The VK_ERROR_DEVICE_LOST crash on AMD/Mali GPUs (barrier srcAccessMask)
was fixed in 2026-03-18. Enable refraction for new sessions so players
get the improved water visuals without needing to touch Settings.
Existing saved configs that explicitly disabled it are preserved.
- AmdFsr3Runtime now probes both the legacy ffxFsr3* API and the newer
generic ffxCreateContext/ffxDispatch API; selects whichever the loaded
runtime library exports (GenericApi takes priority fallback)
- Generic API path implements full upscale + frame-generation context
creation, configure, dispatch, and destroy lifecycle
- dlopen error captured and surfaced in lastError_ on Linux so runtime
initialization failures are actionable
- FSR3 runtime init failure log now includes path kind, error string,
and loaded library path for easier debugging
- tools/generate_ffx_sdk_vk_permutations.sh added: auto-bootstraps
missing VK permutation headers; DXC auto-downloaded on Linux/Windows
MSYS2; macOS reads from PATH (CI installs via brew dxc)
- CMakeLists: add upscalers/include to probe include dirs, invoke
permutation script before SDK build, scope FFX pragma/ODR warning
suppressions to affected TUs, add runtime-copy dependency on wowee
- UI labels updated from "FSR2" → "FSR3" in settings, tuning panel,
performance HUD, and combo boxes
- CI macOS job now installs dxc via Homebrew for permutation codegen
- Add water/lava/lighting rendering features to README
- Add transport riding to movement features
- Update status date and rendering capabilities
- Note interior shadow and lava steam known gaps
New tab lets users explore extracted assets visually: BLP textures
rendered via blp_convert+Pillow, M2/WMO wireframes with mouse-drag
rotation and zoom, DBC/CSV tables with named columns from
dbc_layouts.json, ADT heightmap grids, text file viewer, audio
metadata, and hex dumps. Directory tree lazy-loads from manifest.json
with search and file-type filtering.
Cover all new features: cancel extraction, background override rebuild,
zip-slip protection, extractor auto-detection order, readonly dropdowns,
and selection preservation. Reorganize into per-tab sections with tables.
Use actual WoW 3.3.5a PlayerExertions and Vox sound paths from MPQ
manifests for attack grunts, wounds, and death sounds. Handle Blizzard
naming quirks (HumanFeamle typo, OrcMale no Final suffix, Scourge→Undead).
Add COMBAT_IDLE animation state with ready weapon stance between swings.
Restore deleted MPQ sound manifest docs.
Created WARDEN_COMPLETE.md - the definitive guide to our implementation.
Contents:
- Executive summary of what we built
- Complete architecture diagram (4 layers)
- Detailed how-it-works for each step
- All 9 check types supported
- Platform support matrix (Linux/macOS/Windows/ARM)
- Performance metrics (~120ms first check, ~1-5ms subsequent)
- Complete testing checklist
- Troubleshooting guide
- Production deployment requirements
- Success metrics
Updated MEMORY.md with final status:
- ALL 7 PHASES COMPLETE ✅
- Production-ready for testing
- Cross-platform via Unicorn Engine
- No Wine needed!
- ~2,800 lines of code
- Performance: <100ms overhead
Status: READY FOR REAL MODULE TESTING
Added MD5 hashing and extensive testing documentation for future attempts
at supporting strict Warden servers like Warmane.
Enhancements:
- Added MD5 hash support to Crypto class (OpenSSL-based)
- Tested 6 different module ACK response formats against Warmane
- Analyzed module packet structure (37 bytes: opcode + seed + trailing)
- Enhanced debug logging for plaintext and encrypted Warden data
Documentation:
- WARDEN_IMPLEMENTATION.md: Complete implementation guide with all attempts
- WARDEN_QUICK_REFERENCE.md: Quick troubleshooting and testing guide
Test Results (Warmane):
- Empty ACK (0 bytes): Server silent
- XOR/MD5 checksum (18 bytes): Server silent
- Single byte (1 byte): Server disconnects (rejected)
- Echo trailing (20 bytes): Server silent
- Result + SHA1 (21 bytes): Server silent
Conclusion:
- Current implementation works with permissive/disabled Warden servers
- Warmane requires module execution or undocumented response format
- Full documentation provided for future reverse engineering attempts
Next steps documented:
1. Capture packets from real WoW client (protocol analysis)
2. Implement module execution engine (months of work)
3. Test with local AzerothCore server
Add SkySystem coordinator that follows WoW's actual architecture where skyboxes
are authoritative and procedural elements serve as fallbacks. Integrate lighting
system across all renderers (terrain, WMO, M2, character) with unified parameters.
Sky System:
- SkySystem coordinator manages skybox, celestial bodies, stars, clouds, lens flare
- Skybox is authoritative (baked stars from M2 models, procedural fallback only)
- skyboxHasStars flag gates procedural star rendering (prevents double-star bug)
Celestial Bodies (Lore-Accurate):
- Two moons: White Lady (30-day cycle, pale white) + Blue Child (27-day cycle, pale blue)
- Deterministic moon phases from server gameTime (not deltaTime toys)
- Sun positioning driven by LightingManager directionalDir (DBC-sourced)
- Camera-locked sky dome (translation ignored, rotation applied)
Lighting Integration:
- Apply LightingManager params to WMO, M2, character renderers
- Unified lighting: directional light, diffuse color, ambient color, fog
- Star occlusion by cloud density (70% weight) and fog density (30% weight)
Documentation:
- Add comprehensive SKY_SYSTEM.md technical guide
- Update MEMORY.md with sky system architecture and anti-patterns
- Update README.md with WoW-accurate descriptions
Critical design decisions:
- NO latitude-based star rotation (Azeroth not modeled as spherical planet)
- NO always-on procedural stars (skybox authority prevents zone identity loss)
- NO universal dual-moon setup (map-specific celestial configurations)
Implements full NPC voice interaction system supporting 6 different sound categories
for all playable races/genders. System loads ~450+ voice clips from MPQ archives.
Voice Categories:
- Greeting: Play on NPC right-click interaction
- Farewell: Play when closing gossip/dialog windows
- Vendor: Play when opening merchant/vendor windows
- Pissed: Play after clicking NPC 5+ times (spam protection)
- Aggro: Play when NPC enters combat with player
- Flee: Play when NPC is fleeing (ready for low-health triggers)
Features:
- Race/gender detection from NPC display IDs via CreatureDisplayInfoExtra.dbc
- Intelligent click tracking for pissed sounds
- Combat sounds use player character vocal files for humanoid NPCs
- Cooldown system prevents voice spam (2s default, combat sounds bypass)
- Generic fallback voices for unsupported NPC types
- 3D positional audio support
Voice Support:
- All playable races: Human, Dwarf, Gnome, Night Elf, Orc, Tauren, Troll, Undead
- Male and female variants for each race
- StandardNPC sounds for social interactions
- Character vocal sounds for combat
Technical Changes:
- Refactored NpcVoiceManager to support multiple sound categories
- Added callbacks: NpcFarewell, NpcVendor, NpcAggro
- Extended voice loading to parse both StandardNPC and Character vocal paths
- Integrated with GameHandler for gossip, vendor, and combat events
- Added detailed voice detection logging for debugging
Also includes:
- Sound manifest files added to docs/ for reference
- Blacksmith hammer pitch increased to 1.6x (was 1.4x)
- Blacksmith volume reduced 30% to 0.25 (was 0.35)
