoundRadioSyncPacket.
  • Audio Recordings & Cassettes: ServerBoundRecordingStartPacket, ServerBoundRecordingStopPacket, ServerBoundCassetteCreatePacket, ServerBoundRecordingPlayPacket, ServerBoundRecordingStopPlayPacket, ClientBoundRecordingSyncPacket.
  • DSP Voice Filters: ServerBoundFilterApplyPacket, ServerBoundFilterRemovePacket, ServerBoundFilterClearPacket, ServerBoundFilterTogglePacket, ClientBoundFilterSyncPacket.
  • Acoustic Rooms: ServerBoundRoomCreatePacket, ServerBoundRoomDeletePacket, ServerBoundRoomUpdatePacket, ClientBoundRoomSyncPacket.
  • Transmitters & Projections: ServerBoundTransmitterCreatePacket, ServerBoundTransmitterDeletePacket, ServerBoundTransmitterTogglePacket, ClientBoundTransmitterSyncPacket, ServerBoundProjectionCreatePacket, ServerBoundProjectionDeletePacket, ClientBoundProjectionSyncPacket.
• Zero-Dependency Inter-Mod Integration:
  • Allows external survival mods (craftable megaphones, spy bugs, walkie-talkies, tape recorders) and client HUD mods to control DreamVoice subsystems via JSON custom payloads without hard compile-time dependencies.

---

⚡ Dual-Platform Event Bus (Bukkit & Fabric)

• Unified Voice Events:
  • All event classes extend VoiceEvent and feature self-dispatching via .callEvent().
• Platform Event Dispatching:
  • Paper: Dispatched natively through Bukkit's PluginManager for @EventHandler listeners.
  • Fabric: Dispatched through DreamVoiceEvents.VOICE_EVENT.register(...) callback hook.
• Event Coverage:
  • Cancellable hooks across all lifecycle actions: SpeakerLinkPlayerEvent, SpeakerPlaySoundEvent, RadioChannelJoinEvent, CassetteCreateEvent, CassettePlayEvent, VoiceRecordingStartEvent, VoiceFilterApplyEvent, VoiceWallOcclusionEvent, VoiceKeywordSpokenEvent, WiretapSubscribeEvent, etc.

---

👁️ Real-Time 3D Acoustic Visualizer Mode (/voice visualizer)

• Interactive Acoustic Raycasting:
  • Added /voice visualizer [toggle|on|off] (permission: dreamvoice.visualizer).
  • Renders 3D particle rays illustrating line-of-sight paths, wall collision impacts, aperture diffraction bypasses around doors/openings, and speaker audible coverage spheres.
• Performance-Conscious Hand-Held Requirement:
  • Particle streams are rendered only when the inspecting player has the visualizer tool (default: COMPASS) equipped in their main hand, preventing client lag and visual clutter during normal gameplay.
• Synchronized Visualizer State:
  • State is automatically synced to Fabric clients via ServerBoundVisualizerTogglePacket and ClientBoundVisualizerSyncPacket.

---

🎙️ VoiceWiretap x VoiceProjection Integration

• Cross-Module Audio Propagation:
  • Covert wiretap microphones (VoiceWiretap) now dynamically support and intercept voices emitted through active voice projections (VoiceProjection).
• Dual-Origin Spatial Audio:
  • Physical Emission (isEmitVoiceAtPlayer): Microphones in proximity to the player's physical body capture the direct voice as normal.
  • Anchor Emission (isEmitVoiceAtAnchor): Microphones in proximity to the projection's anchor location (projection.getAnchorLocation()) intercept the projected voice in real time.
  • Acoustic Occlusion & Wall Physics: Line-of-sight raycasting (VoiceRayCast) and material attenuation (VoiceWall) are accurately simulated between the anchor coordinates and the wiretap.
  • Filter Inheritance: Voice projections configured with custom filters (e.g., megaphone, radio, disguise) pass their processed audio downstream to wiretap listeners.

---

🔌 Public Hearing & Attenuation API (VoiceService & VoiceWallService)

• Unified Audibility Check (canHear):
  • Introduced voiceService.canHear(speaker, listener) (with Player and UUID overloads).
  • Evaluates whether a listener can hear a speaker across all communication channels:
    1. Direct point-to-point audio transmitters (VoiceTransmitterService).
    2. Shared radio frequency channels (VoiceRadioService).
    3. Proximity spatial voice chat (canHearProximity).
• Full-Condition Proximity Audibility (canHearProximity):
  • Introduced voiceService.canHearProximity(speaker, listener) (with Player and UUID overloads).
  • Validates active Simple Voice Chat connections for both players.
  • Enforces life states (ALIVE, DEAD, SPECTATE).
  • Evaluates voice projection constraints (emitVoiceAtPlayer, emitVoiceAtAnchor, hearPlayerEnvironment).
  • Respects configured distance limits (Codex.getEffectiveDistance()).
  • Computes total dB attenuation across VoiceWall materials and isolated VoiceRoom cuboids; returns false if sound is blocked ($\ge 99.0\text{ dB}$).
• Direct Decibel Attenuation Metrics:
  • Added voiceService.getEffectiveAttenuationDb(speaker, listener) to measure total acoustic dB loss between two players.
  • Added voiceWallService.getAttenuationDb(speaker, listener) for querying raw wall and room isolation losses.

---

📢 Physical Broadcast System (VoiceBroadcastService)

• Podium & Intercom Broadcast Points:
  • Introduced VoiceBroadcastService and /voice broadcast command suite.
  • Players speaking within a configured capture radius around a location or entity automatically have their voice routed to connected speakers.
• Flexible Audio Routing:
  • All Speakers Routing (allSpeakers = true): Broadcasts to every registered speaker across the server.
  • Selective Target Speakers: Link specific speakers to a broadcast point (/voice broadcast link <name> <speaker>).
• On/Off Toggle & Filter Chaining:
  • Broadcast points can be toggled on/off on demand (/voice broadcast toggle <name>).
  • Optional custom audio filters can be chained to the broadcast (e.g. intercom distortion, PA megaphone filter).
• Modular Persistence:
  • Stored under modules/broadcast/ with automatic saving on server shutdown.

---

🛠️ Interactive Configuration Mode & Tool (/voice config)

• Interactive Configuration Tool (/voice tool):
  • Added /voice config and /voice tool commands (using ECHO_SHARD as the physical tool to prevent conflicts with WorldEdit wands).
  • Config tool supports interactive targeting: entities, blocks, or air/gaze fallback.
• Full InvUI Management GUIs Across All Audio Subsystems:
  • Dashboard (VoiceConfigMainGUI): Visual dashboard covering all audio subsystems with target coordinates/entity indicators.
  • Speakers Manager (SpeakerListGUI, SpeakerEditGUI): Browse speakers, adjust distance/range with +/- buttons, toggle mode (GLOBAL / RESTRICTED), teleport, or create new speakers at current target.
  • Broadcast Manager (BroadcastListGUI, BroadcastEditGUI, BroadcastLinkSpeakersGUI): Browse broadcast points, toggle active state (ON/OFF), adjust radius, toggle all-speakers routing, visually link/unlink individual speakers, cycle filters, or create new points.
  • Wiretaps Manager (WiretapListGUI, WiretapEditGUI): Inspect covert bugs, toggle live eavesdropping, teleport, or deploy new bugs.
  • Acoustic Rooms Inspector (RoomListGUI): View room dimensions, reverberation presets, and acoustic isolation percentages.
  • Radio Channels Manager (RadioListGUI): Browse frequencies, view active listener counts, and tune in/disconnect directly.
  • Transmitters Manager (TransmitterListGUI): Toggle transmitter mode, view linked receiver players, unlink or configure receivers.
  • Projections Inspector (ProjectionListGUI): Inspect active player projections, emission modes (anchor/player), and teleport to anchor coordinates.
  • Acoustic & Wall Engine (AcousticInspectGUI): Live toggle for VoiceWall, cycle occlusion modes (Realistic/Strict/Off), toggle air damping, particle raycast debug, and inspect acoustic properties of targeted blocks/materials.

---

🌐 Complete Localization & Translation System (LangService)

• Adventure Component.translatable Integration:
  • All command messages, items (Cassette, Config Tool), and InvUI configuration menus now use Component.translatable(...) backed by MiniMessage translation stores.
  • Eliminated all hardcoded TextDecoration (<!italic>, <b>) and NamedTextColor calls across UI classes; all styling is defined directly within language files.
• Bilingual English (Default) & French Support:
  • Populated lang-dreamvoice.json with comprehensive translation keys, gradients, formatting tags, and arguments.
  • Configured default plugin language to English (EN) with full client-side French (FR) fallback.

---

📦 Build System & Dependency Centralization

• Centralized gradle.properties:
  • Consolidated all plugin, toolchain, and dependency versions into gradle.properties:
    • minecraft_version = 26.1.2
    • paper_api_version = 26.1.2.build.+
    • fabric_api_version = 0.155.0+26.1.2
    • loader_version = 0.19.3
    • loom_version = 1.17-SNAPSHOT
    • shadow_version = 9.4.1
    • voicechat_api_version = 2.6.20
    • vosk_version = 0.3.45
    • cloud_fabric_version = 2.0.0-beta.16
    • java_version = 25
  • Streamlined settings.gradle and all subproject build files (api, common, paper, fabric) to reference properties exclusively.
• CI/CD Workflow Modernization:
  • Updated GitHub Actions workflows (publish.yml, javadoc.yml, discord-announce.yml) for dual-platform publishing to GitHub Packages and Modrinth.

---

🐛 Bug Fixes & Stability

• Projection Audio Crackling & Transcoding Artifacts:
  • Eliminated audio crackling and high-frequency buzzing occurring when speaking while in voice projection:
    • Removed discontinuous chunk-level 1-pole lowpass smoothing filter that injected 50 Hz boundary discontinuities into audio packets.
    • Prevented multi-decoding of incoming microphone Opus packets on shared decoders across listeners.
    • Implemented encoder.resetState() on speech pauses (> 400ms) to prevent Opus predictor state corruption upon resuming speech.
    • Bypassed PCM decoding and lossy Opus transcoding completely when no wall occlusion or voice filters are present, routing raw Opus packets directly to 3D locational audio channels.
• Speech Recognizer Crash on Reload:
  • Resolved kaldi::KaldiFatalError (Can't add speaker model to already running recognizer) triggered during /voice speech reload by safely closing and cycling active Vosk recognizer instances before applying updated keyword grammars.
• CI / Discord Webhook Release Workflow:
  • Fixed Python variable scope (brand_color) in .github/workflows/discord-announce.yml.
  • Added role mention <@&1546913649308995738> to automated Discord release broadcasts.

---

🧹 Code Quality & Conventions

• Removed redundant control-flow braces on single statements (if, else, for) across fabric, paper, common, and api.
• Cleaned up inline fully-qualified class names in favor of clean header imports.
• Standardized final var declaration across all local variable scopes.