feature: multiplayer (#25)

* build: add asio library

* feat(dev_panel): add Asio to about table

* refactor(gameplay): split Chunk into server/client variants and add networking

* feat(gameplay): add Session class and UUID generation utility

* feat(network): integrate protobuf for player sync and session management

Add Protobuf dependency, define proto messages for player requests, positions, and chunk data. Refactor Session to use strand and async write. Implement player join/exit and position sync in ServerWorld.

* refactor(proto): restructure protobuf definitions and update build system

* feat(protocol): add packet serialization and login handling

Introduce packet header and ID mapping for protobuf messages.
Refactor session and server_world to use new packet wrapper.
Fix missing semicolons in proto files.

* refactor(packet): serialize directly into packet buffer with ByteSizeLong

* feat(world): add player-based chunk loading and UUID support

Restructure world generation to trigger chunk loading based on player
movement. Replace player name with UUID for identification. Implement
chunk request/response protocol. Improve thread management for gen
thread.

* feat(server): add block change packet handling and increase reserved threads

* refactor(server): enhance thread safety and session management

* fix(session): catch std::exception in read_loop and log error

* fix(network-server): stop accept loop on shutdown flag

* fix(gameplay): correct unknown error log condition and missing chunk pos update

* feat(client): add client-side chunk, player, world and network classes

* feat(client-chunk): add greedy meshing, receive_chunk, and biome field

* feat(client_player): add name constructor param, remove chunk transition

* feat(gameplay): integrate network client and thread-safe player pos

* feat(gameplay): add block change sync and coordinate utils

* refactor(server): replace Chunk with ServerChunk

* refactor(gameplay): split world into client and server

Remove monolithic World, Player, Chunk classes. Introduce ClientWorld, ServerWorld, ClientPlayer, and related networked components. Add Abseil dependency for logging and checks. Rename old files to pre_remove_* and update all includes and references accordingly.

* feat(gameplay): add time sync and network improvements

* fix: correct neighbor block indexing and add time update support

* feat(gameplay): optimize chunk request with player chunk tracking

Add `update_player_chunk()` method to `ClientPlayer` that triggers a chunk request when the player moves more than 2 chunks away from the last recorded chunk. Introduce an atomic `m_requesting_chunk` flag in `ClientWorld` to prevent concurrent requests. Rename `m_player_chunk_pos` to `m_last_chunk_pos` for clarity.

* refactor(client_world): split pending queue into data and upload queues

* feat(client): add center-based chunk loading priority

* feat(app): add client/server CLI arguments and init guards

* feat(multiplayer): add rendering of other players as colored cubes

* feat(gameplay): add thread pool for asynchronous chunk data processing

* feat(gameplay): add player logout and disconnect handling

* fix(dev_panel): add protobuf to about table libraries

* build: add ZLIB dependency

* refactor(packet): replace zlib with zstd compression and restructure header

Increase header length to 12 bytes, add CompressType and PacketHeader struct, and implement decode_packet_header. Update CMake to find zstd and link against it, adding Findzstd module.

* refactor(network): extract packet header decoding into function

* feat(packet): implement zstd compression and refactor packet decoding

* perf(gameplay): optimize protobuf message allocation with Arena and limit thread pool size

Switch to Arena allocation for protobuf messages to reduce heap allocations. Pass ChunkDataRsp by value to enable move semantics. Cap thread pool size to hardware concurrency with max 4 threads.

* refactor(gameplay): use steady clock and sleep_until for tick loop timing

* refactor(gameplay): rename remote player types and add interpolation

* refactor(server_world): use thread pool for chunk request handling

* refactor(packet): replace if-constexpr chain with explicit template specializations

Refactor `get_packet_id()` by removing the if-constexpr chain and instead
providing explicit specializations for each packet type. This simplifies the
primary template (which now only contains a static assertion) and improves
compile-time dispatch clarity. The `always_false` helper is no longer needed
in the primary template.

* feat(gameplay): add packet to clear all chunks on server rebuild

Add new packet S2C_ClearAllChunks (ID 3005) to notify clients when the server rebuilds its world. Clients respond by clearing their chunk cache and requesting fresh chunks. Shift UPDATE_TIME packet ID to 3006 to accommodate.

* feat(dev_panel): split world tab; add chunk request logging and fix flag reset

* feat(tools): add RecentQueue to replace std::deque in server_world

* feat(gameplay): add task ID to chunk requests to discard stale responses

* refactor(block): move block source to gameplay directory

* feat(toml): add shared TOML utilities and handle ip.toml

* feat: add player name configuration and rename config file

- Add --player command-line argument and 'player' field in Arguments struct
- Rename ip.toml to server.toml and corresponding internal variable to 'server'
- Fix block.cpp source path in CMakeLists.txt

* build(cmake): modularize build configuration

* fix: build fail on windows

* refactor(renderer): rename Logger level enum value DEBUG to L_DEBUG

* refactor(world): rework chunk state machine and player chunk tracking

Introduce ChunkState enum and ChunkEntity struct to manage chunk lifecycle. Store chunks as shared_ptr to avoid move operations during generation. Add clear_unused_chunks to remove chunks not referenced by any player. Implement deferred chunk request queue for safe processing after generation completes. Update player chunk set during required chunk computation. Improve thread safety with mutexes on chunk and player maps. Fix m_gening flag not reset after generation and add assertions for correctness. Change need_gen to require a player UUID, removing std::optional. Add chunk_size query method for debugging.

* fix(server): ensure chunk is ready before sending or setting block

* refactor(server_world): separate gen and net thread pools

Introduce a second thread pool for network operations and a `ThreadPoolKind` enum to distinguish between gen and net pools. Rename `pool_threads()` to `gen_pool_threads()`, add `change_pool_threads()` overload that accepts the pool kind, and update the dev panel to use the gen pool. Adjust logging and initialization to handle both pools.

* feat(client_world): implement server exit acknowledgment with timeout

* feat(gameplay): implement server stop and client exit handling

Add server_stop flag to LogoutRsp protocol. Modify client_world to check for server stop or own logout to set exit flag. Add ServerWorld::stop() to broadcast stop and cleanly shut down. Refactor chunk ownership to unique_ptr. Remove name parameter from get_look_block_pos.

* fix(app): correct port validation and remove redundant TOML helper

* fix: address unused parameter warnings and missing port assignment

* refactor(gameplay): remove old pre_remove chunk, player, and world files

* refactor(server): replace chunk set with flat_hash_set and add ref count

- Replace std::unordered_set with absl::flat_hash_set for chunk position sets.
- Add ref_count field to ChunkEntity and implement update_ref_count().
- Remove clear_unused_chunks() and use ref counting for chunk lifetime.
- Add get_chunk_pos_set() accessors to ServerPlayer.

* fix(client-world): delay chunk request to wait for server central chunk generation

* refactor(gameplay): migrate chunk storage to tbb::concurrent_hash_map

* refactor(server_world): replace future-based chunk generation with queue

Remove `std::future` per-chunk tracking and poll_finished_chunks().
Instead, generation tasks push completed chunks to a concurrent queue
consumed during update. This eliminates wait_all_chunk_tasks() and
simplifies synchronization.

* refactor(client_world): use concurrent data structures for chunk management

Replace std::shared_mutex and std::unordered_set with tbb::concurrent_hash_map for chunk storage, tbb::concurrent_queue for pending uploads, and absl::flat_hash_set for player chunk positions. This improves thread safety and reduces locking overhead. Refactor receive_chunk and request_chunk to work with the new concurrent model.

* build(windows): add Nvidia and AMD GPU high performance exports

* fix(ci): update header file pattern in clang-format check

* style(clang-format): comment out AlignPPAndNotPP option

* chore: enable AlignPPAndNotPP in clang-format and pin clang version in CI

* ci(workflow): update format check to use LLVM 22

* ci(format-check): install clang-format-22

* perf(world): cull distant players from updates and rendering

Add distance2 utility function. On client, skip rendering other players if distance squared exceeds rendering distance. On server, only send player position and block change updates to players in relevant chunks.

* refactor(math_tools): convert free functions to inline in header

* refactor(dev_panel): remove seed editing and server rebuild world, move client rebuild

- Remove text editing for perlin seed and filter function
- Remove ServerWorld::rebuild_world() method and its atomic flag
- Move 'Rebuild World' button to client world tab
- Remove unused perlin_noise_input_buffer

* fix(server_world): floor block position coordinates in handle_block_change

* refactor(gameplay): pass new_chunks vector as parameter instead of member variable

* perf(client): improve chunk rendering performance with snapshots and threaded generation

* feat(tools): add PriorityThreadPool with priority scheduling

Implement PriorityThreadPool supporting task priorities and FIFO ordering for same priority.
Update ClientWorld to use the new pool with explicit priority for chunk operations.
Fix ThreadPool stop logic with atomic exchange and remove unnecessary lambda capture.

* feat(networking): add priority and sequence ordering to packet send queues

* feat(server-world): use priority thread pool for chunk generation

Introduce PriorityThreadPool to process chunk generation tasks with priorities based on their distance from the player. Closer chunks receive higher priority, improving responsiveness.

* refactor(client): extract AABB creation and add collision check on block place

Make ClientPlayer::get_aabb and ClientWorld::get_block_aabb static.
Change player size to static constexpr M_SIZE.
Use shared_mutex for concurrent reads of other players.
Check for collision with other players before placing a block.

* feat(build): add build-time version system via CMake configure_file

* ci(release): add release build workflow for windows

.github/workflows/format-check.yml

@@ -1,4 +1,3 @@
-# .github/workflows/format-check.yml
 name: Code Format Check
 
 on: [push, pull_request]
@@ -9,9 +8,22 @@ permissions:
 jobs:
   formatting:
     runs-on: ubuntu-latest
-    container: silkeh/clang:latest   
+
     steps:
       - uses: actions/checkout@v4
+
+      - name: Install LLVM 22
+        run: |
+          wget https://apt.llvm.org/llvm.sh
+          chmod +x llvm.sh
+          sudo ./llvm.sh 22
+
+          sudo apt-get install -y clang-format-22
+
+          clang-format-22 --version
+
       - name: Run clang-format
         run: |
-          find src include -name '*.cpp' -o -name '*.h' -print0 | xargs -0 clang-format --dry-run --Werror
+          find src include \
+            \( -name '*.cpp' -o -name '*.hpp' \) \
+            -print0 | xargs -0 clang-format-22 --dry-run --Werror

.github/workflows/release-build.yml

@@ -0,0 +1,153 @@
+name: Release-Build
+
+on:
+  workflow_dispatch:
+    inputs:
+      version:
+        description: Version
+        required: false
+        default: dev
+
+  push:
+    tags:
+      - "v*"
+
+permissions:
+  contents: write
+
+jobs:
+
+  build:
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - os: windows-2022
+            platform: windows
+            
+
+    runs-on: ${{ matrix.os }}
+
+    steps:
+
+      - uses: actions/checkout@v4
+
+
+      - name: Get Version
+        id: version
+        shell: bash
+        run: |
+          if [[ "${GITHUB_REF}" == refs/tags/* ]]; then
+            VERSION="${GITHUB_REF_NAME#v}"
+          else
+            VERSION="${{ github.event.inputs.version }}"
+          fi
+
+          echo "version=$VERSION" >> "$GITHUB_OUTPUT"
+
+      - name: Setup MSVC
+        if: matrix.platform == 'windows'
+        uses: ilammy/msvc-dev-cmd@v1
+
+      - name: Setup vcpkg
+        id: setup-vcpkg
+        if: matrix.platform == 'windows'
+        uses: lukka/run-vcpkg@v11
+        with:
+          runVcpkgInstall: true
+
+      - name: Configure
+        if: matrix.platform == 'windows'
+        run: >
+          cmake
+          -B build
+          -G Ninja
+          -DCMAKE_BUILD_TYPE=Release
+          "-DCUBED_VERSION=${{ steps.version.outputs.version }}"
+          "-DCMAKE_TOOLCHAIN_FILE=${{ steps.setup-vcpkg.outputs.vcpkg-root }}/scripts/buildsystems/vcpkg.cmake"
+
+      - name: Build
+        run: cmake --build build --config Release --target Cubed
+
+      - name: Copy assets
+        if: matrix.platform == 'windows'
+        shell: pwsh
+        run: |
+          if (Test-Path assets) {
+            Copy-Item -Path assets -Destination build/Cubed -Recurse
+          } else {
+            Write-Warning "assets folder not found, skipping copy"
+          }
+
+      - name: Create Zip
+        if: matrix.platform == 'windows'
+        shell: pwsh
+        run: |
+          Compress-Archive `
+            -Path build/Cubed/* `
+            -DestinationPath Cubed-${{ steps.version.outputs.version }}-windows-x64.zip
+      
+      ####################################################################
+      # Upload
+      ####################################################################
+
+      - name: Upload Windows Artifacts
+        if: matrix.platform == 'windows'
+        uses: actions/upload-artifact@v4
+        with:
+          name: windows
+          path: |
+            *.zip
+
+  ##########################################################################
+  # Release
+  ##########################################################################
+
+  release:
+
+    needs: build
+
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 1   
+
+      - name: Download Artifacts
+        uses: actions/download-artifact@v4
+        with:
+          path: artifacts
+
+      - name: Generate SHA256
+        run: |
+          cd artifacts
+
+          find . -type f | while read file
+          do
+            sha256sum "$file"
+          done > SHA256SUMS
+          
+      - name: Determine prerelease
+        id: prerelease
+        run: |
+          MAIN_BRANCH="${{ github.event.repository.default_branch }}"
+          git fetch origin "$MAIN_BRANCH" --depth=1 2>/dev/null || true
+
+          if git merge-base --is-ancestor "${{ github.sha }}" "origin/$MAIN_BRANCH"; then
+            echo "is_prerelease=false" >> $GITHUB_OUTPUT
+            echo "is_prerelease=false"
+          else
+            echo "is_prerelease=true" >> $GITHUB_OUTPUT
+            echo "is_prerelease=true"
+          fi
+
+      - name: Release
+        uses: softprops/action-gh-release@v2
+        with:
+          files: |
+            artifacts/**/*
+            artifacts/SHA256SUMS
+          generate_release_notes: true
+          draft: false
+          prerelease: ${{ steps.prerelease.outputs.is_prerelease }}

.gitignore

@@ -42,4 +42,5 @@ CMakeError.log
 .DS_Store
 assets/config.toml
 .venv/
-pyout/
+pyout/
+vcpkg_installed/

CMakeLists.txt

@@ -1,4 +1,4 @@
-cmake_minimum_required(VERSION 3.14...3.24)
+cmake_minimum_required(VERSION 3.21)
 project(Cubed LANGUAGES C CXX)
 
 set(CMAKE_CXX_STANDARD 23)
@@ -6,167 +6,118 @@ set(CMAKE_CXX_STANDARD_REQUIRED ON)
 set(CMAKE_CXX_EXTENSIONS OFF)
 set(CMAKE_EXPORT_COMPILE_COMMANDS ON)
 
-if(NOT CMAKE_BUILD_TYPE)
-    set(CMAKE_BUILD_TYPE Debug)
+if(NOT CMAKE_CONFIGURATION_TYPES
+   AND NOT CMAKE_BUILD_TYPE)
+
+    set(CMAKE_BUILD_TYPE Debug CACHE STRING "" FORCE)
+
+endif()
+
+if(NOT DEFINED CUBED_VERSION)
+    set(CUBED_VERSION "dev")
 endif()
 
 set(CMAKE_RUNTIME_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/${PROJECT_NAME})
 set(CMAKE_LIBRARY_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/${PROJECT_NAME})
 set(CMAKE_ARCHIVE_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/lib)
 
-if(MSVC)
-    add_compile_options(/utf-8)
-endif()
-
-find_package(OpenGL REQUIRED)
-
-if (UNIX AND NOT APPLE)
-    find_package(Freetype REQUIRED)
-    find_package(PkgConfig REQUIRED)
-    pkg_check_modules(EGL REQUIRED egl)
-    pkg_check_modules(Wayland REQUIRED wayland-client wayland-egl)
-    find_package(glfw3 REQUIRED)   
-endif()
-
-add_library(glad STATIC third_party/glad/src/glad.c)
-target_include_directories(glad PUBLIC third_party/glad/include)
-
-include(FetchContent)
-
-if (WIN32)
-    FetchContent_Declare(
-        glfw
-        GIT_REPOSITORY https://github.com/glfw/glfw.git
-        GIT_TAG 3.4  
-    )
-
-    set(GLFW_BUILD_DOCS OFF CACHE BOOL "" FORCE)
-    set(GLFW_BUILD_TESTS OFF CACHE BOOL "" FORCE)
-    set(GLFW_BUILD_EXAMPLES OFF CACHE BOOL "" FORCE)
-    set(GLFW_INSTALL OFF CACHE BOOL "" FORCE)         
-    set(GLFW_VULKAN_STATIC ON CACHE BOOL "" FORCE)     
-    set(GLFW_STATIC ON CACHE BOOL "" FORCE)          
-
-    FetchContent_MakeAvailable(glfw)
-
-    FetchContent_Declare(
-        freetype
-        GIT_REPOSITORY https://gitlab.freedesktop.org/freetype/freetype.git
-        GIT_TAG VER-2-14-3
-    )
-    FetchContent_MakeAvailable(freetype)
-    if(TARGET freetype)
-    add_library(Freetype::Freetype ALIAS freetype)
-    endif()
-    set(_BUILD_SHARED_LIBS_SAVED ${BUILD_SHARED_LIBS})
-    set(BUILD_SHARED_LIBS ON)
-
-    FetchContent_Declare(
-    onetbb                           
-    GIT_REPOSITORY https://github.com/uxlfoundation/oneTBB.git
-    GIT_TAG v2023.0.0
-    )
-    set(BUILD_TESTING OFF CACHE BOOL "Build tests" FORCE)
-    set(TBB_TEST OFF CACHE BOOL "Build TBB tests" FORCE)
-    FetchContent_MakeAvailable(onetbb)
-
-    set(BUILD_SHARED_LIBS ${_BUILD_SHARED_LIBS_SAVED})
-    unset(_BUILD_SHARED_LIBS_SAVED)
-
-endif()
-
-FetchContent_Declare(
-    glm
-    GIT_REPOSITORY https://github.com/g-truc/glm.git
-    GIT_TAG 1.0.3 
-    
+list(APPEND CMAKE_MODULE_PATH 
+    "${CMAKE_CURRENT_SOURCE_DIR}/cmake"
+    "${CMAKE_CURRENT_SOURCE_DIR}/cmake/modules"
 )
-FetchContent_MakeAvailable(glm)
-FetchContent_Declare(
-    soil2
-    GIT_REPOSITORY https://github.com/SpartanJ/SOIL2.git
-    GIT_TAG 1.31  
-    
-)
-FetchContent_MakeAvailable(soil2)
-FetchContent_Declare(
-    tomlplusplus
-    GIT_REPOSITORY https://github.com/marzer/tomlplusplus.git
-    GIT_TAG v3.4.0
-)
-FetchContent_MakeAvailable(tomlplusplus)
+
+include(Dependencies)
+
+add_subdirectory(third_party/glad)
 
 add_subdirectory(third_party/imgui)
 
-set(INCLUDE_DIR ${PROJECT_SOURCE_DIR}/include)
+add_executable(${PROJECT_NAME})
 
-add_executable(${PROJECT_NAME}
-    src/main.cpp
-    src/app.cpp
-    src/debug_collector.cpp
-    src/camera.cpp
-    src/config.cpp
-    src/dev_panel.cpp
-    src/gameplay/biome.cpp
-    src/gameplay/chunk.cpp
-    src/gameplay/chunk_generator.cpp
-    src/gameplay/player.cpp
-    src/gameplay/tree.cpp
-    src/gameplay/world.cpp
-    src/input.cpp
-    src/map_table.cpp
-    src/renderer.cpp
-    src/shader.cpp
-    src/texture_manager.cpp
-    src/tools/cubed_random.cpp
-    src/tools/math_tools.cpp
-    src/tools/shader_tools.cpp
-    src/tools/font.cpp
-    src/tools/perlin_noise.cpp
-    src/ui/text.cpp
-    src/window.cpp
-    src/gameplay/builders/biome_builder.cpp
-    src/gameplay/builders/plain_builder.cpp
-    src/gameplay/builders/mountain_builder.cpp
-    src/gameplay/builders/river_builder.cpp
-    src/gameplay/builders/desert_builder.cpp
-    src/gameplay/builders/forest_builder.cpp
-    src/gameplay/cave_carver.cpp
-    src/gameplay/cave_path.cpp
-    src/gameplay/builders/snowy_plain_builder.cpp
-    src/gameplay/river_worm.cpp
-    src/gameplay/river_path.cpp
-    src/block.cpp
-    src/gameplay/vertex_data.cpp
-    src/gameplay/builders/ocean_builder.cpp
+add_subdirectory(src)
+
+file(GLOB_RECURSE PROTO_FILES
+    ${CMAKE_CURRENT_SOURCE_DIR}/src/proto/*.proto
 )
 
-if(CMAKE_BUILD_TYPE STREQUAL "Debug")
-    message(STATUS "Building with AddressSanitizer enabled for target: ${PROJECT_NAME}")
+protobuf_generate(
+    TARGET ${PROJECT_NAME}           
+    LANGUAGE cpp         
+    PROTOS ${PROTO_FILES}
+    IMPORT_DIRS ${CMAKE_CURRENT_SOURCE_DIR}/src/proto
+)
 
-    target_compile_options(${PROJECT_NAME} PRIVATE
-        #-fsanitize=address
-        #-fsanitize=thread
-        -fno-omit-frame-pointer   
-        -g                        
-    )
+configure_file(
+    src/version.hpp.in
+    ${CMAKE_BINARY_DIR}/generated/version.hpp
+    @ONLY
+)
 
-    target_link_options(${PROJECT_NAME} PRIVATE
-        #-fsanitize=address
-        #-fsanitize=thread
-    )
+target_compile_options(${PROJECT_NAME}
+    PRIVATE
+        $<$<CONFIG:Debug>:-fno-omit-frame-pointer>
+        $<$<CONFIG:Debug>:-g>
+        #$<$<CONFIG:Debug>:-fsanitize=address>
+        #$<$<CONFIG:Debug>:-fsanitize=thread>
 
-    target_compile_definitions(${PROJECT_NAME} PRIVATE DEBUG_MODE)
-    target_compile_definitions(${PROJECT_NAME} PRIVATE 
-        ASSETS_PATH="${CMAKE_SOURCE_DIR}/assets/"
+        $<$<CXX_COMPILER_ID:GNU,Clang>:-Wall>
+        $<$<CXX_COMPILER_ID:GNU,Clang>:-Wextra>
+        $<$<CXX_COMPILER_ID:GNU,Clang>:-Wpedantic>
+
+        $<$<CXX_COMPILER_ID:MSVC>:/utf-8>
+        $<$<CXX_COMPILER_ID:MSVC>:/W4>
+)
+
+target_link_options(${PROJECT_NAME}
+    PRIVATE
+        #$<$<CONFIG:Debug>:-fsanitize=address>
+        #$<$<CONFIG:Debug>:-fsanitize=thread>
+)
+
+target_compile_definitions(${PROJECT_NAME}
+    PRIVATE
+        ASIO_STANDALONE
+        ASIO_NO_DEPRECATED
+        $<$<CONFIG:Debug>:DEBUG_MODE>
+        $<$<CXX_COMPILER_ID:MSVC>:
+            WIN32_LEAN_AND_MEAN
+            NOMINMAX
+            _CRT_SECURE_NO_WARNINGS
+        >
+)
+
+if(CMAKE_CONFIGURATION_TYPES)
+    # Visual Studio / Xcode multi-configuration generator
+    target_compile_definitions(${PROJECT_NAME}
+        PRIVATE
+            ASSETS_PATH="$<$<CONFIG:Debug>:${PROJECT_SOURCE_DIR}/assets/>$<$<NOT:$<CONFIG:Debug>>:./assets/>"
     )
 else()
-    target_compile_definitions(${PROJECT_NAME} PRIVATE 
-        ASSETS_PATH="./assets/"
-    )
+    # Ninja / Makefiles single-configuration generator
+    if(CMAKE_BUILD_TYPE STREQUAL "Debug")
+        target_compile_definitions(${PROJECT_NAME}
+            PRIVATE
+                ASSETS_PATH="${PROJECT_SOURCE_DIR}/assets/"
+        )
+    else()
+        target_compile_definitions(${PROJECT_NAME}
+            PRIVATE
+                ASSETS_PATH="./assets/"
+        )
+    endif()
 endif()
 
-target_include_directories(${PROJECT_NAME} PUBLIC ${INCLUDE_DIR})
+
+target_include_directories(${PROJECT_NAME}
+    PRIVATE
+        ${PROJECT_SOURCE_DIR}/include
+        ${PROJECT_SOURCE_DIR}/third_party/asio/include
+        ${CMAKE_BINARY_DIR}/generated
+        ${PROJECT_BINARY_DIR}
+        ${PROJECT_BINARY_DIR}/src
+
+)
+
 
 target_link_libraries(${PROJECT_NAME}
     PRIVATE
@@ -179,42 +130,24 @@ target_link_libraries(${PROJECT_NAME}
         tomlplusplus::tomlplusplus
         imgui
         tbb
+        protobuf::libprotobuf
+        absl::log
+        absl::check
+        absl::base
+        absl::strings
+        absl::flat_hash_map
+        zstd::zstd
+        $<$<PLATFORM_ID:Windows>:ws2_32>
+
 )
 
-if (CMAKE_CXX_COMPILER_ID STREQUAL "GNU" OR CMAKE_CXX_COMPILER_ID STREQUAL "Clang")
-#    target_link_libraries(${PROJECT_NAME} PRIVATE tbb)
-endif()
-
-if (UNIX AND NOT APPLE)
-    target_link_libraries(${PROJECT_NAME}
-        PRIVATE
-            ${EGL_LIBRARIES}
-            ${Wayland_LIBRARIES}
+if(WIN32)
+    add_custom_command(
+        TARGET ${PROJECT_NAME}
+        POST_BUILD
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different
+            $<TARGET_RUNTIME_DLLS:${PROJECT_NAME}>
+            $<TARGET_FILE_DIR:${PROJECT_NAME}>
+        COMMAND_EXPAND_LISTS
     )
-
-
-    target_include_directories(${PROJECT_NAME}
-        PRIVATE
-            ${EGL_INCLUDE_DIRS}
-            ${Wayland_INCLUDE_DIRS}
-    )
-
-
-    target_compile_options(${PROJECT_NAME} PRIVATE ${EGL_CFLAGS_OTHER} ${Wayland_CFLAGS_OTHER})
-endif()
-
-if (WIN32)
-    foreach(TBB_LIB IN ITEMS tbb tbbmalloc tbbmalloc_proxy)
-        if(TARGET ${TBB_LIB})
-            add_custom_command(
-                TARGET ${PROJECT_NAME} POST_BUILD
-                COMMAND ${CMAKE_COMMAND} -E copy_if_different
-                    $<TARGET_FILE:${TBB_LIB}>
-                    $<TARGET_FILE_DIR:${PROJECT_NAME}>
-                COMMENT "Copying ${TBB_LIB}.dll"
-            )
-        else()
-            message(STATUS "Target ${TBB_LIB} not found, skipping copy")
-        endif()
-    endforeach()
 endif()

assets/shaders/player_f_shader.glsl

@@ -0,0 +1,8 @@
+#version 460
+
+out vec4 color;
+in vec2 tc;
+
+void main() {
+    color = vec4(0.0, 0.0, 1.0, 1.0);
+}

assets/shaders/player_v_shader.glsl

@@ -0,0 +1,15 @@
+#version 460
+
+layout (location = 0) in vec3 pos;
+layout (location = 1) in vec2 texCoord;
+
+uniform mat4 mv_matrix;
+uniform mat4 proj_matrix;
+
+out vec2 tc;
+
+void main() {
+    vec4 viewPos = mv_matrix * vec4(pos, 1.0);
+    tc = texCoord;
+    gl_Position = proj_matrix * viewPos;
+}

cmake/Dependencies.cmake

@@ -0,0 +1,79 @@
+include(FetchContent)
+
+# System packages
+find_package(OpenGL REQUIRED)
+find_package(Protobuf REQUIRED)
+find_package(absl REQUIRED)
+find_package(zstd REQUIRED)
+if (UNIX AND NOT APPLE)
+    find_package(Freetype REQUIRED)
+    find_package(glfw3 REQUIRED)   
+endif()
+
+
+# Third-party libraries
+FetchContent_Declare(
+    glm
+    GIT_REPOSITORY https://github.com/g-truc/glm.git
+    GIT_TAG 1.0.3 
+    
+)
+FetchContent_MakeAvailable(glm)
+FetchContent_Declare(
+    soil2
+    GIT_REPOSITORY https://github.com/SpartanJ/SOIL2.git
+    GIT_TAG 1.31  
+    
+)
+FetchContent_MakeAvailable(soil2)
+FetchContent_Declare(
+    tomlplusplus
+    GIT_REPOSITORY https://github.com/marzer/tomlplusplus.git
+    GIT_TAG v3.4.0
+)
+FetchContent_MakeAvailable(tomlplusplus)
+
+
+if (WIN32)
+    FetchContent_Declare(
+        glfw
+        GIT_REPOSITORY https://github.com/glfw/glfw.git
+        GIT_TAG 3.4  
+    )
+
+    set(GLFW_BUILD_DOCS OFF CACHE BOOL "" FORCE)
+    set(GLFW_BUILD_TESTS OFF CACHE BOOL "" FORCE)
+    set(GLFW_BUILD_EXAMPLES OFF CACHE BOOL "" FORCE)
+    set(GLFW_INSTALL OFF CACHE BOOL "" FORCE)         
+    set(GLFW_VULKAN_STATIC ON CACHE BOOL "" FORCE)             
+    set(BUILD_SHARED_LIBS ON CACHE BOOL "" FORCE)
+
+    FetchContent_MakeAvailable(glfw)
+
+    FetchContent_Declare(
+        freetype
+        GIT_REPOSITORY https://gitlab.freedesktop.org/freetype/freetype.git
+        GIT_TAG VER-2-14-3
+    )
+    FetchContent_MakeAvailable(freetype)
+    if(TARGET freetype)
+    add_library(Freetype::Freetype ALIAS freetype)
+    endif()
+    set(_BUILD_SHARED_LIBS_SAVED ${BUILD_SHARED_LIBS})
+    set(BUILD_SHARED_LIBS ON)
+
+    FetchContent_Declare(
+    onetbb                           
+    GIT_REPOSITORY https://github.com/uxlfoundation/oneTBB.git
+    GIT_TAG v2023.0.0
+    )
+    
+    set(BUILD_TESTING OFF CACHE BOOL "Build tests" FORCE)
+    set(TBB_TEST OFF CACHE BOOL "Build TBB tests" FORCE)
+    FetchContent_MakeAvailable(onetbb)
+
+    set(BUILD_SHARED_LIBS ${_BUILD_SHARED_LIBS_SAVED})
+    unset(_BUILD_SHARED_LIBS_SAVED)
+
+endif()
+

cmake/modules/Findzstd.cmake

@@ -0,0 +1,22 @@
+find_path(ZSTD_INCLUDE_DIRS
+  NAMES zstd.h
+  HINTS ${zstd_ROOT_DIR}/include)
+
+find_library(ZSTD_LIBRARIES
+  NAMES zstd
+  HINTS ${zstd_ROOT_DIR}/lib)
+
+include(FindPackageHandleStandardArgs)
+find_package_handle_standard_args(zstd DEFAULT_MSG ZSTD_LIBRARIES ZSTD_INCLUDE_DIRS)
+
+mark_as_advanced(
+  ZSTD_LIBRARIES
+  ZSTD_INCLUDE_DIRS)
+
+if(ZSTD_FOUND AND NOT (TARGET zstd::zstd))
+  add_library (zstd::zstd UNKNOWN IMPORTED)
+  set_target_properties(zstd::zstd
+    PROPERTIES
+      IMPORTED_LOCATION ${ZSTD_LIBRARIES}
+      INTERFACE_INCLUDE_DIRECTORIES ${ZSTD_INCLUDE_DIRS})
+endif()

include/Cubed/app.hpp

@@ -1,8 +1,10 @@
 #pragma once
+#include "Cubed/gameplay/client_world.hpp"
+#include "Cubed/gameplay/network_server.hpp"
+#include "Cubed/gameplay/server_world.hpp"
 #define GLFW_INCLUDE_NONE
 #include "Cubed/camera.hpp"
 #include "Cubed/dev_panel.hpp"
-#include "Cubed/gameplay/world.hpp"
 #include "Cubed/renderer.hpp"
 #include "Cubed/texture_manager.hpp"
 #include "Cubed/window.hpp"
@@ -10,6 +12,13 @@ namespace Cubed {
 
 class App {
 public:
+    struct Argument {
+        bool is_client = false;
+        int port = 25530;
+        std::string ip{"127.0.0.1"};
+        std::string player{"Unknown"};
+    };
+
     App();
     ~App();
     static void cursor_position_callback(GLFWwindow* window, double xpos,
@@ -36,14 +45,20 @@ public:
     Renderer& renderer();
     TextureManager& texture_manager();
     Window& window();
-    World& world();
+    ClientWorld& client_world();
+    ServerWorld& server_world();
+    const Argument& argument() const;
 
 private:
     Camera m_camera;
     TextureManager m_texture_manager;
-    World m_world;
+    NetworkServer m_server;
+    std::shared_ptr<NetworkClient> m_client;
+    ClientWorld m_client_world;
+
     DevPanel m_dev_panel{*this};
-    Renderer m_renderer{m_camera, m_world, m_texture_manager, m_dev_panel};
+    Renderer m_renderer{m_camera, m_client_world, m_texture_manager,
+                        m_dev_panel};
 
     Window m_window{m_renderer};
 
@@ -53,9 +68,10 @@ private:
     inline static double fps_time_count = 0.0f;
     inline static int frame_count = 0;
     inline static int fps = 0;
-
-    void init();
-
+    Argument m_argument;
+    void init(int argc, char** argv);
+    void handle_argument(int argc, char** argv);
+    void handle_toml();
     auto init_camera();
     auto init_texture();
     auto init_world();

include/Cubed/camera.hpp

@@ -8,12 +8,12 @@
 
 namespace Cubed {
 
-class Player;
+class ClientPlayer;
 
 class Camera {
 private:
     bool m_firse_mouse = true;
-    Player* m_player;
+    ClientPlayer* m_player;
     float m_last_mouse_x, m_last_mouse_y;
     glm::vec3 m_camera_pos;
     bool m_under_water = false;
@@ -23,7 +23,7 @@ public:
 
     void update_move_camera();
 
-    void camera_init(Player* player);
+    void camera_init(ClientPlayer* player);
     void hot_reload();
     void reset_camera();
     void update_cursor_position_camera(double xpos, double ypos);

include/Cubed/config.hpp

@@ -1,17 +1,9 @@
 #pragma once
 #include "Cubed/tools/cubed_assert.hpp"
-
-#include <toml++/toml.hpp>
+#include "Cubed/tools/toml.utils.hpp"
 
 namespace Cubed {
 
-template <typename T>
-concept TomlValueType =
-    std::same_as<T, int> || std::same_as<T, bool> || std::same_as<T, double> ||
-    std::same_as<T, const char*> || std::same_as<T, toml::date> ||
-    std::same_as<T, toml::time> || std::same_as<T, toml::date_time> ||
-    std::same_as<T, std::string>;
-
 class Config {
 public:
     Config();
@@ -24,7 +16,7 @@ public:
     void load_or_create_config();
     void save_to_file();
 
-    template <TomlValueType T> T get(std::string_view key) const {
+    template <TOML::TomlValueType T> T get(std::string_view key) const {
         size_t cur = 0;
         auto pos = key.find('.');
         const toml::table* table = &m_tbl;
@@ -61,7 +53,7 @@ public:
         }
     }
     template <typename T> void set(std::string_view key, T&& val) {
-        if constexpr (!TomlValueType<std::decay_t<T>>) {
+        if constexpr (!TOML::TomlValueType<std::decay_t<T>>) {
             static_assert(false, "Type Not Support");
         }
         size_t cur = 0;

include/Cubed/constants.hpp

@@ -1,5 +1,4 @@
 #pragma once
-#include "Cubed/gameplay/chunk_pos.hpp"
 
 #include <array>
 namespace Cubed {
@@ -26,13 +25,10 @@ constexpr float DEFAULT_G = 22.5f;
 constexpr int SIZE_X = CHUNK_SIZE;
 constexpr int SIZE_Y = WORLD_SIZE_Y;
 constexpr int SIZE_Z = CHUNK_SIZE;
-constexpr int RESERVED_THREADS = 3;
+constexpr int RESERVED_THREADS = 5;
 
 constexpr float DEFAULT_CAVE_PROBABILITY = 0.035f;
 
-constexpr ChunkPos CHUNK_DIR[]{{1, 0}, {-1, 0}, {0, 1},  {0, -1},
-                               {1, 1}, {-1, 1}, {1, -1}, {-1, -1}};
-
 using HeightMapArray = std::array<std::array<int, CHUNK_SIZE>, CHUNK_SIZE>;
 
 } // namespace Cubed

include/Cubed/dev_panel.hpp

@@ -5,7 +5,7 @@
 namespace Cubed {
 
 class App;
-class Player;
+class ClientPlayer;
 class DevPanel {
     struct ConfigView {
         float fov = 70.0f;
@@ -26,9 +26,6 @@ class DevPanel {
         int gait = 0;
         float pos[3] = {0.0f, 0.0f, 0.0f};
     };
-    struct TextEditing {
-        bool perlin_seed = false;
-    };
 
 public:
     DevPanel(App& app);
@@ -38,9 +35,8 @@ public:
 private:
     App& m_app;
     ConfigView m_config;
-    Player* m_player;
+    ClientPlayer* m_player;
     PlayerProfile m_player_profile;
-    TextEditing m_text_editing;
     bool m_need_save_config = false;
     bool m_gen_thread_running = true;
     int m_theme = 0;
@@ -58,6 +54,8 @@ private:
     void show_chunk_table_bar();
     void show_settings_tab_item();
     void show_world_tab_item();
+    void show_server_world_table_bar();
+    void show_client_world_table_bar();
     void show_player_tab_item();
     void show_items_tab_item();
     void show_shader_tab_item();

include/Cubed/gameplay/biome.hpp

@@ -1,6 +1,9 @@
 #pragma once
+#include "Cubed/tools/cubed_assert.hpp"
+
 #include <array>
 #include <string>
+#include <utility>
 #include <vector>
 
 namespace Cubed {
@@ -77,4 +80,32 @@ DesertParams& desert_params();
 MountainParams& mountain_params();
 RiverParams& river_params();
 
+inline BiomeType get_biome_from_id(int id) {
+    using enum BiomeType;
+    auto to = std::to_underlying<BiomeType>;
+    if (id == to(PLAIN)) {
+        return PLAIN;
+    }
+    if (id == to(FOREST)) {
+        return FOREST;
+    }
+    if (id == to(DESERT)) {
+        return DESERT;
+    }
+    if (id == to(MOUNTAIN)) {
+        return MOUNTAIN;
+    }
+    if (id == to(RIVER)) {
+        return RIVER;
+    }
+    if (id == to(SNOWY_PLAIN)) {
+        return SNOWY_PLAIN;
+    }
+    if (id == to(OCEAN)) {
+        return OCEAN;
+    }
+    ASSERT_MSG(false, "Unknown Biome Id");
+    throw std::invalid_argument("Unknown Biome Id");
+}
+
 } // namespace Cubed

include/Cubed/gameplay/block.hpp

@@ -2,12 +2,15 @@
 
 #include <glad/glad.h>
 #include <glm/glm.hpp>
+#include <optional>
 #include <string>
 #include <vector>
 
 namespace Cubed {
 
 using BlockType = uint8_t;
+using OptionalBlockVectorArray =
+    std::array<std::optional<std::vector<BlockType>>, 4>;
 
 struct BlockTexture {
     std::string name;

include/Cubed/gameplay/cave_carver.hpp

@@ -1,5 +1,6 @@
 #pragma once
 #include "Cubed/constants.hpp"
+#include "Cubed/gameplay/chunk_pos.hpp"
 #include "Cubed/gameplay/path.hpp"
 
 #include <tbb/concurrent_hash_map.h>

include/Cubed/gameplay/chunk_generator.hpp

@@ -11,11 +11,11 @@
 #include <optional>
 namespace Cubed {
 
-class Chunk;
+class ServerChunk;
 
 class ChunkGenerator {
 public:
-    ChunkGenerator(Chunk& chunk);
+    ChunkGenerator(ServerChunk& chunk);
 
     static void init();
     static void reload();
@@ -26,7 +26,7 @@ public:
     void assign_chunk_biome();
     // Adjust Biome
     void resolve_biome_adjacency_conflict(
-        const std::array<const Chunk*, 8>& adj_chunks);
+        const std::array<const ServerChunk*, 8>& adj_chunks);
     // Generate Heightmap
     void generate_heightmap();
     // Adjust Height
@@ -42,7 +42,7 @@ public:
     // Generate Structure
     void generate_vegetation();
     BiomeType get_biome_at(float world_x, float world_z);
-    Chunk& chunk();
+    ServerChunk& chunk();
     Random& random();
     const std::array<BiomeType, 8>& neighbor_biome() const;
     void ocean_build();
@@ -53,7 +53,7 @@ private:
     static inline std::atomic<bool> is_init{false};
     static inline unsigned m_generator_seed{0};
     static inline std::atomic<bool> is_seed_change{false};
-    Chunk& m_chunk;
+    ServerChunk& m_chunk;
     Random m_random;
     std::unique_ptr<BiomeBuilder> m_biome_builder{nullptr};
     bool is_cur_chunk_ins = false;

include/Cubed/gameplay/chunk_pos.hpp

@@ -1,5 +1,7 @@
 #pragma once
 
+#include "Cubed/constants.hpp"
+
 #include <functional>
 
 namespace Cubed {
@@ -34,5 +36,29 @@ struct ChunkPos {
         return *this;
     };
 };
+constexpr ChunkPos CHUNK_DIR[]{{1, 0}, {-1, 0}, {0, 1},  {0, -1},
+                               {1, 1}, {-1, 1}, {1, -1}, {-1, -1}};
+inline ChunkPos get_chunk_pos(int world_x, int world_z) {
+    int chunk_x, chunk_z;
+    if (world_x < 0) {
+        chunk_x = (world_x + 1) / CHUNK_SIZE - 1;
+    }
+    if (world_x >= 0) {
+        chunk_x = world_x / CHUNK_SIZE;
+    }
+    if (world_z < 0) {
+        chunk_z = (world_z + 1) / CHUNK_SIZE - 1;
+    }
+    if (world_z >= 0) {
+        chunk_z = world_z / CHUNK_SIZE;
+    }
+    return {chunk_x, chunk_z};
+}
+
+inline float distance2(const ChunkPos& a, const ChunkPos& b) {
+    float dx = static_cast<float>(a.x) - b.x;
+    float dz = static_cast<float>(a.z) - b.z;
+    return dx * dx + dz * dz;
+}
 
 } // namespace Cubed

include/Cubed/gameplay/client_chunk.hpp

@@ -1,94 +1,42 @@
 #pragma once
-
+#include "Cubed/constants.hpp"
 #include "Cubed/gameplay/biome.hpp"
 #include "Cubed/gameplay/block.hpp"
-#include "Cubed/gameplay/chunk_generator.hpp"
 #include "Cubed/gameplay/chunk_pos.hpp"
 #include "Cubed/gameplay/vertex_data.hpp"
+#include "world/chunk_data.pb.h"
 
 #include <atomic>
+#include <glad/glad.h>
+#include <glm/glm.hpp>
 #include <mutex>
 namespace Cubed {
-
-struct ChunkInfo {
-    ChunkPos pos{0, 0};
-    unsigned seed{0};
-    BiomeType biome{BiomeType::NONE};
-    unsigned first_random{0};
-    bool has_cave_start{false};
-    bool has_cave{false};
+class ClientWorld;
+struct ChunkRenderSnapshot {
+    GLuint normal_vao;
+    size_t normal_vertices_count;
+    GLuint cross_vao;
+    size_t cross_vertices_count;
+    GLuint normal_discard_vao;
+    size_t normal_discard_vertices_count;
+    GLuint normal_blend_vao;
+    size_t normal_blend_vertices_count;
+    GLuint water_vao;
+    size_t water_vertices_count;
+    glm::vec3 center;
+    glm::vec3 half_extents;
 };
-
-class World;
-// if want to use, do init_chunk(), gen_vertex_data() and
-class Chunk {
-private:
-    using OptionalBlockVectorArray =
-        std::array<std::optional<std::vector<BlockType>>, 4>;
-
-    struct FaceKey {
-        BlockType block_id = 0;
-        int face = -1; // 0-5, used to index NORMALS/TANGENTS/TEX_COORDS
-
-        bool valid() const { return block_id != 0; }
-        bool operator==(const FaceKey& o) const {
-            return block_id == o.block_id && face == o.face;
-        }
-        bool operator!=(const FaceKey& o) const { return !(*this == o); }
-    };
-
-    static constexpr int SIZE_X = CHUNK_SIZE;
-    static constexpr int SIZE_Y = WORLD_SIZE_Y;
-    static constexpr int SIZE_Z = CHUNK_SIZE;
-    static constexpr int VERTEX_DATA_SUM = 5;
-    std::atomic<bool> m_dirty{false};
-    std::atomic<bool> m_need_upload{true};
-    std::atomic<bool> m_is_on_gen_vertex_data{false};
-    std::atomic<bool> m_gening{false};
-    std::atomic<bool> m_temp_chunk{false};
-
-    bool m_has_cave{false};
-
-    std::atomic<BiomeType> m_biome = BiomeType::PLAIN;
-    std::mutex m_vertexs_data_mutex;
-
-    std::unique_ptr<ChunkGenerator> m_generator;
-
-    ChunkPos m_chunk_pos;
-    World& m_world;
-    HeightMapArray m_heightmap;
-    // the index is a array of block id
-    std::vector<BlockType> m_blocks;
-
-    /*
-    0 - normal
-    1 - cross_plane
-    2 - normal_discard
-    3 - transparent and blend
-    4 - water
-    */
-    std::vector<VertexData> m_vertex_data;
-    float frequency = 0.01f;
-    float height = 80;
-    unsigned m_seed = 0;
-
-    BiomeConditions m_conditions;
-    ChunkInfo m_info;
-    void clear_dirty();
-    void gen_vertices(const OptionalBlockVectorArray& neighbor_block);
-    void gen_cross_plane_vertices(int world_x, int world_y, int world_z,
-                                  BlockType id);
-    void emit_quad(int axis, int face_dir, int layer, int i, int j, int w,
-                   int h, int u_axis, int v_axis, FaceKey key);
-
+class ClientChunk {
 public:
-    Chunk(World& world, ChunkPos chunk_pos, bool temp_chunk = false);
-    ~Chunk();
-    Chunk(const Chunk&) = delete;
-    Chunk& operator=(const Chunk&) = delete;
-    Chunk(Chunk&&) noexcept;
-    Chunk& operator=(Chunk&&) noexcept;
+    ClientChunk(ClientWorld& world);
+    ~ClientChunk();
+    ClientChunk(const ClientChunk&) = delete;
+    ClientChunk& operator=(const ClientChunk&) = delete;
+    ClientChunk(ClientChunk&&) noexcept;
+    ClientChunk& operator=(ClientChunk&&) noexcept;
 
+    static int index(int x, int y, int z);
+    static int index(const glm::vec3& pos);
     static std::tuple<int, int, int> world_to_block(int world_x, int world_y,
                                                     int world_z, int chunk_x,
                                                     int chunk_z);
@@ -101,33 +49,9 @@ public:
     BiomeType get_biome() const;
     ChunkPos get_chunk_pos() const;
     const std::vector<BlockType>& get_chunk_blocks() const;
-    HeightMapArray get_heightmap() const;
-    static int index(int x, int y, int z);
-    static int index(const glm::vec3& pos);
-    // Init Chunk
-    // Determine biome from temperature and humidity noise
-    void gen_phase_one();
-    // Resolve biome adjacency conflicts with neighbor chunks
-    void gen_phase_two(const std::array<const Chunk*, 8>& adj_chunks);
-    // Generate heightmap using biome-specific noise
-    void gen_phase_three();
-    // Blend heightmap with neighbors for smooth transitions
-    void gen_phase_four(
-        const std::array<std::optional<HeightMapArray>, 8>& neighbor_heightmap,
-        const std::array<BiomeType, 8>& neighbor_biome);
-    // Generate terrain blocks from heightmap and biome
-    void gen_phase_five();
-    // Blend surface blocks at chunk borders with neighbors
-    void gen_phase_six(const std::array<std::optional<std::vector<BlockType>>,
-                                        4>& neighbor_block);
-    // Generate biome-specific vegetation/structures
-    void gen_phase_seven();
-    // void gen_vertex_data();
-    //  0 : (1, 0)
-    //  1 : (-1, 0)
-    //  2 : (0, 1)
-    //  3 : (0, -1)
+    void receive_chunk(const ChunkDataRsp& data);
     void gen_vertex_data(const OptionalBlockVectorArray& neighbor_block);
+    // Can only be called on the render thread
     void upload_to_gpu();
 
     GLuint get_normal_vao() const;
@@ -152,20 +76,58 @@ public:
     void need_upload();
 
     void set_chunk_block(int index, unsigned id);
-    // ensure thread safe!
-    void gen_chunk();
-
     bool is_temp_chunk() const;
     ChunkPos chunk_pos() const;
     BiomeType biome() const;
     void biome(BiomeType b);
-    HeightMapArray& heightmap();
     std::vector<BlockType>& blocks();
-    World& world();
+    ClientWorld& world();
     unsigned seed() const;
-    BiomeConditions& conditions();
-    ChunkInfo get_info() const;
-    bool& has_cave();
-};
+    const ChunkRenderSnapshot* get_render_snapshot() const;
 
+private:
+    struct FaceKey {
+        BlockType block_id = 0;
+        int face = -1; // 0-5, used to index NORMALS/TANGENTS/TEX_COORDS
+
+        bool valid() const { return block_id != 0; }
+        bool operator==(const FaceKey& o) const {
+            return block_id == o.block_id && face == o.face;
+        }
+        bool operator!=(const FaceKey& o) const { return !(*this == o); }
+    };
+
+    static constexpr int SIZE_X = CHUNK_SIZE;
+    static constexpr int SIZE_Y = WORLD_SIZE_Y;
+    static constexpr int SIZE_Z = CHUNK_SIZE;
+    static constexpr int BLOCK_SIZE = SIZE_X * SIZE_Y * SIZE_Z;
+    static constexpr int VERTEX_DATA_SUM = 5;
+    std::atomic<bool> m_dirty{false};
+    std::atomic<bool> m_need_upload{true};
+    std::atomic<bool> m_is_on_gen_vertex_data{false};
+    std::atomic<BiomeType> m_biome = BiomeType::PLAIN;
+    std::mutex m_vertexs_data_mutex;
+    ChunkPos m_chunk_pos;
+    ClientWorld& m_world;
+    // the index is a array of block id
+    std::vector<BlockType> m_blocks;
+    /*
+    0 - normal
+    1 - cross_plane
+    2 - normal_discard
+    3 - transparent and blend
+    4 - water
+    */
+    std::vector<VertexData> m_vertex_data;
+
+    ChunkRenderSnapshot m_render_snapshot;
+
+    unsigned m_seed = 0;
+    void clear_dirty();
+    void gen_vertices(const OptionalBlockVectorArray& neighbor_block);
+    void gen_cross_plane_vertices(int world_x, int world_y, int world_z,
+                                  BlockType id);
+    void emit_quad(int axis, int face_dir, int layer, int i, int j, int w,
+                   int h, int u_axis, int v_axis, FaceKey key);
+};
 } // namespace Cubed

include/Cubed/gameplay/client_player.hpp

@@ -6,17 +6,61 @@
 #include "Cubed/gameplay/game_mode.hpp"
 #include "Cubed/input.hpp"
 
+#include <absl/container/flat_hash_set.h>
 #include <glm/glm.hpp>
 #include <optional>
-#include <string>
-
+#include <shared_mutex>
 namespace Cubed {
-
 enum class Gait { WALK = 0, RUN };
+class ClientWorld;
+class ClientPlayer {
+public:
+    using ChunkPosSet = absl::flat_hash_set<ChunkPos, ChunkPos::Hash>;
+    ClientPlayer(ClientWorld& world);
+    ~ClientPlayer();
 
-class World;
+    void update_chunk_set(const ChunkPosSet& set);
+    const ChunkPosSet& get_chunk_pos_set() const;
+    ChunkPosSet& get_chunk_pos_set();
+
+    static AABB get_aabb(const glm::vec3& pos);
+    const glm::vec3& get_front() const;
+    const Gait& get_gait() const;
+    const std::optional<LookBlock>& get_look_block_pos() const;
+    // thread safe
+    glm::vec3 get_player_pos() const;
+    const MoveState& get_move_state() const;
+
+    void change_mode(GameMode mode);
+    void hot_reload();
+    void set_player_pos(const glm::vec3& pos);
+    void set_place_block(unsigned id);
+    void update(float delta_time);
+    void update_front_vec(float offset_x, float offset_y);
+    void update_player_move_state(int key, int action);
+    void update_scroll(double yoffset);
+
+    float& max_walk_speed();
+    float& max_run_speed();
+    float& max_speed();
+    float& acceleration();
+    float& deceleration();
+    float& g();
+    float& fly_y_speed();
+
+    unsigned place_block() const;
+
+    Gait& gait();
+    GameMode& game_mode();
+
+    const ClientWorld& get_world() const;
+
+    void set_uuid(std::string_view uuid);
+    const std::string& get_uuid() const;
+    const std::string& get_name() const;
+
+    void init(std::string_view name);
 
-class Player {
 private:
     using enum GameMode;
     float m_max_walk_speed = DEFAULT_MAX_WALK_SPEED;
@@ -24,7 +68,6 @@ private:
     float m_acceleration = DEFAULT_ACCELERATION;
     float m_deceleration = DEFAULT_DECELERATION;
     float m_g = DEFAULT_G;
-
     constexpr static float MAX_SPACE_ON_TIME = 0.3f;
 
     float m_yaw = 0.0f;
@@ -50,63 +93,33 @@ private:
     // player is tow block tall, the pos is the lower pos
 
     glm::vec3 m_player_pos{0.0f, 255.0f, 0.0f};
-    ChunkPos m_player_chunk_pos{0, 0};
+    ChunkPos m_last_chunk_pos{0, 0};
 
     glm::vec3 m_front{0, 0, -1};
     glm::vec3 m_right{0, 0, 0};
-    glm::vec3 m_size{0.6f, 1.8f, 0.6f};
+    static constexpr glm::vec3 M_SIZE{0.6f, 1.8f, 0.6f};
 
     Gait m_gait = Gait::WALK;
     MoveState m_move_state{};
     GameMode m_game_mode = CREATIVE;
     std::optional<LookBlock> m_look_block = std::nullopt;
     std::string m_name{};
-    World& m_world;
+    std::string m_uuid;
+    ClientWorld& m_world;
 
+    mutable std::shared_mutex m_player_pos_mutex;
+    mutable std::shared_mutex m_chunk_pos_mutex;
+    ChunkPosSet m_player_chunk_pos_set;
     bool ray_cast(const glm::vec3& start, const glm::vec3& dir,
                   glm::ivec3& block_pos, glm::vec3& normal,
                   float distance = 4.0f);
 
-    void check_player_chunk_transition();
     void update_direction();
     void update_lookup_block();
     void update_move(float delta_time);
-    void update_x_move();
-    void update_y_move();
-    void update_z_move();
-
-public:
-    Player(World& world, const std::string& name);
-    ~Player();
-    AABB get_aabb() const;
-    const glm::vec3& get_front() const;
-    const Gait& get_gait() const;
-    const std::optional<LookBlock>& get_look_block_pos() const;
-    const glm::vec3& get_player_pos() const;
-    const MoveState& get_move_state() const;
-
-    void change_mode(GameMode mode);
-    void hot_reload();
-    void set_player_pos(const glm::vec3& pos);
-    void set_place_block(unsigned id);
-    void update(float delta_time);
-    void update_front_vec(float offset_x, float offset_y);
-    void update_player_move_state(int key, int action);
-    void update_scroll(double yoffset);
-
-    float& max_walk_speed();
-    float& max_run_speed();
-    float& max_speed();
-    float& acceleration();
-    float& deceleration();
-    float& g();
-    float& fly_y_speed();
-
-    unsigned place_block() const;
-
-    Gait& gait();
-    GameMode& game_mode();
-    const World& get_world() const;
+    void update_x_move(glm::vec3& player_pos);
+    void update_y_move(glm::vec3& player_pos);
+    void update_z_move(glm::vec3& player_pos);
+    void update_player_chunk();
 };
-
 } // namespace Cubed

include/Cubed/gameplay/client_world.hpp

@@ -0,0 +1,138 @@
+#pragma once
+#include "Cubed/gameplay/block.hpp"
+#include "Cubed/gameplay/chunk_pos.hpp"
+#include "Cubed/gameplay/client_chunk.hpp"
+#include "Cubed/gameplay/client_player.hpp"
+#include "Cubed/gameplay/game_time.hpp"
+#include "Cubed/gameplay/network_client.hpp"
+#include "Cubed/tools/priority_thread_pool.hpp"
+
+#include <absl/container/flat_hash_set.h>
+#include <deque>
+#include <tbb/concurrent_hash_map.h>
+#include <tbb/concurrent_queue.h>
+#include <tbb/concurrent_unordered_map.h>
+namespace Cubed {
+
+struct RemotePlayerInfo {
+    std::string name;
+    glm::vec3 render_pos;
+    glm::vec3 target_pos;
+};
+
+struct RemotePlayerRenderData {
+    std::string name;
+    glm::vec3 render_pos;
+};
+
+class ClientWorld {
+public:
+    ClientWorld();
+    ~ClientWorld();
+    void init(std::string_view player_name,
+              std::shared_ptr<NetworkClient> client);
+    void update(float delta_time);
+    const std::optional<LookBlock>& get_look_block_pos() const;
+    ClientPlayer& get_player();
+    int get_block(const glm::ivec3& block_pos) const;
+    bool is_solid(const glm::ivec3& block_pos) const;
+    bool can_pass_block(const glm::ivec3& block_pos) const;
+    BlockType get_block_tpye(const glm::ivec3& block_pos) const;
+
+    void rebuild_world();
+
+    void push_delete_vbo(GLuint vbo);
+    void push_delete_vao(GLuint vao);
+    // void hot_reload();
+
+    // void rebuild_world();
+    void report_block_change(const glm::ivec3& pos, unsigned id) const;
+    void receive_block_change(const BlockChangeRsp& rsp);
+    void receive_time(const UpdateTime& rsp);
+
+    void receive_remote_player(const PlayerInfoRsp& rsp);
+    void receive_player_logout(const LogoutRsp& rsp);
+    int rendering_distance() const;
+    void rendering_distance(int rendering_distance);
+    int get_chunk_task_id() const;
+    void start_client_thread(std::string_view uuid);
+    void stop_client_thread();
+
+    void start_thread_pool();
+    void stop_thread_pool();
+    void change_pool_threads(int threads);
+    void hot_reload();
+    void request_chunk();
+    std::vector<glm::vec4>& planes();
+    const std::vector<const ChunkRenderSnapshot*>& render_snapshots() const;
+    const std::vector<RemotePlayerRenderData>& render_player_data() const;
+    glm::vec3 sunlight_dir() const;
+    void receive_chunk(std::vector<uint8_t> data, PacketHeader header);
+    void request_exit();
+    bool is_receive_exit();
+    int chunk_size() const;
+    static AABB get_block_aabb(const glm::ivec3& pos);
+    template <typename Fn>
+    void register_timer(std::string_view id, TickType threshold, Fn&& f) {
+        m_timers.emplace(std::piecewise_construct,
+                         std::forward_as_tuple(std::string(id)),
+                         std::forward_as_tuple(threshold, std::forward<Fn>(f)));
+    }
+
+private:
+    enum class ChunkLoadStyle { RANDOM, CENTER };
+    using ChunkHashMap =
+        tbb::concurrent_hash_map<ChunkPos, std::shared_ptr<ClientChunk>,
+                                 ChunkPos::TBBHash>;
+    using ChunkPosSet = absl::flat_hash_set<ChunkPos, ChunkPos::Hash>;
+    using ChunkPosVector = std::vector<ChunkPos>;
+    using OtherPlayerHashMap =
+        std::unordered_map<std::string, RemotePlayerInfo>;
+    using chunk_acc = ChunkHashMap::accessor;
+    using chunk_cacc = ChunkHashMap::const_accessor;
+    static constexpr int WORLD_EXIT_TIMEOUT = 200;
+    static constexpr int MAX_UPLOAD_CHUNK_SUM = 16;
+    ClientPlayer m_player;
+    OtherPlayerHashMap m_other_players;
+    ChunkHashMap m_chunks;
+    std::vector<glm::vec4> m_planes;
+    std::jthread m_client_thread;
+
+    std::mutex m_delete_vbo_mutex;
+    std::mutex m_delete_vao_mutex;
+    mutable std::shared_mutex m_other_players_mutex;
+
+    tbb::concurrent_queue<std::unique_ptr<ClientChunk>> m_pending_upload_queue;
+    tbb::concurrent_queue<ChunkPos> m_dirty_chunk_queue;
+
+    std::vector<GLuint> m_pending_delete_vbo;
+    std::vector<GLuint> m_pending_delete_vao;
+
+    std::deque<ChunkPos> m_dirty_queue;
+    std::vector<const ChunkRenderSnapshot*> m_render_snapshots;
+    std::vector<RemotePlayerRenderData> m_render_player_data;
+    tbb::concurrent_unordered_map<std::string, Timer> m_timers;
+    std::atomic<bool> m_game_running{false};
+    std::atomic<bool> m_receive_exit{false};
+    std::atomic<int> m_rendering_distance{24};
+    std::atomic<TickType> m_game_ticks{0};
+    std::atomic<TickType> m_day_tick{6000};
+    std::atomic<bool> m_requesting_chunk{false};
+    std::atomic<bool> m_is_rebuilding{false};
+    std::atomic<int> m_chunk_task_id{0};
+    std::shared_ptr<NetworkClient> m_client;
+    ChunkLoadStyle m_chunk_load_style{ChunkLoadStyle::CENTER};
+
+    std::atomic<std::shared_ptr<PriorityThreadPool>> m_thread_pool;
+
+    void client_run(std::stop_token token);
+
+    void set_player_pos();
+
+    void report_player_pos();
+
+    void set_block(const glm::ivec3& pos, unsigned id);
+
+    void update_chunk(const ChunkPosSet& old, const ChunkPosSet& now);
+};
+} // namespace Cubed

include/Cubed/gameplay/game_time.hpp

@@ -1,10 +1,40 @@
 #pragma once
 
 // Prevent unsigned underflow issues in subtraction
+#include "Cubed/tools/cubed_assert.hpp"
+
+#include <functional>
+#include <utility>
+namespace Cubed {
 using TickType = long long;
 
 constexpr int DEFAULT_PER_TICK_TIME = 50;
 
 constexpr TickType DAY_TIME = 24000;
 
-constexpr TickType PER_HOUR = 1000;
+constexpr TickType PER_HOUR = 1000;
+
+class Timer {
+public:
+    template <typename Fn>
+    Timer(TickType threshold, Fn&& f)
+        : m_fn(std::forward<Fn>(f)), m_threshold(threshold) {
+        ASSERT_MSG(threshold > 0, "Threshold Must Rreater Than 0");
+    }
+    bool update() {
+        if (++m_current >= m_threshold) {
+            m_current = 0;
+            m_fn();
+            return true;
+        }
+        return false;
+    }
+    void reset() { m_current = 0; }
+
+private:
+    std::function<void()> m_fn;
+    TickType m_threshold;
+    TickType m_current = 0;
+};
+
+} // namespace Cubed

include/Cubed/gameplay/network_client.hpp

@@ -0,0 +1,65 @@
+#pragma once
+
+#include "Cubed/gameplay/packet.hpp"
+
+#include <asio.hpp>
+#include <queue>
+#include <string>
+#include <thread>
+namespace Cubed {
+using asio::ip::tcp;
+class ClientWorld;
+class NetworkClient : public std::enable_shared_from_this<NetworkClient> {
+public:
+    NetworkClient(ClientWorld& world);
+    ~NetworkClient();
+    void close();
+    void stop();
+    void send(Packet packet, int priority = 10);
+    void start(std::string ip, int port = 25530);
+    bool is_connected() const;
+    bool is_connect_error() const;
+
+private:
+    struct Task {
+        int priority = 10;
+        std::uint64_t sequence = 0;
+        Packet packet;
+        Task(int p, std::uint64_t seq, Packet pac)
+            : priority(p), sequence(seq), packet(std::move(pac)) {}
+    };
+
+    struct TaskCompare {
+        bool operator()(const Task& a, const Task& b) const {
+
+            if (a.priority != b.priority) {
+                return a.priority > b.priority;
+            }
+
+            return a.sequence > b.sequence;
+        }
+    };
+
+    asio::io_context m_io;
+
+    std::thread m_net_thread;
+    static constexpr uint32_t MAX_PACKET_SIZE = 4 * 1024 * 1024;
+    tcp::socket m_socket;
+    std::vector<char> m_read_buffer;
+
+    std::priority_queue<Task, std::vector<Task>, TaskCompare> m_write_queue;
+
+    asio::strand<asio::io_context::executor_type> m_strand;
+    std::atomic<bool> m_closed{false};
+    std::atomic<bool> m_connected{false};
+    std::atomic<bool> m_connect_error{false};
+    // ClientWorld is managed by App
+    ClientWorld& m_world;
+    std::atomic_uint64_t m_sequence{0};
+
+    asio::awaitable<void> connect(std::string ip, int port);
+    asio::awaitable<void> read_loop();
+
+    void do_write();
+};
+} // namespace Cubed

include/Cubed/gameplay/network_server.hpp

@@ -0,0 +1,33 @@
+#pragma once
+#include "Cubed/gameplay/server_world.hpp"
+#include "Cubed/gameplay/session.hpp"
+
+#include <asio.hpp>
+#include <thread>
+namespace Cubed {
+
+class NetworkServer {
+public:
+    NetworkServer(int port = 25530);
+    ~NetworkServer();
+    void stop();
+
+    // Run in another thread after initialization is complete
+    void start_server(int port = 25530);
+
+    int port() const;
+    ServerWorld& server_world();
+
+private:
+    asio::io_context m_io;
+    std::thread m_net_thread;
+    int m_port = 25530;
+    std::atomic<bool> m_stopped{false};
+    std::atomic<bool> m_started{false};
+    ServerWorld m_world;
+    std::mutex m_session_mutex;
+    std::unordered_map<std::string, std::shared_ptr<Session>> m_session;
+    asio::awaitable<void> listen();
+    void net_run();
+};
+} // namespace Cubed

include/Cubed/gameplay/packet.hpp

@@ -0,0 +1,216 @@
+#pragma once
+#include "Cubed/tools/compression.hpp"
+#include "packet.pb.h" // IWYU pragma: keep
+
+#include <concepts>
+#include <cstdint>
+#include <cstring>
+#include <memory>
+#ifdef _WIN32
+#ifndef WIN32_LEAN_AND_MEAN
+#define WIN32_LEAN_AND_MEAN
+#endif
+#include <winsock2.h>
+#else
+#include <netinet/in.h>
+#endif
+#include <span>
+#include <stdexcept>
+#include <type_traits>
+#include <utility>
+namespace Cubed {
+constexpr size_t HEADER_LEN =
+    sizeof(uint16_t) + sizeof(uint16_t) + sizeof(uint32_t) + sizeof(uint32_t);
+constexpr size_t PACKET_COMPRESSION_THRESHOLD = 100;
+using Packet = std::shared_ptr<std::vector<uint8_t>>;
+enum class CompressType : uint16_t {
+    NONE = 0,
+    ZSTD = 1,
+};
+
+inline CompressType get_compress_type(uint16_t id) {
+    using enum CompressType;
+    switch (id) {
+    case std::to_underlying(NONE):
+        return NONE;
+    case std::to_underlying(ZSTD):
+        return ZSTD;
+    }
+    throw std::runtime_error(std::format("Unknown CompressType {}", id));
+}
+
+struct PacketHeader {
+    uint16_t cmd{};
+    CompressType compress_type{}; // 0=none 1=zlib
+    uint32_t uncompressed_size{};
+    uint32_t compressed_size{};
+};
+
+enum class PacketEnum : uint16_t {
+    LOGIN_REQ = 1001,
+    LOGIN_RSP = 1002,
+    LOGOUT_REQ = 1003,
+    LOGOUT_RSP = 1004,
+    PLAYER_INFO = 2001,
+    PLAYER_POS = 2002,
+    PLAYER_INFO_RSP = 2003,
+    CHUNK_DATA_REQ = 3001,
+    CHUNK_DATA_RSP = 3002,
+    BLOCK_CHANGE_REQ = 3003,
+    BLOCK_CHANGE_RSP = 3004,
+    S2C_CLEAR_ALL_CHUNKS = 3005,
+    UPDATE_TIME = 3006,
+    PING = 9001,
+    PONG = 9002
+
+};
+
+template <typename> struct always_false : std::false_type {}; // NOLINT
+
+template <typename T> constexpr uint16_t get_packet_id() {
+    static_assert(always_false<T>::value, "Unknown Type");
+    return 0;
+}
+
+template <> constexpr uint16_t get_packet_id<LoginReq>() {
+    return std::to_underlying(PacketEnum::LOGIN_REQ);
+}
+template <> constexpr uint16_t get_packet_id<LoginRsp>() {
+    return std::to_underlying(PacketEnum::LOGIN_RSP);
+}
+template <> constexpr uint16_t get_packet_id<LogoutReq>() {
+    return std::to_underlying(PacketEnum::LOGOUT_REQ);
+}
+template <> constexpr uint16_t get_packet_id<LogoutRsp>() {
+    return std::to_underlying(PacketEnum::LOGOUT_RSP);
+}
+template <> constexpr uint16_t get_packet_id<PlayerInfo>() {
+    return std::to_underlying(PacketEnum::PLAYER_INFO);
+}
+template <> constexpr uint16_t get_packet_id<PlayerPos>() {
+    return std::to_underlying(PacketEnum::PLAYER_POS);
+}
+template <> constexpr uint16_t get_packet_id<PlayerInfoRsp>() {
+    return std::to_underlying(PacketEnum::PLAYER_INFO_RSP);
+}
+template <> constexpr uint16_t get_packet_id<ChunkDataReq>() {
+    return std::to_underlying(PacketEnum::CHUNK_DATA_REQ);
+}
+template <> constexpr uint16_t get_packet_id<ChunkDataRsp>() {
+    return std::to_underlying(PacketEnum::CHUNK_DATA_RSP);
+}
+template <> constexpr uint16_t get_packet_id<BlockChangeReq>() {
+    return std::to_underlying(PacketEnum::BLOCK_CHANGE_REQ);
+}
+template <> constexpr uint16_t get_packet_id<BlockChangeRsp>() {
+    return std::to_underlying(PacketEnum::BLOCK_CHANGE_RSP);
+}
+template <> constexpr uint16_t get_packet_id<S2C_ClearAllChunks>() {
+    return std::to_underlying(PacketEnum::S2C_CLEAR_ALL_CHUNKS);
+}
+template <> constexpr uint16_t get_packet_id<UpdateTime>() {
+    return std::to_underlying(PacketEnum::UPDATE_TIME);
+}
+template <> constexpr uint16_t get_packet_id<Ping>() {
+    return std::to_underlying(PacketEnum::PING);
+}
+template <> constexpr uint16_t get_packet_id<Pong>() {
+    return std::to_underlying(PacketEnum::PONG);
+}
+
+template <typename T>
+    requires std::derived_from<T, google::protobuf::Message>
+Packet make_packet(const T& msg) {
+    PacketHeader header{};
+    header.cmd = get_packet_id<T>();
+    uint32_t raw_size = static_cast<uint32_t>(msg.ByteSizeLong());
+    std::vector<uint8_t> raw(raw_size);
+
+    if (!msg.SerializeToArray(raw.data(), raw_size)) {
+        return {};
+    }
+    std::vector<uint8_t> payload;
+    if (raw_size >= PACKET_COMPRESSION_THRESHOLD) {
+        std::vector<uint8_t> compressed = compress_data(raw);
+        if (compressed.size() < raw.size()) {
+            payload = std::move(compressed);
+            header.compress_type = CompressType::ZSTD;
+        } else {
+            payload = std::move(raw);
+            header.compress_type = CompressType::NONE;
+        }
+    } else {
+        payload = std::move(raw);
+        header.compress_type = CompressType::NONE;
+    }
+    header.uncompressed_size = raw_size;
+    header.compressed_size = static_cast<uint32_t>(payload.size());
+
+    auto packet =
+        std::make_shared<std::vector<uint8_t>>(HEADER_LEN + payload.size());
+
+    uint16_t cmd_net = htons(header.cmd);
+    uint16_t compress_type_net =
+        htons(std::to_underlying(header.compress_type));
+    uint32_t uncompressed_size_net = htonl(header.uncompressed_size);
+    uint32_t compressed_size_net = htonl(header.compressed_size);
+
+    std::memcpy(packet->data(), &cmd_net, sizeof(cmd_net));
+
+    std::memcpy(packet->data() + 2, &compress_type_net,
+                sizeof(compress_type_net));
+    std::memcpy(packet->data() + 4, &uncompressed_size_net,
+                sizeof(uncompressed_size_net));
+    std::memcpy(packet->data() + 8, &compressed_size_net,
+                sizeof(compressed_size_net));
+    std::memcpy(packet->data() + HEADER_LEN, payload.data(), payload.size());
+
+    return packet;
+}
+
+inline PacketHeader decode_packet_header(std::span<const uint8_t> header) {
+    if (header.size() < HEADER_LEN)
+        throw std::runtime_error("Invalid header");
+    uint16_t cmd_net;
+    uint16_t compress_type_net;
+    uint32_t uncompressed_size_net;
+    uint32_t compressed_size_net;
+    std::memcpy(&cmd_net, header.data(), sizeof(cmd_net));
+    std::memcpy(&compress_type_net, header.data() + 2,
+                sizeof(compress_type_net));
+    std::memcpy(&uncompressed_size_net, header.data() + 4,
+                sizeof(uncompressed_size_net));
+    std::memcpy(&compressed_size_net, header.data() + 8,
+                sizeof(compressed_size_net));
+
+    return {ntohs(cmd_net), get_compress_type(ntohs(compress_type_net)),
+            ntohl(uncompressed_size_net), ntohl(compressed_size_net)};
+}
+template <typename T>
+    requires std::derived_from<T, google::protobuf::Message>
+bool decode_packet(T& message, std::span<const uint8_t> data,
+                   const PacketHeader& header) {
+    if (data.size() != header.compressed_size) {
+        return false;
+    }
+
+    if (header.compress_type == CompressType::NONE &&
+        header.uncompressed_size != header.compressed_size) {
+        return false;
+    }
+
+    switch (header.compress_type) {
+    case CompressType::NONE: {
+        return message.ParseFromArray(
+            data.data(), static_cast<int>(header.uncompressed_size));
+    }
+    case CompressType::ZSTD: {
+        auto raw = decompress_data(data, header.uncompressed_size);
+        return message.ParseFromArray(raw.data(), static_cast<int>(raw.size()));
+    }
+    default:
+        return false;
+    }
+}
+
+} // namespace Cubed

include/Cubed/gameplay/server_chunk.hpp

@@ -0,0 +1,97 @@
+#pragma once
+#include "Cubed/constants.hpp"
+#include "Cubed/gameplay/biome.hpp"
+#include "Cubed/gameplay/block.hpp"
+#include "Cubed/gameplay/chunk_generator.hpp"
+#include "Cubed/gameplay/chunk_pos.hpp"
+
+#include <array>
+#include <atomic>
+#include <optional>
+#include <tuple>
+namespace Cubed {
+class ServerWorld;
+class ServerChunk {
+public:
+    ServerChunk(ServerWorld& world, ChunkPos chunk_pos,
+                bool temp_chunk = false);
+    ServerChunk(const ServerChunk&) = delete;
+    ServerChunk(ServerChunk&&) noexcept;
+    ServerChunk& operator=(const ServerChunk&) = delete;
+    ServerChunk& operator=(ServerChunk&&) noexcept;
+
+    static std::tuple<int, int, int> world_to_block(int world_x, int world_y,
+                                                    int world_z, int chunk_x,
+                                                    int chunk_z);
+    static std::tuple<int, int, int> world_to_block(const glm::ivec3& block_pos,
+                                                    ChunkPos chunk_pos);
+    static std::tuple<int, int, int> block_to_world(int x, int y, int z,
+                                                    int chunk_x, int chunk_z);
+    static std::tuple<int, int, int> block_to_world(const glm::ivec3& block_pos,
+                                                    ChunkPos chunk_pos);
+
+    void set_chunk_block(int index, unsigned id);
+    // ensure thread safe!
+    void gen_chunk();
+
+    BiomeType get_biome() const;
+    ChunkPos get_chunk_pos() const;
+    const std::vector<BlockType>& get_chunk_blocks() const;
+    HeightMapArray get_heightmap() const;
+    bool is_temp_chunk() const;
+    ChunkPos chunk_pos() const;
+    BiomeType biome() const;
+    void biome(BiomeType b);
+    HeightMapArray& heightmap();
+    std::vector<BlockType>& blocks();
+    ServerWorld& world();
+    unsigned seed() const;
+    BiomeConditions& conditions();
+    bool& has_cave();
+    const OptionalBlockVectorArray& get_neightbor_blocks() const;
+    static int index(int x, int y, int z);
+    static int index(const glm::vec3& pos);
+
+private:
+    static constexpr int SIZE_X = CHUNK_SIZE;
+    static constexpr int SIZE_Y = WORLD_SIZE_Y;
+    static constexpr int SIZE_Z = CHUNK_SIZE;
+
+    std::atomic<bool> m_gening{false};
+    std::atomic<bool> m_temp_chunk{false};
+
+    bool m_has_cave{false};
+
+    std::atomic<BiomeType> m_biome = BiomeType::PLAIN;
+
+    ChunkPos m_chunk_pos;
+    ServerWorld& m_world;
+    HeightMapArray m_heightmap;
+    // the index is a array of block id
+    std::vector<BlockType> m_blocks;
+    OptionalBlockVectorArray m_neightbor_blocks;
+    float frequency = 0.01f;
+    float height = 80;
+    unsigned m_seed = 0;
+
+    BiomeConditions m_conditions;
+
+    std::unique_ptr<ChunkGenerator> m_generator;
+
+    // Init Chunk
+    // Determine biome from temperature and humidity noise
+    void gen_phase_one();
+
+    // Generate heightmap using biome-specific noise
+    void gen_phase_two();
+
+    // Generate terrain blocks from heightmap and biome
+    void gen_phase_three();
+    // Blend surface blocks at chunk borders with neighbors
+    void gen_phase_four(const std::array<std::optional<std::vector<BlockType>>,
+                                         4>& neighbor_block);
+    // Generate biome-specific vegetation/structures
+    void gen_phase_five();
+};
+
+} // namespace Cubed

include/Cubed/gameplay/server_player.hpp

@@ -0,0 +1,54 @@
+#pragma once
+#include "Cubed/gameplay/chunk_pos.hpp"
+#include "Cubed/gameplay/game_time.hpp"
+
+#include <absl/container/flat_hash_set.h>
+#include <atomic>
+#include <glm/glm.hpp>
+#include <memory>
+#include <shared_mutex>
+#include <string>
+#include <string_view>
+namespace Cubed {
+class ServerWorld;
+class Session;
+class ServerPlayer {
+
+public:
+    using ChunkPosSet = absl::flat_hash_set<ChunkPos, ChunkPos::Hash>;
+    ServerPlayer(const ServerPlayer&) = delete;
+    ServerPlayer(ServerPlayer&&) = delete;
+    ServerPlayer& operator=(const ServerPlayer&) = delete;
+    ServerPlayer& operator=(ServerPlayer&&) = delete;
+    ServerPlayer(std::string_view name, std::string_view uuid,
+                 ServerWorld& m_world, std::shared_ptr<Session> session,
+                 TickType gametick);
+
+    const glm::vec3& get_pos() const;
+    const std::string& get_name() const;
+    const std::string& get_uuid() const;
+    std::shared_ptr<Session> get_session() const;
+    void update_pos(float x, float y, float z);
+    void update_sync_gametick(TickType gametick);
+    bool is_disconnect(TickType current_gametick) const;
+    int task_id() const;
+    void task_id(int id);
+    bool has_player(ChunkPos pos) const;
+    void update_chunk_set(const ChunkPosSet& set);
+    const ChunkPosSet& get_chunk_pos_set() const;
+    ChunkPosSet& get_chunk_pos_set();
+
+private:
+    static constexpr TickType TIMEOUT = 200;
+    std::string m_name;
+    std::string m_uuid;
+    glm::vec3 m_pos{0.0f};
+    ServerWorld& m_world;
+    ChunkPos m_last_chunk_pos{0, 0};
+    std::shared_ptr<Session> m_session;
+    std::atomic<TickType> m_last_gametick{0};
+    std::atomic<int> m_chunk_task_id{0};
+    mutable std::shared_mutex m_chunk_pos_mutex;
+    ChunkPosSet m_player_chunk_pos_set;
+};
+} // namespace Cubed

include/Cubed/gameplay/server_world.hpp

@@ -0,0 +1,187 @@
+#pragma once
+
+#include "Cubed/gameplay/cave_carver.hpp"
+#include "Cubed/gameplay/chunk_pos.hpp"
+#include "Cubed/gameplay/game_time.hpp"
+#include "Cubed/gameplay/river_worm.hpp"
+#include "Cubed/gameplay/server_chunk.hpp"
+#include "Cubed/gameplay/server_player.hpp"
+#include "Cubed/tools/priority_thread_pool.hpp"
+#include "Cubed/tools/recent_queue.hpp"
+#include "Cubed/tools/thread_pool.hpp"
+#include "world/block_change.pb.h"
+
+#include <absl/container/flat_hash_set.h>
+#include <shared_mutex>
+#include <tbb/concurrent_hash_map.h>
+#include <tbb/concurrent_queue.h>
+#include <tbb/concurrent_unordered_map.h>
+#include <unordered_map>
+#include <utility>
+#include <vector>
+namespace Cubed {
+class Session;
+class ServerWorld {
+public:
+    enum class ThreadPoolKind { NET, GEN };
+    ServerWorld();
+    ~ServerWorld();
+    void stop();
+    void handle_player_exit(const std::string& uuid);
+    void init_world();
+    void need_gen(std::string uuid);
+    void update();
+    void hot_reload();
+
+    int rendering_distance() const;
+    void rendering_distance(int rendering_distance);
+    void start_gen_thread();
+    void start_server_thread();
+
+    void stop_gen_thread();
+    void stop_server_thread();
+
+    void stop_thread_pool();
+    void start_thread_pool();
+
+    void serever_run(std::stop_token stoken);
+
+    CaveCarver& cave_carcer();
+    RiverWorm& river_worm();
+
+    TickType game_tick() const;
+    TickType day_tick() const;
+
+    void day_tick(TickType tick);
+
+    int per_tick_time() const;
+    void per_tick_time(int ms);
+    bool is_tick_running() const;
+    void tick_running(bool run);
+
+    int gen_pool_threads() const;
+    int max_threads() const;
+
+    void change_pool_threads(ThreadPoolKind kind, int threads);
+
+    int chunk_load_style() const;
+    void set_chunk_load_style(int id);
+
+    bool set_block(const glm::ivec3& block_pos, unsigned id);
+
+    void sync_player_pos(const std::string& uuid, float x, float y, float z);
+    void handle_player_login(const std::string& player_name,
+                             std::shared_ptr<Session> session);
+    glm::vec3 get_player_pos(const std::string& uuid) const;
+
+    void handle_chunk_req(int task_id, const std::string& uuid, ChunkPos pos);
+    void handle_block_change(const BlockChangeReq& req);
+
+    int chunk_size() const;
+    template <typename Fn>
+    void register_timer(std::string_view id, TickType threshold, Fn&& f) {
+        m_timers.emplace(std::piecewise_construct,
+                         std::forward_as_tuple(std::string(id)),
+                         std::forward_as_tuple(threshold, std::forward<Fn>(f)));
+    }
+
+private:
+    enum class ChunkState { NONE, GENERATING, READY, PENDING_DELETE };
+    struct ChunkEntity {
+        ChunkState state;
+        std::shared_ptr<ServerChunk> chunk;
+        uint32_t ref_count = 0;
+    };
+
+    enum class ChunkLoadStyle { RANDOM, CENTER };
+    struct PendingRequest {
+        std::string uuid;
+        int task_id;
+        ChunkPos pos;
+    };
+    struct PendingChunk {
+        ChunkPos pos;
+        std::unique_ptr<ServerChunk> chunk;
+    };
+
+    using ChunkHashMap =
+        tbb::concurrent_hash_map<ChunkPos, ChunkEntity, ChunkPos::TBBHash>;
+    using PlayerHashMap = std::unordered_map<std::string, ServerPlayer>;
+    using NewChunkVector = std::vector<PendingChunk>;
+    using ChunkPosSet = absl::flat_hash_set<ChunkPos, ChunkPos::Hash>;
+    using PlayerUUIDMap = tbb::concurrent_hash_map<std::string, std::string>;
+
+    using chunk_acc = ChunkHashMap::accessor;
+    using chunk_caac = ChunkHashMap::const_accessor;
+
+    using uuid_acc = PlayerUUIDMap::accessor;
+    using uuid_cacc = PlayerUUIDMap::const_accessor;
+    // key = uuid
+    PlayerHashMap m_players;
+    ChunkHashMap m_chunks;
+
+    CaveCarver m_cave_carcer;
+    RiverWorm m_river_worm;
+
+    std::jthread m_gen_thread;
+    std::jthread m_server_thread;
+
+    std::atomic<bool> m_chunk_gen_finished{false};
+    std::atomic<bool> m_could_gen{true};
+    std::atomic<bool> m_gen_running{false};
+    std::atomic<bool> m_need_gen_chunk{false};
+    std::atomic<bool> m_init{false};
+    std::atomic<bool> m_stopped{false};
+    std::atomic<int> m_rendering_distance{24};
+    std::atomic<int> m_gen_pool_threads{0};
+    std::atomic<int> m_net_pool_threads{0};
+    std::atomic<int> m_max_threads{1};
+
+    std::atomic<TickType> m_game_ticks{0};
+    std::atomic<TickType> m_day_tick{6000};
+    std::atomic<bool> m_tick_running{true};
+    std::atomic<int> m_per_tick_time = DEFAULT_PER_TICK_TIME; // ms
+
+    mutable std::shared_mutex m_player_mutex;
+    std::mutex m_need_gen_queue_mutex;
+    std::condition_variable_any m_gen_cv;
+
+    RecentQueue<std::string> m_need_gen_queue;
+
+    std::atomic<std::shared_ptr<PriorityThreadPool>> m_gen_thread_pool;
+    std::atomic<std::shared_ptr<ThreadPool>> m_net_thread_pool;
+
+    std::atomic<ChunkLoadStyle> m_chunk_load_style{ChunkLoadStyle::CENTER};
+
+    PlayerUUIDMap m_uuid_to_name;
+
+    tbb::concurrent_unordered_map<std::string, Timer> m_timers;
+    tbb::concurrent_queue<PendingRequest> m_waiting_chunk_requests;
+    tbb::concurrent_queue<std::unique_ptr<ServerChunk>> m_finished_queue;
+
+    void init_chunks();
+
+    void gen_chunks_internal(const std::string& uuid);
+
+    void compute_required_chunks(ChunkPosSet& required_chunks,
+                                 const std::optional<std::string>& uuid);
+    void sync_and_collect_missing_chunks(std::vector<ChunkPos>&,
+                                         const ChunkPosSet&);
+    void submit_new_chunks(const std::string& uuid, NewChunkVector& new_chunks);
+    // void wait_all_chunk_tasks();
+
+    void update_ref_count(const ChunkPosSet& old, const ChunkPosSet& now);
+
+    void send_time();
+
+    void send_chunk(int task_id, const std::string& uuid, ChunkPos pos);
+
+    int
+    change_pool_threads(std::atomic<std::shared_ptr<ThreadPool>>& thread_pool,
+                        int threads);
+    int change_pool_threads(
+        std::atomic<std::shared_ptr<PriorityThreadPool>>& thread_pool,
+        int threads);
+    void send_server_stop();
+};
+} // namespace Cubed

include/Cubed/gameplay/session.hpp

@@ -0,0 +1,60 @@
+#pragma once
+
+#include "Cubed/gameplay/packet.hpp"
+
+#include <asio.hpp>
+#include <memory>
+#include <queue>
+#include <string>
+namespace Cubed {
+
+using asio::ip::tcp;
+class ServerWorld;
+class Session : public std::enable_shared_from_this<Session> {
+
+public:
+    Session(tcp::socket socket, ServerWorld& server_world,
+            asio::io_context& io);
+    ~Session();
+    void start();
+    void send(Packet packet, int priority = 10);
+
+    void close();
+    const std::string& uuid() const;
+
+private:
+    struct Task {
+        int priority = 10;
+        std::uint64_t sequence = 0;
+        Packet packet;
+        Task(int p, std::uint64_t seq, Packet pac)
+            : priority(p), sequence(seq), packet(std::move(pac)) {}
+    };
+
+    struct TaskCompare {
+        bool operator()(const Task& a, const Task& b) const {
+
+            if (a.priority != b.priority) {
+                return a.priority > b.priority;
+            }
+
+            return a.sequence > b.sequence;
+        }
+    };
+
+    static constexpr uint32_t MAX_PACKET_SIZE = 4 * 1024 * 1024;
+    tcp::socket m_socket;
+    std::vector<char> m_read_buffer;
+    std::priority_queue<Task, std::vector<Task>, TaskCompare> m_write_queue;
+    asio::strand<asio::io_context::executor_type> m_strand;
+    std::string m_uuid;
+    ServerWorld& m_server_world;
+    std::atomic<bool> m_closed{false};
+
+    std::atomic_uint64_t m_sequence{0};
+
+    asio::awaitable<void> read_loop();
+
+    void do_write();
+};
+} // namespace Cubed

include/Cubed/gameplay/tree.hpp

@@ -4,13 +4,13 @@
 
 namespace Cubed {
 
-class Chunk;
+class ServerChunk;
 
 struct TreeStructNode {
     glm::ivec3 offset{0, 0, 0};
     unsigned id = 0;
 };
 
-bool build_tree(Chunk& chunk, const glm::ivec3& pos);
+bool build_tree(ServerChunk& chunk, const glm::ivec3& pos);
 
 } // namespace Cubed

include/Cubed/gameplay/vertex_data.hpp

@@ -5,14 +5,14 @@
 #include <glad/glad.h>
 #include <vector>
 namespace Cubed {
-class World;
+class ClientWorld;
 struct VertexData {
     std::vector<Vertex3D> m_vertices;
     GLuint m_vbo = 0;
     GLuint m_vao = 0;
     std::atomic<std::size_t> m_sum{0};
-    World& m_world;
-    VertexData(World& world);
+    ClientWorld& m_world;
+    VertexData(ClientWorld& world);
     ~VertexData();
     VertexData(const VertexData&) = delete;
     VertexData(VertexData&&) noexcept;

include/Cubed/gameplay/world.hpp

@@ -1,173 +0,0 @@
-#pragma once
-#include "Cubed/AABB.hpp"
-#include "Cubed/gameplay/cave_carver.hpp"
-#include "Cubed/gameplay/chunk.hpp"
-#include "Cubed/gameplay/game_time.hpp"
-#include "Cubed/gameplay/river_worm.hpp"
-#include "Cubed/tools/thread_pool.hpp"
-
-#include <atomic>
-#include <condition_variable>
-#include <mutex>
-#include <optional>
-#include <shared_mutex>
-#include <thread>
-#include <unordered_map>
-#include <unordered_set>
-
-namespace Cubed {
-
-struct ChunkRenderSnapshot {
-    GLuint normal_vao;
-    size_t normal_vertices_count;
-    GLuint cross_vao;
-    size_t cross_vertices_count;
-    GLuint normal_discard_vao;
-    size_t normal_discard_vertices_count;
-    GLuint normal_blend_vao;
-    size_t normal_blend_vertices_count;
-    GLuint water_vao;
-    size_t water_vertices_count;
-    glm::vec3 center;
-    glm::vec3 half_extents;
-};
-
-class Player;
-class TextureManager;
-class World {
-private:
-    enum class ChunkLoadStyle { RANDOM, CENTER };
-
-    struct PendingChunk {
-        Chunk chunk;
-        std::future<void> future;
-    };
-
-    using OptionalBlockVectorArray =
-        std::array<std::optional<std::vector<BlockType>>, 4>;
-    using ChunkPtrUpdateList = std::vector<std::pair<ChunkPos, Chunk*>>;
-    using ChunkPairVector = std::vector<std::pair<ChunkPos, Chunk>>;
-    using ChunkPairQueue = std::queue<std::pair<ChunkPos, Chunk>>;
-    using ConstChunkMap =
-        std::unordered_map<ChunkPos, const Chunk*, ChunkPos::Hash>;
-    using ChunkPosSet = std::unordered_set<ChunkPos, ChunkPos::Hash>;
-    using ChunkHashMap = std::unordered_map<ChunkPos, Chunk, ChunkPos::Hash>;
-    using PendingChunkHashMap =
-        std::unordered_map<ChunkPos, PendingChunk, ChunkPos::Hash>;
-    glm::vec3 m_gen_player_pos{0.0f, 0.0f, 0.0f};
-    ChunkHashMap m_chunks;
-    std::unordered_map<std::size_t, Player> m_players;
-    std::vector<glm::vec4> m_planes;
-
-    std::thread m_gen_thread;
-    std::thread m_server_thread;
-    std::atomic<std::shared_ptr<ThreadPool>> m_gen_thread_pool;
-    std::stop_source m_server_stop_source;
-
-    std::atomic<int> m_per_tick_time = DEFAULT_PER_TICK_TIME; // ms
-
-    std::atomic<TickType> m_day_tick = 6000;
-
-    mutable std::shared_mutex m_chunks_mutex;
-    std::mutex m_gen_signal_mutex;
-    std::mutex m_new_chunk_mutex;
-    std::mutex m_delete_vbo_mutex;
-    std::mutex m_delete_vao_mutex;
-    std::mutex m_gen_player_pos_mutex;
-    std::vector<GLuint> m_pending_delete_vbo;
-    std::vector<GLuint> m_pending_delete_vao;
-    std::condition_variable m_gen_cv;
-    std::atomic<bool> m_gen_running{false};
-    std::atomic<bool> m_need_gen_chunk{false};
-    std::atomic<bool> m_is_rebuilding{false};
-    std::atomic<bool> m_chunk_gen_finished{false};
-    std::atomic<bool> m_could_gen{true};
-    std::atomic<bool> m_tick_running{true};
-    std::atomic<int> m_rendering_distance{24};
-    std::atomic<int> m_pool_threads{0};
-    std::atomic<int> m_max_threads{1};
-    std::atomic<TickType> m_game_ticks{0};
-    std::atomic<ChunkLoadStyle> m_chunk_load_style{ChunkLoadStyle::RANDOM};
-    std::vector<ChunkPos> m_dirty_queue;
-    std::vector<ChunkRenderSnapshot> m_render_snapshots;
-    std::vector<std::pair<ChunkPos, Chunk>> m_new_finished_chunk;
-    // Can only be used in the gen thread
-    PendingChunkHashMap new_chunks;
-
-    CaveCarver m_cave_carcer;
-    RiverWorm m_river_worm;
-    void init_chunks();
-
-    void gen_chunks_internal();
-    void sync_player_pos(glm::vec3& player_pos);
-    void compute_required_chunks(ChunkPosSet& required_chunks);
-    void sync_and_collect_missing_chunks(std::vector<ChunkPos>&,
-                                         const ChunkPosSet&);
-
-    void submit_new_chunks();
-    void poll_finished_chunks();
-    void wait_all_chunk_tasks();
-
-public:
-    World();
-    ~World();
-
-    bool can_move(const AABB& player_box) const;
-    // const BlockRenderData& get_block_render_data(int x, int y ,int z);
-
-    const std::optional<LookBlock>&
-    get_look_block_pos(const std::string& name) const;
-    // const Chunk* get_chunk(const ChunkPos& pos) const;
-
-    Player& get_player(const std::string& name);
-    void init_world();
-    int get_block(const glm::ivec3& block_pos) const;
-    bool is_solid(const glm::ivec3& block_pos) const;
-    bool can_pass_block(const glm::ivec3& block_pos) const;
-    BlockType get_block_tpye(const glm::ivec3& block_pos) const;
-    static ChunkPos get_chunk_pos(int world_x, int world_z);
-
-    void need_gen();
-
-    void set_block(const glm::ivec3& pos, unsigned id);
-    void update(float delta_time);
-
-    void push_delete_vbo(GLuint vbo);
-    void push_delete_vao(GLuint vao);
-    void hot_reload();
-
-    void rebuild_world();
-
-    int rendering_distance() const;
-    void rendering_distance(int rendering_distance);
-    void start_gen_thread();
-    void start_server_thread();
-    void stop_gen_thread();
-    void stop_server_thread();
-    void stop_thread_pool();
-    void start_thread_pool();
-    void serever_run(std::stop_token stoken);
-
-    CaveCarver& cave_carcer();
-    RiverWorm& river_worm();
-    std::vector<glm::vec4>& planes();
-    std::vector<ChunkRenderSnapshot>& render_snapshots();
-
-    glm::vec3 sunlight_dir() const;
-    TickType game_tick() const;
-    TickType day_tick() const;
-    void day_tick(TickType tick);
-    int per_tick_time() const;
-    void per_tick_time(int ms);
-
-    bool is_tick_running() const;
-    void tick_running(bool run);
-    int pool_threads() const;
-    int max_threads() const;
-    void change_pool_threads(int threads);
-    int chunk_load_style() const;
-    void set_chunk_load_style(int id);
-    ChunkInfo get_chunk_info(const glm::vec3& world_pos) const;
-};
-
-} // namespace Cubed

include/Cubed/primitive_data.hpp

@@ -272,6 +272,55 @@ constexpr float CROSS_TANGENTS[2][6][3] = {
      {-0.7071f, 0.0f, 0.7071f}}};
 
 #pragma endregion
+
+#pragma region Player
+
+constexpr float VERTICES_PLAYER[6][6][3] = {
+    // ===== front (z = +1) =====
+    {{0.0f, 0.0f, 1.0f},  // bottom left
+     {0.0f, 2.0f, 1.0f},  // top left
+     {1.0f, 2.0f, 1.0f},  // top right
+     {1.0f, 2.0f, 1.0f},  // top right
+     {1.0f, 0.0f, 1.0f},  // bottom right
+     {0.0f, 0.0f, 1.0f}}, // bottom left
+    // ===== right (x = +1) =====
+    {{1.0f, 0.0f, 1.0f},  // bottom front
+     {1.0f, 0.0f, 0.0f},  // bottom back
+     {1.0f, 2.0f, 0.0f},  // top back
+     {1.0f, 2.0f, 0.0f},  // top back
+     {1.0f, 2.0f, 1.0f},  // top front
+     {1.0f, 0.0f, 1.0f}}, // bottom front
+    // ===== back (z = -1) =====
+    {{0.0f, 0.0f, 0.0f},  // bottom left
+     {1.0f, 0.0f, 0.0f},  // bottom right
+     {1.0f, 2.0f, 0.0f},  // top right
+     {1.0f, 2.0f, 0.0f},  // top right
+     {0.0f, 2.0f, 0.0f},  // top left
+     {0.0f, 0.0f, 0.0f}}, // bottom left
+    // ===== left (x = -1) =====
+    {{0.0f, 0.0f, 0.0f},  // bottom back
+     {0.0f, 0.0f, 1.0f},  // bottom front
+     {0.0f, 2.0f, 1.0f},  // top front
+     {0.0f, 2.0f, 1.0f},  // top front
+     {0.0f, 2.0f, 0.0f},  // top back
+     {0.0f, 0.0f, 0.0f}}, // bottom back
+    // ===== top (y = +2) =====
+    {{0.0f, 2.0f, 0.0f},  // back left
+     {1.0f, 2.0f, 0.0f},  // back right
+     {1.0f, 2.0f, 1.0f},  // front right
+     {1.0f, 2.0f, 1.0f},  // front right
+     {0.0f, 2.0f, 1.0f},  // front left
+     {0.0f, 2.0f, 0.0f}}, // back left
+    // ===== bottom (y = -1) =====
+    {{0.0f, 0.0f, 1.0f}, // front left
+     {1.0f, 0.0f, 1.0f}, // front right
+     {1.0f, 0.0f, 0.0f}, // back right
+     {1.0f, 0.0f, 0.0f}, // back right
+     {0.0f, 0.0f, 0.0f}, // back left
+     {0.0f, 0.0f, 1.0f}} // front left
+};
+#pragma endregion
+
 // [-1, 1]
 constexpr float QUAD_VERTICES[] = {
     // postion        // texcoorlds

include/Cubed/renderer.hpp

@@ -11,13 +11,13 @@ namespace Cubed {
 
 class Camera;
 class TextureManager;
-class World;
+class ClientWorld;
 class DevPanel;
 class Renderer {
 public:
     constexpr static int NUM_VAO = 7;
 
-    Renderer(const Camera& camera, World& world,
+    Renderer(const Camera& camera, ClientWorld& world,
              const TextureManager& texture_manager, DevPanel& dev_panel);
     ~Renderer();
     void hot_reload();
@@ -91,7 +91,7 @@ private:
     const Camera& m_camera;
     DevPanel& m_dev_panel;
     const TextureManager& m_texture_manager;
-    World& m_world;
+    ClientWorld& m_world;
 
     bool m_discard_tranparent = true;
     bool m_shader_on = true;
@@ -99,6 +99,9 @@ private:
     bool m_water_depth_fade = true;
     bool m_pbr = true;
     bool m_flip_y = false;
+
+    bool m_init = false;
+
     int m_shadow_mode = 0;
     int m_light_cull_face = 0;
     float m_aspect = 0.0f;
@@ -119,7 +122,7 @@ private:
     GLuint m_outline_indices_vbo = 0;
     GLuint m_outline_vbo = 0;
     GLuint m_ui_vbo = 0;
-
+    GLuint m_player_vbo = 0;
     GLuint m_fbo = 0;
     GLuint m_screen_texture = 0;
     GLuint m_screen_depth_texture = 0;
@@ -171,7 +174,7 @@ private:
     2 - outline vao
     3 - ui vao
     4 - text vao
-
+    5 - player vao
     */
     std::vector<GLuint> m_vao;
     std::vector<Vertex2D> m_ui;
@@ -186,6 +189,7 @@ private:
     void render_text();
     void render_ui();
     void render_world();
+    void render_player();
     void render_underwater();
     void render_dev_panel();
 

include/Cubed/texture_manager.hpp

@@ -8,6 +8,7 @@ namespace Cubed {
 class TextureManager {
 private:
     bool m_need_reload = false;
+    bool m_init = false;
     GLuint m_block_status_array = 0;
     GLuint m_texture_array = 0;
     GLuint m_cross_plane_array = 0;

include/Cubed/tools/arg_parser.hpp

@@ -0,0 +1,35 @@
+#pragma once
+
+#include <format>
+#include <span>
+#include <stdexcept>
+#include <string_view>
+
+namespace Cubed {
+class ArgParser {
+public:
+    ArgParser(int argc, char** argv) : m_args(argv, argc) {};
+    ArgParser(std::span<char*> args) : m_args(args) {}
+
+    bool has_next() const { return m_index < m_args.size(); }
+
+    std::string_view next() {
+        if (!has_next()) {
+            throw std::runtime_error("No more arguments");
+        }
+        return m_args[m_index++];
+    }
+
+    std::string_view require_next(std::string_view option) {
+        if (!has_next()) {
+            throw std::runtime_error(
+                std::format("{} requires an argument", option));
+        }
+        return next();
+    }
+
+private:
+    std::span<char*> m_args;
+    size_t m_index = 1;
+};
+} // namespace Cubed

include/Cubed/tools/compression.hpp

@@ -0,0 +1,40 @@
+#pragma once
+#include <cstdint>
+#include <format>
+#include <span>
+#include <stdexcept>
+#include <vector>
+#include <zstd.h>
+namespace Cubed {
+constexpr int DEFAULT_ZSTD_LEVEL = 3;
+inline std::vector<uint8_t> compress_data(std::span<const uint8_t> data) {
+    size_t max_size = ZSTD_compressBound(data.size());
+    std::vector<uint8_t> compressed_data(max_size);
+    size_t compressed_bytes =
+        ZSTD_compress(compressed_data.data(), max_size, data.data(),
+                      data.size(), DEFAULT_ZSTD_LEVEL);
+    if (ZSTD_isError(compressed_bytes)) {
+        throw std::runtime_error(std::format(
+            "Compress Fail {}", ZSTD_getErrorName(compressed_bytes)));
+    }
+    compressed_data.resize(compressed_bytes);
+    return compressed_data;
+}
+
+inline std::vector<uint8_t> decompress_data(std::span<const uint8_t> data,
+                                            uint32_t original_size) {
+    std::vector<uint8_t> decompressed_data(original_size);
+    size_t decompressed_bytes = ZSTD_decompress(
+        decompressed_data.data(), original_size, data.data(), data.size());
+    if (ZSTD_isError(decompressed_bytes)) {
+        throw std::runtime_error(std::format(
+            "Decompress Fail {}", ZSTD_getErrorName(decompressed_bytes)));
+    }
+    if (decompressed_bytes != original_size) {
+        throw std::runtime_error("Unexpected decompressed size");
+    }
+
+    return decompressed_data;
+}
+
+} // namespace Cubed

include/Cubed/tools/log.hpp

@@ -9,7 +9,7 @@
 namespace Cubed {
 
 namespace Logger {
-enum class Level { TRACE, DEBUG, INFO, ERROR, WARN };
+enum class Level { L_TRACE, L_DEBUG, L_INFO, L_ERROR, L_WARN };
 
 template <typename... Args>
 inline void info(std::format_string<Args...> fmt, Args&&... args) {
@@ -53,7 +53,7 @@ inline void log(Level level, std::source_location loc,
         std::chrono::system_clock::now());
     std::string msg = std::vformat(fmt.get(), std::make_format_args(args...));
     switch (level) {
-    case Logger::Level::TRACE:
+    case Logger::Level::L_TRACE:
         std::osyncstream(std::cout)
             << "\033[1;34m"
             << std::format("[TRACE][{:%Y-%m-%d %H:%M:%S}]", now_time) << "["
@@ -61,20 +61,20 @@ inline void log(Level level, std::source_location loc,
             << "[" << loc.function_name() << "]" << msg << "\033[0m"
             << "\n";
         break;
-    case Logger::Level::DEBUG:
+    case Logger::Level::L_DEBUG:
         std::osyncstream(std::cout)
             << "\033[1;34m"
             << std::format("[DEBUG][{:%Y-%m-%d %H:%M:%S}]", now_time) << msg
             << "\033[0m"
             << "\n";
         break;
-    case Logger::Level::INFO:
+    case Logger::Level::L_INFO:
         info(fmt, std::forward<Args>(args)...);
         break;
-    case Logger::Level::WARN:
+    case Logger::Level::L_WARN:
         warn(fmt, std::forward<Args>(args)...);
         break;
-    case Logger::Level::ERROR:
+    case Logger::Level::L_ERROR:
         error(fmt, std::forward<Args>(args)...);
         break;
     }

include/Cubed/tools/math_tools.hpp

@@ -1,18 +1,108 @@
 #pragma once
+#include <algorithm>
 #include <glm/glm.hpp>
-
+#include <glm/gtc/type_ptr.hpp>
 namespace Cubed {
 
 namespace Math {
 
-void extract_frustum_planes(const glm::mat4& mvp_matrix,
-                            std::vector<glm::vec4>& planes);
+inline void extract_frustum_planes(const glm::mat4& mvp_matrix,
+                                   std::vector<glm::vec4>& planes) {
+    if (planes.size() != 6) {
+        planes.resize(6);
+    }
 
-float smootherstep(float edge0, float edge1, float x);
-bool is_aabb_in_frustum(const glm::vec3& center, const glm::vec3& half_extents,
-                        const std::vector<glm::vec4>& planes);
-float deterministic_random(int x, int z, uint64_t seed);
-glm::vec3 slerp(const glm::vec3& from, const glm::vec3& to, float t);
+    const float* m = glm::value_ptr(mvp_matrix);
+
+    // left plane
+    planes[0] =
+        glm::vec4(m[3] + m[0], m[7] + m[4], m[11] + m[8], m[15] + m[12]);
+    // right plane
+    planes[1] =
+        glm::vec4(m[3] - m[0], m[7] - m[4], m[11] - m[8], m[15] - m[12]);
+    // bottom plane
+    planes[2] =
+        glm::vec4(m[3] + m[1], m[7] + m[5], m[11] + m[9], m[15] + m[13]);
+    // top plane
+    planes[3] =
+        glm::vec4(m[3] - m[1], m[7] - m[5], m[11] - m[9], m[15] - m[13]);
+    // near plane
+    planes[4] =
+        glm::vec4(m[3] + m[2], m[7] + m[6], m[11] + m[10], m[15] + m[14]);
+    // far plane
+    planes[5] =
+        glm::vec4(m[3] - m[2], m[7] - m[6], m[11] - m[10], m[15] - m[14]);
+
+    for (auto& p : planes) {
+        p = glm::normalize(p);
+    }
+}
+
+inline float smootherstep(float edge0, float edge1, float x) {
+
+    x = std::clamp((x - edge0) / (edge1 - edge0), 0.0f, 1.0f);
+
+    return x * x * x * (x * (6.0f * x - 15.0f) + 10.0f);
+}
+
+inline bool is_aabb_in_frustum(const glm::vec3& center,
+                               const glm::vec3& half_extents,
+                               const std::vector<glm::vec4>& planes) {
+    for (const auto& plane : planes) {
+        // distance
+        float d = glm::dot(glm::vec3(plane), center) + plane.w;
+        float r = half_extents.x * std::abs(plane.x) +
+                  half_extents.y * std::abs(plane.y) +
+                  half_extents.z * std::abs(plane.z);
+        if (d + r < 0) {
+            return false;
+        }
+    }
+    return true;
+}
+inline float deterministic_random(int x, int z, uint64_t seed) {
+    uint64_t h = seed;
+    h = h * 6364136223846793005ULL + (uint64_t)x;
+    h = h * 6364136223846793005ULL + (uint64_t)z;
+    return (float)(h >> 40) / (float)(1 << 24);
+}
+
+inline glm::vec3 slerp(const glm::vec3& from, const glm::vec3& to, float t) {
+
+    float cos_theta = glm::clamp(glm::dot(from, to), -1.0f, 1.0f);
+
+    if (cos_theta > 0.9995f) {
+        return glm::normalize(glm::mix(from, to, t));
+    }
+
+    if (cos_theta < -0.9995f) {
+
+        glm::vec3 axis = (std::fabs(from.x) < 0.9f)
+                             ? glm::vec3(1.0f, 0.0f, 0.0f)
+                             : glm::vec3(0.0f, 1.0f, 0.0f);
+        glm::vec3 ortho = glm::normalize(glm::cross(from, axis));
+
+        float angle = glm::pi<float>() * t;
+
+        glm::vec3 rotated =
+            from * std::cos(angle) + glm::cross(ortho, from) * std::sin(angle);
+
+        return glm::normalize(rotated);
+    }
+
+    float theta = std::acos(cos_theta);
+    float sin_theta = std::sin(theta);
+
+    float a = std::sin((1.0f - t) * theta) / sin_theta;
+    float b = std::sin(t * theta) / sin_theta;
+
+    return glm::normalize(a * from + b * to);
+}
+
+inline float distance2(const glm::vec3& a, const glm::vec3& b) {
+    glm::vec3 diff = a - b;
+    return glm::dot(diff, diff);
+}
 
 } // namespace Math
 

include/Cubed/tools/priority_thread_pool.hpp

@@ -0,0 +1,151 @@
+#pragma once
+
+#include <condition_variable>
+#include <cstddef>
+#include <functional>
+#include <future>
+#include <mutex>
+#include <queue>
+#include <thread>
+#include <vector>
+namespace Cubed {
+class PriorityThreadPool {
+private:
+    struct Task {
+        int priority = 10;
+        std::uint64_t sequence;
+        std::function<void()> task;
+        Task(int p, std::uint64_t seq, std::function<void()> t)
+            : priority(p), sequence(seq), task(std::move(t)) {}
+    };
+
+    struct TaskCompare {
+        bool operator()(const Task& a, const Task& b) const {
+
+            if (a.priority != b.priority) {
+                return a.priority > b.priority;
+            }
+
+            return a.sequence > b.sequence;
+        }
+    };
+
+    std::vector<std::jthread> m_workers;
+    std::priority_queue<Task, std::vector<Task>, TaskCompare> m_tasks;
+    std::mutex m_mtx;
+    std::condition_variable_any m_cv;
+    std::atomic<bool> m_stopping{false};
+    std::atomic<size_t> m_thread_sum{0};
+    std::atomic_uint64_t m_sequence{0};
+
+public:
+    PriorityThreadPool(const PriorityThreadPool&) = delete;
+    PriorityThreadPool(PriorityThreadPool&&) = delete;
+    PriorityThreadPool& operator=(const PriorityThreadPool&) = delete;
+    PriorityThreadPool& operator=(PriorityThreadPool&&) = delete;
+    explicit PriorityThreadPool(size_t thread_sum) : m_thread_sum(thread_sum) {
+        for (size_t i = 0; i < thread_sum; i++) {
+            m_workers.emplace_back([this](std::stop_token stoken) {
+                while (true) {
+                    std::function<void()> task;
+                    {
+                        std::unique_lock lock(m_mtx);
+                        m_cv.wait(lock, stoken,
+                                  [this] { return !m_tasks.empty(); });
+                        if (stoken.stop_requested() && m_tasks.empty()) {
+                            return;
+                        }
+                        task = std::move(m_tasks.top().task);
+                        m_tasks.pop();
+                    }
+                    task();
+                }
+            });
+        }
+    }
+    ~PriorityThreadPool() { stop(); }
+    template <typename F> auto enqueue(int priority, F&& f) {
+
+        using R = std::invoke_result_t<F>;
+
+        auto task =
+            std::make_shared<std::packaged_task<R()>>(std::forward<F>(f));
+        auto fut = task->get_future();
+
+        {
+            std::lock_guard lock(m_mtx);
+            if (m_stopping)
+                throw std::runtime_error("thread pool stopped");
+            m_tasks.emplace(priority, m_sequence++, [task] { (*task)(); });
+        }
+        m_cv.notify_one();
+        return fut;
+    }
+
+    template <typename F> auto enqueue(F&& f) {
+        return enqueue(10, std::forward<F>(f));
+    }
+
+    void stop() {
+        if (m_stopping.exchange(true)) {
+            return;
+        }
+
+        for (auto& w : m_workers) {
+            w.request_stop();
+        }
+
+        m_cv.notify_all();
+
+        for (auto& w : m_workers) {
+            if (w.joinable()) {
+                w.join();
+            }
+        }
+    }
+    size_t thread_sum() const { return m_thread_sum.load(); }
+};
+
+template <std::random_access_iterator Iter, typename F>
+void parallel_do(PriorityThreadPool& pool, Iter first, Iter last,
+                 size_t max_threads, F&& f) {
+    max_threads = std::max<size_t>(1, max_threads);
+    max_threads = std::min(max_threads, pool.thread_sum());
+    std::decay_t<F> fn(std::forward<F>(f));
+    size_t length = std::distance(first, last);
+    if (!length) {
+        return;
+    }
+
+    constexpr size_t MIN_PER_THREAD = 25;
+    size_t num_blocks =
+        std::min(max_threads, (length + MIN_PER_THREAD - 1) / MIN_PER_THREAD);
+    num_blocks = std::max<size_t>(1, num_blocks);
+    size_t block_size = (length + num_blocks - 1) / num_blocks;
+
+    std::vector<std::future<void>> futures;
+    futures.reserve(num_blocks - 1);
+    Iter block_start = first;
+    for (size_t i = 0; i < num_blocks - 1; ++i) {
+        Iter block_end = block_start;
+        auto remain = std::distance(block_start, last);
+        std::advance(block_end, std::min<size_t>(block_size, remain));
+
+        futures.emplace_back(pool.enqueue([block_start, block_end, &fn]() {
+            for (auto it = block_start; it != block_end; ++it) {
+                fn(*it);
+            }
+        }));
+
+        block_start = block_end;
+    }
+    for (auto it = block_start; it != last; ++it) {
+        fn(*it);
+    }
+
+    for (auto& fut : futures) {
+        fut.get();
+    }
+};
+
+} // namespace Cubed

include/Cubed/tools/recent_queue.hpp

@@ -0,0 +1,55 @@
+#pragma once
+#include <list>
+#include <unordered_map>
+namespace Cubed {
+template <typename T> class RecentQueue {
+private:
+    std::list<T> m_list;
+    std::unordered_map<T, typename std::list<T>::iterator> m_map;
+
+public:
+    void enqueue(T key) {
+        auto it = m_map.find(key);
+        if (it != m_map.end()) {
+            m_list.splice(m_list.end(), m_list, it->second);
+            return;
+        }
+        m_list.emplace_back(std::move(key));
+        auto iter = std::prev(m_list.end());
+        m_map.emplace(*iter, iter);
+    }
+
+    void pop() {
+        if (m_list.empty()) {
+            return;
+        }
+        m_map.erase(m_list.front());
+        m_list.pop_front();
+    }
+    void clear() {
+        m_list.clear();
+        m_map.clear();
+    }
+
+    const T& front() const {
+        assert(!empty());
+        return m_list.front();
+    }
+    const T& back() const {
+        assert(!empty());
+        return m_list.back();
+    }
+    [[nodiscard]]
+    bool empty() const {
+        return m_list.empty();
+    }
+    [[nodiscard]]
+    size_t size() const {
+        return m_list.size();
+    }
+    [[nodiscard]]
+    bool contains(const T& key) const {
+        return m_map.find(key) != m_map.end();
+    }
+};
+} // namespace Cubed

include/Cubed/tools/thread_pool.hpp

@@ -31,7 +31,7 @@ public:
                     {
                         std::unique_lock lock(m_mtx);
                         m_cv.wait(lock, stoken,
-                                  [this, stoken] { return !m_tasks.empty(); });
+                                  [this] { return !m_tasks.empty(); });
                         if (stoken.stop_requested() && m_tasks.empty()) {
                             return;
                         }
@@ -62,7 +62,9 @@ public:
         return fut;
     }
     void stop() {
-        m_stopping = true;
+        if (m_stopping.exchange(true)) {
+            return;
+        }
         for (auto& w : m_workers) {
             w.request_stop();
         }

include/Cubed/tools/toml.utils.hpp

@@ -0,0 +1,33 @@
+#pragma once
+
+#include "Cubed/tools/log.hpp"
+
+#include <toml++/toml.hpp>
+namespace Cubed {
+namespace TOML {
+
+template <typename T>
+concept TomlValueType =
+    std::same_as<std::decay_t<T>, int> || std::same_as<std::decay_t<T>, bool> ||
+    std::same_as<std::decay_t<T>, double> ||
+    std::same_as<std::decay_t<T>, char> ||
+    std::same_as<std::decay_t<T>, toml::date> ||
+    std::same_as<std::decay_t<T>, toml::time> ||
+    std::same_as<std::decay_t<T>, toml::date_time> ||
+    std::same_as<std::decay_t<T>, std::string>;
+
+template <TomlValueType T>
+std::optional<T> safe_get_value(const toml::table& table, std::string_view key,
+                                const T& default_value) {
+    auto value = table[key].value<T>();
+    if (value == std::nullopt) {
+        Logger::warn("Key {} Is Not Find, Wiil Set the Default Value {}", key,
+                     default_value);
+        value = default_value;
+    }
+    return value;
+}
+
+} // namespace TOML
+
+} // namespace Cubed

include/Cubed/tools/uuid.hpp

@@ -0,0 +1,42 @@
+#pragma once
+
+#include <array>
+#include <chrono>
+#include <cstdint>
+#include <iomanip>
+#include <iostream>
+#include <random>
+#include <sstream>
+#include <string>
+namespace Cubed {
+inline std::string generate_uuid() {
+
+    static std::mt19937_64 rng(
+        std::chrono::steady_clock::now().time_since_epoch().count() ^
+        (std::random_device{}()));
+    std::uniform_int_distribution<uint64_t> dist(0, UINT64_MAX);
+
+    uint64_t a = dist(rng);
+    uint64_t b = dist(rng);
+
+    std::array<uint8_t, 16> bytes{};
+    for (int i = 0; i < 8; ++i) {
+        bytes[i] = (a >> (56 - 8 * i)) & 0xFF;
+        bytes[8 + i] = (b >> (56 - 8 * i)) & 0xFF;
+    }
+
+    bytes[6] = (bytes[6] & 0x0F) | 0x40;
+
+    bytes[8] = (bytes[8] & 0x3F) | 0x80;
+
+    std::ostringstream ss;
+    ss << std::hex << std::setfill('0');
+    for (size_t i = 0; i < bytes.size(); ++i) {
+        if (i == 4 || i == 6 || i == 8 || i == 10) {
+            ss << '-';
+        }
+        ss << std::setw(2) << static_cast<int>(bytes[i]);
+    }
+    return ss.str();
+}
+} // namespace Cubed

include/Cubed/window.hpp

@@ -23,6 +23,7 @@ public:
 
 private:
     bool m_mouse_enable = false;
+    bool m_imgui_init = false;
     float m_aspect;
     GLFWwindow* m_window;
     int m_width;

src/CMakeLists.txt

@@ -0,0 +1,46 @@
+target_sources(${PROJECT_NAME}
+    PRIVATE
+        main.cpp
+        app.cpp
+        debug_collector.cpp
+        camera.cpp
+        config.cpp
+        dev_panel.cpp
+        gameplay/biome.cpp
+        gameplay/chunk_generator.cpp
+        gameplay/tree.cpp
+        input.cpp
+        map_table.cpp
+        renderer.cpp
+        shader.cpp
+        texture_manager.cpp
+        tools/cubed_random.cpp
+        tools/shader_tools.cpp
+        tools/font.cpp
+        tools/perlin_noise.cpp
+        ui/text.cpp
+        window.cpp
+        gameplay/builders/biome_builder.cpp
+        gameplay/builders/plain_builder.cpp
+        gameplay/builders/mountain_builder.cpp
+        gameplay/builders/river_builder.cpp
+        gameplay/builders/desert_builder.cpp
+        gameplay/builders/forest_builder.cpp
+        gameplay/cave_carver.cpp
+        gameplay/cave_path.cpp
+        gameplay/builders/snowy_plain_builder.cpp
+        gameplay/river_worm.cpp
+        gameplay/river_path.cpp
+        gameplay/block.cpp
+        gameplay/vertex_data.cpp
+        gameplay/builders/ocean_builder.cpp
+        gameplay/network_server.cpp
+        gameplay/server_world.cpp
+        gameplay/client_world.cpp
+        gameplay/server_chunk.cpp
+        gameplay/client_chunk.cpp
+        gameplay/server_player.cpp
+        gameplay/client_player.cpp
+        gameplay/session.cpp
+        gameplay/network_client.cpp
+)

src/app.cpp

@@ -2,19 +2,23 @@
 
 #include "Cubed/config.hpp"
 #include "Cubed/debug_collector.hpp"
-#include "Cubed/gameplay/player.hpp"
+#include "Cubed/tools/arg_parser.hpp"
 #include "Cubed/tools/cubed_assert.hpp"
 #include "Cubed/tools/log.hpp"
 #include "Cubed/tools/system_info.hpp"
+#include "version.hpp"
 
 #include <exception>
 #include <imgui_impl_glfw.h>
-
 namespace Cubed {
 
 App::App() {}
 
-App::~App() {}
+App::~App() {
+    if (m_client) {
+        m_client->stop();
+    }
+}
 void App::cursor_position_callback(GLFWwindow* window, double xpos,
                                    double ypos) {
     ImGuiIO& io = ImGui::GetIO();
@@ -30,9 +34,13 @@ void App::cursor_position_callback(GLFWwindow* window, double xpos,
         app->m_camera.update_cursor_position_camera(xpos, ypos);
     }
 }
-void App::init() {
+void App::init(int argc, char** argv) {
+    handle_toml();
+    handle_argument(argc, argv);
+
     m_window.init();
     m_window.imgui_init();
+
     Logger::info("Window Init Success");
 
     glfwSetWindowUserPointer(m_window.get_glfw_window(), this);
@@ -50,6 +58,7 @@ void App::init() {
     glfwSetCursorEnterCallback(m_window.get_glfw_window(),
                                cursor_enter_callback);
     glfwSetCharCallback(m_window.get_glfw_window(), char_callback);
+
     ChunkGenerator::init();
     BlockManager::init();
     m_renderer.init();
@@ -58,13 +67,94 @@ void App::init() {
     // MapTable::init_map();
     m_texture_manager.init_texture();
     Logger::info("Texture Load Success");
-    m_world.init_world();
+    if (!m_argument.is_client) {
+        m_server.start_server(m_argument.port);
+    }
+
+    m_client = std::make_shared<NetworkClient>(m_client_world);
+
+    m_client->start(m_argument.ip, m_argument.port);
+    // init will send packet
+    m_client_world.init(m_argument.player, m_client);
+
     Logger::info("World Init Success");
 
-    m_camera.camera_init(&m_world.get_player("TestPlayer"));
+    m_camera.camera_init(&m_client_world.get_player());
     m_dev_panel.init();
 }
 
+void App::handle_argument(int argc, char** argv) {
+
+    static const std::unordered_map<std::string_view,
+                                    std::function<void(ArgParser&)>>
+        HANDLERS{
+
+            {"--client", [&](ArgParser&) { m_argument.is_client = true; }},
+
+            {"--host", [&](ArgParser&) { m_argument.is_client = false; }},
+
+            {"-p",
+             [&](ArgParser& p) {
+                 auto arg = p.require_next("-p");
+
+                 auto r = std::from_chars(arg.data(), arg.data() + arg.size(),
+                                          m_argument.port);
+
+                 if (r.ec != std::errc{} || r.ptr != arg.data() + arg.size()) {
+                     throw std::runtime_error(
+                         std::format("Invalid port: {}", arg));
+                 }
+
+                 if (m_argument.port > 65535) {
+                     throw std::runtime_error(
+                         std::format("Port {} out of range", m_argument.port));
+                 }
+             }},
+
+            {"--ip",
+             [&](ArgParser& p) {
+                 auto arg = p.require_next("--ip");
+                 m_argument.ip = arg;
+             }},
+            {"--player",
+             [&](ArgParser& p) {
+                 auto arg = p.require_next("--player");
+                 m_argument.player = arg;
+             }},
+            {"-V",
+             [&](ArgParser) {
+                 std::cout << CUBED_VERSION << "\n";
+                 exit(EXIT_SUCCESS);
+             }}
+
+        };
+    ArgParser parser(argc, argv);
+
+    while (parser.has_next()) {
+        auto arg = parser.next();
+        if (auto it = HANDLERS.find(arg); it != HANDLERS.end()) {
+            it->second(parser);
+        } else {
+            Logger::warn("Unknown argument: {}", arg);
+        }
+    }
+}
+
+void App::handle_toml() {
+    toml::table server;
+    try {
+        server = toml::parse_file("server.toml");
+    } catch (const toml::parse_error& e) {
+        // Logger::warn("Ip toml parse error {}", e.what());
+        return;
+    }
+
+    m_argument.ip =
+        *TOML::safe_get_value(server, "ip", std::string("127.0.01"));
+    m_argument.port = *TOML::safe_get_value(server, "port", 25530);
+    m_argument.is_client = *TOML::safe_get_value(server, "client", false);
+}
+
 void App::key_callback(GLFWwindow* window, int key, int scancode, int action,
                        int mods) {
     ImGuiIO& io = ImGui::GetIO();
@@ -111,7 +201,7 @@ void App::key_callback(GLFWwindow* window, int key, int scancode, int action,
         break;
     }
 
-    app->m_world.get_player("TestPlayer").update_player_move_state(key, action);
+    app->m_client_world.get_player().update_player_move_state(key, action);
 }
 
 void App::mouse_button_callback(GLFWwindow* window, int button, int action,
@@ -162,8 +252,7 @@ void App::window_focus_callback(GLFWwindow* window, int focused) {
     }
 }
 
-void App::window_reshape_callback(GLFWwindow* window, int new_width,
-                                  int new_height) {
+void App::window_reshape_callback(GLFWwindow* window, int, int) {
 
     App* app = static_cast<App*>(glfwGetWindowUserPointer(window));
     ASSERT_MSG(app, "nullptr");
@@ -180,7 +269,7 @@ void App::mouse_scroll_callback(GLFWwindow* window, double xoffset,
         ImGui_ImplGlfw_ScrollCallback(window, xoffset, yoffset);
         return;
     }
-    auto& player = app->m_world.get_player("TestPlayer");
+    auto& player = app->m_client_world.get_player();
     player.update_scroll(yoffset);
 }
 
@@ -218,10 +307,16 @@ void App::run() {
 
     last_time = glfwGetTime();
     while (!glfwWindowShouldClose(m_window.get_glfw_window())) {
-
+        if (m_client_world.is_receive_exit()) {
+            break;
+        }
         update();
         render();
     }
+    m_client_world.request_exit();
+    if (!m_argument.is_client) {
+        m_server.server_world().stop();
+    }
 }
 static Gait player_gait = Gait::WALK;
 void App::update() {
@@ -244,9 +339,9 @@ void App::update() {
             std::format("RSS: {}mb", Tools::get_current_rss() / (1024 * 1024)));
     }
     m_texture_manager.update();
-    m_world.update(delta_time);
+    m_client_world.update(delta_time);
     m_camera.update_move_camera();
-    const auto& player = m_world.get_player("TestPlayer");
+    const auto& player = m_client_world.get_player();
     if (player_gait != player.get_gait()) {
         player_gait = player.get_gait();
         float fov = static_cast<float>(Config::get().get<double>("player.fov"));
@@ -265,7 +360,7 @@ int App::start_cubed_application(int argc, char** argv) {
     App app;
 
     try {
-        app.init();
+        app.init(argc, argv);
         Logger::info("Game Init Finish Start Run...");
         app.run();
 
@@ -288,6 +383,7 @@ DevPanel& App::dev_panel() { return m_dev_panel; }
 Renderer& App::renderer() { return m_renderer; }
 TextureManager& App::texture_manager() { return m_texture_manager; }
 Window& App::window() { return m_window; }
-World& App::world() { return m_world; }
-
+ClientWorld& App::client_world() { return m_client_world; }
+ServerWorld& App::server_world() { return m_server.server_world(); }
+const App::Argument& App::argument() const { return m_argument; }
 } // namespace Cubed

src/camera.cpp

@@ -1,7 +1,7 @@
 #include "Cubed/camera.hpp"
 
-#include "Cubed/gameplay/player.hpp"
-#include "Cubed/gameplay/world.hpp"
+#include "Cubed/gameplay/client_player.hpp"
+#include "Cubed/gameplay/client_world.hpp"
 #include "Cubed/tools/cubed_assert.hpp"
 
 namespace Cubed {
@@ -22,7 +22,7 @@ void Camera::update_move_camera() {
     }
 }
 
-void Camera::camera_init(Player* player) {
+void Camera::camera_init(ClientPlayer* player) {
     m_player = player;
     update_move_camera();
     reset_camera();

src/config.cpp

@@ -24,7 +24,6 @@ toml::table& Config::table() { return m_tbl; }
 
 void Config::create_config() {
     static constexpr auto SOURCE = R"(
-        version = "0.0.1"
 
         [window]
         width = 800

src/debug_collector.cpp

@@ -1,8 +1,8 @@
 #include "Cubed/debug_collector.hpp"
 
-#include "Cubed/config.hpp"
 #include "Cubed/tools/cubed_hash.hpp"
 #include "Cubed/tools/system_info.hpp"
+#include "version.hpp"
 
 namespace Cubed {
 
@@ -24,10 +24,16 @@ void DebugCollector::init_text() {
     Text opengl_version_text("opengl_version");
     Text biome_text("biome");
     Text speed_text("speed");
+    std::string version{"Version: " CUBED_VERSION};
+#ifdef DEBUG_MODE
+    version.append("-debug");
+#else
+    version.append("-release");
+#endif
     version_text.position(0.0f, 100.0f)
         .scale(0.8f)
         .color(Color::WHITE)
-        .text("Version: " + Config::get().get<std::string>("version"));
+        .text(version);
     fps_text.position(0.0f, 50.0f).text("FPS: 0");
     player_pos_text.position(0.0f, 150.0f)
         .scale(0.8f)

src/dev_panel.cpp

@@ -3,7 +3,7 @@
 #include "Cubed/app.hpp"
 #include "Cubed/config.hpp"
 #include "Cubed/gameplay/cave_path.hpp"
-#include "Cubed/gameplay/player.hpp"
+#include "Cubed/gameplay/client_player.hpp"
 #include "Cubed/gameplay/river.path.hpp"
 #include "Cubed/tools/log.hpp"
 
@@ -16,7 +16,7 @@ namespace Cubed {
 static constexpr const char* THEMES[] = {"Dark", "Light"};
 static constexpr const char* GAITS[] = {"Walk", "Run"};
 static constexpr const char* GAME_MODES[] = {"Creative", "Spectator"};
-static char perlin_noise_input_buffer[64];
+static constexpr const char* CHUNK_LOAD_STYLE[] = {"Random", "Center"};
 
 constexpr float TEMP_MIN = 0.0f;
 constexpr float TEMP_MAX = 1.0f;
@@ -46,20 +46,10 @@ constexpr float DELTA_ANGLE_MAX = 30.0f;
 constexpr int PATH_STEP_MIN = 1;
 constexpr int PATH_STEP_MAX = 1000;
 
-static int filter_unsigned(ImGuiInputTextCallbackData* data) {
-    if (data->EventFlag == ImGuiInputTextFlags_CallbackCharFilter) {
-        char c = data->EventChar;
-        if (c < '0' || c > '9') {
-            return 1;
-        }
-    }
-    return 0;
-}
-
 DevPanel::DevPanel(App& app) : m_app(app) {}
 
 void DevPanel::init() {
-    m_player = &m_app.world().get_player("TestPlayer");
+    m_player = &m_app.client_world().get_player();
     update_config_view();
     update_player_profile();
 }
@@ -111,6 +101,9 @@ void DevPanel::show_about_table_bar() {
         ImGui::Text("toml++");
         ImGui::Text("Dear ImGui");
         ImGui::Text("Tbb");
+        ImGui::Text("Asio");
+        ImGui::Text("protobuf");
+        ImGui::Text("zstd");
         ImGui::Separator();
         ImGui::Text("Special Thanks");
         ImGui::Text("TANGERIME");
@@ -268,7 +261,7 @@ void DevPanel::show_biome_table_bar() {
 }
 
 void DevPanel::show_time_table_bar() {
-    World& world = m_app.world();
+    ServerWorld& world = m_app.server_world();
     ImGui::Text("Game Tick %llu", world.game_tick());
     ImGui::SameLine();
     ImGui::Text("Day Tick %llu", world.day_tick());
@@ -341,17 +334,17 @@ void DevPanel::show_river_table_bar() {
 }
 
 void DevPanel::show_chunk_table_bar() {
-    auto& world = m_app.world();
-    auto& player = world.get_player("TestPlayer");
-    auto info = world.get_chunk_info(player.get_player_pos());
+    /*
+   auto& world = m_app.client_world();
+   auto& player = world.get_player();
 
-    ImGui::Text("Chunk X: %d Z: %d Info", info.pos.x, info.pos.z);
-    ImGui::Text("Seed: %u", info.seed);
-    ImGui::Text("%s", ("Biome " + get_biome_str(info.biome)).c_str());
-    ImGui::Text("First Random %u", info.first_random);
-    ImGui::Text("%s",
-                std::format("Has Cave Start {}", info.has_cave_start).c_str());
-    ImGui::Text("%s", std::format("Has Cave {}", info.has_cave).c_str());
+   ImGui::Text("Chunk X: %d Z: %d Info", info.pos.x, info.pos.z);
+   ImGui::Text("Seed: %u", info.seed);
+   ImGui::Text("%s", ("Biome " + get_biome_str(info.biome)).c_str());
+   ImGui::Text("First Random %u", info.first_random);
+   ImGui::Text("%s",
+               std::format("Has Cave Start {}", info.has_cave_start).c_str());
+   ImGui::Text("%s", std::format("Has Cave {}", info.has_cave).c_str());*/
 }
 
 void DevPanel::show_settings_tab_item() {
@@ -384,7 +377,7 @@ void DevPanel::show_settings_tab_item() {
                              128)) {
             Config::get().set("world.rendering_distance",
                               m_config.rendering_distance);
-            m_app.world().hot_reload();
+            m_app.client_world().hot_reload();
         }
         if (ImGui::Checkbox("Fullscreen", &m_config.fullscreen)) {
             Config::get().set("window.fullscreen", m_config.fullscreen);
@@ -455,106 +448,116 @@ void DevPanel::show_settings_tab_item() {
 
 void DevPanel::show_world_tab_item() {
     if (ImGui::BeginTabItem("world")) {
-        if (m_text_editing.perlin_seed) {
-            if (ImGui::InputText("ChunkGenerator Seed",
-                                 perlin_noise_input_buffer,
-                                 sizeof(perlin_noise_input_buffer),
-                                 ImGuiInputTextFlags_CallbackCharFilter |
-                                     ImGuiInputTextFlags_EnterReturnsTrue,
-                                 filter_unsigned)) {
-                ChunkGenerator::seed(static_cast<unsigned int>(
-                    std::strtoul(perlin_noise_input_buffer, nullptr, 10)));
-                m_text_editing.perlin_seed = false;
-                m_player->set_player_pos({0.0f, 255.0f, 0.0f});
-                m_app.world().rebuild_world();
-            }
-        }
-        if (!m_text_editing.perlin_seed) {
-            ImGui::Text("ChunkGenerator Seed: %u", ChunkGenerator::seed());
-            if (ImGui::IsItemClicked()) {
-                m_text_editing.perlin_seed = true;
-            }
-        }
-        static int rendering_distance = m_app.world().rendering_distance();
-        if (ImGui::SliderInt("Render Distance", &rendering_distance, 2, 128)) {
-            m_app.world().rendering_distance(rendering_distance);
-        }
-        ImGui::Text(
-            "Pool Threads %d  Max Support Threads %d  Reserved Threads %d",
-            m_app.world().pool_threads(), m_app.world().max_threads(),
-            RESERVED_THREADS);
-        ImGui::SliderInt("Set Pool Threads", &m_threads, 1,
-                         m_app.world().max_threads());
-        ImGui::SameLine();
-        if (ImGui::Button("Set")) {
-            m_app.world().change_pool_threads(m_threads);
-        }
-        if (m_threads > m_app.world().max_threads() - RESERVED_THREADS) {
-            ImGui::TextColored(
-                ImVec4(1.0f, 1.0f, 0.0f, 1.0f),
-                "Waring: When the threads in the thread pool exceed \n(maximum "
-                "threads minus reserved threads), \nit may cause stuttering.");
-        }
 
-        static const char* chunk_load_style[] = {"Random", "Center"};
-        m_chunk_style = m_app.world().chunk_load_style();
-        if (ImGui::Combo("ChunkLoadStyle", &m_chunk_style, chunk_load_style,
-                         IM_ARRAYSIZE(chunk_load_style))) {
-            m_app.world().set_chunk_load_style(m_chunk_style);
-        }
-        if (ImGui::Button("Rebuild World")) {
-            m_app.world().rebuild_world();
-        }
-        ImGui::SameLine();
-        if (ImGui::Button("Request Chunk Build")) {
-            m_app.world().need_gen();
-        }
-        ImGui::SameLine();
-        if (ImGui::Button("Spawn Point")) {
-            m_player->set_player_pos({0.0f, 255.0f, 0.0f});
-        }
-        ImGui::SameLine();
-        if (ImGui::Checkbox("Gen Thread", &m_gen_thread_running)) {
-            if (m_gen_thread_running) {
-                m_app.world().start_gen_thread();
-            } else {
-                m_app.world().stop_gen_thread();
+        if (ImGui::BeginTabBar("World Kind")) {
+            if (!m_app.argument().is_client) {
+                if (ImGui::BeginTabItem("ServerWorld")) {
+                    show_server_world_table_bar();
+                    ImGui::EndTabItem();
+                }
             }
-        }
-        // ImGui::Text("Chunk Build Progress\n");
-        // ImGui::ProgressBar(m_app.world().chunk_gen_fraction());
-        show_chunk_table_bar();
-        if (ImGui::BeginTabBar("World Settings")) {
-            if (ImGui::BeginTabItem("Time")) {
-                show_time_table_bar();
+            if (ImGui::BeginTabItem("Client World")) {
+                show_client_world_table_bar();
                 ImGui::EndTabItem();
             }
-            /*
-            if (ImGui::BeginTabItem("Cave")) {
-                show_cave_table_bar();
-                ImGui::EndTabItem();
-            }
-            if (ImGui::BeginTabItem("River")) {
-                show_river_table_bar();
-                ImGui::EndTabItem();
-            }*/
-            if (ImGui::BeginTabItem("Biome")) {
-                show_biome_table_bar();
-                ImGui::EndTabItem();
-            }
-
             ImGui::EndTabBar();
         }
+
+        // ImGui::Text("Chunk Build Progress\n");
+        // ImGui::ProgressBar(m_app.world().chunk_gen_fraction());
+        // show_chunk_table_bar();
+
         ImGui::EndTabItem();
     }
 }
 
+void DevPanel::show_server_world_table_bar() {
+
+    ImGui::Text("ChunkGenerator Seed: %u", ChunkGenerator::seed());
+
+    ImGui::Text("Pool Threads %d  Max Support Threads %d  Reserved Threads %d",
+                m_app.server_world().gen_pool_threads(),
+                m_app.server_world().max_threads(), RESERVED_THREADS);
+    ImGui::SliderInt("Set Pool Threads", &m_threads, 1,
+                     m_app.server_world().max_threads());
+    ImGui::SameLine();
+    if (ImGui::Button("Set")) {
+        m_app.server_world().change_pool_threads(
+            ServerWorld::ThreadPoolKind::GEN, m_threads);
+    }
+    if (m_threads > m_app.server_world().max_threads() - RESERVED_THREADS) {
+        ImGui::TextColored(
+            ImVec4(1.0f, 1.0f, 0.0f, 1.0f),
+            "Waring: When the threads in the thread pool exceed \n(maximum "
+            "threads minus reserved threads), \nit may cause stuttering.");
+    }
+
+    m_chunk_style = m_app.server_world().chunk_load_style();
+    if (ImGui::Combo("ChunkLoadStyle", &m_chunk_style, CHUNK_LOAD_STYLE,
+                     IM_ARRAYSIZE(CHUNK_LOAD_STYLE))) {
+        m_app.server_world().set_chunk_load_style(m_chunk_style);
+    }
+
+    if (ImGui::Button("Request Chunk Build")) {
+        m_app.server_world().need_gen(m_player->get_uuid());
+    }
+    ImGui::SameLine();
+    if (ImGui::Checkbox("Gen Thread", &m_gen_thread_running)) {
+        if (m_gen_thread_running) {
+            m_app.server_world().start_gen_thread();
+        } else {
+            m_app.server_world().stop_gen_thread();
+        }
+    }
+    ImGui::Text("Server Chunk Size %d", m_app.server_world().chunk_size());
+
+    if (ImGui::BeginTabBar("World Settings")) {
+        if (ImGui::BeginTabItem("Time")) {
+            show_time_table_bar();
+            ImGui::EndTabItem();
+        }
+        /*
+        if (ImGui::BeginTabItem("Cave")) {
+            show_cave_table_bar();
+            ImGui::EndTabItem();
+        }
+        if (ImGui::BeginTabItem("River")) {
+            show_river_table_bar();
+            ImGui::EndTabItem();
+        }*/
+        if (ImGui::BeginTabItem("Biome")) {
+            show_biome_table_bar();
+            ImGui::EndTabItem();
+        }
+
+        ImGui::EndTabBar();
+    }
+}
+void DevPanel::show_client_world_table_bar() {
+
+    static int rendering_distance = m_app.client_world().rendering_distance();
+    if (ImGui::SliderInt("Render Distance", &rendering_distance, 2, 128)) {
+        m_app.client_world().rendering_distance(rendering_distance);
+        // Config::get().set("world.rendering_distance", rendering_distance);
+    }
+    if (ImGui::Button("Rebuild World")) {
+        m_app.client_world().rebuild_world();
+    }
+    ImGui::SameLine();
+    if (ImGui::Button("Spawn Point")) {
+        m_player->set_player_pos({0.0f, 255.0f, 0.0f});
+    }
+    ImGui::Text("Chunk Task Id %d", m_app.client_world().get_chunk_task_id());
+    ImGui::Text("Client World Chunk %d", m_app.client_world().chunk_size());
+}
+
 void DevPanel::show_player_tab_item() {
     if (!m_player) {
         Logger::error("Player is Nullptr");
         return;
     }
     if (ImGui::BeginTabItem("player")) {
+        ImGui::Text("Player %s", m_player->get_name().c_str());
         if (ImGui::Combo("GameMode", &m_player_profile.game_mode, GAME_MODES,
                          IM_ARRAYSIZE(GAME_MODES))) {
             if (m_player_profile.game_mode == 0) {
@@ -667,8 +670,9 @@ void DevPanel::show_shader_tab_item() {
         ImGui::Checkbox("Flip Y", &m_app.renderer().flip_y());
         if (ImGui::SliderFloat("AmbientStrength",
                                &m_app.renderer().ambient_strength(), 0.0f,
-                               0.35f))
-            ;
+                               0.35f)) {
+        }
+
         ImGui::SliderFloat("SpecularStrength",
                            &m_app.renderer().specular_strength(), 0.0f, 2.0f);
         ImGui::Checkbox("Discard Transparent",

src/gameplay/block.cpp

@@ -1,31 +1,18 @@
 #include "Cubed/gameplay/block.hpp"
 
-#include "Cubed/config.hpp"
 #include "Cubed/tools/cubed_assert.hpp"
 #include "Cubed/tools/log.hpp"
+#include "Cubed/tools/toml.utils.hpp"
 
 #include <filesystem>
-#include <toml++/toml.hpp>
 
 namespace fs = std::filesystem;
 
 using namespace std::string_literals;
-
+using namespace Cubed::TOML;
 namespace {
 std::string block_data_dir = ASSETS_PATH + "data/block"s;
 
-template <Cubed::TomlValueType T>
-std::optional<T> safe_get_value(const toml::table& table, std::string_view key,
-                                const T& default_value) {
-    auto value = table[key].value<T>();
-    if (value == std::nullopt) {
-        Cubed::Logger::warn("Key {} Is Not Find, Wiil Set the Default Value {}",
-                            key, default_value);
-        value = default_value;
-    }
-    return value;
-}
-
 } // namespace
 
 namespace Cubed {

src/gameplay/builders/biome_builder.cpp

@@ -1,39 +1,39 @@
 #include "Cubed/gameplay/builders/biome_builder.hpp"
 
-#include "Cubed/gameplay/chunk.hpp"
 #include "Cubed/gameplay/chunk_generator.hpp"
+#include "Cubed/gameplay/server_chunk.hpp"
 namespace Cubed {
 void BiomeBuilder::build_bottom() {
     ChunkGenerator& chunk_generator = get_chunk_generator();
-    Chunk& chunk = chunk_generator.chunk();
+    ServerChunk& chunk = chunk_generator.chunk();
     auto& m_blocks = chunk.blocks();
     for (int x = 0; x < CHUNK_SIZE; x++) {
         for (int y = 0; y < 5; y++) {
             for (int z = 0; z < CHUNK_SIZE; z++) {
-                m_blocks[Chunk::index(x, y, z)] = 3;
+                m_blocks[ServerChunk::index(x, y, z)] = 3;
             }
         }
     }
 }
 void BiomeBuilder::place_grass() {
     ChunkGenerator& chunk_generator = get_chunk_generator();
-    Chunk& chunk = chunk_generator.chunk();
+    ServerChunk& chunk = chunk_generator.chunk();
     auto& blocks = chunk.blocks();
     const auto& heightmap = chunk.get_heightmap();
     auto& random = chunk_generator.random();
     for (int x = 0; x < SIZE_X; ++x) {
         for (int z = 0; z < SIZE_Z; ++z) {
             int y = heightmap[x][z];
-            BlockType top_id = blocks[Chunk::index(x, y, z)];
+            BlockType top_id = blocks[ServerChunk::index(x, y, z)];
             if (top_id != 1) {
                 continue;
             }
-            if (blocks[Chunk::index(x, y + 1, z)] != 0) {
+            if (blocks[ServerChunk::index(x, y + 1, z)] != 0) {
                 continue;
             }
             if (random.random_bool(0.2)) {
                 if (y + 1 < SIZE_Y) {
-                    blocks[Chunk::index(x, y + 1, z)] = 9;
+                    blocks[ServerChunk::index(x, y + 1, z)] = 9;
                 }
             }
         }
@@ -42,7 +42,7 @@ void BiomeBuilder::place_grass() {
 
 void BiomeBuilder::ocean_water_build() {
     ChunkGenerator& chunk_generator = get_chunk_generator();
-    Chunk& chunk = chunk_generator.chunk();
+    ServerChunk& chunk = chunk_generator.chunk();
     auto& blocks = chunk.blocks();
     const auto& heightmap = chunk.get_heightmap();
 
@@ -51,7 +51,7 @@ void BiomeBuilder::ocean_water_build() {
             int height = heightmap[x][z];
             if (height <= SEA_LEVEL) {
                 for (int y = height; y <= SEA_LEVEL; y++) {
-                    blocks[Chunk::index(x, y, z)] = 7;
+                    blocks[ServerChunk::index(x, y, z)] = 7;
                 }
             }
         }

src/gameplay/builders/desert_builder.cpp

@@ -1,7 +1,7 @@
 #include "Cubed/gameplay/builders/desert_builder.hpp"
 
-#include "Cubed/gameplay/chunk.hpp"
 #include "Cubed/gameplay/chunk_generator.hpp"
+#include "Cubed/gameplay/server_chunk.hpp"
 namespace Cubed {
 DesertBuilder::DesertBuilder(ChunkGenerator& chunk_generator)
     : m_chunk_generator(chunk_generator) {}
@@ -19,11 +19,11 @@ void DesertBuilder::build_blocks() {
         for (int z = 0; z < CHUNK_SIZE; z++) {
             int height = static_cast<int>(m_heightmap[x][z]);
             for (int y = 5; y < height - 5; y++) {
-                m_blocks[Chunk::index(x, y, z)] = 3;
+                m_blocks[ServerChunk::index(x, y, z)] = 3;
             }
 
             for (int y = height - 5; y <= height; y++) {
-                m_blocks[Chunk::index(x, y, z)] = 4;
+                m_blocks[ServerChunk::index(x, y, z)] = 4;
             }
         }
     }

src/gameplay/builders/forest_builder.cpp

@@ -1,7 +1,7 @@
 #include "Cubed/gameplay/builders/forest_builder.hpp"
 
-#include "Cubed/gameplay/chunk.hpp"
 #include "Cubed/gameplay/chunk_generator.hpp"
+#include "Cubed/gameplay/server_chunk.hpp"
 #include "Cubed/gameplay/tree.hpp"
 
 #include <algorithm>
@@ -24,12 +24,12 @@ void ForestBuilder::build_blocks() {
         for (int z = 0; z < CHUNK_SIZE; z++) {
             int height = static_cast<int>(m_heightmap[x][z]);
             for (int y = 5; y < height - 5; y++) {
-                m_blocks[Chunk::index(x, y, z)] = 3;
+                m_blocks[ServerChunk::index(x, y, z)] = 3;
             }
             for (int y = height - 5; y < height; y++) {
-                m_blocks[Chunk::index(x, y, z)] = 2;
+                m_blocks[ServerChunk::index(x, y, z)] = 2;
             }
-            m_blocks[Chunk::index(x, height, z)] = 1;
+            m_blocks[ServerChunk::index(x, height, z)] = 1;
         }
     }
 }

src/gameplay/builders/mountain_builder.cpp

@@ -1,7 +1,7 @@
 #include "Cubed/gameplay/builders/mountain_builder.hpp"
 
-#include "Cubed/gameplay/chunk.hpp"
 #include "Cubed/gameplay/chunk_generator.hpp"
+#include "Cubed/gameplay/server_chunk.hpp"
 namespace Cubed {
 MountainBuilder::MountainBuilder(ChunkGenerator& chunk_generator)
     : m_chunk_generator(chunk_generator) {}
@@ -19,7 +19,7 @@ void MountainBuilder::build_blocks() {
         for (int z = 0; z < CHUNK_SIZE; z++) {
             int height = static_cast<int>(m_heightmap[x][z]);
             for (int y = 5; y <= height; y++) {
-                m_blocks[Chunk::index(x, y, z)] = 3;
+                m_blocks[ServerChunk::index(x, y, z)] = 3;
             }
         }
     }

src/gameplay/builders/ocean_builder.cpp

@@ -1,7 +1,7 @@
 #include "Cubed/gameplay/builders/ocean_builder.hpp"
 
-#include "Cubed/gameplay/chunk.hpp"
 #include "Cubed/gameplay/chunk_generator.hpp"
+#include "Cubed/gameplay/server_chunk.hpp"
 namespace Cubed {
 OceanBuilder::OceanBuilder(ChunkGenerator& chunk_generator)
     : m_chunk_generator(chunk_generator) {}
@@ -19,7 +19,7 @@ void OceanBuilder::build_blocks() {
         for (int z = 0; z < CHUNK_SIZE; z++) {
             int height = static_cast<int>(m_heightmap[x][z]);
             for (int y = 5; y <= height; y++) {
-                m_blocks[Chunk::index(x, y, z)] = 3;
+                m_blocks[ServerChunk::index(x, y, z)] = 3;
             }
         }
     }

src/gameplay/builders/plain_builder.cpp

@@ -1,7 +1,7 @@
 #include "Cubed/gameplay/builders/plain_builder.hpp"
 
-#include "Cubed/gameplay/chunk.hpp"
 #include "Cubed/gameplay/chunk_generator.hpp"
+#include "Cubed/gameplay/server_chunk.hpp"
 namespace Cubed {
 PlainBuilder::PlainBuilder(ChunkGenerator& chunk_generator)
     : m_chunk_generator(chunk_generator) {}
@@ -19,12 +19,12 @@ void PlainBuilder::build_blocks() {
         for (int z = 0; z < CHUNK_SIZE; z++) {
             int height = static_cast<int>(m_heightmap[x][z]);
             for (int y = 5; y < height - 5; y++) {
-                m_blocks[Chunk::index(x, y, z)] = 3;
+                m_blocks[ServerChunk::index(x, y, z)] = 3;
             }
             for (int y = height - 5; y < height; y++) {
-                m_blocks[Chunk::index(x, y, z)] = 2;
+                m_blocks[ServerChunk::index(x, y, z)] = 2;
             }
-            m_blocks[Chunk::index(x, height, z)] = 1;
+            m_blocks[ServerChunk::index(x, height, z)] = 1;
         }
     }
 }

src/gameplay/builders/snowy_plain_builder.cpp

@@ -1,7 +1,7 @@
 #include "Cubed/gameplay/builders/snowy_plain_builder.hpp"
 
-#include "Cubed/gameplay/chunk.hpp"
 #include "Cubed/gameplay/chunk_generator.hpp"
+#include "Cubed/gameplay/server_chunk.hpp"
 namespace Cubed {
 SnowyPlainBuilder::SnowyPlainBuilder(ChunkGenerator& chunk_generator)
     : m_chunk_generator(chunk_generator) {}
@@ -19,12 +19,12 @@ void SnowyPlainBuilder::build_blocks() {
         for (int z = 0; z < CHUNK_SIZE; z++) {
             int height = static_cast<int>(m_heightmap[x][z]);
             for (int y = 5; y < height - 5; y++) {
-                m_blocks[Chunk::index(x, y, z)] = 3;
+                m_blocks[ServerChunk::index(x, y, z)] = 3;
             }
             for (int y = height - 5; y < height; y++) {
-                m_blocks[Chunk::index(x, y, z)] = 2;
+                m_blocks[ServerChunk::index(x, y, z)] = 2;
             }
-            m_blocks[Chunk::index(x, height, z)] = 8;
+            m_blocks[ServerChunk::index(x, height, z)] = 8;
         }
     }
 }

src/gameplay/chunk_generator.cpp

@@ -8,10 +8,10 @@
 #include "Cubed/gameplay/builders/river_builder.hpp"
 #include "Cubed/gameplay/builders/snowy_plain_builder.hpp"
 #include "Cubed/gameplay/cave_path.hpp"
-#include "Cubed/gameplay/chunk.hpp"
 #include "Cubed/gameplay/river.path.hpp"
+#include "Cubed/gameplay/server_chunk.hpp"
+#include "Cubed/gameplay/server_world.hpp"
 #include "Cubed/gameplay/tree.hpp"
-#include "Cubed/gameplay/world.hpp"
 #include "Cubed/tools/cubed_assert.hpp"
 #include "Cubed/tools/cubed_hash.hpp"
 #include "Cubed/tools/math_tools.hpp"
@@ -99,7 +99,7 @@ void carve_worm(const std::vector<PathPoint>& points, const ChunkPos& chunk_pos,
 using enum BiomeType;
 
 constexpr int BLEND_RADIUS = 8;
-ChunkGenerator::ChunkGenerator(Chunk& chunk) : m_chunk(chunk) {
+ChunkGenerator::ChunkGenerator(ServerChunk& chunk) : m_chunk(chunk) {
     ASSERT_MSG(is_init, "ChunksGenerator is not init");
     ChunkPos pos = m_chunk.get_chunk_pos();
     unsigned seed = HASH::chunk_seed_hash(pos.x, pos.z, m_generator_seed);
@@ -157,7 +157,7 @@ void ChunkGenerator::assign_chunk_biome() {
 }
 
 void ChunkGenerator::resolve_biome_adjacency_conflict(
-    const std::array<const Chunk*, 8>& adj_chunks) {
+    const std::array<const ServerChunk*, 8>& adj_chunks) {
     auto m_biome = m_chunk.biome();
     for (int i = 0; i < 8; i++) {
         auto& chunk = adj_chunks[i];
@@ -555,7 +555,8 @@ void ChunkGenerator::blend_surface_blocks_borders(
                                            int nx, int nz) -> BlockType {
         // Search from topmost y downwards for the first non-zero block
         for (int y = WORLD_HEIGHT - 1; y >= 0; --y) {
-            int idx = Chunk::index(nx, y,
+            int idx =
+                ServerChunk::index(nx, y,
                                    nz); // linear index: y * area + z * size + x
             if (idx >= 0 && idx < static_cast<int>(blocks.size())) {
                 BlockType neighbor_type = blocks[idx];
@@ -574,7 +575,7 @@ void ChunkGenerator::blend_surface_blocks_borders(
             BlockType type_self = 0;
             int top_y = -1;
             top_y = m_heightmap[x][z];
-            type_self = m_blocks[Chunk::index(x, top_y, z)];
+            type_self = m_blocks[ServerChunk::index(x, top_y, z)];
 
             if (top_y == -1)
                 continue; // no block? skip
@@ -665,7 +666,7 @@ void ChunkGenerator::blend_surface_blocks_borders(
             if (final_type != type_self) {
                 // top block
                 BlockType new_surface = final_type;
-                m_blocks[Chunk::index(x, top_y, z)] = new_surface;
+                m_blocks[ServerChunk::index(x, top_y, z)] = new_surface;
                 // bottom block
                 unsigned fill_type = 2;
                 if (final_type == 1 || final_type == 8) {
@@ -674,7 +675,7 @@ void ChunkGenerator::blend_surface_blocks_borders(
                     fill_type = final_type;
                 }
                 for (int y = std::max(0, top_y - 5); y < top_y; y++) {
-                    m_blocks[Chunk::index(x, y, z)] = fill_type;
+                    m_blocks[ServerChunk::index(x, y, z)] = fill_type;
                 }
             }
         }
@@ -741,12 +742,12 @@ void ChunkGenerator::generate_cave() {
 
             carve_worm(path.points(), chunk_pos,
                        [&](int x, int y, int z) -> void {
-                           int idx = Chunk::index(x, y, z);
+                           int idx = ServerChunk::index(x, y, z);
                            m_chunk.has_cave() = true;
                            if (blocks[idx] == 7)
                                return;
                            if (y < WORLD_SIZE_Y - 1 &&
-                               blocks[Chunk::index(x, y + 1, z)] == 7)
+                               blocks[ServerChunk::index(x, y + 1, z)] == 7)
                                return;
                            blocks[idx] = 0;
                        });
@@ -780,7 +781,7 @@ void ChunkGenerator::generate_river() {
 
             carve_worm(path.points(), chunk_pos,
                        [&](int x, int y, int z) -> void {
-                           int idx = Chunk::index(x, y, z);
+                           int idx = ServerChunk::index(x, y, z);
                            if (y > SEA_LEVEL) {
                                blocks[idx] = 0;
                                return;
@@ -799,7 +800,7 @@ void ChunkGenerator::generate_river() {
     }
 }
 
-Chunk& ChunkGenerator::chunk() { return m_chunk; }
+ServerChunk& ChunkGenerator::chunk() { return m_chunk; }
 
 Random& ChunkGenerator::random() { return m_random; }
 const std::array<BiomeType, 8>& ChunkGenerator::neighbor_biome() const {

src/gameplay/client_chunk.cpp

@@ -1,10 +1,6 @@
-#include "Cubed/gameplay/chunk.hpp"
+#include "Cubed/gameplay/client_chunk.hpp"
 
-#include "Cubed/gameplay/world.hpp"
 #include "Cubed/tools/cubed_assert.hpp"
-#include "Cubed/tools/log.hpp"
-
-#include <utility>
 
 namespace Cubed {
 using OptionalBlockVectorArray =
@@ -39,14 +35,14 @@ get_block_safe(int lx, int ly, int lz, ChunkPos& chunk_pos,
                const OptionalBlockVectorArray& neighbor_block) {
     if (lx >= 0 && lx < CHUNK_SIZE && ly >= 0 && ly < WORLD_SIZE_Y && lz >= 0 &&
         lz < CHUNK_SIZE) {
-        return blocks[Chunk::index(lx, ly, lz)];
+        return blocks[ClientChunk::index(lx, ly, lz)];
     }
 
     // Out of bounds: check neighbors
     int world_x = lx + chunk_pos.x * CHUNK_SIZE;
     int world_z = lz + chunk_pos.z * CHUNK_SIZE;
 
-    auto [nb_cx, nb_cz] = World::get_chunk_pos(world_x, world_z);
+    auto [nb_cx, nb_cz] = get_chunk_pos(world_x, world_z);
 
     const std::optional<std::vector<BlockType>>* nb = nullptr;
     if (nb_cx == chunk_pos.x + 1)
@@ -69,7 +65,7 @@ get_block_safe(int lx, int ly, int lz, ChunkPos& chunk_pos,
         nby >= WORLD_SIZE_Y || nbz >= CHUNK_SIZE)
         return 0;
 
-    int idx = Chunk::index(nbx, nby, nbz);
+    int idx = ClientChunk::index(nbx, nby, nbz);
     if (static_cast<size_t>(idx) >= (*nb)->size()) {
         return 0;
     }
@@ -100,32 +96,28 @@ inline int choose_buf(BlockType id) {
 }
 
 } // namespace
-
-Chunk::Chunk(World& world, ChunkPos chunk_pos, bool temp_chunk)
-    : m_temp_chunk(temp_chunk), m_chunk_pos(chunk_pos), m_world(world) {
+ClientChunk::ClientChunk(ClientWorld& world) : m_world(world) {
     for (int i = 0; i < VERTEX_DATA_SUM; i++) {
         m_vertex_data.emplace_back(m_world);
     }
 }
+ClientChunk::~ClientChunk() {}
 
-Chunk::~Chunk() {}
-
-Chunk::Chunk(Chunk&& other) noexcept
+ClientChunk::ClientChunk(ClientChunk&& other) noexcept
     : m_dirty(other.is_dirty()), m_need_upload(other.m_need_upload.load()),
       m_is_on_gen_vertex_data(other.m_is_on_gen_vertex_data.load()),
       m_biome(other.m_biome.load()), m_chunk_pos(std::move(other.m_chunk_pos)),
-      m_world(other.m_world), m_heightmap(std::move(other.m_heightmap)),
-      m_blocks(std::move(other.m_blocks)),
-      m_vertex_data(std::move(other.m_vertex_data)), m_seed(other.m_seed),
-      m_conditions(other.m_conditions), m_info(std::move(other.m_info)) {}
+      m_world(other.m_world), m_blocks(std::move(other.m_blocks)),
+      m_vertex_data(std::move(other.m_vertex_data)), m_seed(other.m_seed) {}
 
-Chunk& Chunk::operator=(Chunk&& other) noexcept {
+ClientChunk& ClientChunk::operator=(ClientChunk&& other) noexcept {
     // Logger::info("other Chunk pos {} {} in Chunk& Chunk::operator=(Chunk&&
     // other) this {}", other.m_chunk_pos.x, other.m_chunk_pos.z,
     // static_cast<const void*>(&other));
-
+    if (this == &other) {
+        return *this;
+    }
     m_chunk_pos = std::move(other.m_chunk_pos);
-    m_heightmap = std::move(other.m_heightmap);
     m_blocks = std::move(other.m_blocks);
     m_dirty = other.is_dirty();
     m_vertex_data = std::move(other.m_vertex_data);
@@ -133,55 +125,10 @@ Chunk& Chunk::operator=(Chunk&& other) noexcept {
     m_is_on_gen_vertex_data = other.m_is_on_gen_vertex_data.load();
     m_need_upload = other.m_need_upload.load();
     m_seed = other.m_seed;
-    m_conditions = other.m_conditions;
-    m_info = std::move(other.m_info);
     return *this;
 }
 
-std::tuple<int, int, int> Chunk::world_to_block(int world_x, int world_y,
-                                                int world_z, int chunk_x,
-                                                int chunk_z) {
-    int x, y, z;
-    y = world_y;
-    x = world_x - chunk_x * CHUNK_SIZE;
-    z = world_z - chunk_z * CHUNK_SIZE;
-    return {x, y, z};
-}
-
-std::tuple<int, int, int> Chunk::world_to_block(const glm::ivec3& block_pos,
-                                                ChunkPos chunk_pos) {
-    return world_to_block(block_pos.x, block_pos.y, block_pos.z, chunk_pos.x,
-                          chunk_pos.z);
-}
-
-std::tuple<int, int, int> Chunk::block_to_world(int x, int y, int z,
-                                                int chunk_x, int chunk_z) {
-    int world_x = x + chunk_x * CHUNK_SIZE;
-    int world_z = z + chunk_z * CHUNK_SIZE;
-    int world_y = y;
-    return {world_x, world_y, world_z};
-}
-std::tuple<int, int, int> Chunk::block_to_world(const glm::ivec3& block_pos,
-                                                ChunkPos chunk_pos) {
-    return block_to_world(block_pos.x, block_pos.y, block_pos.z, chunk_pos.x,
-                          chunk_pos.z);
-}
-
-BiomeType Chunk::get_biome() const { return m_biome.load(); }
-
-ChunkPos Chunk::get_chunk_pos() const { return m_chunk_pos; }
-
-const std::vector<BlockType>& Chunk::get_chunk_blocks() const {
-    return m_blocks;
-}
-
-HeightMapArray Chunk::get_heightmap() const {
-    // Logger::info("Chunk pos {} {} in get_heightmap this {}", m_chunk_pos.x,
-    // m_chunk_pos.z, static_cast<const void*>(this));
-    return m_heightmap;
-}
-
-int Chunk::index(int x, int y, int z) {
+int ClientChunk::index(int x, int y, int z) {
     ASSERT(!(x < 0 || y < 0 || z < 0 || x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
              z >= CHUNK_SIZE));
     if ((x * WORLD_SIZE_Y + y) * CHUNK_SIZE + z < 0 ||
@@ -193,15 +140,50 @@ int Chunk::index(int x, int y, int z) {
     return (x * WORLD_SIZE_Y + y) * CHUNK_SIZE + z;
 }
 
-int Chunk::index(const glm::vec3& pos) {
-    return Chunk::index(pos.x, pos.y, pos.z);
+int ClientChunk::index(const glm::vec3& pos) {
+    return ClientChunk::index(pos.x, pos.y, pos.z);
+}
+std::tuple<int, int, int> ClientChunk::world_to_block(int world_x, int world_y,
+                                                      int world_z, int chunk_x,
+                                                      int chunk_z) {
+    int x, y, z;
+    y = world_y;
+    x = world_x - chunk_x * CHUNK_SIZE;
+    z = world_z - chunk_z * CHUNK_SIZE;
+    return {x, y, z};
 }
 
-void Chunk::gen_vertex_data(const OptionalBlockVectorArray& neighbor_block) {
-    if (m_is_on_gen_vertex_data) {
+std::tuple<int, int, int>
+ClientChunk::world_to_block(const glm::ivec3& block_pos, ChunkPos chunk_pos) {
+    return world_to_block(block_pos.x, block_pos.y, block_pos.z, chunk_pos.x,
+                          chunk_pos.z);
+}
+
+std::tuple<int, int, int>
+ClientChunk::block_to_world(int x, int y, int z, int chunk_x, int chunk_z) {
+    int world_x = x + chunk_x * CHUNK_SIZE;
+    int world_z = z + chunk_z * CHUNK_SIZE;
+    int world_y = y;
+    return {world_x, world_y, world_z};
+}
+std::tuple<int, int, int>
+ClientChunk::block_to_world(const glm::ivec3& block_pos, ChunkPos chunk_pos) {
+    return block_to_world(block_pos.x, block_pos.y, block_pos.z, chunk_pos.x,
+                          chunk_pos.z);
+}
+BiomeType ClientChunk::get_biome() const { return m_biome.load(); }
+
+ChunkPos ClientChunk::get_chunk_pos() const { return m_chunk_pos; }
+
+const std::vector<BlockType>& ClientChunk::get_chunk_blocks() const {
+    return m_blocks;
+}
+
+void ClientChunk::gen_vertex_data(
+    const OptionalBlockVectorArray& neighbor_block) {
+    if (m_is_on_gen_vertex_data.exchange(true)) {
         return;
     }
-    m_is_on_gen_vertex_data = true;
     std::lock_guard lk(m_vertexs_data_mutex);
 
     for (auto& data : m_vertex_data) {
@@ -216,108 +198,44 @@ void Chunk::gen_vertex_data(const OptionalBlockVectorArray& neighbor_block) {
     m_is_on_gen_vertex_data = false;
 }
 
-GLuint Chunk::get_normal_vao() const { return m_vertex_data[0].m_vao; }
+GLuint ClientChunk::get_normal_vao() const { return m_vertex_data[0].m_vao; }
 
-size_t Chunk::get_normal_vertices_sum() const {
+size_t ClientChunk::get_normal_vertices_sum() const {
     if (m_vertex_data[0].m_sum == 0) {
         Logger::warn("m_normal_vertices_sum is 0");
     }
     return m_vertex_data[0].m_sum.load();
 }
 
-GLuint Chunk::get_cross_vao() const { return m_vertex_data[1].m_vao; }
-size_t Chunk::get_cross_vertices_sum() const {
+GLuint ClientChunk::get_cross_vao() const { return m_vertex_data[1].m_vao; }
+size_t ClientChunk::get_cross_vertices_sum() const {
     return m_vertex_data[1].m_sum.load();
 }
 
-GLuint Chunk::get_normal_discard_vao() const { return m_vertex_data[2].m_vao; }
-size_t Chunk::get_normal_discard_vertices_sum() const {
+GLuint ClientChunk::get_normal_discard_vao() const {
+    return m_vertex_data[2].m_vao;
+}
+size_t ClientChunk::get_normal_discard_vertices_sum() const {
     return m_vertex_data[2].m_sum.load();
 }
 
-GLuint Chunk::get_normal_blend_vao() const { return m_vertex_data[3].m_vao; }
-size_t Chunk::get_normal_blend_vertices_sum() const {
+GLuint ClientChunk::get_normal_blend_vao() const {
+    return m_vertex_data[3].m_vao;
+}
+size_t ClientChunk::get_normal_blend_vertices_sum() const {
     return m_vertex_data[3].m_sum.load();
 }
 
-GLuint Chunk::get_water_vao() const { return m_vertex_data[4].m_vao; }
-size_t Chunk::get_water_vertices_sum() const {
+GLuint ClientChunk::get_water_vao() const { return m_vertex_data[4].m_vao; }
+size_t ClientChunk::get_water_vertices_sum() const {
     return m_vertex_data[4].m_sum.load();
 }
 
-void Chunk::gen_phase_one() {
-    m_generator = std::make_unique<ChunkGenerator>(*this);
-    if (!m_generator) {
-        Logger::error("ChunkGenerator is Nullptr");
+void ClientChunk::upload_to_gpu() {
+
+    if (!is_need_upload()) {
         return;
     }
-    m_generator->assign_chunk_biome();
-    m_seed = m_generator->chunk_seed();
-}
-
-void Chunk::gen_phase_two(const std::array<const Chunk*, 8>& adj_chunks) {
-    if (!m_generator) {
-        Logger::error("ChunkGenerator is Nullptr");
-        return;
-    }
-    // m_generator->resolve_biome_adjacency_conflict(adj_chunks);
-}
-
-void Chunk::gen_phase_three() {
-    if (!m_generator) {
-        Logger::error("ChunkGenerator is Nullptr");
-        return;
-    }
-    m_generator->generate_heightmap();
-}
-
-void Chunk::gen_phase_four(
-    const std::array<std::optional<HeightMapArray>, 8>& neighbor_heightmap,
-    const std::array<BiomeType, 8>& neighbor_biome) {
-    if (!m_generator) {
-        Logger::error("ChunkGenerator is Nullptr");
-        return;
-    }
-    // m_generator->blend_heightmap_boundaries(neighbor_heightmap,
-    // neighbor_biome);
-}
-
-void Chunk::gen_phase_five() {
-    if (!m_generator) {
-        Logger::error("ChunkGenerator is Nullptr");
-        return;
-    }
-    m_generator->generate_terrain_blocks();
-}
-
-void Chunk::gen_phase_six(
-    const std::array<std::optional<std::vector<BlockType>>, 4>&
-        neighbor_block) {
-    if (!m_generator) {
-        Logger::error("ChunkGenerator is Nullptr");
-        return;
-    }
-    // This must be fully completed before any other operations can proceed!
-    m_generator->blend_surface_blocks_borders(neighbor_block);
-}
-
-void Chunk::gen_phase_seven() {
-    if (!m_generator) {
-        Logger::error("ChunkGenerator is Nullptr");
-        return;
-    }
-    m_generator->ocean_build();
-    m_generator->generate_river();
-    m_generator->generate_cave();
-
-    m_generator->generate_vegetation();
-    mark_dirty();
-    m_generator = nullptr;
-}
-
-void Chunk::upload_to_gpu() {
-
-    ASSERT(is_need_upload());
 
     std::lock_guard lk(m_vertexs_data_mutex);
 
@@ -327,206 +245,64 @@ void Chunk::upload_to_gpu() {
 
     // after fininshed it, can use
     clear_dirty();
+
+    m_render_snapshot = {
+        get_normal_vao(),
+        get_normal_vertices_sum(),
+        get_cross_vao(),
+        get_cross_vertices_sum(),
+        get_normal_discard_vao(),
+        get_normal_discard_vertices_sum(),
+        get_normal_blend_vao(),
+        get_normal_blend_vertices_sum(),
+        get_water_vao(),
+        get_water_vertices_sum(),
+        glm::vec3(static_cast<float>(m_chunk_pos.x * CHUNK_SIZE) +
+                      static_cast<float>(CHUNK_SIZE / 2),
+                  static_cast<float>(WORLD_SIZE_Y / 2),
+                  static_cast<float>(m_chunk_pos.z * CHUNK_SIZE) +
+                      static_cast<float>(CHUNK_SIZE / 2)),
+        glm::vec3(static_cast<float>(CHUNK_SIZE / 2),
+                  static_cast<float>(WORLD_SIZE_Y / 2),
+                  static_cast<float>(CHUNK_SIZE / 2))};
+
     m_need_upload = false;
 }
 
-bool Chunk::is_dirty() const { return m_dirty.load(); }
+bool ClientChunk::is_dirty() const { return m_dirty.load(); }
 
-void Chunk::mark_dirty() { m_dirty = true; }
+void ClientChunk::mark_dirty() { m_dirty = true; }
 
-void Chunk::clear_dirty() { m_dirty = false; }
+void ClientChunk::clear_dirty() { m_dirty = false; }
 
-bool Chunk::is_need_upload() const { return m_need_upload.load(); }
+bool ClientChunk::is_need_upload() const { return m_need_upload.load(); }
 
-void Chunk::need_upload() { m_need_upload = true; }
+void ClientChunk::need_upload() { m_need_upload = true; }
 
-void Chunk::set_chunk_block(int index, unsigned id) {
+void ClientChunk::set_chunk_block(int index, unsigned id) {
     m_blocks[index] = id;
-    mark_dirty();
 }
 
-ChunkPos Chunk::chunk_pos() const { return m_chunk_pos; }
+ChunkPos ClientChunk::chunk_pos() const { return m_chunk_pos; }
 
-BiomeType Chunk::biome() const { return m_biome; }
+BiomeType ClientChunk::biome() const { return m_biome; }
 
-void Chunk::biome(BiomeType b) { m_biome = b; }
+void ClientChunk::biome(BiomeType b) { m_biome = b; }
 
-HeightMapArray& Chunk::heightmap() { return m_heightmap; }
-std::vector<BlockType>& Chunk::blocks() { return m_blocks; }
-World& Chunk::world() { return m_world; }
-unsigned Chunk::seed() const {
+std::vector<BlockType>& ClientChunk::blocks() { return m_blocks; }
+ClientWorld& ClientChunk::world() { return m_world; }
+unsigned ClientChunk::seed() const {
     if (m_seed == 0) {
         Logger::warn("Seed Not Generator");
     }
     return m_seed;
 }
 
-BiomeConditions& Chunk::conditions() { return m_conditions; }
-
-ChunkInfo Chunk::get_info() const {
-    if (m_gening) {
-        return ChunkInfo{};
-    }
-    return m_info;
+const ChunkRenderSnapshot* ClientChunk::get_render_snapshot() const {
+    return &m_render_snapshot;
 }
-/*
-void Chunk::gen_vertices(const OptionalBlockVectorArray& neighbor_block) {
-    static const glm::ivec3 DIR[6] = {{0, 0, 1},  {1, 0, 0}, {0, 0, -1},
-                                      {-1, 0, 0}, {0, 1, 0}, {0, -1, 0}};
 
-    for (int x = 0; x < SIZE_X; x++) {
-        for (int y = 0; y < SIZE_Y; y++) {
-            for (int z = 0; z < SIZE_Z; z++) {
-                int world_x = x + m_chunk_pos.x * CHUNK_SIZE;
-                int world_z = z + m_chunk_pos.z * CHUNK_SIZE;
-                int world_y = y;
-                int cur_id = m_blocks[index(x, y, z)];
-                // air
-                if (cur_id == 0) {
-                    continue;
-                }
-
-                for (int face = 0; face < 6; face++) {
-                    int nx = x + DIR[face].x;
-                    int ny = y + DIR[face].y;
-                    int nz = z + DIR[face].z;
-                    bool neighbor_culled = false;
-
-                    if (nx < 0 || nx >= SIZE_X || ny < 0 || ny >= SIZE_Y ||
-                        nz < 0 || nz >= SIZE_Z) {
-
-                        int world_nx = world_x + DIR[face].x;
-                        int world_ny = world_y + DIR[face].y;
-                        int world_nz = world_z + DIR[face].z;
-
-                        auto [neighbor_x, neighbor_z] =
-                            World::get_chunk_pos(world_nx, world_nz);
-
-                        auto is_culled =
-                            [&](const std::optional<std::vector<BlockType>>&
-                                    chunk_blocks) {
-                                if (chunk_blocks == std::nullopt) {
-                                    return true;
-                                }
-                                int x, y, z;
-                                y = world_ny;
-                                x = world_nx - neighbor_x * CHUNK_SIZE;
-                                z = world_nz - neighbor_z * CHUNK_SIZE;
-                                if (x < 0 || y < 0 || z < 0 ||
-                                    x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
-                                    z >= CHUNK_SIZE) {
-                                    return false;
-                                }
-
-                                int idx = Chunk::index(x, y, z);
-                                // not init
-                                if (static_cast<size_t>(idx) >=
-                                    chunk_blocks->size()) {
-                                    // Logger::warn("not init");
-                                    return true;
-                                }
-                                auto id = (*chunk_blocks)[idx];
-                                // transparent
-                                if (BlockManager::is_transparent(id)) {
-                                    if (id == cur_id) {
-                                        return true;
-                                    } else {
-                                        return false;
-                                    }
-
-                                } else {
-                                    return true;
-                                }
-                            };
-
-                        if (m_chunk_pos.x + 1 == neighbor_x) {
-                            neighbor_culled = is_culled(neighbor_block[0]);
-                        } else if (m_chunk_pos.x - 1 == neighbor_x) {
-                            neighbor_culled = is_culled(neighbor_block[1]);
-                        } else if (m_chunk_pos.z + 1 == neighbor_z) {
-                            neighbor_culled = is_culled(neighbor_block[2]);
-                        } else if (m_chunk_pos.z - 1 == neighbor_z) {
-                            neighbor_culled = is_culled(neighbor_block[3]);
-                        }
-                        // neighbor_cull = m_world.is_block(glm::ivec3(world_x,
-                        // world_y, world_z) + DIR[face]);
-                    } else {
-                        auto neighbor_id = m_blocks[index(nx, ny, nz)];
-                        // transparent block
-                        if (!BlockManager::is_transparent(neighbor_id)) {
-                            neighbor_culled = true;
-                        } else {
-                            if (neighbor_id == cur_id) {
-                                neighbor_culled = true;
-                            } else {
-                                neighbor_culled = false;
-                            }
-                        }
-                    }
-
-                    if (neighbor_culled) {
-                        continue;
-                    }
-                    if (BlockManager::is_cross_plane(cur_id)) {
-                        gen_cross_plane_vertices(world_x, world_y, world_z,
-                                                 cur_id);
-                    }
-                    for (int i = 0; i < 6; i++) {
-                        Vertex3D vex = {
-                            VERTICES_POS[face][i][0] + (float)world_x * 1.0f,
-                            VERTICES_POS[face][i][1] + (float)world_y * 1.0f,
-                            VERTICES_POS[face][i][2] + (float)world_z * 1.0f,
-                            TEX_COORDS[face][i][0],
-                            TEX_COORDS[face][i][1],
-
-                            static_cast<float>(cur_id * 6 + face),
-
-                            NORMALS[face][i][0],
-                            NORMALS[face][i][1],
-                            NORMALS[face][i][2],
-                            BlockManager::roughness(cur_id),
-                            TANGENTS[face][i][0],
-                            TANGENTS[face][i][1],
-                            TANGENTS[face][i][2]
-
-                        };
-                        if (BlockManager::is_transparent(cur_id)) {
-                            if (BlockManager::is_discard(cur_id) &&
-                                BlockManager::is_blend(cur_id)) {
-                                Logger::warn(
-                                    "Block id {} is both discard and blend is "
-                                    "must only one can true !!!",
-                                    cur_id);
-                            }
-                            if (BlockManager::is_discard(cur_id)) {
-                                m_vertex_data[2].m_vertices.emplace_back(vex);
-                            } else if (BlockManager::is_blend(cur_id)) {
-                                if (cur_id == 7) {
-                                    m_vertex_data[4].m_vertices.emplace_back(
-                                        vex);
-                                } else {
-                                    m_vertex_data[3].m_vertices.emplace_back(
-                                        vex);
-                                }
-
-                            } else {
-                                Logger::warn("Id {} is transparent but not "
-                                             "discard or blend",
-                                             cur_id);
-                                m_vertex_data[3].m_vertices.emplace_back(vex);
-                            }
-
-                        } else {
-                            m_vertex_data[0].m_vertices.emplace_back(vex);
-                        }
-                    }
-                }
-            }
-        }
-    }
-}
-*/
-void Chunk::gen_vertices(const OptionalBlockVectorArray& neighbor_block) {
+void ClientChunk::gen_vertices(const OptionalBlockVectorArray& neighbor_block) {
 
     // SIZE_X=SIZE_Z=CHUNK_SIZE=16, SIZE_Y=WORLD_SIZE_Y=256
     // Axis order: axis 0=X, 1=Y, 2=Z
@@ -648,8 +424,8 @@ void Chunk::gen_vertices(const OptionalBlockVectorArray& neighbor_block) {
         }
     }
 }
-void Chunk::emit_quad(int axis, int face_dir, int layer, int i, int j, int w,
-                      int h, int u_axis, int v_axis, FaceKey key) {
+void ClientChunk::emit_quad(int axis, int face_dir, int layer, int i, int j,
+                            int w, int h, int u_axis, int v_axis, FaceKey key) {
     float axis_val = (float)(layer + (face_dir > 0 ? 1 : 0));
     float wx_base = (float)(m_chunk_pos.x * CHUNK_SIZE);
     float wz_base = (float)(m_chunk_pos.z * CHUNK_SIZE);
@@ -720,8 +496,8 @@ void Chunk::emit_quad(int axis, int face_dir, int layer, int i, int j, int w,
     }
 }
 
-void Chunk::gen_cross_plane_vertices(int world_x, int world_y, int world_z,
-                                     BlockType id) {
+void ClientChunk::gen_cross_plane_vertices(int world_x, int world_y,
+                                           int world_z, BlockType id) {
 
     if (!BlockManager::is_cross_plane(id)) {
         Logger::warn("Block {} {} {} id {} is not cross plane", world_x,
@@ -751,55 +527,47 @@ void Chunk::gen_cross_plane_vertices(int world_x, int world_y, int world_z,
     }
 }
 
-void Chunk::gen_chunk() {
-    if (m_gening.exchange(true))
+void ClientChunk::receive_chunk(const ChunkDataRsp& data) {
+    OptionalBlockVectorArray neighbor;
+
+    m_chunk_pos.x = data.pos().x();
+    m_chunk_pos.z = data.pos().z();
+    m_seed = data.chunk_seed();
+    m_biome = get_biome_from_id(data.biome_type());
+
+    for (int i = 0; i < 4; i++) {
+        neighbor[i] = std::nullopt;
+    }
+
+    if (data.chunk_blocks_size() != BLOCK_SIZE) {
+        Logger::error("Bad Chunk, size {}", data.chunk_blocks_size());
         return;
-    m_gening = true;
-    if (m_blocks.size() != 0) {
-        Logger::warn(
-            "Request Generator Chunk {} {} ,but the Blocks size is Not 0",
-            m_chunk_pos.x, m_chunk_pos.z);
     }
-    std::vector<Chunk> neighbor;
-    for (int i = 0; i < 4; i++) {
-        neighbor.emplace_back(m_world, m_chunk_pos + CHUNK_DIR[i], true);
-    }
-    for (auto& chunk : neighbor) {
-        chunk.gen_phase_one();
-        chunk.gen_phase_three();
-        chunk.gen_phase_five();
-        chunk.gen_phase_seven();
-    }
-    gen_phase_one();
-    gen_phase_three();
-    gen_phase_five();
+    m_blocks.reserve(BLOCK_SIZE);
 
-    OptionalBlockVectorArray neightbor_blocks;
-    for (int i = 0; i < 4; i++) {
-        neightbor_blocks[i] = neighbor[i].get_chunk_blocks();
+    for (const auto& b : data.chunk_blocks()) {
+        m_blocks.push_back(static_cast<BlockType>(b));
     }
-    gen_phase_six(neightbor_blocks);
-    gen_phase_seven();
-    for (int i = 0; i < 4; i++) {
-        neightbor_blocks[i] = neighbor[i].get_chunk_blocks();
-    }
-    gen_vertex_data(neightbor_blocks);
+    // temp neighbor block data
+    auto load_neighbor = [&](int idx, const auto& blocks) {
+        if (blocks.size() != BLOCK_SIZE)
+            return;
 
-    // collect chunk info for debugging
-    m_info.biome = m_biome;
-    m_info.pos = m_chunk_pos;
-    m_info.seed = m_seed;
-    Random r(m_seed);
-    unsigned first = r.engine()();
-    m_info.first_random = first;
-    r.init(m_seed);
-    m_info.has_cave_start = r.random_bool(DEFAULT_CAVE_PROBABILITY);
-    m_info.has_cave = m_has_cave;
+        neighbor[idx].emplace();
+        neighbor[idx]->reserve(BLOCK_SIZE);
+
+        for (auto b : blocks) {
+            neighbor[idx]->push_back(static_cast<BlockType>(b));
+        }
+    };
+
+    load_neighbor(0, data.neighbor_blocks_1());
+    load_neighbor(1, data.neighbor_blocks_2());
+    load_neighbor(2, data.neighbor_blocks_3());
+    load_neighbor(3, data.neighbor_blocks_4());
+
+    gen_vertex_data(neighbor);
+    mark_dirty();
 }
-// Logger::info("Cross Sum {}", m_cross_vertices_sum.load());
 
-bool Chunk::is_temp_chunk() const { return m_temp_chunk.load(); }
-
-bool& Chunk::has_cave() { return m_has_cave; }
-
-} // namespace Cubed
+} // namespace Cubed

src/gameplay/client_player.cpp

@@ -1,48 +1,41 @@
-#include "Cubed/gameplay/player.hpp"
+#include "Cubed/gameplay/client_player.hpp"
 
 #include "Cubed/config.hpp"
 #include "Cubed/debug_collector.hpp"
-#include "Cubed/gameplay/world.hpp"
-#include "Cubed/tools/log.hpp"
-
-#include <GLFW/glfw3.h>
+#include "Cubed/gameplay/client_world.hpp"
 
 namespace Cubed {
+ClientPlayer::ClientPlayer(ClientWorld& world) : m_world(world) {}
+ClientPlayer::~ClientPlayer() {}
 
-Player::Player(World& world, const std::string& name)
-    : m_name(name), m_world(world) {
-    hot_reload();
-}
-Player::~Player() {}
+AABB ClientPlayer::get_aabb(const glm::vec3& pos) {
+    float half_width = M_SIZE.x / 2.0f;
+    float half_depth = M_SIZE.z / 2.0f;
 
-AABB Player::get_aabb() const {
-    float half_width = m_size.x / 2.0f;
-    float half_depth = m_size.z / 2.0f;
+    glm::vec3 min{pos.x - half_width, pos.y, pos.z - half_depth};
 
-    glm::vec3 min{m_player_pos.x - half_width, m_player_pos.y,
-                  m_player_pos.z - half_depth};
-
-    glm::vec3 max{m_player_pos.x + half_width, m_player_pos.y + m_size.y,
-                  m_player_pos.z + half_depth};
+    glm::vec3 max{pos.x + half_width, pos.y + M_SIZE.y, pos.z + half_depth};
 
     return AABB{min, max};
 }
+const glm::vec3& ClientPlayer::get_front() const { return m_front; }
 
-const glm::vec3& Player::get_front() const { return m_front; }
+const Gait& ClientPlayer::get_gait() const { return m_gait; }
 
-const Gait& Player::get_gait() const { return m_gait; }
-
-const std::optional<LookBlock>& Player::get_look_block_pos() const {
+const std::optional<LookBlock>& ClientPlayer::get_look_block_pos() const {
     return m_look_block;
 }
+glm::vec3 ClientPlayer::get_player_pos() const {
 
-const glm::vec3& Player::get_player_pos() const { return m_player_pos; }
+    std::shared_lock lock(m_player_pos_mutex);
+    return m_player_pos;
+}
 
-const MoveState& Player::get_move_state() const { return m_move_state; }
+const MoveState& ClientPlayer::get_move_state() const { return m_move_state; }
 
-bool Player::ray_cast(const glm::vec3& start, const glm::vec3& front,
-                      glm::ivec3& block_pos, glm::vec3& normal,
-                      float distance) {
+bool ClientPlayer::ray_cast(const glm::vec3& start, const glm::vec3& front,
+                            glm::ivec3& block_pos, glm::vec3& normal,
+                            float distance) {
     glm::vec3 dir = glm::normalize(front);
     // float step = 0.1f;
     glm::ivec3 cur = glm::floor(start);
@@ -113,7 +106,7 @@ bool Player::ray_cast(const glm::vec3& start, const glm::vec3& front,
     return false;
 }
 
-void Player::change_mode(GameMode mode) {
+void ClientPlayer::change_mode(GameMode mode) {
     m_game_mode = mode;
     Logger::info("Change GameMode to {}", to_str(mode));
     if (mode == CREATIVE) {
@@ -125,22 +118,19 @@ void Player::change_mode(GameMode mode) {
         m_max_speed = m_max_run_speed;
     }
 }
-
-void Player::hot_reload() {
+void ClientPlayer::hot_reload() {
     auto& config = Config::get();
     m_sensitivity =
         static_cast<float>(config.get<double>("player.mouse_sensitivity"));
 }
+void ClientPlayer::set_player_pos(const glm::vec3& pos) { m_player_pos = pos; }
 
-void Player::set_player_pos(const glm::vec3& pos) { m_player_pos = pos; }
+void ClientPlayer::set_place_block(unsigned id) { m_place_block = id; }
 
-void Player::set_place_block(unsigned id) { m_place_block = id; }
-
-void Player::update(float delta_time) {
+void ClientPlayer::update(float delta_time) {
 
     update_move(delta_time);
     update_lookup_block();
-    check_player_chunk_transition();
 
     DebugCollector::get().report("player_pos",
                                  std::format("x: {:.2f} y: {:.2f} z: {:.2f}",
@@ -150,8 +140,7 @@ void Player::update(float delta_time) {
     DebugCollector::get().report("speed",
                                  std::format("Speed: {:.2} m/s", m_xz_speed));
 }
-
-void Player::update_player_move_state(int key, int action) {
+void ClientPlayer::update_player_move_state(int key, int action) {
     switch (key) {
     case GLFW_KEY_W:
         if (action == GLFW_PRESS) {
@@ -231,7 +220,7 @@ void Player::update_player_move_state(int key, int action) {
     }
 }
 
-void Player::update_front_vec(float offset_x, float offset_y) {
+void ClientPlayer::update_front_vec(float offset_x, float offset_y) {
     m_yaw += offset_x * m_sensitivity;
     m_pitch += offset_y * m_sensitivity;
 
@@ -246,15 +235,7 @@ void Player::update_front_vec(float offset_x, float offset_y) {
     m_front = glm::normalize(m_front);
 }
 
-void Player::check_player_chunk_transition() {
-    ChunkPos cur_pos = m_world.get_chunk_pos(m_player_pos.x, m_player_pos.z);
-    if (cur_pos != m_player_chunk_pos) {
-        m_world.need_gen();
-        m_player_chunk_pos = cur_pos;
-    }
-}
-
-void Player::update_direction() {
+void ClientPlayer::update_direction() {
     m_right = glm::normalize(glm::cross(m_front, glm::vec3(0.0f, 1.0f, 0.0f)));
 
     glm::vec3 move_dir_front = glm::vec3(0.0f);
@@ -279,7 +260,7 @@ void Player::update_direction() {
     }
 }
 
-void Player::update_lookup_block() {
+void ClientPlayer::update_lookup_block() {
     // calculate the block that is looked
     glm::ivec3 block_pos;
     glm::vec3 block_normal;
@@ -294,25 +275,17 @@ void Player::update_lookup_block() {
     if (m_look_block != std::nullopt) {
         if (Input::get_input_state().mouse_state.left) {
             if (m_world.is_solid(m_look_block->pos)) {
-                m_world.set_block(m_look_block->pos, 0);
+                m_world.report_block_change(m_look_block->pos, 0);
             }
             Input::get_input_state().mouse_state.left = false;
         }
         if (Input::get_input_state().mouse_state.right) {
             glm::ivec3 near_pos = m_look_block->pos + m_look_block->normal;
             if (!m_world.is_solid(near_pos)) {
-                auto x = near_pos.x;
-                auto y = near_pos.y;
-                auto z = near_pos.z;
-                AABB block_box = {glm::vec3{static_cast<float>(x),
-                                            static_cast<float>(y),
-                                            static_cast<float>(z)},
-                                  glm::vec3{static_cast<float>(x + 1),
-                                            static_cast<float>(y + 1),
-                                            static_cast<float>(z + 1)}};
-                AABB player_box = get_aabb();
+                AABB block_box = ClientWorld::get_block_aabb(near_pos);
+                AABB player_box = get_aabb(get_player_pos());
                 if (!player_box.intersects(block_box)) {
-                    m_world.set_block(near_pos, m_place_block);
+                    m_world.report_block_change(near_pos, m_place_block);
                 }
             }
             Input::get_input_state().mouse_state.right = false;
@@ -320,11 +293,19 @@ void Player::update_lookup_block() {
     }
 }
 
-void Player::update_move(float delta_time) {
+void ClientPlayer::update_move(float delta_time) {
     // if frame rate less than 1 frame per second, don't update
     if (delta_time > 1.0f) {
         return;
     }
+    // ensure the thread safe
+    glm::vec3 player_pos;
+
+    {
+        std::shared_lock lock(m_player_pos_mutex);
+        player_pos = m_player_pos;
+    }
+
     if (m_game_mode != SPECTATOR) {
         if (m_gait == Gait::RUN) {
             m_max_speed = m_max_run_speed;
@@ -386,24 +367,30 @@ void Player::update_move(float delta_time) {
 
     move_distance.y = m_y_speed * delta_time;
     // y
-    update_y_move();
+    update_y_move(player_pos);
     // x
-    update_x_move();
+    update_x_move(player_pos);
 
-    update_z_move();
+    update_z_move(player_pos);
 
-    if (m_player_pos.y < -15.0f) {
+    if (player_pos.y < -15.0f) {
         Logger::warn("y is tow low");
-        m_player_pos += glm::vec3(1.0f, 100.0f, 1.0f);
+        player_pos += glm::vec3(1.0f, 100.0f, 1.0f);
     }
+
+    {
+        std::lock_guard lock(m_player_pos_mutex);
+        m_player_pos = player_pos;
+    }
+    update_player_chunk();
 }
 
-void Player::update_x_move() {
-    m_player_pos.x += move_distance.x;
+void ClientPlayer::update_x_move(glm::vec3& player_pos) {
+    player_pos.x += move_distance.x;
     if (m_game_mode == SPECTATOR) {
         return;
     }
-    AABB player_box = get_aabb();
+    AABB player_box = get_aabb(player_pos);
     int minx = std::floor(player_box.min.x);
     int maxx = std::floor(player_box.max.x);
     int miny = std::floor(player_box.min.y);
@@ -414,16 +401,12 @@ void Player::update_x_move() {
     for (int x = minx; x <= maxx; ++x) {
         for (int y = miny; y <= maxy; ++y) {
             for (int z = minz; z <= maxz; ++z) {
-                if (!m_world.can_pass_block(glm::ivec3{x, y, z})) {
-                    AABB block_box = {glm::vec3{static_cast<float>(x),
-                                                static_cast<float>(y),
-                                                static_cast<float>(z)},
-                                      glm::vec3{static_cast<float>(x + 1),
-                                                static_cast<float>(y + 1),
-                                                static_cast<float>(z + 1)}};
+                glm::ivec3 block_pos{x, y, z};
+                if (!m_world.can_pass_block(block_pos)) {
+                    AABB block_box = ClientWorld::get_block_aabb(block_pos);
                     if (player_box.intersects(block_box)) {
                         m_gait = Gait::WALK;
-                        m_player_pos.x -= move_distance.x;
+                        player_pos.x -= move_distance.x;
                         return;
                     }
                 }
@@ -432,12 +415,12 @@ void Player::update_x_move() {
     }
 }
 
-void Player::update_y_move() {
-    m_player_pos.y += move_distance.y;
+void ClientPlayer::update_y_move(glm::vec3& player_pos) {
+    player_pos.y += move_distance.y;
     if (m_game_mode == SPECTATOR) {
         return;
     }
-    AABB player_box = get_aabb();
+    AABB player_box = get_aabb(player_pos);
     int minx = std::floor(player_box.min.x);
     int maxx = std::floor(player_box.max.x);
     int miny = std::floor(player_box.min.y);
@@ -448,15 +431,11 @@ void Player::update_y_move() {
     for (int x = minx; x <= maxx; ++x) {
         for (int y = miny; y <= maxy; ++y) {
             for (int z = minz; z <= maxz; ++z) {
-                if (!m_world.can_pass_block(glm::ivec3{x, y, z})) {
-                    AABB block_box = {glm::vec3{static_cast<float>(x),
-                                                static_cast<float>(y),
-                                                static_cast<float>(z)},
-                                      glm::vec3{static_cast<float>(x + 1),
-                                                static_cast<float>(y + 1),
-                                                static_cast<float>(z + 1)}};
+                glm::ivec3 block_pos{x, y, z};
+                if (!m_world.can_pass_block(block_pos)) {
+                    AABB block_box = ClientWorld::get_block_aabb(block_pos);
                     if (player_box.intersects(block_box)) {
-                        m_player_pos.y -= move_distance.y;
+                        player_pos.y -= move_distance.y;
                         m_y_speed = 0.0f;
                         if (move_distance.y < 0) {
                             can_up = true;
@@ -470,12 +449,12 @@ void Player::update_y_move() {
     }
 }
 
-void Player::update_z_move() {
-    m_player_pos.z += move_distance.z;
+void ClientPlayer::update_z_move(glm::vec3& player_pos) {
+    player_pos.z += move_distance.z;
     if (m_game_mode == SPECTATOR) {
         return;
     }
-    AABB player_box = get_aabb();
+    AABB player_box = get_aabb(player_pos);
     int minx = std::floor(player_box.min.x);
     int maxx = std::floor(player_box.max.x);
     int miny = std::floor(player_box.min.y);
@@ -486,16 +465,12 @@ void Player::update_z_move() {
     for (int x = minx; x <= maxx; ++x) {
         for (int y = miny; y <= maxy; ++y) {
             for (int z = minz; z <= maxz; ++z) {
-                if (!m_world.can_pass_block(glm::ivec3{x, y, z})) {
-                    AABB block_box = {glm::vec3{static_cast<float>(x),
-                                                static_cast<float>(y),
-                                                static_cast<float>(z)},
-                                      glm::vec3{static_cast<float>(x + 1),
-                                                static_cast<float>(y + 1),
-                                                static_cast<float>(z + 1)}};
+                glm::ivec3 block_pos{x, y, z};
+                if (!m_world.can_pass_block(block_pos)) {
+                    AABB block_box = ClientWorld::get_block_aabb(block_pos);
                     if (player_box.intersects(block_box)) {
                         m_gait = Gait::WALK;
-                        m_player_pos.z -= move_distance.z;
+                        player_pos.z -= move_distance.z;
                         return;
                     }
                 }
@@ -504,7 +479,23 @@ void Player::update_z_move() {
     }
 }
 
-void Player::update_scroll(double yoffset) {
+void ClientPlayer::update_player_chunk() {
+    float x, z;
+    {
+        std::shared_lock lock(m_player_pos_mutex);
+        x = m_player_pos.x;
+        z = m_player_pos.z;
+    }
+    ChunkPos chunk_pos = get_chunk_pos(x, z);
+    float dist = distance2(chunk_pos, m_last_chunk_pos);
+    if (dist > 2) {
+        Logger::info("Player request new chunk");
+        m_world.request_chunk();
+        m_last_chunk_pos = chunk_pos;
+    }
+}
+
+void ClientPlayer::update_scroll(double yoffset) {
     if (m_game_mode == SPECTATOR) {
         if (yoffset > 0) {
             if (m_max_speed < 500.0f) {
@@ -531,15 +522,36 @@ void Player::update_scroll(double yoffset) {
     }
 }
 
-float& Player::max_walk_speed() { return m_max_walk_speed; }
-float& Player::max_run_speed() { return m_max_run_speed; }
-float& Player::max_speed() { return m_max_speed; }
-float& Player::acceleration() { return m_acceleration; }
-float& Player::deceleration() { return m_deceleration; }
-float& Player::g() { return m_g; }
-float& Player::fly_y_speed() { return m_fly_y_speed; }
-unsigned Player::place_block() const { return m_place_block; };
-Gait& Player::gait() { return m_gait; }
-GameMode& Player::game_mode() { return m_game_mode; }
-const World& Player::get_world() const { return m_world; }
-} // namespace Cubed
+void ClientPlayer::update_chunk_set(const ChunkPosSet& set) {
+    std::lock_guard lock(m_chunk_pos_mutex);
+    m_player_chunk_pos_set.clear();
+    m_player_chunk_pos_set.insert(set.begin(), set.end());
+}
+
+const ClientPlayer::ChunkPosSet& ClientPlayer::get_chunk_pos_set() const {
+    std::shared_lock lock(m_chunk_pos_mutex);
+    return m_player_chunk_pos_set;
+}
+
+ClientPlayer::ChunkPosSet& ClientPlayer::get_chunk_pos_set() {
+    std::lock_guard lock(m_chunk_pos_mutex);
+    return m_player_chunk_pos_set;
+}
+
+float& ClientPlayer::max_walk_speed() { return m_max_walk_speed; }
+float& ClientPlayer::max_run_speed() { return m_max_run_speed; }
+float& ClientPlayer::max_speed() { return m_max_speed; }
+float& ClientPlayer::acceleration() { return m_acceleration; }
+float& ClientPlayer::deceleration() { return m_deceleration; }
+float& ClientPlayer::g() { return m_g; }
+float& ClientPlayer::fly_y_speed() { return m_fly_y_speed; }
+unsigned ClientPlayer::place_block() const { return m_place_block; };
+Gait& ClientPlayer::gait() { return m_gait; }
+GameMode& ClientPlayer::game_mode() { return m_game_mode; }
+const ClientWorld& ClientPlayer::get_world() const { return m_world; }
+
+void ClientPlayer::set_uuid(std::string_view uuid) { m_uuid = uuid; }
+const std::string& ClientPlayer::get_uuid() const { return m_uuid; }
+const std::string& ClientPlayer::get_name() const { return m_name; }
+void ClientPlayer::init(std::string_view name) { m_name = name; }
+} // namespace Cubed

src/gameplay/client_world.cpp

@@ -0,0 +1,746 @@
+#include "Cubed/gameplay/client_world.hpp"
+
+#include "Cubed/config.hpp"
+#include "Cubed/gameplay/game_time.hpp"
+#include "Cubed/gameplay/packet.hpp"
+#include "Cubed/tools/math_tools.hpp"
+
+#include <absl/container/inlined_vector.h>
+#include <numbers>
+
+using namespace std::chrono;
+using namespace std::chrono_literals;
+using namespace google::protobuf;
+namespace Cubed {
+
+namespace {
+struct ChunkRenderData {
+    std::array<const std::vector<BlockType>*, 4> neighbor_block;
+    ClientChunk* chunk;
+};
+} // namespace
+
+ClientWorld::ClientWorld() : m_player(*this) {}
+
+ClientWorld::~ClientWorld() {
+    stop_client_thread();
+    stop_thread_pool();
+
+    m_chunks.clear();
+
+    {
+        std::lock_guard lk(m_delete_vbo_mutex);
+        for (auto x : m_pending_delete_vbo) {
+            glDeleteBuffers(1, &x);
+        }
+        m_pending_delete_vbo.clear();
+    }
+    {
+        std::lock_guard lk(m_delete_vao_mutex);
+        for (auto x : m_pending_delete_vao) {
+            glDeleteVertexArrays(1, &x);
+        }
+        m_pending_delete_vao.clear();
+    }
+    m_timers.clear();
+}
+
+const std::optional<LookBlock>& ClientWorld::get_look_block_pos() const {
+
+    return m_player.get_look_block_pos();
+}
+
+ClientPlayer& ClientWorld::get_player() { return m_player; }
+
+int ClientWorld::get_block(const glm::ivec3& block_pos) const {
+    auto [chunk_x, chunk_z] = get_chunk_pos(block_pos.x, block_pos.z);
+    chunk_cacc cacc;
+
+    if (!m_chunks.find(cacc, ChunkPos{chunk_x, chunk_z})) {
+        return 0;
+    }
+
+    const auto& chunk_blocks = cacc->second->get_chunk_blocks();
+    auto [x, y, z] = ClientChunk::world_to_block(block_pos, {chunk_x, chunk_z});
+    if (x < 0 || y < 0 || z < 0 || x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
+        z >= CHUNK_SIZE) {
+        return 0;
+    }
+    return chunk_blocks[ClientChunk::index(x, y, z)];
+}
+bool ClientWorld::is_solid(const glm::ivec3& block_pos) const {
+    auto [chunk_x, chunk_z] = get_chunk_pos(block_pos.x, block_pos.z);
+    chunk_cacc cacc;
+
+    if (!m_chunks.find(cacc, ChunkPos{chunk_x, chunk_z})) {
+        return false;
+    }
+    const auto& chunk_blocks = cacc->second->get_chunk_blocks();
+    auto [x, y, z] = ClientChunk::world_to_block(block_pos, {chunk_x, chunk_z});
+    if (x < 0 || y < 0 || z < 0 || x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
+        z >= CHUNK_SIZE) {
+        return false;
+    }
+    auto id = chunk_blocks[ClientChunk::index(x, y, z)];
+    if (BlockManager::is_gas(id) || BlockManager::is_liquid(id)) {
+        return false;
+    } else {
+        return true;
+    }
+}
+bool ClientWorld::can_pass_block(const glm::ivec3& block_pos) const {
+    auto [chunk_x, chunk_z] = get_chunk_pos(block_pos.x, block_pos.z);
+    chunk_cacc cacc;
+
+    if (!m_chunks.find(cacc, ChunkPos{chunk_x, chunk_z})) {
+        return true;
+    }
+    const auto& chunk_blocks = cacc->second->get_chunk_blocks();
+    auto [x, y, z] = ClientChunk::world_to_block(block_pos, {chunk_x, chunk_z});
+    if (x < 0 || y < 0 || z < 0 || x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
+        z >= CHUNK_SIZE) {
+        return true;
+    }
+    auto id = chunk_blocks[ClientChunk::index(x, y, z)];
+    return BlockManager::is_passable(id);
+}
+
+void ClientWorld::rebuild_world() {
+    if (m_is_rebuilding.exchange(true)) {
+        return;
+    }
+
+    stop_client_thread();
+    stop_thread_pool();
+
+    m_chunks.clear();
+
+    m_pending_upload_queue.clear();
+
+    start_thread_pool();
+    start_client_thread(m_player.get_uuid());
+    request_chunk();
+    m_is_rebuilding = false;
+}
+
+BlockType ClientWorld::get_block_tpye(const glm::ivec3& block_pos) const {
+    auto [chunk_x, chunk_z] = get_chunk_pos(block_pos.x, block_pos.z);
+    chunk_cacc cacc;
+    ;
+
+    if (!m_chunks.find(cacc, ChunkPos{chunk_x, chunk_z})) {
+        // Logger::error("Can't Find Block {} {} {}", block_pos.x, block_pos.y,
+        //               block_pos.z);
+        return 0;
+    }
+    const auto& chunk_blocks = cacc->second->get_chunk_blocks();
+    auto [x, y, z] = ClientChunk::world_to_block(block_pos, {chunk_x, chunk_z});
+    if (x < 0 || y < 0 || z < 0 || x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
+        z >= CHUNK_SIZE) {
+        // Logger::error("Can't Find Block {} {} {}", block_pos.x, block_pos.y,
+        //               block_pos.z);
+        return 0;
+    }
+    return chunk_blocks[ClientChunk::index(x, y, z)];
+}
+void ClientWorld::set_block(const glm::ivec3& block_pos, unsigned id) {
+    int world_x, world_y, world_z;
+    world_x = block_pos.x;
+    world_y = block_pos.y;
+    world_z = block_pos.z;
+
+    auto [chunk_x, chunk_z] = get_chunk_pos(world_x, world_z);
+    ChunkPos pos{chunk_x, chunk_z};
+    {
+        chunk_acc acc;
+
+        if (!m_chunks.find(acc, pos)) {
+            return;
+        }
+
+        auto [x, y, z] = ClientChunk::world_to_block(world_x, world_y, world_z,
+                                                     chunk_x, chunk_z);
+        if (x < 0 || y < 0 || z < 0 || x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
+            z >= CHUNK_SIZE) {
+            return;
+        }
+
+        acc->second->set_chunk_block(ClientChunk::index(x, y, z), id);
+        acc->second->mark_dirty();
+    }
+
+    auto pool = m_thread_pool.load();
+
+    pool->enqueue(0, [this, pos]() {
+        std::shared_ptr<ClientChunk> chunk;
+
+        {
+            chunk_acc acc;
+            if (m_chunks.find(acc, pos)) {
+                chunk = acc->second;
+            }
+        }
+
+        if (!chunk) {
+            return;
+        }
+
+        OptionalBlockVectorArray neighbor_block;
+        for (int i = 0; i < 4; i++) {
+            chunk_cacc cacc;
+            if (m_chunks.find(cacc, pos + CHUNK_DIR[i])) {
+                neighbor_block[i] = (cacc->second->get_chunk_blocks());
+            } else {
+                neighbor_block[i] = std::nullopt;
+            }
+        }
+
+        chunk->gen_vertex_data(neighbor_block);
+        m_dirty_chunk_queue.emplace(pos);
+    });
+
+    static const glm::ivec3 NEIGHBOR_DIRS[] = {
+        {1, 0, 0}, {-1, 0, 0}, {0, 0, -1}, {0, 0, 1}};
+    static constexpr int NPOS_SUM = sizeof(NEIGHBOR_DIRS);
+
+    absl::InlinedVector<ChunkPos, NPOS_SUM> nposes;
+
+    for (const auto& dir : NEIGHBOR_DIRS) {
+        glm::ivec3 neighbor = block_pos + dir;
+
+        auto [cx, cz] = get_chunk_pos(neighbor.x, neighbor.z);
+        {
+            chunk_acc acc;
+            if (m_chunks.find(acc, {cx, cz})) {
+                if (acc->second->is_dirty()) {
+                    continue;
+                }
+                nposes.emplace_back(acc->first);
+            }
+        }
+    }
+
+    for (auto& npos : nposes) {
+        pool->enqueue(0, [this, npos]() {
+            std::shared_ptr<ClientChunk> chunk;
+
+            {
+                chunk_acc acc;
+                if (m_chunks.find(acc, npos)) {
+                    chunk = acc->second;
+                }
+            }
+
+            if (!chunk) {
+                return;
+            }
+
+            OptionalBlockVectorArray neighbor_block;
+            for (int i = 0; i < 4; i++) {
+                chunk_cacc cacc;
+                if (m_chunks.find(cacc, npos + CHUNK_DIR[i])) {
+                    neighbor_block[i] = (cacc->second->get_chunk_blocks());
+                } else {
+                    neighbor_block[i] = std::nullopt;
+                }
+            }
+
+            chunk->gen_vertex_data(neighbor_block);
+
+            m_dirty_chunk_queue.emplace(npos);
+        });
+    }
+}
+void ClientWorld::push_delete_vbo(GLuint vbo) {
+    std::lock_guard lk(m_delete_vbo_mutex);
+    m_pending_delete_vbo.push_back(vbo);
+}
+void ClientWorld::push_delete_vao(GLuint vao) {
+    std::lock_guard lk(m_delete_vao_mutex);
+    m_pending_delete_vao.push_back(vao);
+}
+
+void ClientWorld::report_block_change(const glm::ivec3& pos,
+                                      unsigned id) const {
+    {
+        AABB block_box = get_block_aabb(pos);
+        std::shared_lock lock(m_other_players_mutex);
+
+        for (auto& [uuid, player] : m_other_players) {
+            AABB box = ClientPlayer::get_aabb(player.target_pos);
+            if (box.intersects(block_box)) {
+                return;
+            }
+        }
+    }
+
+    Arena arena;
+    auto* req = Arena::Create<BlockChangeReq>(&arena);
+    req->set_uuid(m_player.get_uuid());
+    req->set_block(id);
+    auto* p = req->mutable_pos();
+    p->set_x(pos.x);
+    p->set_y(pos.y);
+    p->set_z(pos.z);
+    m_client->send(make_packet(*req), 0);
+}
+
+void ClientWorld::receive_block_change(const BlockChangeRsp& rsp) {
+    glm::vec3 pos{rsp.pos().x(), rsp.pos().y(), rsp.pos().z()};
+    set_block(pos, rsp.block());
+}
+
+void ClientWorld::receive_time(const UpdateTime& rsp) {
+    m_game_ticks = rsp.game_tick();
+    m_day_tick = rsp.day_tick();
+}
+
+void ClientWorld::receive_remote_player(const PlayerInfoRsp& rsp) {
+    {
+        std::lock_guard lock(m_other_players_mutex);
+        glm::vec3 pos{rsp.pos().x(), rsp.pos().y(), rsp.pos().z()};
+        auto it = m_other_players.find(rsp.uuid());
+        if (it == m_other_players.end()) {
+            m_other_players.emplace(
+                std::piecewise_construct, std::forward_as_tuple(rsp.uuid()),
+                std::forward_as_tuple(rsp.name(), pos, pos));
+        } else {
+            it->second.target_pos = pos;
+        }
+        // Logger::info("Player {} pos Update", rsp.name());
+    }
+}
+
+void ClientWorld::receive_player_logout(const LogoutRsp& rsp) {
+    if (rsp.server_stop()) {
+        m_receive_exit = true;
+        return;
+    }
+    if (rsp.uuid() == m_player.get_uuid()) {
+        m_receive_exit = true;
+        return;
+    }
+    {
+        std::lock_guard lock(m_other_players_mutex);
+        int sum = m_other_players.erase(rsp.uuid());
+        if (sum == 0) {
+            Logger::warn("Player {} not find", rsp.uuid());
+        } else {
+            Logger::info("Player {} erase", rsp.uuid());
+        }
+    }
+}
+
+void ClientWorld::init(std::string_view player_name,
+                       std::shared_ptr<NetworkClient> client) {
+    m_player.init(player_name);
+    m_client = client;
+    // timer
+    register_timer("player_pos", 1, [this]() { report_player_pos(); });
+    LoginReq req;
+    req.set_name(m_player.get_name());
+    while (!client->is_connected()) {
+        if (client->is_connect_error()) {
+            throw std::runtime_error("Can't connect to the server");
+        }
+        std::this_thread::sleep_for(milliseconds(200));
+    }
+    start_thread_pool();
+    // request login
+    Logger::info("Send Login Request");
+    m_client->send(make_packet(req), 0);
+}
+
+void ClientWorld::start_client_thread(std::string_view uuid) {
+    if (m_game_running) {
+        Logger::error("Game Already Running");
+        return;
+    }
+    // response
+    m_player.set_uuid(uuid);
+    m_client_thread = std::jthread([this](std::stop_token token) {
+        m_game_running = true;
+        client_run(token);
+    });
+
+    // Wait for 20 ticks, after the server's central chunk is generated, then
+    // request chunks
+
+    std::this_thread::sleep_for(milliseconds(20 * DEFAULT_PER_TICK_TIME));
+
+    request_chunk();
+}
+
+void ClientWorld::stop_client_thread() {
+    m_client_thread.request_stop();
+    if (m_client_thread.joinable()) {
+        m_client_thread.join();
+    }
+    m_game_running = false;
+}
+void ClientWorld::start_thread_pool() {
+    int max_threads = std::thread::hardware_concurrency();
+    int threads = std::min<size_t>(max_threads, 4);
+    change_pool_threads(threads);
+}
+void ClientWorld::stop_thread_pool() {
+    auto pool_ptr = m_thread_pool.load();
+    if (pool_ptr) {
+        pool_ptr->stop();
+    }
+    m_thread_pool.store(nullptr);
+    Logger::info("Thread Pool Stopped");
+}
+
+void ClientWorld::change_pool_threads(int threads) {
+    int m_max_threads = std::thread::hardware_concurrency();
+    if (m_max_threads < 1) {
+        Logger::warn("Can't Get Max Support Threads, Set Max Threads to 4");
+        m_max_threads = 1;
+    }
+    int used_thread = std::clamp(threads, 1, m_max_threads);
+    Logger::info("Create New Thread Pool Use {} Threads", used_thread);
+    m_thread_pool.store(std::make_shared<PriorityThreadPool>(used_thread));
+}
+
+void ClientWorld::hot_reload() {
+    auto& config = Config::get();
+    int dist = config.get<int>("world.rendering_distance");
+    Logger::info("Get Config Randering dist {}", dist);
+    m_rendering_distance = dist <= MAX_DISTANCE ? dist : MAX_DISTANCE;
+    request_chunk();
+}
+
+void ClientWorld::client_run(std::stop_token stoken) {
+    Logger::info("Client Thread Started");
+    using Clock = std::chrono::steady_clock;
+
+    constexpr auto TICK = std::chrono::milliseconds(DEFAULT_PER_TICK_TIME);
+
+    auto next = Clock::now();
+    while (!stoken.stop_requested()) {
+        next += TICK;
+        for (auto& x : m_timers) {
+            x.second.update();
+        }
+        std::this_thread::sleep_until(next);
+    }
+}
+
+void ClientWorld::report_player_pos() {
+    if (!m_client) {
+        return;
+    }
+    Arena arena;
+    auto* pos = Arena::Create<PlayerPos>(&arena);
+    pos->set_uuid(m_player.get_uuid());
+    glm::vec3 player_pos = m_player.get_player_pos();
+    auto* v3 = pos->mutable_pos();
+    v3->set_x(player_pos.x);
+    v3->set_y(player_pos.y);
+    v3->set_z(player_pos.z);
+    m_client->send(make_packet(*pos), 0);
+}
+
+void ClientWorld::update_chunk(const ChunkPosSet& old, const ChunkPosSet& now) {
+
+    // Elements in the old set that are not contained in now are not needed by
+    // the current player.
+
+    for (auto& pos : old) {
+        if (!now.contains(pos)) {
+
+            chunk_acc acc;
+            if (!m_chunks.find(acc, pos)) {
+                Logger::warn("Update Ref Count Error, can't Find old pos "
+                             "in m_chunks");
+                continue;
+            }
+
+            m_chunks.erase(acc);
+        }
+    }
+}
+
+void ClientWorld::request_chunk() {
+    if (m_requesting_chunk.exchange(true)) {
+        Logger::warn("It is requesting new chunk!");
+        return;
+    }
+    ChunkPosSet required_chunks;
+
+    glm::vec3 player_pos = m_player.get_player_pos();
+
+    int x = std::floor(player_pos.x);
+    int z = std::floor(player_pos.z);
+    auto [chunk_x, chunk_z] = get_chunk_pos(x, z);
+    int radius = m_rendering_distance;
+    Logger::info("Client Chunk Radius {}", radius);
+    int r2 = radius * radius;
+    required_chunks.reserve(radius * radius);
+
+    for (int dx = -radius; dx <= radius; ++dx) {
+        for (int dz = -radius; dz <= radius; ++dz) {
+            if (dx * dx + dz * dz <= r2) {
+                required_chunks.emplace(chunk_x + dx, chunk_z + dz);
+            }
+        }
+    }
+
+    ChunkPosSet old = std::move(m_player.get_chunk_pos_set());
+    m_player.update_chunk_set(required_chunks);
+
+    ChunkPosVector need_send_pos;
+
+    for (auto pos : required_chunks) {
+        chunk_cacc cacc;
+        if (!m_chunks.find(cacc, pos)) {
+            need_send_pos.emplace_back(pos);
+        }
+    }
+
+    update_chunk(old, required_chunks);
+
+    if (need_send_pos.empty()) {
+        m_requesting_chunk = false;
+        return;
+    }
+    using enum ChunkLoadStyle;
+    switch (m_chunk_load_style) {
+    case RANDOM:
+
+        break;
+    case CENTER: {
+
+        glm::vec3 player_pos = m_player.get_player_pos();
+        ChunkPos player_chunk_pos = get_chunk_pos(player_pos.x, player_pos.z);
+        auto dist2 = [player_chunk_pos](ChunkPos chunk_pos) {
+            float dx = player_chunk_pos.x - chunk_pos.x;
+            float dz = player_chunk_pos.z - chunk_pos.z;
+            return dx * dx + dz * dz;
+        };
+
+        std::sort(need_send_pos.begin(), need_send_pos.end(),
+                  [&dist2](const auto& a, const auto& b) {
+                      return dist2(a) < dist2(b);
+                  });
+    }
+    }
+    auto uuid = m_player.get_uuid();
+    Arena arena;
+    ++m_chunk_task_id;
+    auto* req = Arena::Create<ChunkDataReq>(&arena);
+    for (const auto& pos : need_send_pos) {
+        req->set_task_id(m_chunk_task_id.load());
+        req->set_uuid(uuid);
+        auto* p = req->mutable_pos();
+        p->set_x(pos.x);
+        p->set_z(pos.z);
+        m_client->send(make_packet(*req));
+    }
+    Logger::info("Send Chunk Request Success");
+    m_requesting_chunk = false;
+}
+
+void ClientWorld::receive_chunk(std::vector<uint8_t> raw_data,
+                                PacketHeader header) {
+
+    // vertex data will genrator in  client thread pool instead of net thread;
+    auto pool = m_thread_pool.load();
+    if (!pool) {
+        Logger::error("Client Thread Pool is nullptr");
+        return;
+    }
+    pool->enqueue(
+        [this, raw_data = std::move(raw_data), header = std::move(header)]() {
+            Arena arena;
+            auto* data = Arena::Create<ChunkDataRsp>(&arena);
+            if (!decode_packet(*data, raw_data, header)) {
+                return;
+            }
+
+            if (data->task_id() < m_chunk_task_id) {
+                return;
+            }
+
+            {
+                chunk_cacc cacc;
+                ChunkPos pos{data->pos().x(), data->pos().z()};
+                if (m_chunks.find(cacc, pos)) {
+                    Logger::warn("Chunk {} {} has already in client world",
+                                 pos.x, pos.z);
+                    return;
+                }
+            }
+
+            std::unique_ptr<ClientChunk> chunk =
+                std::make_unique<ClientChunk>(*this);
+            chunk->receive_chunk(*data);
+
+            m_pending_upload_queue.emplace(std::move(chunk));
+        });
+}
+bool ClientWorld::is_receive_exit() { return m_receive_exit; }
+
+int ClientWorld::chunk_size() const { return m_chunks.size(); }
+
+AABB ClientWorld::get_block_aabb(const glm::ivec3& pos) {
+    auto x = pos.x;
+    auto y = pos.y;
+    auto z = pos.z;
+    return {glm::vec3{static_cast<float>(x), static_cast<float>(y),
+                      static_cast<float>(z)},
+            glm::vec3{static_cast<float>(x + 1), static_cast<float>(y + 1),
+                      static_cast<float>(z + 1)}};
+}
+
+void ClientWorld::request_exit() {
+    if (m_receive_exit) {
+        return;
+    }
+    Arena arena;
+    auto* req = Arena::Create<LogoutReq>(&arena);
+    req->set_uuid(m_player.get_uuid());
+    m_client->send(make_packet(*req));
+    int cnt = 0;
+    while (!m_receive_exit) {
+        std::this_thread::sleep_for(milliseconds(DEFAULT_PER_TICK_TIME));
+        ++cnt;
+        if (cnt >= WORLD_EXIT_TIMEOUT) {
+            Logger::warn("Can't Receive Server Exit Sign");
+            break;
+        }
+    }
+}
+
+void ClientWorld::update(float delta_time) {
+    m_player.update(delta_time);
+    {
+        std::lock_guard lk(m_delete_vbo_mutex);
+        for (auto x : m_pending_delete_vbo) {
+            glDeleteBuffers(1, &x);
+        }
+        m_pending_delete_vbo.clear();
+    }
+
+    {
+        std::lock_guard lk(m_delete_vao_mutex);
+        for (auto x : m_pending_delete_vao) {
+            glDeleteVertexArrays(1, &x);
+        }
+        m_pending_delete_vao.clear();
+    }
+
+    std::vector<std::unique_ptr<ClientChunk>> new_chunks;
+    {
+        std::unique_ptr<ClientChunk> chunk;
+        int sum = 0;
+        while (m_pending_upload_queue.try_pop(chunk)) {
+            new_chunks.emplace_back(std::move(chunk));
+            ++sum;
+            if (sum >= MAX_UPLOAD_CHUNK_SUM) {
+                break; // Limit the maximum number of uploads per frame to
+                       // improve frame rate performance
+            }
+        }
+    }
+
+    for (auto& c : new_chunks) {
+        c->upload_to_gpu();
+    }
+
+    for (auto& c : new_chunks) {
+        m_chunks.emplace(c->get_chunk_pos(), std::move(c));
+    }
+    m_render_snapshots.clear();
+
+    ChunkPos pos;
+
+    while (m_dirty_chunk_queue.try_pop(pos)) {
+        std::shared_ptr<ClientChunk> chunk;
+        {
+            chunk_acc acc;
+            if (m_chunks.find(acc, pos)) {
+                chunk = acc->second;
+            }
+        }
+        if (!chunk) {
+            continue;
+        }
+
+        chunk->upload_to_gpu();
+    }
+
+    auto chunk_pos_set = m_player.get_chunk_pos_set();
+
+    for (auto& pos : chunk_pos_set) {
+        std::shared_ptr<ClientChunk> chunk;
+        {
+            chunk_acc acc;
+            if (m_chunks.find(acc, pos)) {
+                chunk = acc->second;
+            }
+        }
+        if (!chunk) {
+            continue;
+        }
+
+        m_render_snapshots.push_back(chunk->get_render_snapshot());
+    }
+
+    m_render_player_data.clear();
+    {
+        std::lock_guard lock(m_other_players_mutex);
+        for (auto& [uuid, player] : m_other_players) {
+            player.render_pos =
+                glm::mix(player.render_pos, player.target_pos, 0.15f);
+            if (Math::distance2(player.render_pos, m_player.get_player_pos()) >
+                m_rendering_distance * CHUNK_SIZE * m_rendering_distance *
+                    CHUNK_SIZE) {
+                continue;
+            }
+            m_render_player_data.emplace_back(player.name, player.render_pos);
+        }
+    }
+}
+
+glm::vec3 ClientWorld::sunlight_dir() const {
+    float altitude = sin((m_day_tick - 6 * PER_HOUR) /
+                         static_cast<float>(DAY_TIME / 2) * std::numbers::pi) *
+                     90.0f;
+
+    float t = static_cast<float>(m_day_tick) / DAY_TIME;
+    float azimuth = 90.0f - 360.0f * (t - 0.25f);
+
+    float alt = glm::radians(altitude);
+    float az = glm::radians(azimuth);
+    glm::vec3 dir;
+    dir.x = cos(alt) * sin(az);
+    dir.y = sin(alt);
+    dir.z = cos(alt) * cos(az);
+
+    return glm::normalize(-dir);
+}
+int ClientWorld::rendering_distance() const {
+    return m_rendering_distance.load();
+}
+
+void ClientWorld::rendering_distance(int rendering_distance) {
+    m_rendering_distance = rendering_distance;
+    Logger::info("Set Rendering dist {} , the value is {}", rendering_distance,
+                 m_rendering_distance.load());
+    request_chunk();
+}
+
+int ClientWorld::get_chunk_task_id() const { return m_chunk_task_id.load(); }
+
+const std::vector<const ChunkRenderSnapshot*>&
+ClientWorld::render_snapshots() const {
+    return m_render_snapshots;
+};
+const std::vector<RemotePlayerRenderData>&
+ClientWorld::render_player_data() const {
+    return m_render_player_data;
+}
+std::vector<glm::vec4>& ClientWorld::planes() { return m_planes; }
+} // namespace Cubed

src/gameplay/network_client.cpp

@@ -0,0 +1,205 @@
+#include "Cubed/gameplay/network_client.hpp"
+
+#include "Cubed/gameplay/client_world.hpp"
+#include "Cubed/tools/log.hpp"
+
+#include <utility>
+
+using namespace google::protobuf;
+namespace Cubed {
+NetworkClient::NetworkClient(ClientWorld& world)
+    : m_socket(m_io), m_strand(asio::make_strand(m_io)), m_world(world) {}
+
+NetworkClient::~NetworkClient() { close(); }
+
+void NetworkClient::start(std::string ip, int port) {
+    if (m_net_thread.joinable()) {
+        return;
+    }
+    m_net_thread = std::thread([self = shared_from_this(), ip, port]() {
+        asio::co_spawn(self->m_strand, self->connect(ip, port), asio::detached);
+        self->m_io.run();
+    });
+    Logger::info("NetworkClient Started");
+}
+
+bool NetworkClient::is_connected() const { return m_connected.load(); }
+bool NetworkClient::is_connect_error() const { return m_connect_error.load(); }
+asio::awaitable<void> NetworkClient::connect(std::string ip, int port) {
+    Logger::info("Connect Begin");
+    try {
+        auto ex = co_await asio::this_coro::executor;
+        tcp::resolver resolver(ex);
+        auto eps = co_await resolver.async_resolve(ip, std::to_string(port),
+                                                   asio::use_awaitable);
+        Logger::info("Resolve Success");
+        co_await async_connect(m_socket, eps, asio::use_awaitable);
+        Logger::info("Connect Success, Server ip {} port {}", ip, port);
+        asio::co_spawn(m_strand, read_loop(), asio::detached);
+        Logger::info("NetworkClient Read Loop Started");
+        m_connected = true;
+        co_return;
+
+    } catch (const std::exception& e) {
+        Logger::error("Client Error {}", e.what());
+        m_connect_error = true;
+    }
+}
+
+asio::awaitable<void> NetworkClient::read_loop() {
+    try {
+        while (true) {
+            std::array<uint8_t, HEADER_LEN> header_buffer;
+            co_await asio::async_read(m_socket, asio::buffer(header_buffer),
+                                      asio::use_awaitable);
+            auto header = decode_packet_header(header_buffer);
+            uint32_t total_len = HEADER_LEN + header.compressed_size;
+
+            if (total_len < HEADER_LEN || total_len > MAX_PACKET_SIZE) {
+
+                throw std::runtime_error("invalid packet length");
+            }
+            // maybe move, don't use it after switch!
+            std::vector<uint8_t> body_data(header.compressed_size);
+            if (header.compressed_size > 0) {
+                co_await asio::async_read(m_socket, asio::buffer(body_data),
+                                          asio::use_awaitable);
+            }
+
+            using std::to_underlying;
+            Arena arena;
+            switch (header.cmd) {
+            case std::to_underlying(PacketEnum::LOGIN_RSP): {
+                auto* rsp = Arena::Create<LoginRsp>(&arena);
+                Logger::info("Client: Receive Login rsp");
+                if (decode_packet(*rsp, body_data, header)) {
+                    if (rsp->success()) {
+                        m_world.start_client_thread(rsp->uuid());
+                    } else {
+                        Logger::error("Connected Server Fail");
+                    }
+                }
+            } break;
+            case std::to_underlying(PacketEnum::CHUNK_DATA_RSP): {
+                // Logger::info("Client: Receive Chunk Data rsp, size {}mb",
+                //              body_data.size() / 1024.0f / 1024);
+                m_world.receive_chunk(std::move(body_data), header);
+            } break;
+            case std::to_underlying(PacketEnum::BLOCK_CHANGE_RSP): {
+                auto* rsp = Arena::Create<BlockChangeRsp>(&arena);
+                Logger::info("Client: Receive Block Change rsp");
+                if (decode_packet(*rsp, body_data, header)) {
+                    m_world.receive_block_change(*rsp);
+                }
+            } break;
+            case std::to_underlying(PacketEnum::UPDATE_TIME): {
+                auto* rsp = Arena::Create<UpdateTime>(&arena);
+                if (decode_packet(*rsp, body_data, header)) {
+                    m_world.receive_time(*rsp);
+                }
+            } break;
+            case std::to_underlying(PacketEnum::PLAYER_INFO_RSP): {
+                auto* rsp = Arena::Create<PlayerInfoRsp>(&arena);
+                if (decode_packet(*rsp, body_data, header)) {
+                    m_world.receive_remote_player(*rsp);
+                }
+            } break;
+            case std::to_underlying(PacketEnum::LOGOUT_RSP): {
+                auto* rsp = Arena::Create<LogoutRsp>(&arena);
+                if (decode_packet(*rsp, body_data, header)) {
+                    m_world.receive_player_logout(*rsp);
+                }
+            } break;
+            case std::to_underlying(PacketEnum::S2C_CLEAR_ALL_CHUNKS): {
+                auto* rsp = Arena::Create<S2C_ClearAllChunks>(&arena);
+                if (decode_packet(*rsp, body_data, header)) {
+                    if (rsp->clear()) {
+                        Logger::info("Client Clear All Chunk");
+                        m_world.rebuild_world();
+                    }
+                }
+            } break;
+            }
+        }
+    } catch (const asio::system_error& e) {
+        auto ec = e.code();
+
+        if (ec == asio::error::eof || ec == asio::error::operation_aborted) {
+
+            Logger::info("Client disconnected");
+        } else {
+            Logger::warn("Asio Error {}", e.what());
+        }
+
+        close();
+    } catch (const std::exception& e) {
+        Logger::error("Session Error {}", e.what());
+        close();
+    } catch (...) {
+        Logger::error("Unknow Error");
+        close();
+    }
+    co_return;
+}
+
+void NetworkClient::send(Packet packet, int priority) {
+    if (m_closed.load()) {
+        return;
+    }
+    asio::post(m_strand, [self = shared_from_this(), packet = std::move(packet),
+                          priority]() mutable {
+        bool idle = self->m_write_queue.empty();
+        self->m_write_queue.emplace(priority, self->m_sequence++,
+                                    std::move(packet));
+        if (idle) {
+            self->do_write();
+        }
+    });
+}
+
+void NetworkClient::do_write() {
+    if (m_closed.load()) {
+        return;
+    }
+
+    auto self = shared_from_this();
+    auto packet = std::move(m_write_queue.top().packet);
+    asio::async_write(
+        m_socket, asio::buffer(*packet),
+        asio::bind_executor(m_strand, [self](std::error_code ec, size_t) {
+            if (ec) {
+                Logger::warn("Write Ec {}", ec.message());
+                self->close();
+                return;
+            }
+            self->m_write_queue.pop();
+            if (!self->m_write_queue.empty()) {
+                self->do_write();
+            }
+        }));
+}
+
+void NetworkClient::close() {
+    if (m_closed.exchange(true)) {
+        return;
+    }
+
+    std::error_code ec;
+
+    m_socket.shutdown(tcp::socket::shutdown_both, ec);
+
+    m_socket.close(ec);
+    Logger::info("NetworkClient Closed");
+    m_connected = false;
+    m_io.stop();
+}
+
+void NetworkClient::stop() {
+    close();
+
+    if (m_net_thread.joinable()) {
+        m_net_thread.join();
+    }
+}
+
+} // namespace Cubed

src/gameplay/network_server.cpp

@@ -0,0 +1,94 @@
+#include "Cubed/gameplay/network_server.hpp"
+
+#include "Cubed/tools/log.hpp"
+using asio::ip::tcp;
+namespace Cubed {
+
+NetworkServer::NetworkServer(int port) : m_port(port) {}
+
+NetworkServer::~NetworkServer() { stop(); }
+
+void NetworkServer::stop() {
+    if (!m_started) {
+        return;
+    }
+    if (m_stopped.exchange(true)) {
+        return;
+    }
+
+    m_io.stop();
+
+    std::vector<std::shared_ptr<Session>> sessions;
+
+    {
+        std::lock_guard lock(m_session_mutex);
+
+        for (auto& [id, s] : m_session) {
+            sessions.push_back(s);
+        }
+
+        m_session.clear();
+    }
+
+    for (auto& s : sessions) {
+        s->close();
+    }
+
+    if (m_net_thread.joinable()) {
+        Logger::info("Server join thread={}, current={}", m_net_thread.get_id(),
+                     std::this_thread::get_id());
+        m_net_thread.join();
+    }
+    Logger::info("Server Net Thread Stopped!");
+}
+
+asio::awaitable<void> NetworkServer::listen() {
+
+    try {
+        tcp::acceptor acceptor(m_io, tcp::endpoint(tcp::v4(), m_port));
+        while (!m_stopped) {
+            tcp::socket socket =
+                co_await acceptor.async_accept(asio::use_awaitable);
+
+            std::shared_ptr<Session> s =
+                std::make_shared<Session>(std::move(socket), m_world, m_io);
+            {
+                std::lock_guard lock(m_session_mutex);
+                m_session.emplace(s->uuid(), s);
+            }
+            s->start();
+        }
+    } catch (const std::exception& e) {
+        if (!m_stopped) {
+            Logger::error("accept error {}", e.what());
+        }
+    } catch (...) {
+        if (!m_stopped) {
+            Logger::error("Network Server: Unknown Error");
+        }
+    }
+
+    co_return;
+}
+
+void NetworkServer::net_run() {
+    if (m_net_thread.joinable()) {
+        return;
+    }
+    m_net_thread = std::thread([this]() {
+        asio::co_spawn(m_io, listen(), asio::detached);
+        m_io.run();
+    });
+    Logger::info("Server Started!");
+}
+
+void NetworkServer::start_server(int port) {
+    m_port = port;
+    m_world.init_world();
+    net_run();
+    m_started = true;
+}
+
+int NetworkServer::port() const { return m_port; }
+ServerWorld& NetworkServer::server_world() { return m_world; }
+} // namespace Cubed

src/gameplay/server_chunk.cpp

@@ -0,0 +1,211 @@
+#include "Cubed/gameplay/server_chunk.hpp"
+
+#include "Cubed/tools/cubed_assert.hpp"
+
+namespace Cubed {
+ServerChunk::ServerChunk(ServerWorld& world, ChunkPos chunk_pos,
+                         bool temp_chunk)
+    : m_temp_chunk(temp_chunk), m_chunk_pos(chunk_pos), m_world(world) {}
+
+ServerChunk::ServerChunk(ServerChunk&& other) noexcept
+    : m_gening(other.m_gening.load()), m_has_cave(other.m_has_cave),
+      m_biome(other.m_biome.load()), m_chunk_pos(std::move(other.m_chunk_pos)),
+      m_world(other.m_world), m_heightmap(std::move(other.m_heightmap)),
+      m_blocks(std::move(other.m_blocks)),
+      m_neightbor_blocks(std::move(other.m_neightbor_blocks)),
+      m_seed(other.m_seed), m_conditions(other.m_conditions) {
+    ASSERT_MSG(!other.m_gening, "Other is Gening Can't Move");
+}
+
+ServerChunk& ServerChunk::operator=(ServerChunk&& other) noexcept {
+    // Logger::info("other Chunk pos {} {} in Chunk& Chunk::operator=(Chunk&&
+    // other) this {}", other.m_chunk_pos.x, other.m_chunk_pos.z,
+    // static_cast<const void*>(&other));
+    if (this == &other) {
+        return *this;
+    }
+    ASSERT_MSG(!other.m_gening, "Other is Gening Can't Move");
+    m_chunk_pos = std::move(other.m_chunk_pos);
+    m_heightmap = std::move(other.m_heightmap);
+    m_blocks = std::move(other.m_blocks);
+    m_biome = other.m_biome.load();
+    m_seed = other.m_seed;
+    m_conditions = other.m_conditions;
+    m_neightbor_blocks = std::move(other.m_neightbor_blocks);
+    m_has_cave = other.m_has_cave;
+    m_gening = other.m_gening.load();
+    return *this;
+}
+
+std::tuple<int, int, int> ServerChunk::world_to_block(int world_x, int world_y,
+                                                      int world_z, int chunk_x,
+                                                      int chunk_z) {
+    int x, y, z;
+    y = world_y;
+    x = world_x - chunk_x * CHUNK_SIZE;
+    z = world_z - chunk_z * CHUNK_SIZE;
+    return {x, y, z};
+}
+
+std::tuple<int, int, int>
+ServerChunk::world_to_block(const glm::ivec3& block_pos, ChunkPos chunk_pos) {
+    return world_to_block(block_pos.x, block_pos.y, block_pos.z, chunk_pos.x,
+                          chunk_pos.z);
+}
+
+std::tuple<int, int, int>
+ServerChunk::block_to_world(int x, int y, int z, int chunk_x, int chunk_z) {
+    int world_x = x + chunk_x * CHUNK_SIZE;
+    int world_z = z + chunk_z * CHUNK_SIZE;
+    int world_y = y;
+    return {world_x, world_y, world_z};
+}
+std::tuple<int, int, int>
+ServerChunk::block_to_world(const glm::ivec3& block_pos, ChunkPos chunk_pos) {
+    return block_to_world(block_pos.x, block_pos.y, block_pos.z, chunk_pos.x,
+                          chunk_pos.z);
+}
+
+BiomeType ServerChunk::get_biome() const { return m_biome.load(); }
+
+ChunkPos ServerChunk::get_chunk_pos() const { return m_chunk_pos; }
+
+const std::vector<BlockType>& ServerChunk::get_chunk_blocks() const {
+    return m_blocks;
+}
+
+HeightMapArray ServerChunk::get_heightmap() const {
+    // Logger::info("Chunk pos {} {} in get_heightmap this {}", m_chunk_pos.x,
+    // m_chunk_pos.z, static_cast<const void*>(this));
+    return m_heightmap;
+}
+
+int ServerChunk::index(int x, int y, int z) {
+    ASSERT(!(x < 0 || y < 0 || z < 0 || x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
+             z >= CHUNK_SIZE));
+    if ((x * WORLD_SIZE_Y + y) * CHUNK_SIZE + z < 0 ||
+        (x * WORLD_SIZE_Y + y) * CHUNK_SIZE + z >=
+            CHUNK_SIZE * CHUNK_SIZE * WORLD_SIZE_Y) {
+        Logger::error("block pos x {} y {} z {} range error", x, y, z);
+        ASSERT(0);
+    }
+    return (x * WORLD_SIZE_Y + y) * CHUNK_SIZE + z;
+}
+
+int ServerChunk::index(const glm::vec3& pos) {
+    return ServerChunk::index(pos.x, pos.y, pos.z);
+}
+
+void ServerChunk::gen_phase_one() {
+    m_generator = std::make_unique<ChunkGenerator>(*this);
+    if (!m_generator) {
+        Logger::error("ChunkGenerator is Nullptr");
+        return;
+    }
+    m_generator->assign_chunk_biome();
+    m_seed = m_generator->chunk_seed();
+}
+void ServerChunk::gen_phase_two() {
+    if (!m_generator) {
+        Logger::error("ChunkGenerator is Nullptr");
+        return;
+    }
+    m_generator->generate_heightmap();
+}
+void ServerChunk::gen_phase_three() {
+    if (!m_generator) {
+        Logger::error("ChunkGenerator is Nullptr");
+        return;
+    }
+    m_generator->generate_terrain_blocks();
+}
+
+void ServerChunk::gen_phase_four(
+    const std::array<std::optional<std::vector<BlockType>>, 4>&
+        neighbor_block) {
+    if (!m_generator) {
+        Logger::error("ChunkGenerator is Nullptr");
+        return;
+    }
+    // This must be fully completed before any other operations can proceed!
+    m_generator->blend_surface_blocks_borders(neighbor_block);
+}
+
+void ServerChunk::gen_phase_five() {
+    if (!m_generator) {
+        Logger::error("ChunkGenerator is Nullptr");
+        return;
+    }
+    m_generator->ocean_build();
+    m_generator->generate_river();
+    m_generator->generate_cave();
+
+    m_generator->generate_vegetation();
+    m_generator = nullptr;
+}
+
+void ServerChunk::gen_chunk() {
+    if (m_gening.exchange(true))
+        return;
+    m_gening = true;
+    ASSERT_MSG(m_blocks.empty(),
+               "Blocks isn't Empty, chunk already generated!");
+    if (m_blocks.size() != 0) {
+        Logger::warn(
+            "Request Generator Chunk {} {} ,but the Blocks size is Not 0",
+            m_chunk_pos.x, m_chunk_pos.z);
+        return;
+    }
+    std::vector<ServerChunk> neighbor;
+    for (int i = 0; i < 4; i++) {
+        neighbor.emplace_back(m_world, m_chunk_pos + CHUNK_DIR[i], true);
+    }
+    for (auto& chunk : neighbor) {
+        chunk.gen_phase_one();
+        chunk.gen_phase_two();
+        chunk.gen_phase_three();
+        chunk.gen_phase_five();
+    }
+    gen_phase_one();
+    gen_phase_two();
+    gen_phase_three();
+
+    for (int i = 0; i < 4; i++) {
+        m_neightbor_blocks[i] = neighbor[i].get_chunk_blocks();
+    }
+    gen_phase_four(m_neightbor_blocks);
+    gen_phase_five();
+    m_gening = false;
+}
+// Logger::info("Cross Sum {}", m_cross_vertices_sum.load());
+
+bool ServerChunk::is_temp_chunk() const { return m_temp_chunk.load(); }
+
+bool& ServerChunk::has_cave() { return m_has_cave; }
+
+const OptionalBlockVectorArray& ServerChunk::get_neightbor_blocks() const {
+    return m_neightbor_blocks;
+}
+
+void ServerChunk::set_chunk_block(int index, unsigned id) {
+    m_blocks[index] = id;
+}
+ChunkPos ServerChunk::chunk_pos() const { return m_chunk_pos; }
+
+BiomeType ServerChunk::biome() const { return m_biome; }
+
+void ServerChunk::biome(BiomeType b) { m_biome = b; }
+
+HeightMapArray& ServerChunk::heightmap() { return m_heightmap; }
+std::vector<BlockType>& ServerChunk::blocks() { return m_blocks; }
+ServerWorld& ServerChunk::world() { return m_world; }
+unsigned ServerChunk::seed() const {
+    if (m_seed == 0) {
+        Logger::warn("Seed Not Generator");
+    }
+    return m_seed;
+}
+
+BiomeConditions& ServerChunk::conditions() { return m_conditions; }
+
+} // namespace Cubed

src/gameplay/server_player.cpp

@@ -0,0 +1,56 @@
+#include "Cubed/gameplay/server_player.hpp"
+
+#include "Cubed/gameplay/server_world.hpp"
+namespace Cubed {
+ServerPlayer::ServerPlayer(std::string_view name, std::string_view uuid,
+                           ServerWorld& world, std::shared_ptr<Session> session,
+                           TickType gametick)
+    : m_name(name), m_uuid(uuid), m_world(world), m_session(session),
+      m_last_gametick(gametick) {}
+const glm::vec3& ServerPlayer::get_pos() const { return m_pos; }
+const std::string& ServerPlayer::get_name() const { return m_name; }
+const std::string& ServerPlayer::get_uuid() const { return m_uuid; }
+std::shared_ptr<Session> ServerPlayer::get_session() const { return m_session; }
+void ServerPlayer::update_pos(float x, float y, float z) {
+    m_pos = glm::vec3{x, y, z};
+    ChunkPos chunk_pos = get_chunk_pos(x, z);
+    float dist = distance2(chunk_pos, m_last_chunk_pos);
+    if (dist > 2) {
+        m_world.need_gen(m_uuid);
+        m_last_chunk_pos = chunk_pos;
+    }
+}
+
+void ServerPlayer::update_sync_gametick(TickType gametick) {
+    m_last_gametick = gametick;
+}
+bool ServerPlayer::is_disconnect(TickType current_gametick) const {
+    if (current_gametick - m_last_gametick > TIMEOUT) {
+        return true;
+    }
+    return false;
+}
+
+int ServerPlayer::task_id() const { return m_chunk_task_id.load(); }
+void ServerPlayer::task_id(int id) { m_chunk_task_id = id; }
+
+bool ServerPlayer::has_player(ChunkPos pos) const {
+    std::shared_lock lock(m_chunk_pos_mutex);
+    return m_player_chunk_pos_set.find(pos) != m_player_chunk_pos_set.end();
+}
+void ServerPlayer::update_chunk_set(const ChunkPosSet& set) {
+    std::lock_guard lock(m_chunk_pos_mutex);
+    m_player_chunk_pos_set.clear();
+    m_player_chunk_pos_set.insert(set.begin(), set.end());
+}
+
+const ServerPlayer::ChunkPosSet& ServerPlayer::get_chunk_pos_set() const {
+    std::shared_lock lock(m_chunk_pos_mutex);
+    return m_player_chunk_pos_set;
+}
+
+ServerPlayer::ChunkPosSet& ServerPlayer::get_chunk_pos_set() {
+    std::lock_guard lock(m_chunk_pos_mutex);
+    return m_player_chunk_pos_set;
+}
+} // namespace Cubed

src/gameplay/server_world.cpp

@@ -0,0 +1,829 @@
+#include "Cubed/gameplay/server_world.hpp"
+
+#include "Cubed/config.hpp"
+#include "Cubed/gameplay/packet.hpp"
+#include "Cubed/gameplay/session.hpp"
+#include "Cubed/tools/cubed_assert.hpp"
+#include "Cubed/tools/log.hpp"
+#include "Cubed/tools/uuid.hpp"
+
+#include <ranges>
+#include <utility>
+using namespace std::chrono;
+using namespace std::chrono_literals;
+using namespace google::protobuf;
+
+namespace Cubed {
+ServerWorld::ServerWorld() {}
+
+ServerWorld::~ServerWorld() { stop(); }
+
+void ServerWorld::stop() {
+    if (!m_init) {
+        return;
+    }
+    if (m_stopped.exchange(true)) {
+        return;
+    }
+    send_server_stop();
+    stop_gen_thread();
+    stop_server_thread();
+    // wait_all_chunk_tasks();
+    stop_thread_pool();
+
+    m_finished_queue.clear();
+    m_chunks.clear();
+}
+
+void ServerWorld::update_ref_count(const ChunkPosSet& old,
+                                   const ChunkPosSet& now) {
+
+    // Elements in the old set that are not contained in now are not needed by
+    // the current player.
+
+    for (auto& pos : old) {
+        if (!now.contains(pos)) {
+
+            chunk_acc acc;
+            if (!m_chunks.find(acc, pos)) {
+                Logger::warn("Update Ref Count Error, can't Find old pos "
+                             "in m_chunks");
+                continue;
+            }
+            if (acc->second.ref_count == 0) {
+                Logger::error("Chunk {} {} error, ref count is 0", pos.x,
+                              pos.z);
+                m_chunks.erase(acc);
+                continue;
+            }
+            if (--acc->second.ref_count == 0) {
+                m_chunks.erase(acc);
+            }
+        }
+    }
+
+    for (auto& pos : now) {
+
+        chunk_acc acc;
+        if (!m_chunks.find(acc, pos)) {
+            Logger::warn(
+                "Update Ref Count Error, can't Find now pos in m_chunks");
+            continue;
+        }
+        if (!old.contains(pos)) {
+            ++acc->second.ref_count;
+        }
+    }
+}
+
+void ServerWorld::send_time() {
+    Arena arena;
+    auto* rsp = Arena::Create<UpdateTime>(&arena);
+
+    rsp->set_day_tick(m_day_tick);
+    rsp->set_game_tick(m_game_ticks);
+
+    for (auto& [uuid, player] : m_players) {
+        player.get_session()->send(make_packet(*rsp), 3);
+    }
+}
+
+void ServerWorld::send_chunk(int task_id, const std::string& uuid,
+                             ChunkPos pos) {
+
+    {
+        std::shared_lock lock(m_player_mutex);
+        auto it = m_players.find(uuid);
+        if (it == m_players.end()) {
+            return;
+        }
+        if (task_id < it->second.task_id()) {
+            // Old chunk requests are simply discarded
+            return;
+        }
+    }
+
+    Arena arean;
+    ChunkDataRsp* rsp = Arena::Create<ChunkDataRsp>(&arean);
+    auto* rsq_pos = rsp->mutable_pos();
+    rsq_pos->set_x(pos.x);
+    rsq_pos->set_z(pos.z);
+    {
+        chunk_caac cacc;
+        if (!m_chunks.find(cacc, pos)) {
+            // No chunk found and not generating
+            Logger::error("Chunk {} {} neither pending nor ready", pos.x,
+                          pos.z);
+            return;
+        }
+
+        if (cacc->second.state == ChunkState::GENERATING) {
+
+            m_waiting_chunk_requests.emplace(uuid, task_id, pos);
+            return;
+        }
+        if (cacc->second.state != ChunkState::READY) {
+            Logger::error("Chunk {} {} is invaild", pos.x, pos.z);
+            return;
+        }
+
+        rsp->set_chunk_seed(cacc->second.chunk->seed());
+        rsp->set_biome_type(std::to_underlying(cacc->second.chunk->biome()));
+        auto* blocks = rsp->mutable_chunk_blocks();
+        auto& chunk_blocks = cacc->second.chunk->get_chunk_blocks();
+        blocks->Assign(chunk_blocks.begin(), chunk_blocks.end());
+        auto& neighbor_blocks = cacc->second.chunk->get_neightbor_blocks();
+
+        auto assign = [](auto* nb,
+                         const std::optional<std::vector<BlockType>>& blocks) {
+            if (!blocks) {
+                return;
+            }
+            if (!nb) {
+                return;
+            }
+            nb->Assign(blocks->begin(), blocks->end());
+        };
+        auto* nb1 = rsp->mutable_neighbor_blocks_1();
+        auto* nb2 = rsp->mutable_neighbor_blocks_2();
+        auto* nb3 = rsp->mutable_neighbor_blocks_3();
+        auto* nb4 = rsp->mutable_neighbor_blocks_4();
+        assign(nb1, neighbor_blocks[0]);
+        assign(nb2, neighbor_blocks[1]);
+        assign(nb3, neighbor_blocks[2]);
+        assign(nb4, neighbor_blocks[3]);
+    }
+    std::shared_ptr<Session> s;
+    {
+        std::shared_lock lock(m_player_mutex);
+        auto it = m_players.find(uuid);
+        if (it != m_players.end()) {
+            s = it->second.get_session();
+            it->second.update_sync_gametick(m_game_ticks);
+        }
+    }
+    if (!s) {
+        Logger::error("Player {} session not exist", uuid);
+        return;
+    }
+    rsp->set_task_id(task_id);
+    s->send(make_packet(*rsp));
+}
+
+void ServerWorld::init_world() {
+
+    register_timer("player disconnect", 5, [this]() {
+        std::vector<std::string> disconnect;
+        {
+            std::shared_lock lock(m_player_mutex);
+            for (auto& [uuid, player] : m_players) {
+                if (player.is_disconnect(m_game_ticks)) {
+                    disconnect.emplace_back(uuid);
+                }
+            }
+        }
+        for (auto& uuid : disconnect) {
+            handle_player_exit(uuid);
+        }
+    });
+    // Periodically process pending players
+    register_timer("player chunk send", 1, [this]() {
+        PendingRequest request;
+        if (m_waiting_chunk_requests.try_pop(request)) {
+            handle_chunk_req(request.task_id, request.uuid, request.pos);
+        }
+    });
+
+    m_cave_carcer.init(ChunkGenerator::seed());
+    m_river_worm.init(ChunkGenerator::seed());
+    // m_chunks.reserve(MAX_DISTANCE * MAX_DISTANCE * 4);
+    start_thread_pool();
+
+    auto t1 = std::chrono::system_clock::now();
+
+    start_gen_thread();
+    init_chunks();
+    auto t2 = std::chrono::system_clock::now();
+    auto d = std::chrono::duration_cast<std::chrono::milliseconds>(t2 - t1);
+    Logger::info("Chunk Block Init Finish, Time Consuming: {}", d);
+
+    start_server_thread();
+    m_init = true;
+}
+
+void ServerWorld::init_chunks() { hot_reload(); }
+
+void ServerWorld::gen_chunks_internal(const std::string& uuid) {
+    // Logger::info("gen_chunks_internal");
+    m_chunk_gen_finished = false;
+
+    ChunkPosSet required_chunks_set;
+    compute_required_chunks(required_chunks_set, uuid);
+    std::vector<ChunkPos> need_gen_chunks_pos;
+
+    ChunkPosSet old_set;
+    sync_and_collect_missing_chunks(need_gen_chunks_pos, required_chunks_set);
+    {
+        std::lock_guard lock(m_player_mutex);
+        auto it = m_players.find(uuid);
+        if (it == m_players.end()) {
+            return;
+        }
+        old_set = std::move(it->second.get_chunk_pos_set());
+        it->second.update_chunk_set(required_chunks_set);
+    }
+
+    update_ref_count(old_set, required_chunks_set);
+    ASSERT_MSG(!required_chunks_set.empty(), "required chunks is empty!!");
+
+    Logger::info("New Gen Chunks Sum: {}", need_gen_chunks_pos.size());
+
+    if (need_gen_chunks_pos.empty()) {
+        m_could_gen = true;
+
+        return;
+    }
+    NewChunkVector new_chunks;
+
+    // Create new chunk
+
+    for (auto& pos : need_gen_chunks_pos) {
+        new_chunks.emplace_back(
+            pos, std::make_unique<ServerChunk>(ServerChunk(*this, pos)));
+    }
+
+    submit_new_chunks(uuid, new_chunks);
+    m_chunk_gen_finished = true;
+}
+
+void ServerWorld::compute_required_chunks(
+    ChunkPosSet& required_chunks, const std::optional<std::string>& uuid) {
+    glm::vec3 player_pos;
+    if (uuid == std::nullopt) {
+        player_pos = glm::vec3{0.0f};
+    } else {
+        player_pos = get_player_pos(uuid.value());
+    }
+    int x = std::floor(player_pos.x);
+    int z = std::floor(player_pos.z);
+    auto [chunk_x, chunk_z] = get_chunk_pos(x, z);
+    int radius = m_rendering_distance;
+    int r2 = radius * radius;
+    required_chunks.reserve(radius * radius);
+
+    for (int dx = -radius; dx <= radius; ++dx) {
+        for (int dz = -radius; dz <= radius; ++dz) {
+            if (dx * dx + dz * dz <= r2) {
+                required_chunks.emplace(chunk_x + dx, chunk_z + dz);
+            }
+        }
+    }
+}
+
+void ServerWorld::sync_and_collect_missing_chunks(
+    std::vector<ChunkPos>& need_gen_chunks_pos,
+    const ChunkPosSet& required_chunks) {
+
+    for (auto pos : required_chunks) {
+        chunk_acc acc;
+        if (m_chunks.insert(acc, pos)) {
+            need_gen_chunks_pos.push_back(pos);
+            acc->second = ChunkEntity{ChunkState::GENERATING, nullptr, 0};
+        }
+    }
+}
+
+void ServerWorld::submit_new_chunks(const std::string& uuid,
+                                    NewChunkVector& new_chunks) {
+    using enum ChunkLoadStyle;
+    auto pool_ptr = m_gen_thread_pool.load();
+    if (!pool_ptr) {
+        return;
+    }
+    switch (m_chunk_load_style) {
+    case RANDOM:
+        // Enqueue directly in random order
+        for (auto& task : new_chunks) {
+
+            pool_ptr->enqueue([&task, this]() {
+                std::unique_ptr<ServerChunk> chunk{std::move(task.chunk)};
+                chunk->gen_chunk();
+                m_finished_queue.push(std::move(chunk));
+            });
+        }
+        break;
+    case CENTER: {
+        std::vector<std::pair<ChunkPos, PendingChunk*>> tasks;
+        for (auto& task : new_chunks) {
+
+            tasks.emplace_back(task.pos, &task);
+        }
+        glm::vec3 player_pos = get_player_pos(uuid);
+        ChunkPos player_chunk_pos = get_chunk_pos(player_pos.x, player_pos.z);
+        auto dist2 = [player_chunk_pos](ChunkPos chunk_pos) {
+            float dx = player_chunk_pos.x - chunk_pos.x;
+            float dz = player_chunk_pos.z - chunk_pos.z;
+            return dx * dx + dz * dz;
+        };
+
+        std::sort(tasks.begin(), tasks.end(),
+                  [&dist2](const auto& a, const auto& b) {
+                      return dist2(a.first) < dist2(b.first);
+                  });
+
+        const int CHUNKS_PER_PRIORITY = m_gen_pool_threads;
+
+        for (size_t i = 0; i < tasks.size(); ++i) {
+            int priority = 10 + static_cast<int>(i / CHUNKS_PER_PRIORITY);
+            auto* task = tasks[i].second;
+            pool_ptr->enqueue(priority,
+                              [this, chunk = std::move(task->chunk)]() mutable {
+                                  chunk->gen_chunk();
+                                  m_finished_queue.push(std::move(chunk));
+                              });
+        }
+    } break;
+    }
+}
+
+void ServerWorld::start_gen_thread() {
+    m_gen_running = true;
+    Logger::info("Gen Thread Started");
+    m_gen_thread = std::jthread([this](std::stop_token token) {
+        while (!token.stop_requested()) {
+            std::unique_lock<std::mutex> lk(m_need_gen_queue_mutex);
+
+            m_gen_cv.wait(lk, token, [this]() {
+                return m_need_gen_chunk.load() || !m_gen_running ||
+                       !m_need_gen_queue.empty();
+            });
+            if (!m_gen_running) {
+                break;
+            }
+            if (token.stop_requested()) {
+                break;
+            }
+            m_need_gen_chunk = false;
+            std::string uuid;
+            if (!m_need_gen_queue.empty()) {
+                uuid = m_need_gen_queue.front();
+                m_need_gen_queue.pop();
+            }
+            lk.unlock();
+            gen_chunks_internal(uuid);
+        }
+    });
+}
+
+void ServerWorld::start_server_thread() {
+    m_server_thread =
+        std::jthread([this](std::stop_token token) { serever_run(token); });
+}
+
+void ServerWorld::start_thread_pool() {
+    int max_thread = std::thread::hardware_concurrency();
+    if (m_gen_pool_threads == 0) {
+        m_gen_pool_threads = change_pool_threads(m_gen_thread_pool,
+                                                 max_thread - RESERVED_THREADS);
+    } else {
+        m_gen_pool_threads =
+            change_pool_threads(m_gen_thread_pool, m_gen_pool_threads);
+    }
+
+    if (m_net_pool_threads == 0) {
+        m_net_pool_threads = change_pool_threads(m_net_thread_pool, 4);
+    } else {
+        m_net_pool_threads =
+            change_pool_threads(m_net_thread_pool, m_net_pool_threads);
+    }
+}
+
+void ServerWorld::stop_gen_thread() {
+    m_gen_running = false;
+    m_gen_cv.notify_all();
+    m_gen_thread.request_stop();
+    if (m_gen_thread.joinable()) {
+        m_gen_thread.join();
+    }
+    Logger::info("Gen Thread Stopped");
+}
+
+void ServerWorld::stop_server_thread() {
+    m_server_thread.request_stop();
+    if (m_server_thread.joinable()) {
+        m_server_thread.join();
+    }
+}
+
+void ServerWorld::stop_thread_pool() {
+    auto pool_ptr = m_gen_thread_pool.load();
+    if (pool_ptr) {
+        pool_ptr->stop();
+    }
+    m_gen_thread_pool.store(nullptr);
+    Logger::info("Gen Thread Pool Stopped");
+
+    auto p = m_net_thread_pool.load();
+    if (p) {
+        p->stop();
+    }
+    m_net_thread_pool.store(nullptr);
+    Logger::info("Net Thread Pool Stopped");
+}
+
+void ServerWorld::serever_run(std::stop_token stoken) {
+    Logger::info("Server Thread Started!");
+
+    using Clock = std::chrono::steady_clock;
+    constexpr auto TICK = std::chrono::milliseconds(DEFAULT_PER_TICK_TIME);
+
+    auto next = Clock::now();
+    while (!stoken.stop_requested()) {
+        next += TICK;
+        if (m_tick_running) {
+            ++m_game_ticks;
+            m_day_tick = (m_day_tick + 1) % DAY_TIME;
+        }
+        update();
+        std::this_thread::sleep_until(next);
+    }
+    Logger::info("Server Thread Stopped!");
+}
+
+void ServerWorld::need_gen(std::string uuid) {
+
+    // if (!m_could_gen) {
+    //     Logger::warn("It is generating or consuming new chunks");
+    //     return;
+    // }
+
+    m_could_gen = false;
+
+    {
+        std::lock_guard lock(m_need_gen_queue_mutex);
+        m_need_gen_queue.enqueue(std::move(uuid));
+    }
+
+    // m_gen_player_pos = get_player("TestPlayer").get_player_pos();
+
+    m_need_gen_chunk = true;
+
+    m_gen_cv.notify_one();
+}
+
+bool ServerWorld::set_block(const glm::ivec3& block_pos, unsigned id) {
+
+    int world_x, world_y, world_z;
+    world_x = block_pos.x;
+    world_y = block_pos.y;
+    world_z = block_pos.z;
+
+    auto [chunk_x, chunk_z] = get_chunk_pos(world_x, world_z);
+    chunk_acc acc;
+
+    if (!m_chunks.find(acc, ChunkPos{chunk_x, chunk_z})) {
+        return false;
+    }
+    if (acc->second.state != ChunkState::READY) {
+        return false;
+    }
+    auto [x, y, z] = ServerChunk::world_to_block(world_x, world_y, world_z,
+                                                 chunk_x, chunk_z);
+    if (x < 0 || y < 0 || z < 0 || x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
+        z >= CHUNK_SIZE) {
+        return false;
+    }
+
+    acc->second.chunk->set_chunk_block(ServerChunk::index(x, y, z), id);
+    return true;
+}
+
+void ServerWorld::hot_reload() {
+    auto& config = Config::get();
+    int dist = config.get<int>("world.rendering_distance");
+    m_rendering_distance = dist <= MAX_DISTANCE ? dist : MAX_DISTANCE;
+}
+
+void ServerWorld::update() {
+    // poll_finished_chunks();
+    {
+        bool consumed = false;
+        std::unique_ptr<ServerChunk> chunk;
+        while (m_finished_queue.try_pop(chunk)) {
+            if (!chunk) {
+                Logger::error("Finished Queue has nullptr Chunk");
+                return;
+            }
+            chunk_acc acc;
+            auto pos = chunk->get_chunk_pos();
+            if (!m_chunks.find(acc, pos)) {
+                Logger::error(
+                    "New Chunk {} {} not Find, don't move to m_chunks", pos.x,
+                    pos.z);
+                continue;
+            }
+            acc->second.chunk = std::move(chunk);
+            acc->second.state = ChunkState::READY;
+            consumed = true;
+        }
+        if (consumed) {
+            m_could_gen = true;
+        }
+    }
+
+    send_time();
+    for (auto& [id, timer] : m_timers) {
+        timer.update();
+    }
+}
+
+void ServerWorld::sync_player_pos(const std::string& uuid, float x, float y,
+                                  float z) {
+    std::string name;
+    {
+        std::lock_guard lock(m_player_mutex);
+        auto it = m_players.find(uuid);
+        if (it == m_players.end()) {
+            Logger::warn("Player {} is not in this Server", uuid);
+            return;
+        }
+        it->second.update_pos(x, y, z);
+        it->second.update_sync_gametick(m_game_ticks);
+        name = it->second.get_name();
+    }
+    ChunkPos pos = get_chunk_pos(x, z);
+    // update other player pos;
+    std::vector<std::shared_ptr<Session>> other;
+    {
+        std::shared_lock lock(m_player_mutex);
+        for (auto& [o_uuid, player] : m_players) {
+            if (o_uuid == uuid) {
+                continue;
+            }
+            if (player.has_player(pos)) {
+                other.emplace_back(player.get_session());
+            }
+        }
+    }
+
+    for (auto& session : other) {
+        if (!session) {
+            continue;
+        }
+        Arena arena;
+        auto* rsp = Arena::Create<PlayerInfoRsp>(&arena);
+        rsp->set_uuid(uuid);
+        rsp->set_name(name);
+        auto* pos = rsp->mutable_pos();
+        pos->set_x(x);
+        pos->set_y(y);
+        pos->set_z(z);
+        session->send(make_packet(*rsp), 0);
+    }
+}
+
+void ServerWorld::handle_player_login(const std::string& name,
+                                      std::shared_ptr<Session> session) {
+    std::string uuid = generate_uuid();
+    Logger::info("Player {} (uuid {}) join the world", name, uuid);
+    bool sucess = true;
+    {
+        std::lock_guard lock(m_player_mutex);
+        auto [_, inserted] = m_players.emplace(
+            std::piecewise_construct, std::forward_as_tuple(std::string(uuid)),
+            std::forward_as_tuple(name, uuid, *this, session, m_game_ticks));
+        if (!inserted) {
+            Logger::error("Player insert Fail");
+        }
+        sucess = inserted;
+    }
+
+    Arena arena;
+    if (!sucess) {
+        auto* rsp = Arena::Create<LoginRsp>(&arena);
+        rsp->set_success(false);
+        session->send(make_packet(*rsp), 0);
+        return;
+    }
+
+    m_uuid_to_name.emplace(uuid, name);
+    // Pre-insert into new_chunks to ensure correct addition to waiting_player
+    /*ChunkPosSet required_chunks;
+    compute_required_chunks(required_chunks, uuid);
+    std::vector<ChunkPos> need_gen_chunks_pos;
+
+    sync_and_collect_missing_chunks(need_gen_chunks_pos, required_chunks);
+
+    {
+        std::lock_guard lock(m_new_chunk_mutex);
+        for (auto& pos : need_gen_chunks_pos) {
+            m_new_chunks.emplace(pos, ServerChunk(*this, pos));
+        }
+    }
+    */
+    need_gen(uuid);
+
+    auto* rsp = Arena::Create<LoginRsp>(&arena);
+    rsp->set_success(true);
+    rsp->set_uuid(uuid);
+    session->send(make_packet(*rsp), 0);
+}
+
+void ServerWorld::handle_player_exit(const std::string& uuid) {
+    std::shared_ptr<Session> exit_session;
+    ChunkPosSet old_set;
+    {
+        std::lock_guard lock(m_player_mutex);
+        auto it = m_players.find(uuid);
+        if (it != m_players.end()) {
+            Logger::info("Player {} Exit the Server", it->second.get_name());
+            exit_session = it->second.get_session();
+            old_set = std::move(it->second.get_chunk_pos_set());
+            m_players.erase(it);
+        } else {
+            Logger::error("Player {} isn't in Server", uuid);
+            return;
+        }
+    }
+
+    m_uuid_to_name.erase(uuid);
+
+    update_ref_count(old_set, {});
+
+    Arena arena;
+    auto* rsp = Arena::Create<LogoutRsp>(&arena);
+    rsp->set_uuid(uuid);
+    rsp->set_server_stop(false);
+    exit_session->send(make_packet(*rsp), 0);
+
+    std::vector<std::shared_ptr<Session>> sessions;
+    {
+        std::shared_lock lock(m_player_mutex);
+        for (auto& [uuid, player] : m_players) {
+            sessions.emplace_back(player.get_session());
+        }
+    }
+
+    for (auto& s : sessions) {
+        if (s) {
+            s->send(make_packet(*rsp), 0);
+        }
+    }
+}
+
+glm::vec3 ServerWorld::get_player_pos(const std::string& uuid) const {
+    std::shared_lock lock(m_player_mutex);
+    auto it = m_players.find(uuid);
+    if (it == m_players.end()) {
+        Logger::error("Can't find player uuid {}", uuid);
+        return glm::vec3{0.0f};
+    }
+    return it->second.get_pos();
+}
+
+void ServerWorld::handle_chunk_req(int task_id, const std::string& uuid,
+                                   ChunkPos pos) {
+    {
+        std::shared_lock lock(m_player_mutex);
+        auto it = m_players.find(uuid);
+        if (it == m_players.end()) {
+            return;
+        }
+        if (it->second.task_id() < task_id) {
+            // task_id is an atomic variable, can be operated on directly
+            it->second.task_id(task_id);
+        }
+    }
+    auto pool = m_net_thread_pool.load();
+    pool->enqueue(
+        [task_id, uuid, pos, this]() { send_chunk(task_id, uuid, pos); });
+}
+
+void ServerWorld::handle_block_change(const BlockChangeReq& req) {
+    float x = std::floor(req.pos().x());
+    float y = std::floor(req.pos().y());
+    float z = std::floor(req.pos().z());
+    if (!set_block(glm::ivec3(x, y, z), req.block())) {
+        return;
+    }
+
+    Arena arena;
+    BlockChangeRsp* rsp = Arena::Create<BlockChangeRsp>(&arena);
+    auto* pos = rsp->mutable_pos();
+    pos->set_x(x);
+    pos->set_y(y);
+    pos->set_z(z);
+    rsp->set_block(req.block());
+    std::vector<std::shared_ptr<Session>> sessions;
+    auto chunk_pos = get_chunk_pos(x, z);
+    {
+        std::shared_lock lock(m_player_mutex);
+        for (auto& [uuid, player] : m_players) {
+            if (player.has_player(chunk_pos)) {
+                auto session = player.get_session();
+                sessions.emplace_back(std::move(session));
+            }
+        }
+    }
+
+    for (auto& x : sessions) {
+        if (x) {
+            x->send(make_packet(*rsp), 1);
+        }
+    }
+}
+
+int ServerWorld::rendering_distance() const {
+    return m_rendering_distance.load();
+}
+
+void ServerWorld::rendering_distance(int rendering_distance) {
+    m_rendering_distance = rendering_distance;
+}
+
+CaveCarver& ServerWorld::cave_carcer() { return m_cave_carcer; }
+RiverWorm& ServerWorld::river_worm() { return m_river_worm; }
+
+TickType ServerWorld::game_tick() const { return m_game_ticks.load(); }
+TickType ServerWorld::day_tick() const { return m_day_tick.load(); }
+void ServerWorld::day_tick(TickType tick) {
+    tick %= DAY_TIME;
+    m_day_tick = tick;
+}
+int ServerWorld::per_tick_time() const { return m_per_tick_time.load(); }
+void ServerWorld::per_tick_time(int ms) { m_per_tick_time = ms; }
+
+bool ServerWorld::is_tick_running() const { return m_tick_running.load(); }
+void ServerWorld::tick_running(bool run) { m_tick_running = run; }
+int ServerWorld::gen_pool_threads() const { return m_gen_pool_threads.load(); }
+int ServerWorld::max_threads() const { return m_max_threads.load(); }
+
+void ServerWorld::change_pool_threads(ThreadPoolKind kind, int threads) {
+    switch (kind) {
+    case ThreadPoolKind::NET:
+        m_net_pool_threads = change_pool_threads(m_net_thread_pool, threads);
+        break;
+    case ThreadPoolKind::GEN:
+        m_gen_pool_threads = change_pool_threads(m_gen_thread_pool, threads);
+        break;
+    }
+}
+
+int ServerWorld::change_pool_threads(
+    std::atomic<std::shared_ptr<ThreadPool>>& thread_pool, int threads) {
+    m_max_threads = std::thread::hardware_concurrency();
+    if (m_max_threads < 1) {
+        Logger::warn("Can't Get Max Support Threads, Set Max Threads to 4");
+        m_max_threads = 1;
+    }
+    int used_thread = std::clamp(threads, 1, m_max_threads.load());
+    Logger::info("Create New Thread Pool Use {} Threads", used_thread);
+    thread_pool.store(std::make_shared<ThreadPool>(used_thread));
+    return used_thread;
+}
+
+int ServerWorld::change_pool_threads(
+    std::atomic<std::shared_ptr<PriorityThreadPool>>& thread_pool,
+    int threads) {
+    m_max_threads = std::thread::hardware_concurrency();
+    if (m_max_threads < 1) {
+        Logger::warn("Can't Get Max Support Threads, Set Max Threads to 4");
+        m_max_threads = 1;
+    }
+    int used_thread = std::clamp(threads, 1, m_max_threads.load());
+    Logger::info("Create New Thread Pool Use {} Threads", used_thread);
+    thread_pool.store(std::make_shared<PriorityThreadPool>(used_thread));
+    return used_thread;
+}
+
+void ServerWorld::send_server_stop() {
+    Arena arena;
+    auto* rsp = Arena::Create<LogoutRsp>(&arena);
+    rsp->set_server_stop(true);
+    std::shared_lock lock(m_player_mutex);
+    for (auto& [uuid, player] : m_players) {
+        player.get_session()->send(make_packet(*rsp), 0);
+    }
+    Logger::info("Send Server Mesaage Success");
+}
+
+int ServerWorld::chunk_load_style() const {
+    return std::to_underlying(m_chunk_load_style.load());
+}
+void ServerWorld::set_chunk_load_style(int id) {
+    using enum ChunkLoadStyle;
+
+    switch (id) {
+    case std::to_underlying(RANDOM):
+        m_chunk_load_style = RANDOM;
+        return;
+    case std::to_underlying(CENTER):
+        m_chunk_load_style = CENTER;
+        return;
+    }
+    Logger::error("Can,t Find Chunk Load Style Id {}, Nothing Will Do", id);
+}
+
+int ServerWorld::chunk_size() const { return m_chunks.size(); }
+
+} // namespace Cubed

src/gameplay/session.cpp

@@ -0,0 +1,150 @@
+#include "Cubed/gameplay/session.hpp"
+
+#include "Cubed/gameplay/server_world.hpp"
+#include "Cubed/tools/log.hpp"
+#include "Cubed/tools/uuid.hpp"
+using asio::ip::tcp;
+using namespace google::protobuf;
+namespace Cubed {
+Session::Session(tcp::socket socket, ServerWorld& server_world,
+                 asio::io_context& io)
+    : m_socket(std::move(socket)), m_strand(asio::make_strand(io)),
+      m_uuid(generate_uuid()), m_server_world(server_world) {}
+
+Session::~Session() {}
+
+void Session::start() {
+    auto self = shared_from_this();
+    asio::co_spawn(
+        m_strand,
+        [self]() -> asio::awaitable<void> { co_await self->read_loop(); },
+        asio::detached);
+}
+
+void Session::send(std::shared_ptr<std::vector<uint8_t>> packet, int priority) {
+    asio::post(m_strand, [self = shared_from_this(), packet = std::move(packet),
+                          priority]() mutable {
+        bool idle = self->m_write_queue.empty();
+        self->m_write_queue.emplace(priority, self->m_sequence++,
+                                    std::move(packet));
+        if (idle) {
+            self->do_write();
+        }
+    });
+}
+
+const std::string& Session::uuid() const { return m_uuid; }
+
+asio::awaitable<void> Session::read_loop() {
+    try {
+        while (true) {
+            std::array<uint8_t, HEADER_LEN> header_buffer;
+            co_await asio::async_read(m_socket, asio::buffer(header_buffer),
+                                      asio::use_awaitable);
+
+            auto header = decode_packet_header(header_buffer);
+            uint32_t total_len = HEADER_LEN + header.compressed_size;
+
+            if (total_len < HEADER_LEN || total_len > MAX_PACKET_SIZE) {
+
+                throw std::runtime_error("invalid packet length");
+            }
+            std::vector<uint8_t> body_data(header.compressed_size);
+            if (header.compressed_size > 0) {
+                co_await asio::async_read(m_socket, asio::buffer(body_data),
+                                          asio::use_awaitable);
+            }
+            auto cmd_id = header.cmd;
+            Arena arena;
+            if (cmd_id == std::to_underlying(PacketEnum::LOGIN_REQ)) {
+                auto* req = Arena::Create<LoginReq>(&arena);
+                Logger::info("Session: Receive Login req");
+                if (decode_packet(*req, body_data, header)) {
+                    m_server_world.handle_player_login(req->name(),
+                                                       shared_from_this());
+                }
+            }
+            if (cmd_id == std::to_underlying(PacketEnum::PLAYER_POS)) {
+                auto* pos = Arena::Create<PlayerPos>(&arena);
+                if (decode_packet(*pos, body_data, header)) {
+                    m_server_world.sync_player_pos(pos->uuid(), pos->pos().x(),
+                                                   pos->pos().y(),
+                                                   pos->pos().z());
+                }
+            }
+            if (cmd_id == std::to_underlying(PacketEnum::CHUNK_DATA_REQ)) {
+                auto* req = Arena::Create<ChunkDataReq>(&arena);
+                // Logger::info("Session: Receive Chunk Data req");
+                if (decode_packet(*req, body_data, header)) {
+                    m_server_world.handle_chunk_req(
+                        req->task_id(), req->uuid(),
+                        ChunkPos(req->pos().x(), req->pos().z()));
+                }
+            }
+            if (cmd_id == std::to_underlying(PacketEnum::BLOCK_CHANGE_REQ)) {
+                auto* req = Arena::Create<BlockChangeReq>(&arena);
+                Logger::info("Session: Receive Block Change req");
+                if (decode_packet(*req, body_data, header)) {
+                    m_server_world.handle_block_change(*req);
+                }
+            }
+            if (cmd_id == std::to_underlying(PacketEnum::LOGOUT_REQ)) {
+                auto* req = Arena::Create<LogoutReq>(&arena);
+                if (decode_packet(*req, body_data, header)) {
+                    m_server_world.handle_player_exit(req->uuid());
+                }
+            }
+        }
+    } catch (const asio::system_error& e) {
+        auto ec = e.code();
+
+        if (ec == asio::error::eof || ec == asio::error::operation_aborted) {
+
+            Logger::info("Client disconnected");
+        } else {
+            Logger::warn("Asio Error {}", e.what());
+        }
+
+        close();
+    } catch (const std::exception& e) {
+        Logger::error("Session Error {}", e.what());
+        close();
+    } catch (...) {
+        Logger::error("Unknow Error");
+        close();
+    }
+    co_return;
+}
+
+void Session::do_write() {
+
+    auto self = shared_from_this();
+    auto packet = std::move(m_write_queue.top().packet);
+    asio::async_write(
+        m_socket, asio::buffer(*packet),
+        asio::bind_executor(m_strand, [self](std::error_code ec, size_t) {
+            if (ec) {
+                Logger::warn("Write Ec {}", ec.message());
+                self->close();
+                return;
+            }
+            self->m_write_queue.pop();
+            if (!self->m_write_queue.empty()) {
+                self->do_write();
+            }
+        }));
+}
+
+void Session::close() {
+    if (m_closed.exchange(true)) {
+        return;
+    }
+
+    std::error_code ec;
+
+    m_socket.shutdown(tcp::socket::shutdown_both, ec);
+
+    m_socket.close(ec);
+}
+
+} // namespace Cubed

src/gameplay/tree.cpp

@@ -1,6 +1,6 @@
 #include "Cubed/gameplay/tree.hpp"
 
-#include "Cubed/gameplay/chunk.hpp"
+#include "Cubed/gameplay/server_chunk.hpp"
 
 #include <array>
 
@@ -27,10 +27,10 @@ static constexpr std::array<TreeStructNode, 62> TREE{{
     {{-1, 3, -2}, 6}, {{-2, 3, -1}, 6},
 }};
 
-bool build_tree(Chunk& chunk, const glm::ivec3& pos) {
+bool build_tree(ServerChunk& chunk, const glm::ivec3& pos) {
     auto& block = chunk.get_chunk_blocks();
 
-    if (block[Chunk::index(pos)] != 1) {
+    if (block[ServerChunk::index(pos)] != 1) {
         return false;
     }
     for (const auto& d : TREE) {
@@ -42,13 +42,13 @@ bool build_tree(Chunk& chunk, const glm::ivec3& pos) {
             z >= CHUNK_SIZE) {
             return false;
         }
-        if (block[Chunk::index(tree_node)] != 0) {
+        if (block[ServerChunk::index(tree_node)] != 0) {
             return false;
         }
     }
     for (const auto& d : TREE) {
         auto tree_node = pos + d.offset;
-        chunk.set_chunk_block(Chunk::index(tree_node), d.id);
+        chunk.set_chunk_block(ServerChunk::index(tree_node), d.id);
     }
     return true;
 }

src/gameplay/vertex_data.cpp

@@ -1,9 +1,9 @@
 #include "Cubed/gameplay/vertex_data.hpp"
 
-#include "Cubed/gameplay/world.hpp"
+#include "Cubed/gameplay/client_world.hpp"
 
 namespace Cubed {
-VertexData::VertexData(World& world) : m_world(world) {}
+VertexData::VertexData(ClientWorld& world) : m_world(world) {}
 VertexData::~VertexData() {
     if (m_vbo != 0) {
         m_world.push_delete_vbo(m_vbo);

src/gameplay/world.cpp

@@ -1,703 +0,0 @@
-#include "Cubed/gameplay/world.hpp"
-
-#include "Cubed/config.hpp"
-#include "Cubed/gameplay/player.hpp"
-#include "Cubed/tools/cubed_assert.hpp"
-#include "Cubed/tools/cubed_hash.hpp"
-
-#include <glm/gtc/constants.hpp>
-#include <numbers>
-#include <utility>
-using namespace std::chrono;
-using namespace std::chrono_literals;
-
-namespace Cubed {
-
-struct ChunkRenderData {
-    std::array<const std::vector<BlockType>*, 4> neighbor_block;
-    Chunk* chunk;
-};
-
-World::World() {}
-
-World::~World() {
-    stop_gen_thread();
-    stop_server_thread();
-    wait_all_chunk_tasks();
-    stop_thread_pool();
-
-    m_chunks.clear();
-    {
-        std::lock_guard lk(m_delete_vbo_mutex);
-        for (auto x : m_pending_delete_vbo) {
-            glDeleteBuffers(1, &x);
-        }
-        m_pending_delete_vbo.clear();
-    }
-    {
-        std::lock_guard lk(m_delete_vao_mutex);
-        for (auto x : m_pending_delete_vao) {
-            glDeleteVertexArrays(1, &x);
-        }
-        m_pending_delete_vao.clear();
-    }
-}
-
-void World::wait_all_chunk_tasks() {
-    for (auto& [pos, task] : new_chunks) {
-        task.future.get();
-    }
-}
-
-bool World::can_move(const AABB& player_box) const { return true; }
-
-const std::optional<LookBlock>&
-World::get_look_block_pos(const std::string& name) const {
-    static std::optional<LookBlock> null_look_block = std::nullopt;
-    auto it = m_players.find(HASH::str(name));
-    if (it == m_players.end()) {
-        Logger::error("Can't find player {}", name);
-        ASSERT(0);
-        return null_look_block;
-    }
-
-    return it->second.get_look_block_pos();
-}
-/*
-const Chunk* World::get_chunk(const ChunkPos& pos) const {
-    std::lock_guard lk(m_chunks_mutex);
-    auto it = m_chunks.find(pos);
-    if (it == m_chunks.end()) {
-        return nullptr;
-    }
-    return &it->second;
-}*/
-
-Player& World::get_player(const std::string& name) {
-    auto it = m_players.find(HASH::str(name));
-    if (it == m_players.end()) {
-        Logger::error("Can't find player {}", name);
-        ASSERT(0);
-    }
-
-    return it->second;
-}
-
-void World::init_world() {
-    m_cave_carcer.init(ChunkGenerator::seed());
-    m_river_worm.init(ChunkGenerator::seed());
-    m_chunks.reserve(MAX_DISTANCE * MAX_DISTANCE * 4);
-    start_thread_pool();
-
-    auto t1 = std::chrono::system_clock::now();
-
-    // init players
-    m_players.emplace(HASH::str("TestPlayer"), Player(*this, "TestPlayer"));
-
-    start_gen_thread();
-    init_chunks();
-    auto t2 = std::chrono::system_clock::now();
-    auto d = std::chrono::duration_cast<std::chrono::milliseconds>(t2 - t1);
-    Logger::info("Chunk Block Init Finish, Time Consuming: {}", d);
-
-    start_server_thread();
-
-    Logger::info("TestPlayer Create Finish");
-}
-void World::init_chunks() {
-    hot_reload();
-    while (!m_chunk_gen_finished) {
-        // Logger::info("World Spawn: {:.2f}%", m_chunk_gen_fraction.load());
-        std::this_thread::sleep_for(std::chrono::microseconds(200));
-    }
-}
-
-ChunkPos World::get_chunk_pos(int world_x, int world_z) {
-    int chunk_x, chunk_z;
-    if (world_x < 0) {
-        chunk_x = (world_x + 1) / CHUNK_SIZE - 1;
-    }
-    if (world_x >= 0) {
-        chunk_x = world_x / CHUNK_SIZE;
-    }
-    if (world_z < 0) {
-        chunk_z = (world_z + 1) / CHUNK_SIZE - 1;
-    }
-    if (world_z >= 0) {
-        chunk_z = world_z / CHUNK_SIZE;
-    }
-    return {chunk_x, chunk_z};
-}
-
-#pragma region ChunkGenerate
-
-void World::gen_chunks_internal() {
-    // Logger::info("gen_chunks_internal");
-    m_chunk_gen_finished = false;
-
-    ChunkPosSet required_chunks;
-    compute_required_chunks(required_chunks);
-
-    ASSERT_MSG(!required_chunks.empty(), "required chunks is empty!!");
-
-    std::vector<ChunkPos> need_gen_chunks_pos;
-
-    sync_and_collect_missing_chunks(need_gen_chunks_pos, required_chunks);
-
-    Logger::info("New Gen Chunks Sum: {}", need_gen_chunks_pos.size());
-
-    if (need_gen_chunks_pos.empty()) {
-        m_could_gen = true;
-
-        return;
-    }
-
-    for (auto& pos : need_gen_chunks_pos) {
-        new_chunks.emplace(pos, Chunk(*this, pos));
-    }
-
-    submit_new_chunks();
-    m_chunk_gen_finished = true;
-}
-
-void World::sync_player_pos(glm::vec3& player_pos) {
-    std::lock_guard lk(m_gen_player_pos_mutex);
-    player_pos = m_gen_player_pos;
-}
-
-void World::compute_required_chunks(ChunkPosSet& required_chunks) {
-    glm::vec3 player_pos;
-    sync_player_pos(player_pos);
-
-    int x = std::floor(player_pos.x);
-    int z = std::floor(player_pos.z);
-    auto [chunk_x, chunk_z] = get_chunk_pos(x, z);
-    int radius = m_rendering_distance;
-    int r2 = radius * radius;
-    required_chunks.reserve(radius * radius);
-
-    for (int dx = -radius; dx <= radius; ++dx) {
-        for (int dz = -radius; dz <= radius; ++dz) {
-            if (dx * dx + dz * dz <= r2) {
-                required_chunks.emplace(chunk_x + dx, chunk_z + dz);
-            }
-        }
-    }
-}
-
-void World::sync_and_collect_missing_chunks(
-    std::vector<ChunkPos>& need_gen_chunks_pos,
-    const ChunkPosSet& required_chunks) {
-    std::lock_guard lk(m_chunks_mutex);
-    for (auto it = m_chunks.begin(); it != m_chunks.end();) {
-        if (required_chunks.find(it->first) == required_chunks.end()) {
-            it = m_chunks.erase(it);
-        } else {
-            ++it;
-        }
-    }
-
-    for (auto pos : required_chunks) {
-        auto it = m_chunks.find(pos);
-        if (it == m_chunks.end()) {
-            need_gen_chunks_pos.push_back(pos);
-        }
-    }
-}
-
-void World::submit_new_chunks() {
-    using enum ChunkLoadStyle;
-    std::lock_guard lock(m_new_chunk_mutex);
-    auto pool_ptr = m_gen_thread_pool.load();
-    if (!pool_ptr) {
-        return;
-    }
-    switch (m_chunk_load_style) {
-    case RANDOM:
-        for (auto& [pos, task] : new_chunks) {
-            if (!task.future.valid()) {
-                task.future =
-                    pool_ptr->enqueue([&task]() { task.chunk.gen_chunk(); });
-            }
-        }
-        break;
-    case CENTER: {
-        std::vector<std::pair<ChunkPos, PendingChunk*>> tasks;
-        for (auto& [pos, task] : new_chunks) {
-            if (!task.future.valid()) {
-                tasks.emplace_back(pos, &task);
-            }
-        }
-        glm::vec3 player_pos;
-        sync_player_pos(player_pos);
-        auto dist2 = [player_pos](ChunkPos chunk_pos) {
-            ChunkPos player_chunk_pos =
-                get_chunk_pos(player_pos.x, player_pos.z);
-            float dx = player_chunk_pos.x - chunk_pos.x;
-            float dz = player_chunk_pos.z - chunk_pos.z;
-            return dx * dx + dz * dz;
-        };
-
-        std::sort(tasks.begin(), tasks.end(),
-                  [&dist2](const auto& a, const auto& b) {
-                      return dist2(a.first) < dist2(b.first);
-                  });
-        for (auto& [pos, task] : tasks) {
-            if (!task->future.valid()) {
-                task->future =
-                    pool_ptr->enqueue([task]() { task->chunk.gen_chunk(); });
-            }
-        }
-    }
-    }
-}
-
-void World::poll_finished_chunks() {
-    m_new_finished_chunk.clear();
-    std::lock_guard lock(m_new_chunk_mutex);
-    std::erase_if(
-        new_chunks, [&](std::pair<const ChunkPos, PendingChunk>& pair) {
-            auto& pending = pair.second;
-            if (!pending.future.valid()) {
-                return false;
-            }
-            if (pending.future.wait_for(0ms) != std::future_status::ready) {
-                return false;
-            }
-            pending.future.get();
-
-            m_new_finished_chunk.emplace_back(pair.first,
-                                              std::move(pending.chunk));
-            return true;
-        });
-}
-
-#pragma endregion
-
-void World::start_gen_thread() {
-    m_gen_running = true;
-    Logger::info("Gen Thread Started");
-    m_gen_thread = std::thread([this]() {
-        while (m_gen_running) {
-            std::unique_lock<std::mutex> lk(m_gen_signal_mutex);
-
-            m_gen_cv.wait(lk, [this]() {
-                return m_need_gen_chunk.load() || !m_gen_running;
-            });
-            if (!m_gen_running) {
-                break;
-            }
-            m_need_gen_chunk = false;
-            lk.unlock();
-
-            gen_chunks_internal();
-        }
-    });
-}
-
-void World::start_server_thread() {
-    m_server_thread = std::thread(
-        [this]() { serever_run(m_server_stop_source.get_token()); });
-}
-
-void World::stop_gen_thread() {
-    m_gen_running = false;
-    m_gen_cv.notify_all();
-    if (m_gen_thread.joinable()) {
-        m_gen_thread.join();
-    }
-    Logger::info("Gen Thread Stopped");
-}
-
-void World::stop_server_thread() {
-    m_server_stop_source.request_stop();
-    if (m_server_thread.joinable()) {
-        m_server_thread.join();
-    }
-}
-
-void World::stop_thread_pool() {
-    auto pool_ptr = m_gen_thread_pool.load();
-    if (pool_ptr) {
-        pool_ptr->stop();
-    }
-    m_gen_thread_pool.store(nullptr);
-    Logger::info("Thread Pool Stopped");
-}
-
-void World::start_thread_pool() {
-    int max_thread = std::thread::hardware_concurrency();
-    if (m_pool_threads == 0) {
-        change_pool_threads(max_thread - RESERVED_THREADS);
-    } else {
-        change_pool_threads(m_pool_threads);
-    }
-}
-
-void World::serever_run(std::stop_token stoken) {
-    Logger::info("Server Thread Started!");
-    while (!stoken.stop_requested()) {
-        std::this_thread::sleep_for(milliseconds(m_per_tick_time));
-        if (m_tick_running) {
-            ++m_game_ticks;
-            m_day_tick = (m_day_tick + 1) % DAY_TIME;
-        }
-    }
-    Logger::info("Server Thread Stopped!");
-}
-
-void World::need_gen() {
-
-    if (!m_could_gen) {
-        Logger::warn("It is generating or consuming new chunks");
-        return;
-    }
-
-    m_could_gen = false;
-    {
-        std::lock_guard lk(m_gen_player_pos_mutex);
-        m_gen_player_pos = get_player("TestPlayer").get_player_pos();
-    }
-
-    m_need_gen_chunk = true;
-
-    m_gen_cv.notify_one();
-}
-
-int World::get_block(const glm::ivec3& block_pos) const {
-    auto [chunk_x, chunk_z] = get_chunk_pos(block_pos.x, block_pos.z);
-    std::shared_lock lk(m_chunks_mutex);
-    auto it = m_chunks.find(ChunkPos{chunk_x, chunk_z});
-
-    if (it == m_chunks.end()) {
-        return 0;
-    }
-
-    const auto& chunk_blocks = it->second.get_chunk_blocks();
-    auto [x, y, z] = Chunk::world_to_block(block_pos, {chunk_x, chunk_z});
-    if (x < 0 || y < 0 || z < 0 || x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
-        z >= CHUNK_SIZE) {
-        return 0;
-    }
-    return chunk_blocks[Chunk::index(x, y, z)];
-}
-
-bool World::is_solid(const glm::ivec3& block_pos) const {
-    auto [chunk_x, chunk_z] = get_chunk_pos(block_pos.x, block_pos.z);
-    std::shared_lock lk(m_chunks_mutex);
-    auto it = m_chunks.find(ChunkPos{chunk_x, chunk_z});
-
-    if (it == m_chunks.end()) {
-        return false;
-    }
-    const auto& chunk_blocks = it->second.get_chunk_blocks();
-    auto [x, y, z] = Chunk::world_to_block(block_pos, {chunk_x, chunk_z});
-    if (x < 0 || y < 0 || z < 0 || x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
-        z >= CHUNK_SIZE) {
-        return false;
-    }
-    auto id = chunk_blocks[Chunk::index(x, y, z)];
-    if (BlockManager::is_gas(id) || BlockManager::is_liquid(id)) {
-        return false;
-    } else {
-        return true;
-    }
-}
-
-bool World::can_pass_block(const glm::ivec3& block_pos) const {
-    auto [chunk_x, chunk_z] = get_chunk_pos(block_pos.x, block_pos.z);
-    std::shared_lock lk(m_chunks_mutex);
-    auto it = m_chunks.find(ChunkPos{chunk_x, chunk_z});
-
-    if (it == m_chunks.end()) {
-        return true;
-    }
-    const auto& chunk_blocks = it->second.get_chunk_blocks();
-    auto [x, y, z] = Chunk::world_to_block(block_pos, {chunk_x, chunk_z});
-    if (x < 0 || y < 0 || z < 0 || x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
-        z >= CHUNK_SIZE) {
-        return true;
-    }
-    auto id = chunk_blocks[Chunk::index(x, y, z)];
-    return BlockManager::is_passable(id);
-}
-
-BlockType World::get_block_tpye(const glm::ivec3& block_pos) const {
-    auto [chunk_x, chunk_z] = get_chunk_pos(block_pos.x, block_pos.z);
-    std::shared_lock lk(m_chunks_mutex);
-    auto it = m_chunks.find(ChunkPos{chunk_x, chunk_z});
-
-    if (it == m_chunks.end()) {
-        // Logger::error("Can't Find Block {} {} {}", block_pos.x, block_pos.y,
-        //               block_pos.z);
-        return 0;
-    }
-    const auto& chunk_blocks = it->second.get_chunk_blocks();
-    auto [x, y, z] = Chunk::world_to_block(block_pos, {chunk_x, chunk_z});
-    if (x < 0 || y < 0 || z < 0 || x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
-        z >= CHUNK_SIZE) {
-        // Logger::error("Can't Find Block {} {} {}", block_pos.x, block_pos.y,
-        //               block_pos.z);
-        return 0;
-    }
-    return chunk_blocks[Chunk::index(x, y, z)];
-}
-
-void World::set_block(const glm::ivec3& block_pos, unsigned id) {
-
-    int world_x, world_y, world_z;
-    world_x = block_pos.x;
-    world_y = block_pos.y;
-    world_z = block_pos.z;
-
-    auto [chunk_x, chunk_z] = get_chunk_pos(world_x, world_z);
-    std::lock_guard lk(m_chunks_mutex);
-    auto it = m_chunks.find(ChunkPos{chunk_x, chunk_z});
-
-    if (it == m_chunks.end()) {
-        return;
-    }
-
-    auto [x, y, z] =
-        Chunk::world_to_block(world_x, world_y, world_z, chunk_x, chunk_z);
-    if (x < 0 || y < 0 || z < 0 || x >= CHUNK_SIZE || y >= WORLD_SIZE_Y ||
-        z >= CHUNK_SIZE) {
-        return;
-    }
-
-    it->second.set_chunk_block(Chunk::index(x, y, z), id);
-
-    static const glm::ivec3 NEIGHBOR_DIRS[] = {
-        {1, 0, 0}, {-1, 0, 0}, {0, 0, -1}, {0, 0, 1}};
-
-    for (const auto& dir : NEIGHBOR_DIRS) {
-        glm::ivec3 neighbor = block_pos + dir;
-
-        auto [cx, cz] = get_chunk_pos(neighbor.x, neighbor.z);
-        auto it = m_chunks.find({cx, cz});
-        if (it != m_chunks.end()) {
-            it->second.mark_dirty();
-        }
-    }
-}
-
-void World::update(float delta_time) {
-    for (auto& player : m_players) {
-        player.second.update(delta_time);
-    }
-    {
-        std::lock_guard lk(m_delete_vbo_mutex);
-        for (auto x : m_pending_delete_vbo) {
-            glDeleteBuffers(1, &x);
-        }
-        m_pending_delete_vbo.clear();
-    }
-
-    {
-        std::lock_guard lk(m_delete_vao_mutex);
-        for (auto x : m_pending_delete_vao) {
-            glDeleteVertexArrays(1, &x);
-        }
-        m_pending_delete_vao.clear();
-    }
-
-    poll_finished_chunks();
-
-    for (auto& x : m_new_finished_chunk) {
-        x.second.upload_to_gpu();
-    }
-
-    // unified compute vertex data before rendering
-    {
-        std::lock_guard lk(m_chunks_mutex);
-        bool consumed = false;
-
-        for (auto& x : m_new_finished_chunk) {
-            m_chunks.insert_or_assign(x.first, std::move(x.second));
-            consumed = true;
-        }
-        if (consumed) {
-            m_could_gen = true;
-        }
-
-        m_render_snapshots.clear();
-        for (auto& [pos, chunk] : m_chunks) {
-            if (chunk.is_dirty()) {
-                // the curial fator influence
-                OptionalBlockVectorArray neighbor_block;
-                for (int i = 0; i < 4; i++) {
-                    auto it = m_chunks.find(pos + CHUNK_DIR[i]);
-                    if (it != m_chunks.end()) {
-                        neighbor_block[i] = (it->second.get_chunk_blocks());
-                    } else {
-                        neighbor_block[i] = std::nullopt;
-                    }
-                }
-                chunk.gen_vertex_data(neighbor_block);
-                chunk.upload_to_gpu();
-            }
-            if (!chunk.is_dirty()) {
-                if (chunk.is_need_upload()) {
-                    chunk.upload_to_gpu();
-                }
-                m_render_snapshots.push_back(
-                    {chunk.get_normal_vao(), chunk.get_normal_vertices_sum(),
-                     chunk.get_cross_vao(), chunk.get_cross_vertices_sum(),
-                     chunk.get_normal_discard_vao(),
-                     chunk.get_normal_discard_vertices_sum(),
-                     chunk.get_normal_blend_vao(),
-                     chunk.get_normal_blend_vertices_sum(),
-                     chunk.get_water_vao(), chunk.get_water_vertices_sum(),
-                     glm::vec3(static_cast<float>(pos.x * CHUNK_SIZE) +
-                                   static_cast<float>(CHUNK_SIZE / 2),
-                               static_cast<float>(WORLD_SIZE_Y / 2),
-                               static_cast<float>(pos.z * CHUNK_SIZE) +
-                                   static_cast<float>(CHUNK_SIZE / 2)),
-                     glm::vec3(static_cast<float>(CHUNK_SIZE / 2),
-                               static_cast<float>(WORLD_SIZE_Y / 2),
-                               static_cast<float>(CHUNK_SIZE / 2))});
-            }
-        }
-    }
-}
-
-void World::push_delete_vbo(GLuint vbo) {
-    std::lock_guard lk(m_delete_vbo_mutex);
-    m_pending_delete_vbo.push_back(vbo);
-}
-
-void World::push_delete_vao(GLuint vao) {
-    std::lock_guard lk(m_delete_vao_mutex);
-    m_pending_delete_vao.push_back(vao);
-}
-
-void World::hot_reload() {
-    auto& config = Config::get();
-    int dist = config.get<int>("world.rendering_distance");
-    m_rendering_distance = dist <= MAX_DISTANCE ? dist : MAX_DISTANCE;
-    need_gen();
-}
-
-void World::rebuild_world() {
-    if (m_is_rebuilding) {
-        return;
-    }
-    m_is_rebuilding = true;
-    stop_gen_thread();
-    stop_thread_pool();
-    m_cave_carcer.reload(ChunkGenerator::seed());
-    m_river_worm.reload(ChunkGenerator::seed());
-    {
-        std::lock_guard lk(m_chunks_mutex);
-        m_chunks.clear();
-        m_new_finished_chunk.clear();
-    }
-    m_could_gen = true;
-    ChunkGenerator::reload();
-    start_thread_pool();
-    start_gen_thread();
-    need_gen();
-
-    m_is_rebuilding = false;
-}
-
-/*
-glm::vec3 World::sunlight_dir() const {
-    float t = static_cast<float>(m_day_tick) / DAY_TIME;
-
-    float azimuth = glm::radians(90.0f - t * 360.0f);
-
-    float altitude =
-        glm::half_pi<float>() * sin((t - 0.25f) * glm::two_pi<float>());
-
-    glm::vec3 dir{cos(altitude) * cos(azimuth), sin(altitude),
-                  cos(altitude) * sin(azimuth)};
-
-    return glm::normalize(-dir);
-}
-*/
-
-glm::vec3 World::sunlight_dir() const {
-    float altitude = sin((m_day_tick - 6 * PER_HOUR) /
-                         static_cast<float>(DAY_TIME / 2) * std::numbers::pi) *
-                     90.0f;
-
-    float t = static_cast<float>(m_day_tick) / DAY_TIME;
-    float azimuth = 90.0f - 360.0f * (t - 0.25f);
-
-    float alt = glm::radians(altitude);
-    float az = glm::radians(azimuth);
-    glm::vec3 dir;
-    dir.x = cos(alt) * sin(az);
-    dir.y = sin(alt);
-    dir.z = cos(alt) * cos(az);
-
-    return glm::normalize(-dir);
-}
-
-int World::rendering_distance() const { return m_rendering_distance.load(); }
-
-void World::rendering_distance(int rendering_distance) {
-    m_rendering_distance = rendering_distance;
-}
-
-CaveCarver& World::cave_carcer() { return m_cave_carcer; }
-RiverWorm& World::river_worm() { return m_river_worm; }
-std::vector<glm::vec4>& World::planes() { return m_planes; }
-std::vector<ChunkRenderSnapshot>& World::render_snapshots() {
-    return m_render_snapshots;
-};
-
-TickType World::game_tick() const { return m_game_ticks.load(); }
-TickType World::day_tick() const { return m_day_tick.load(); }
-void World::day_tick(TickType tick) {
-    tick %= DAY_TIME;
-    m_day_tick = tick;
-}
-int World::per_tick_time() const { return m_per_tick_time.load(); }
-void World::per_tick_time(int ms) { m_per_tick_time = ms; }
-
-bool World::is_tick_running() const { return m_tick_running.load(); }
-void World::tick_running(bool run) { m_tick_running = run; }
-int World::pool_threads() const { return m_pool_threads.load(); }
-int World::max_threads() const { return m_max_threads.load(); }
-void World::change_pool_threads(int threads) {
-    m_max_threads = std::thread::hardware_concurrency();
-    if (m_max_threads < 1) {
-        Logger::warn("Can't Get Max Support Threads, Set Max Threads to 4");
-        m_max_threads = 4;
-    }
-    int used_thread = std::clamp(threads, 1, m_max_threads.load());
-    Logger::info("Create New Thread Pool Use {} Threads", used_thread);
-    m_gen_thread_pool.store(std::make_shared<ThreadPool>(used_thread));
-    m_pool_threads = used_thread;
-}
-int World::chunk_load_style() const {
-    return std::to_underlying(m_chunk_load_style.load());
-}
-void World::set_chunk_load_style(int id) {
-    using enum ChunkLoadStyle;
-
-    switch (id) {
-    case std::to_underlying(RANDOM):
-        m_chunk_load_style = RANDOM;
-        return;
-    case std::to_underlying(CENTER):
-        m_chunk_load_style = CENTER;
-        return;
-    }
-    Logger::error("Can,t Find Chunk Load Style Id {}, Nothing Will Do", id);
-}
-
-ChunkInfo World::get_chunk_info(const glm::vec3& world_pos) const {
-    ChunkPos pos = get_chunk_pos(world_pos.x, world_pos.z);
-
-    std::shared_lock lock(m_chunks_mutex);
-    auto it = m_chunks.find(pos);
-    if (it == m_chunks.end()) {
-        return ChunkInfo{};
-    }
-    return it->second.get_info();
-}
-
-} // namespace Cubed

src/main.cpp

@@ -1,5 +1,14 @@
 #include "Cubed/app.hpp"
 
+#ifdef _WIN32
+
+extern "C" {
+__declspec(dllexport) unsigned long NvOptimusEnablement = 1;
+__declspec(dllexport) int AmdPowerXpressRequestHighPerformance = 1;
+}
+
+#endif
+
 int main(int argc, char** argv) {
 
     static_assert(sizeof(int) == sizeof(int32_t));

src/proto/auth/auth.proto

@@ -0,0 +1,20 @@
+syntax = "proto3";
+
+message LoginReq {
+    string name = 1;
+}
+
+message LoginRsp {
+    bool success = 1;
+    string uuid = 2;
+}
+
+message LogoutReq {
+    string uuid = 1;
+}
+
+message LogoutRsp {
+    string uuid = 1;
+    bool server_stop = 2;
+}
+

src/proto/common/chunk_pos.proto

@@ -0,0 +1,6 @@
+syntax = "proto3";
+
+message ChunkPosNet {
+    int32 x = 1;
+    int32 z = 2;
+}

src/proto/common/error.proto

@@ -0,0 +1,6 @@
+syntax = "proto3";
+
+message Error {
+    int32 code = 1;
+    string mes = 2;
+}

src/proto/common/player_info.proto

@@ -0,0 +1,5 @@
+syntax = "proto3";
+
+message PlayerInfo {
+    string name = 1;
+}

src/proto/common/vector3.proto

@@ -0,0 +1,7 @@
+syntax = "proto3";
+
+message Vec3 {
+    float x = 1;
+    float y = 2;
+    float z = 3;
+}

src/proto/packet.proto

@@ -0,0 +1,12 @@
+syntax = "proto3";
+
+import "common/error.proto";
+import "common/player_info.proto";
+import "common/vector3.proto";
+import "player/player.proto";
+import "auth/auth.proto";
+import "system/ping.proto";
+import "system/pong.proto";
+import "world/chunk_data.proto";
+import "world/block_change.proto";
+import "world/time.proto";

src/proto/player/player.proto

@@ -0,0 +1,14 @@
+syntax = "proto3";
+
+import "common/vector3.proto";
+
+message PlayerPos {
+    string uuid = 1;
+    Vec3 pos = 2;
+}
+
+message PlayerInfoRsp {
+    string uuid = 1;
+    string name = 2;
+    Vec3 pos = 3;
+}

src/proto/system/ping.proto

@@ -0,0 +1,5 @@
+syntax = "proto3";
+
+message Ping {
+    uint64 timestamp = 1;
+}

src/proto/system/pong.proto

@@ -0,0 +1,5 @@
+syntax = "proto3";
+
+message Pong {
+    uint64 timestamp = 1;
+}

src/proto/world/block_change.proto

@@ -0,0 +1,14 @@
+syntax = "proto3";
+
+import "common/vector3.proto";
+
+message BlockChangeReq {
+    string uuid = 1;
+    Vec3 pos = 2;
+    uint32 block = 3;
+}
+
+message BlockChangeRsp {
+    Vec3 pos = 1;
+    uint32 block = 2;
+}

src/proto/world/chunk_data.proto

@@ -0,0 +1,25 @@
+syntax = "proto3";
+
+import "common/chunk_pos.proto";
+
+message ChunkDataReq {
+    int32 task_id = 1;
+    string uuid = 2;
+    ChunkPosNet pos = 3;
+}
+
+message ChunkDataRsp {
+    int32 task_id = 1;
+    ChunkPosNet pos = 2;
+    uint32 chunk_seed = 3;
+    int32 biome_type = 4;
+    repeated uint32 chunk_blocks = 5 [packed=true];
+    repeated uint32 neighbor_blocks_1 = 6 [packed=true];
+    repeated uint32 neighbor_blocks_2 = 7 [packed=true];
+    repeated uint32 neighbor_blocks_3 = 8 [packed=true];
+    repeated uint32 neighbor_blocks_4 = 9 [packed=true];
+}
+
+message S2C_ClearAllChunks {
+    bool clear = 1;
+}

src/proto/world/time.proto

@@ -0,0 +1,6 @@
+syntax = "proto3";
+
+message UpdateTime {
+    int64 game_tick = 1;
+    int64 day_tick = 2;
+}

src/renderer.cpp

@@ -4,8 +4,8 @@
 #include "Cubed/config.hpp"
 #include "Cubed/debug_collector.hpp"
 #include "Cubed/dev_panel.hpp"
-#include "Cubed/gameplay/player.hpp"
-#include "Cubed/gameplay/world.hpp"
+#include "Cubed/gameplay/client_player.hpp"
+#include "Cubed/gameplay/client_world.hpp"
 #include "Cubed/primitive_data.hpp"
 #include "Cubed/texture_manager.hpp"
 #include "Cubed/tools/cubed_assert.hpp"
@@ -21,31 +21,34 @@
 
 namespace Cubed {
 
-Renderer::Renderer(const Camera& camera, World& world,
+Renderer::Renderer(const Camera& camera, ClientWorld& world,
                    const TextureManager& texture_manager, DevPanel& dev_panel)
     : m_camera(camera), m_dev_panel(dev_panel),
       m_texture_manager(texture_manager), m_world(world) {}
 
 Renderer::~Renderer() {
-    glBindBuffer(GL_ARRAY_BUFFER, 0);
-    glDeleteBuffers(1, &m_outline_vbo);
-    glDeleteBuffers(1, &m_outline_indices_vbo);
-    glDeleteBuffers(1, &m_sky_vbo);
-    glDeleteBuffers(1, &m_ui_vbo);
-    glDeleteBuffers(1, &m_text_vbo);
-    glBindVertexArray(0);
-    glDeleteVertexArrays(NUM_VAO, m_vao.data());
-    glDeleteFramebuffers(1, &m_fbo);
-    glDeleteTextures(1, &m_screen_texture);
-    glDeleteTextures(1, &m_screen_depth_texture);
+    if (m_init) {
+        glBindBuffer(GL_ARRAY_BUFFER, 0);
+        glDeleteBuffers(1, &m_outline_vbo);
+        glDeleteBuffers(1, &m_outline_indices_vbo);
+        glDeleteBuffers(1, &m_sky_vbo);
+        glDeleteBuffers(1, &m_ui_vbo);
+        glDeleteBuffers(1, &m_text_vbo);
+        glDeleteBuffers(1, &m_player_vbo);
+        glBindVertexArray(0);
+        glDeleteVertexArrays(NUM_VAO, m_vao.data());
+        glDeleteFramebuffers(1, &m_fbo);
+        glDeleteTextures(1, &m_screen_texture);
+        glDeleteTextures(1, &m_screen_depth_texture);
 
-    glDeleteFramebuffers(1, &m_oit_fbo);
-    glDeleteTextures(1, &m_accum_texture);
-    glDeleteTextures(1, &m_reveal_texture);
-    glDeleteTextures(1, &m_oit_depth_texture);
+        glDeleteFramebuffers(1, &m_oit_fbo);
+        glDeleteTextures(1, &m_accum_texture);
+        glDeleteTextures(1, &m_reveal_texture);
+        glDeleteTextures(1, &m_oit_depth_texture);
 
-    glDeleteFramebuffers(1, &m_depth_map_fbo);
-    glDeleteTextures(1, &m_depth_map_texture);
+        glDeleteFramebuffers(1, &m_depth_map_fbo);
+        glDeleteTextures(1, &m_depth_map_texture);
+    }
 }
 
 void Renderer::hot_reload() {
@@ -86,7 +89,9 @@ void Renderer::init() {
                      "shaders/billboard_f_shader.glsl"};
     Shader water_shader{"water", "shaders/water_v_shader.glsl",
                         "shaders/water_f_shader.glsl"};
-
+    Shader player_shader{"player", "shaders/player_v_shader.glsl",
+                         "shaders/player_f_shader.glsl"};
+    m_shaders.insert({player_shader.hash(), std::move(player_shader)});
     m_shaders.insert({world_shader.hash(), std::move(world_shader)});
     m_shaders.insert({outline_shader.hash(), std::move(outline_shader)});
     m_shaders.insert({sky_shdaer.hash(), std::move(sky_shdaer)});
@@ -100,6 +105,7 @@ void Renderer::init() {
     m_shaders.insert({depth_shader.hash(), std::move(depth_shader)});
     m_shaders.insert({billboard.hash(), std::move(billboard)});
     m_shaders.insert({water_shader.hash(), std::move(water_shader)});
+
     glEnable(GL_DEPTH_TEST);
     glDepthFunc(GL_LEQUAL);
 
@@ -109,9 +115,9 @@ void Renderer::init() {
 #ifdef DEBUG_MODE
     glEnable(GL_DEBUG_OUTPUT);
     glDebugMessageCallback(
-        [](GLenum source, GLenum type, GLuint id, GLenum severity,
-           GLsizei length, const GLchar* message, const void* user_param) {
-            Logger::log(Logger::Level::DEBUG, std::source_location::current(),
+        [](GLenum, GLenum, GLuint, GLenum, GLsizei, const GLchar* message,
+           const void*) {
+            Logger::log(Logger::Level::L_DEBUG, std::source_location::current(),
                         "GL Debug: {}", reinterpret_cast<const char*>(message));
         },
         nullptr);
@@ -165,12 +171,22 @@ void Renderer::init() {
     glEnableVertexAttribArray(1);
     glEnableVertexAttribArray(2);
 
+    glBindVertexArray(m_vao[5]);
+    glGenBuffers(1, &m_player_vbo);
+    glBindBuffer(GL_ARRAY_BUFFER, m_player_vbo);
+    glBufferData(GL_ARRAY_BUFFER, sizeof(VERTICES_PLAYER), VERTICES_PLAYER,
+                 GL_STATIC_DRAW);
+
+    glVertexAttribPointer(0, 3, GL_FLOAT, GL_FALSE, 0, 0);
+    glEnableVertexAttribArray(0);
+
     init_quad();
     init_text();
     hot_reload();
 
     glBindVertexArray(0);
     glBindBuffer(GL_ARRAY_BUFFER, 0);
+    m_init = true;
 }
 
 const Shader& Renderer::get_shader(const std::string& name) const {
@@ -214,7 +230,7 @@ void Renderer::render() {
     render_sky();
     render_world();
     render_outline();
-
+    render_player();
     glBindFramebuffer(GL_FRAMEBUFFER, 0);
     glEnable(GL_FRAMEBUFFER_SRGB);
     glDisable(GL_DEPTH_TEST);
@@ -260,7 +276,7 @@ void Renderer::render_outline() {
     const auto& shader = get_shader("outline");
     shader.use();
 
-    const auto& block_pos = m_world.get_look_block_pos("TestPlayer");
+    const auto& block_pos = m_world.get_look_block_pos();
 
     if (block_pos != std::nullopt) {
 
@@ -660,37 +676,42 @@ void Renderer::render_world() {
         glActiveTexture(GL_TEXTURE1);
         glEnable(GL_DEPTH_TEST);
         for (const auto& snapshot : m_render_snapshots) {
+            if (!snapshot) {
+                continue;
+            }
             glBindTexture(GL_TEXTURE_2D_ARRAY,
                           m_texture_manager.get_texture_array());
-            glBindVertexArray(snapshot.normal_vao);
+            glBindVertexArray(snapshot->normal_vao);
 
-            glDrawArrays(GL_TRIANGLES, 0, snapshot.normal_vertices_count);
+            glDrawArrays(GL_TRIANGLES, 0, snapshot->normal_vertices_count);
         }
 
         // cross_plane and discard
 
         for (const auto& snapshot : m_render_snapshots) {
-
+            if (!snapshot) {
+                continue;
+            }
             glm::vec2 camera_pos_xz{camera_pos.x, camera_pos.z};
-            if (snapshot.cross_vertices_count != 0) {
-                glm::vec2 center_xz{snapshot.center.x, snapshot.center.z};
+            if (snapshot->cross_vertices_count != 0) {
+                glm::vec2 center_xz{snapshot->center.x, snapshot->center.z};
                 float dist2d = glm::distance(camera_pos_xz, center_xz);
                 if (dist2d <= CROSS_PLANE_DISTANCE * 16) {
                     glBindTexture(GL_TEXTURE_2D_ARRAY,
                                   m_texture_manager.get_cross_plane_array());
-                    glBindVertexArray(snapshot.cross_vao);
+                    glBindVertexArray(snapshot->cross_vao);
 
                     glDrawArrays(GL_TRIANGLES, 0,
-                                 snapshot.cross_vertices_count);
+                                 snapshot->cross_vertices_count);
                 }
             }
-            if (snapshot.normal_discard_vertices_count != 0) {
+            if (snapshot->normal_discard_vertices_count != 0) {
                 glBindTexture(GL_TEXTURE_2D_ARRAY,
                               m_texture_manager.get_texture_array());
-                glBindVertexArray(snapshot.normal_discard_vao);
+                glBindVertexArray(snapshot->normal_discard_vao);
 
                 glDrawArrays(GL_TRIANGLES, 0,
-                             snapshot.normal_discard_vertices_count);
+                             snapshot->normal_discard_vertices_count);
             }
         }
     }
@@ -746,28 +767,33 @@ void Renderer::render_world() {
     glBindTexture(GL_TEXTURE_2D_ARRAY, m_texture_manager.get_pbr_texture());
     normal_block_shader.set_loc("enablePBR", m_pbr);
     for (const auto& snapshot : m_render_snapshots) {
-
-        if (Math::is_aabb_in_frustum(snapshot.center, snapshot.half_extents,
+        if (!snapshot) {
+            continue;
+        }
+        if (Math::is_aabb_in_frustum(snapshot->center, snapshot->half_extents,
                                      m_planes)) {
 
-            glBindVertexArray(snapshot.normal_vao);
+            glBindVertexArray(snapshot->normal_vao);
 
-            glDrawArrays(GL_TRIANGLES, 0, snapshot.normal_vertices_count);
+            glDrawArrays(GL_TRIANGLES, 0, snapshot->normal_vertices_count);
 
             rendered_sum++;
         }
     }
     // discard
     for (const auto& snapshot : m_render_snapshots) {
-        if (!Math::is_aabb_in_frustum(snapshot.center, snapshot.half_extents,
+        if (!snapshot) {
+            continue;
+        }
+        if (!Math::is_aabb_in_frustum(snapshot->center, snapshot->half_extents,
                                       m_planes)) {
             continue;
         }
-        if (snapshot.normal_discard_vertices_count != 0) {
-            glBindVertexArray(snapshot.normal_discard_vao);
+        if (snapshot->normal_discard_vertices_count != 0) {
+            glBindVertexArray(snapshot->normal_discard_vao);
 
             glDrawArrays(GL_TRIANGLES, 0,
-                         snapshot.normal_discard_vertices_count);
+                         snapshot->normal_discard_vertices_count);
         }
     }
     // cross_plane
@@ -776,18 +802,21 @@ void Renderer::render_world() {
                   m_texture_manager.get_cross_plane_array());
     normal_block_shader.set_loc("enablePBR", false);
     for (const auto& snapshot : m_render_snapshots) {
-        if (!Math::is_aabb_in_frustum(snapshot.center, snapshot.half_extents,
+        if (!snapshot) {
+            continue;
+        }
+        if (!Math::is_aabb_in_frustum(snapshot->center, snapshot->half_extents,
                                       m_planes)) {
             continue;
         }
         glm::vec2 camera_pos_xz{camera_pos.x, camera_pos.z};
-        if (snapshot.cross_vertices_count != 0) {
-            glm::vec2 center_xz{snapshot.center.x, snapshot.center.z};
+        if (snapshot->cross_vertices_count != 0) {
+            glm::vec2 center_xz{snapshot->center.x, snapshot->center.z};
             float dist2d = glm::distance(camera_pos_xz, center_xz);
             if (dist2d <= CROSS_PLANE_DISTANCE * 16) {
-                glBindVertexArray(snapshot.cross_vao);
+                glBindVertexArray(snapshot->cross_vao);
 
-                glDrawArrays(GL_TRIANGLES, 0, snapshot.cross_vertices_count);
+                glDrawArrays(GL_TRIANGLES, 0, snapshot->cross_vertices_count);
             }
         }
     }
@@ -837,16 +866,20 @@ void Renderer::render_world() {
     glActiveTexture(GL_TEXTURE0);
     glBindTexture(GL_TEXTURE_2D_ARRAY, m_texture_manager.get_texture_array());
     for (const auto& snapshot : m_render_snapshots) {
-        if (!Math::is_aabb_in_frustum(snapshot.center, snapshot.half_extents,
+        if (!snapshot) {
+            continue;
+        }
+        if (!Math::is_aabb_in_frustum(snapshot->center, snapshot->half_extents,
                                       m_planes)) {
             continue;
         }
 
-        if (snapshot.normal_blend_vertices_count != 0) {
+        if (snapshot->normal_blend_vertices_count != 0) {
 
-            glBindVertexArray(snapshot.normal_blend_vao);
+            glBindVertexArray(snapshot->normal_blend_vao);
 
-            glDrawArrays(GL_TRIANGLES, 0, snapshot.normal_blend_vertices_count);
+            glDrawArrays(GL_TRIANGLES, 0,
+                         snapshot->normal_blend_vertices_count);
         }
     }
 
@@ -886,16 +919,19 @@ void Renderer::render_world() {
     glActiveTexture(GL_TEXTURE0);
     glBindTexture(GL_TEXTURE_2D_ARRAY, m_texture_manager.get_texture_array());
     for (const auto& snapshot : m_render_snapshots) {
-        if (!Math::is_aabb_in_frustum(snapshot.center, snapshot.half_extents,
+        if (!snapshot) {
+            continue;
+        }
+        if (!Math::is_aabb_in_frustum(snapshot->center, snapshot->half_extents,
                                       m_planes)) {
             continue;
         }
 
-        if (snapshot.water_vertices_count != 0) {
+        if (snapshot->water_vertices_count != 0) {
 
-            glBindVertexArray(snapshot.water_vao);
+            glBindVertexArray(snapshot->water_vao);
 
-            glDrawArrays(GL_TRIANGLES, 0, snapshot.water_vertices_count);
+            glDrawArrays(GL_TRIANGLES, 0, snapshot->water_vertices_count);
         }
     }
 
@@ -924,6 +960,26 @@ void Renderer::render_world() {
     DebugCollector::get().report(
         "rendered_chunk", "Rendered Chunk: " + std::to_string(rendered_sum));
 }
+
+void Renderer::render_player() {
+    auto& shader = get_shader("player");
+    shader.use();
+    m_v_mat = m_camera.get_camera_lookat();
+
+    auto& players = m_world.render_player_data();
+
+    for (auto& player : players) {
+        m_m_mat = glm::translate(
+            glm::mat4(1.0f), player.render_pos + glm::vec3(-0.5f, 0.0f, -0.5f));
+        m_mv_mat = m_v_mat * m_m_mat;
+        shader.set_loc("mv_matrix", m_mv_mat);
+        shader.set_loc("proj_matrix", m_p_mat);
+        glBindVertexArray(m_vao[5]);
+        glEnable(GL_DEPTH_TEST);
+        glDrawArrays(GL_TRIANGLES, 0, 36);
+    }
+}
+
 #pragma endregion
 void Renderer::render_dev_panel() {
     glDisable(GL_DEPTH_TEST);

src/texture_manager.cpp

@@ -36,14 +36,16 @@ TextureManager::TextureManager() {}
 TextureManager::~TextureManager() { delet_texture(); }
 
 void TextureManager::delet_texture() {
-    glDeleteTextures(1, &m_texture_array);
-    glDeleteTextures(1, &m_block_status_array);
-    glDeleteTextures(1, &m_cross_plane_array);
-    glDeleteTextures(1, &m_normal_texture_array);
-    for (auto& id : m_item_textures) {
-        glDeleteTextures(1, &id);
+    if (m_init) {
+        glDeleteTextures(1, &m_texture_array);
+        glDeleteTextures(1, &m_block_status_array);
+        glDeleteTextures(1, &m_cross_plane_array);
+        glDeleteTextures(1, &m_normal_texture_array);
+        for (auto& id : m_item_textures) {
+            glDeleteTextures(1, &id);
+        }
+        Logger::info("Successfully delete all texture");
     }
-    Logger::info("Successfully delete all texture");
 }
 
 GLuint TextureManager::get_block_status_array() const {
@@ -313,6 +315,7 @@ void TextureManager::init_texture() {
     init_block();
     init_block_status();
     init_ui();
+    m_init = true;
 }
 
 void TextureManager::update() {

src/tools/math_tools.cpp

@@ -1,105 +0,0 @@
-#include "Cubed/tools/math_tools.hpp"
-
-#include <algorithm>
-#include <glm/glm.hpp>
-#include <glm/gtc/type_ptr.hpp>
-
-namespace Cubed {
-
-namespace Math {
-
-void extract_frustum_planes(const glm::mat4& mvp_matrix,
-                            std::vector<glm::vec4>& planes) {
-    if (planes.size() != 6) {
-        planes.resize(6);
-    }
-
-    const float* m = glm::value_ptr(mvp_matrix);
-
-    // left plane
-    planes[0] =
-        glm::vec4(m[3] + m[0], m[7] + m[4], m[11] + m[8], m[15] + m[12]);
-    // right plane
-    planes[1] =
-        glm::vec4(m[3] - m[0], m[7] - m[4], m[11] - m[8], m[15] - m[12]);
-    // bottom plane
-    planes[2] =
-        glm::vec4(m[3] + m[1], m[7] + m[5], m[11] + m[9], m[15] + m[13]);
-    // top plane
-    planes[3] =
-        glm::vec4(m[3] - m[1], m[7] - m[5], m[11] - m[9], m[15] - m[13]);
-    // near plane
-    planes[4] =
-        glm::vec4(m[3] + m[2], m[7] + m[6], m[11] + m[10], m[15] + m[14]);
-    // far plane
-    planes[5] =
-        glm::vec4(m[3] - m[2], m[7] - m[6], m[11] - m[10], m[15] - m[14]);
-
-    for (auto& p : planes) {
-        p = glm::normalize(p);
-    }
-}
-
-float smootherstep(float edge0, float edge1, float x) {
-
-    x = std::clamp((x - edge0) / (edge1 - edge0), 0.0f, 1.0f);
-
-    return x * x * x * (x * (6.0f * x - 15.0f) + 10.0f);
-}
-
-bool is_aabb_in_frustum(const glm::vec3& center, const glm::vec3& half_extents,
-                        const std::vector<glm::vec4>& planes) {
-    for (const auto& plane : planes) {
-        // distance
-        float d = glm::dot(glm::vec3(plane), center) + plane.w;
-        float r = half_extents.x * std::abs(plane.x) +
-                  half_extents.y * std::abs(plane.y) +
-                  half_extents.z * std::abs(plane.z);
-        if (d + r < 0) {
-            return false;
-        }
-    }
-    return true;
-}
-float deterministic_random(int x, int z, uint64_t seed) {
-    uint64_t h = seed;
-    h = h * 6364136223846793005ULL + (uint64_t)x;
-    h = h * 6364136223846793005ULL + (uint64_t)z;
-    return (float)(h >> 40) / (float)(1 << 24);
-}
-
-glm::vec3 slerp(const glm::vec3& from, const glm::vec3& to, float t) {
-
-    float cos_theta = glm::clamp(glm::dot(from, to), -1.0f, 1.0f);
-
-    if (cos_theta > 0.9995f) {
-        return glm::normalize(glm::mix(from, to, t));
-    }
-
-    if (cos_theta < -0.9995f) {
-
-        glm::vec3 axis = (std::fabs(from.x) < 0.9f)
-                             ? glm::vec3(1.0f, 0.0f, 0.0f)
-                             : glm::vec3(0.0f, 1.0f, 0.0f);
-        glm::vec3 ortho = glm::normalize(glm::cross(from, axis));
-
-        float angle = glm::pi<float>() * t;
-
-        glm::vec3 rotated =
-            from * std::cos(angle) + glm::cross(ortho, from) * std::sin(angle);
-
-        return glm::normalize(rotated);
-    }
-
-    float theta = std::acos(cos_theta);
-    float sin_theta = std::sin(theta);
-
-    float a = std::sin((1.0f - t) * theta) / sin_theta;
-    float b = std::sin(t * theta) / sin_theta;
-
-    return glm::normalize(a * from + b * to);
-}
-
-} // namespace Math
-
-} // namespace Cubed

src/version.hpp.in

@@ -0,0 +1,3 @@
+#pragma once
+
+#define CUBED_VERSION "@CUBED_VERSION@"

src/window.cpp

@@ -18,10 +18,14 @@ static int windowed_width = 800, windowed_height = 600;
 Window::Window(Renderer& renderer) : m_renderer(renderer) {}
 
 Window::~Window() {
+    if (m_imgui_init) {
+        ImGui_ImplOpenGL3_Shutdown();
+        ImGui_ImplGlfw_Shutdown();
+    }
 
-    ImGui_ImplOpenGL3_Shutdown();
-    ImGui_ImplGlfw_Shutdown();
-    ImGui::DestroyContext();
+    if (ImGui::GetCurrentContext() != nullptr) {
+        ImGui::DestroyContext();
+    }
 
     if (m_window) {
         glfwDestroyWindow(m_window);
@@ -212,6 +216,8 @@ void Window::imgui_init() {
     // Setup Platform/Renderer backends
     ImGui_ImplGlfw_InitForOpenGL(m_window, false);
     ImGui_ImplOpenGL3_Init();
+
+    m_imgui_init = true;
 }
 
 } // namespace Cubed

third_party/asio/include/.gitignore

@@ -0,0 +1,2 @@
+Makefile
+Makefile.in

third_party/asio/include/Makefile.am

@@ -0,0 +1,640 @@
+# find . -name "*.*pp" | sed -e 's/^\.\///' | sed -e 's/^.*$/  & \\/' | sort
+nobase_include_HEADERS = \
+	asio/any_completion_executor.hpp \
+	asio/any_completion_handler.hpp \
+	asio/any_io_executor.hpp \
+	asio/append.hpp \
+	asio/as_tuple.hpp \
+	asio/associated_allocator.hpp \
+	asio/associated_cancellation_slot.hpp \
+	asio/associated_executor.hpp \
+	asio/associated_immediate_executor.hpp \
+	asio/associator.hpp \
+	asio/async_result.hpp \
+	asio/awaitable.hpp \
+	asio/basic_datagram_socket.hpp \
+	asio/basic_deadline_timer.hpp \
+	asio/basic_file.hpp \
+	asio/basic_io_object.hpp \
+	asio/basic_random_access_file.hpp \
+	asio/basic_raw_socket.hpp \
+	asio/basic_readable_pipe.hpp \
+	asio/basic_seq_packet_socket.hpp \
+	asio/basic_serial_port.hpp \
+	asio/basic_signal_set.hpp \
+	asio/basic_socket_acceptor.hpp \
+	asio/basic_socket.hpp \
+	asio/basic_socket_iostream.hpp \
+	asio/basic_socket_streambuf.hpp \
+	asio/basic_streambuf_fwd.hpp \
+	asio/basic_streambuf.hpp \
+	asio/basic_stream_file.hpp \
+	asio/basic_stream_socket.hpp \
+	asio/basic_waitable_timer.hpp \
+	asio/basic_writable_pipe.hpp \
+	asio/bind_allocator.hpp \
+	asio/bind_cancellation_slot.hpp \
+	asio/bind_executor.hpp \
+	asio/bind_immediate_executor.hpp \
+	asio/buffered_read_stream_fwd.hpp \
+	asio/buffered_read_stream.hpp \
+	asio/buffered_stream_fwd.hpp \
+	asio/buffered_stream.hpp \
+	asio/buffered_write_stream_fwd.hpp \
+	asio/buffered_write_stream.hpp \
+	asio/buffer.hpp \
+	asio/buffer_registration.hpp \
+	asio/buffers_iterator.hpp \
+	asio/cancel_after.hpp \
+	asio/cancel_at.hpp \
+	asio/cancellation_signal.hpp \
+	asio/cancellation_state.hpp \
+	asio/cancellation_type.hpp \
+	asio/co_composed.hpp \
+	asio/co_spawn.hpp \
+	asio/completion_condition.hpp \
+	asio/compose.hpp \
+	asio/composed.hpp \
+	asio/config.hpp \
+	asio/connect.hpp \
+	asio/connect_pipe.hpp \
+	asio/consign.hpp \
+	asio/coroutine.hpp \
+	asio/deadline_timer.hpp \
+	asio/defer.hpp \
+	asio/deferred.hpp \
+	asio/default_completion_token.hpp \
+	asio/detached.hpp \
+	asio/detail/array_fwd.hpp \
+	asio/detail/array.hpp \
+	asio/detail/assert.hpp \
+	asio/detail/atomic_count.hpp \
+	asio/detail/atomic_slim_mutex.hpp \
+	asio/detail/base_from_cancellation_state.hpp \
+	asio/detail/base_from_completion_cond.hpp \
+	asio/detail/bind_handler.hpp \
+	asio/detail/blocking_executor_op.hpp \
+	asio/detail/buffered_stream_storage.hpp \
+	asio/detail/buffer_resize_guard.hpp \
+	asio/detail/buffer_sequence_adapter.hpp \
+	asio/detail/call_stack.hpp \
+	asio/detail/chrono.hpp \
+	asio/detail/chrono_time_traits.hpp \
+	asio/detail/completion_handler.hpp \
+	asio/detail/completion_message.hpp \
+	asio/detail/completion_payload.hpp \
+	asio/detail/completion_payload_handler.hpp \
+	asio/detail/composed_work.hpp \
+	asio/detail/concurrency_hint.hpp \
+	asio/detail/conditionally_enabled_event.hpp \
+	asio/detail/conditionally_enabled_mutex.hpp \
+	asio/detail/config.hpp \
+	asio/detail/consuming_buffers.hpp \
+	asio/detail/cstddef.hpp \
+	asio/detail/cstdint.hpp \
+	asio/detail/date_time_fwd.hpp \
+	asio/detail/deadline_timer_service.hpp \
+	asio/detail/dependent_type.hpp \
+	asio/detail/descriptor_ops.hpp \
+	asio/detail/descriptor_read_op.hpp \
+	asio/detail/descriptor_write_op.hpp \
+	asio/detail/dev_poll_reactor.hpp \
+	asio/detail/epoll_reactor.hpp \
+	asio/detail/eventfd_select_interrupter.hpp \
+	asio/detail/event.hpp \
+	asio/detail/exception.hpp \
+	asio/detail/executor_function.hpp \
+	asio/detail/executor_op.hpp \
+	asio/detail/fd_set_adapter.hpp \
+	asio/detail/fenced_block.hpp \
+	asio/detail/functional.hpp \
+	asio/detail/future.hpp \
+	asio/detail/global.hpp \
+	asio/detail/handler_alloc_helpers.hpp \
+	asio/detail/handler_cont_helpers.hpp \
+	asio/detail/handler_tracking.hpp \
+	asio/detail/handler_type_requirements.hpp \
+	asio/detail/handler_work.hpp \
+	asio/detail/hash_map.hpp \
+	asio/detail/impl/buffer_sequence_adapter.ipp \
+	asio/detail/impl/descriptor_ops.ipp \
+	asio/detail/impl/dev_poll_reactor.hpp \
+	asio/detail/impl/dev_poll_reactor.ipp \
+	asio/detail/impl/epoll_reactor.hpp \
+	asio/detail/impl/epoll_reactor.ipp \
+	asio/detail/impl/eventfd_select_interrupter.ipp \
+	asio/detail/impl/handler_tracking.ipp \
+	asio/detail/impl/io_uring_descriptor_service.ipp \
+	asio/detail/impl/io_uring_file_service.ipp \
+	asio/detail/impl/io_uring_service.hpp \
+	asio/detail/impl/io_uring_service.ipp \
+	asio/detail/impl/io_uring_socket_service_base.ipp \
+	asio/detail/impl/kqueue_reactor.hpp \
+	asio/detail/impl/kqueue_reactor.ipp \
+	asio/detail/impl/null_event.ipp \
+	asio/detail/impl/pipe_select_interrupter.ipp \
+	asio/detail/impl/posix_event.ipp \
+	asio/detail/impl/posix_mutex.ipp \
+	asio/detail/impl/posix_serial_port_service.ipp \
+	asio/detail/impl/posix_thread.ipp \
+	asio/detail/impl/posix_tss_ptr.ipp \
+	asio/detail/impl/reactive_descriptor_service.ipp \
+	asio/detail/impl/reactive_socket_service_base.ipp \
+	asio/detail/impl/resolver_service_base.ipp \
+	asio/detail/impl/resolver_thread_pool.ipp \
+	asio/detail/impl/scheduler.ipp \
+	asio/detail/impl/select_reactor.hpp \
+	asio/detail/impl/select_reactor.ipp \
+	asio/detail/impl/service_registry.hpp \
+	asio/detail/impl/service_registry.ipp \
+	asio/detail/impl/signal_set_service.ipp \
+	asio/detail/impl/socket_ops.ipp \
+	asio/detail/impl/socket_select_interrupter.ipp \
+	asio/detail/impl/strand_executor_service.hpp \
+	asio/detail/impl/strand_executor_service.ipp \
+	asio/detail/impl/strand_service.hpp \
+	asio/detail/impl/strand_service.ipp \
+	asio/detail/impl/thread_context.ipp \
+	asio/detail/impl/throw_error.ipp \
+	asio/detail/impl/timer_queue_set.ipp \
+	asio/detail/impl/win_critsec_mutex.ipp \
+	asio/detail/impl/win_event.ipp \
+	asio/detail/impl/win_iocp_file_service.ipp \
+	asio/detail/impl/win_iocp_handle_service.ipp \
+	asio/detail/impl/win_iocp_io_context.hpp \
+	asio/detail/impl/win_iocp_io_context.ipp \
+	asio/detail/impl/win_iocp_serial_port_service.ipp \
+	asio/detail/impl/win_iocp_socket_service_base.ipp \
+	asio/detail/impl/win_object_handle_service.ipp \
+	asio/detail/impl/winrt_ssocket_service_base.ipp \
+	asio/detail/impl/winrt_timer_scheduler.hpp \
+	asio/detail/impl/winrt_timer_scheduler.ipp \
+	asio/detail/impl/winsock_init.ipp \
+	asio/detail/impl/win_static_mutex.ipp \
+	asio/detail/impl/win_thread.ipp \
+	asio/detail/impl/win_tss_ptr.ipp \
+	asio/detail/initiate_defer.hpp \
+	asio/detail/initiate_dispatch.hpp \
+	asio/detail/initiate_post.hpp \
+	asio/detail/initiation_base.hpp \
+	asio/detail/io_control.hpp \
+	asio/detail/io_object_impl.hpp \
+	asio/detail/io_uring_descriptor_read_at_op.hpp \
+	asio/detail/io_uring_descriptor_read_op.hpp \
+	asio/detail/io_uring_descriptor_service.hpp \
+	asio/detail/io_uring_descriptor_write_at_op.hpp \
+	asio/detail/io_uring_descriptor_write_op.hpp \
+	asio/detail/io_uring_file_service.hpp \
+	asio/detail/io_uring_null_buffers_op.hpp \
+	asio/detail/io_uring_operation.hpp \
+	asio/detail/io_uring_service.hpp \
+	asio/detail/io_uring_socket_accept_op.hpp \
+	asio/detail/io_uring_socket_connect_op.hpp \
+	asio/detail/io_uring_socket_recvfrom_op.hpp \
+	asio/detail/io_uring_socket_recvmsg_op.hpp \
+	asio/detail/io_uring_socket_recv_op.hpp \
+	asio/detail/io_uring_socket_send_op.hpp \
+	asio/detail/io_uring_socket_sendto_op.hpp \
+	asio/detail/io_uring_socket_service_base.hpp \
+	asio/detail/io_uring_socket_service.hpp \
+	asio/detail/io_uring_wait_op.hpp \
+	asio/detail/is_buffer_sequence.hpp \
+	asio/detail/is_executor.hpp \
+	asio/detail/keyword_tss_ptr.hpp \
+	asio/detail/kqueue_reactor.hpp \
+	asio/detail/limits.hpp \
+	asio/detail/local_free_on_block_exit.hpp \
+	asio/detail/memory.hpp \
+	asio/detail/mutex.hpp \
+	asio/detail/non_const_lvalue.hpp \
+	asio/detail/noncopyable.hpp \
+	asio/detail/null_event.hpp \
+	asio/detail/null_fenced_block.hpp \
+	asio/detail/null_global.hpp \
+	asio/detail/null_mutex.hpp \
+	asio/detail/null_reactor.hpp \
+	asio/detail/null_signal_blocker.hpp \
+	asio/detail/null_socket_service.hpp \
+	asio/detail/null_static_mutex.hpp \
+	asio/detail/null_thread.hpp \
+	asio/detail/null_tss_ptr.hpp \
+	asio/detail/object_pool.hpp \
+	asio/detail/old_win_sdk_compat.hpp \
+	asio/detail/operation.hpp \
+	asio/detail/op_queue.hpp \
+	asio/detail/pipe_select_interrupter.hpp \
+	asio/detail/pop_options.hpp \
+	asio/detail/posix_event.hpp \
+	asio/detail/posix_fd_set_adapter.hpp \
+	asio/detail/posix_global.hpp \
+	asio/detail/posix_mutex.hpp \
+	asio/detail/posix_serial_port_service.hpp \
+	asio/detail/posix_signal_blocker.hpp \
+	asio/detail/posix_static_mutex.hpp \
+	asio/detail/posix_thread.hpp \
+	asio/detail/posix_tss_ptr.hpp \
+	asio/detail/push_options.hpp \
+	asio/detail/reactive_descriptor_service.hpp \
+	asio/detail/reactive_null_buffers_op.hpp \
+	asio/detail/reactive_socket_accept_op.hpp \
+	asio/detail/reactive_socket_connect_op.hpp \
+	asio/detail/reactive_socket_recvfrom_op.hpp \
+	asio/detail/reactive_socket_recvmsg_op.hpp \
+	asio/detail/reactive_socket_recv_op.hpp \
+	asio/detail/reactive_socket_send_op.hpp \
+	asio/detail/reactive_socket_sendto_op.hpp \
+	asio/detail/reactive_socket_service_base.hpp \
+	asio/detail/reactive_socket_service.hpp \
+	asio/detail/reactive_wait_op.hpp \
+	asio/detail/reactor.hpp \
+	asio/detail/reactor_op.hpp \
+	asio/detail/reactor_op_queue.hpp \
+	asio/detail/recycling_allocator.hpp \
+	asio/detail/regex_fwd.hpp \
+	asio/detail/resolve_endpoint_op.hpp \
+	asio/detail/resolve_op.hpp \
+	asio/detail/resolve_query_op.hpp \
+	asio/detail/resolver_service_base.hpp \
+	asio/detail/resolver_thread_pool.hpp \
+	asio/detail/resolver_service.hpp \
+	asio/detail/scheduler.hpp \
+	asio/detail/scheduler_operation.hpp \
+	asio/detail/scheduler_task.hpp \
+	asio/detail/scheduler_thread_info.hpp \
+	asio/detail/scoped_lock.hpp \
+	asio/detail/scoped_ptr.hpp \
+	asio/detail/select_interrupter.hpp \
+	asio/detail/select_reactor.hpp \
+	asio/detail/service_registry.hpp \
+	asio/detail/signal_blocker.hpp \
+	asio/detail/signal_handler.hpp \
+	asio/detail/signal_init.hpp \
+	asio/detail/signal_op.hpp \
+	asio/detail/signal_set_service.hpp \
+	asio/detail/slim_mutex.hpp \
+	asio/detail/socket_holder.hpp \
+	asio/detail/socket_ops.hpp \
+	asio/detail/socket_option.hpp \
+	asio/detail/socket_select_interrupter.hpp \
+	asio/detail/socket_types.hpp \
+	asio/detail/source_location.hpp \
+	asio/detail/static_mutex.hpp \
+	asio/detail/std_event.hpp \
+	asio/detail/std_fenced_block.hpp \
+	asio/detail/std_global.hpp \
+	asio/detail/std_mutex.hpp \
+	asio/detail/std_static_mutex.hpp \
+	asio/detail/std_thread.hpp \
+	asio/detail/strand_executor_service.hpp \
+	asio/detail/strand_service.hpp \
+	asio/detail/string_view.hpp \
+	asio/detail/thread_context.hpp \
+	asio/detail/thread_group.hpp \
+	asio/detail/thread.hpp \
+	asio/detail/thread_info_base.hpp \
+	asio/detail/throw_error.hpp \
+	asio/detail/throw_exception.hpp \
+	asio/detail/timed_cancel_op.hpp \
+	asio/detail/timer_queue_base.hpp \
+	asio/detail/timer_queue.hpp \
+	asio/detail/timer_queue_set.hpp \
+	asio/detail/timer_scheduler_fwd.hpp \
+	asio/detail/timer_scheduler.hpp \
+	asio/detail/tss_ptr.hpp \
+	asio/detail/type_traits.hpp \
+	asio/detail/utility.hpp \
+	asio/detail/wait_handler.hpp \
+	asio/detail/wait_op.hpp \
+	asio/detail/winapp_thread.hpp \
+	asio/detail/wince_thread.hpp \
+	asio/detail/win_critsec_mutex.hpp \
+	asio/detail/win_event.hpp \
+	asio/detail/win_fd_set_adapter.hpp \
+	asio/detail/win_global.hpp \
+	asio/detail/win_iocp_file_service.hpp \
+	asio/detail/win_iocp_handle_read_op.hpp \
+	asio/detail/win_iocp_handle_service.hpp \
+	asio/detail/win_iocp_handle_write_op.hpp \
+	asio/detail/win_iocp_io_context.hpp \
+	asio/detail/win_iocp_null_buffers_op.hpp \
+	asio/detail/win_iocp_operation.hpp \
+	asio/detail/win_iocp_overlapped_op.hpp \
+	asio/detail/win_iocp_overlapped_ptr.hpp \
+	asio/detail/win_iocp_serial_port_service.hpp \
+	asio/detail/win_iocp_socket_accept_op.hpp \
+	asio/detail/win_iocp_socket_connect_op.hpp \
+	asio/detail/win_iocp_socket_recvfrom_op.hpp \
+	asio/detail/win_iocp_socket_recvmsg_op.hpp \
+	asio/detail/win_iocp_socket_recv_op.hpp \
+	asio/detail/win_iocp_socket_send_op.hpp \
+	asio/detail/win_iocp_socket_service_base.hpp \
+	asio/detail/win_iocp_socket_service.hpp \
+	asio/detail/win_iocp_thread_info.hpp \
+	asio/detail/win_iocp_wait_op.hpp \
+	asio/detail/win_mutex.hpp \
+	asio/detail/win_object_handle_service.hpp \
+	asio/detail/winrt_async_manager.hpp \
+	asio/detail/winrt_async_op.hpp \
+	asio/detail/winrt_resolve_op.hpp \
+	asio/detail/winrt_resolver_service.hpp \
+	asio/detail/winrt_socket_connect_op.hpp \
+	asio/detail/winrt_socket_recv_op.hpp \
+	asio/detail/winrt_socket_send_op.hpp \
+	asio/detail/winrt_ssocket_service_base.hpp \
+	asio/detail/winrt_ssocket_service.hpp \
+	asio/detail/winrt_timer_scheduler.hpp \
+	asio/detail/winrt_utils.hpp \
+	asio/detail/winsock_init.hpp \
+	asio/detail/win_static_mutex.hpp \
+	asio/detail/win_thread.hpp \
+	asio/detail/win_tss_ptr.hpp \
+	asio/detail/work_dispatcher.hpp \
+	asio/detail/wrapped_handler.hpp \
+	asio/dispatch.hpp \
+	asio/disposition.hpp \
+	asio/error_code.hpp \
+	asio/error.hpp \
+	asio/execution.hpp \
+	asio/execution_context.hpp \
+	asio/execution/allocator.hpp \
+	asio/execution/any_executor.hpp \
+	asio/execution/bad_executor.hpp \
+	asio/execution/blocking.hpp \
+	asio/execution/blocking_adaptation.hpp \
+	asio/execution/context.hpp \
+	asio/execution/context_as.hpp \
+	asio/execution/executor.hpp \
+	asio/execution/impl/bad_executor.ipp \
+	asio/execution/inline_exception_handling.hpp \
+	asio/execution/invocable_archetype.hpp \
+	asio/execution/mapping.hpp \
+	asio/execution/occupancy.hpp \
+	asio/execution/outstanding_work.hpp \
+	asio/execution/prefer_only.hpp \
+	asio/execution/relationship.hpp \
+	asio/executor.hpp \
+	asio/executor_work_guard.hpp \
+	asio/experimental/as_single.hpp \
+	asio/experimental/awaitable_operators.hpp \
+	asio/experimental/basic_channel.hpp \
+	asio/experimental/basic_concurrent_channel.hpp \
+	asio/experimental/cancellation_condition.hpp \
+	asio/experimental/channel.hpp \
+	asio/experimental/channel_error.hpp \
+	asio/experimental/channel_traits.hpp \
+	asio/experimental/co_composed.hpp \
+	asio/experimental/co_spawn.hpp \
+	asio/experimental/concurrent_channel.hpp \
+	asio/experimental/coro.hpp \
+	asio/experimental/coro_traits.hpp \
+	asio/experimental/detail/channel_operation.hpp \
+	asio/experimental/detail/channel_receive_op.hpp \
+	asio/experimental/detail/channel_send_functions.hpp \
+	asio/experimental/detail/channel_send_op.hpp \
+	asio/experimental/detail/channel_service.hpp \
+	asio/experimental/detail/coro_completion_handler.hpp \
+	asio/experimental/detail/coro_promise_allocator.hpp \
+	asio/experimental/detail/has_signature.hpp \
+	asio/experimental/detail/impl/channel_service.hpp \
+	asio/experimental/detail/partial_promise.hpp \
+	asio/experimental/impl/as_single.hpp \
+	asio/experimental/impl/channel_error.ipp \
+	asio/experimental/impl/coro.hpp \
+	asio/experimental/impl/parallel_group.hpp \
+	asio/experimental/impl/promise.hpp \
+	asio/experimental/impl/use_coro.hpp \
+	asio/experimental/impl/use_promise.hpp \
+	asio/experimental/parallel_group.hpp \
+	asio/experimental/promise.hpp \
+	asio/experimental/use_coro.hpp \
+	asio/experimental/use_promise.hpp \
+	asio/file_base.hpp \
+	asio/generic/basic_endpoint.hpp \
+	asio/generic/datagram_protocol.hpp \
+	asio/generic/detail/endpoint.hpp \
+	asio/generic/detail/impl/endpoint.ipp \
+	asio/generic/raw_protocol.hpp \
+	asio/generic/seq_packet_protocol.hpp \
+	asio/generic/stream_protocol.hpp \
+	asio/handler_continuation_hook.hpp \
+	asio/high_resolution_timer.hpp \
+	asio.hpp \
+	asio/immediate.hpp \
+	asio/impl/any_completion_executor.ipp \
+	asio/impl/any_io_executor.ipp \
+	asio/impl/append.hpp \
+	asio/impl/as_tuple.hpp \
+	asio/impl/awaitable.hpp \
+	asio/impl/awaitable.ipp \
+	asio/impl/buffered_read_stream.hpp \
+	asio/impl/buffered_write_stream.hpp \
+	asio/impl/cancel_after.hpp \
+	asio/impl/cancel_at.hpp \
+	asio/impl/cancellation_signal.ipp \
+	asio/impl/co_spawn.hpp \
+	asio/impl/config.hpp \
+	asio/impl/config.ipp \
+	asio/impl/connect.hpp \
+	asio/impl/connect_pipe.hpp \
+	asio/impl/connect_pipe.ipp \
+	asio/impl/consign.hpp \
+	asio/impl/deferred.hpp \
+	asio/impl/detached.hpp \
+	asio/impl/error_code.ipp \
+	asio/impl/error.ipp \
+	asio/impl/execution_context.hpp \
+	asio/impl/execution_context.ipp \
+	asio/impl/executor.hpp \
+	asio/impl/executor.ipp \
+	asio/impl/io_context.hpp \
+	asio/impl/io_context.ipp \
+	asio/impl/multiple_exceptions.ipp \
+	asio/impl/prepend.hpp \
+	asio/impl/read_at.hpp \
+	asio/impl/read.hpp \
+	asio/impl/read_until.hpp \
+	asio/impl/redirect_disposition.hpp \
+	asio/impl/redirect_error.hpp \
+	asio/impl/serial_port_base.hpp \
+	asio/impl/serial_port_base.ipp \
+	asio/impl/spawn.hpp \
+	asio/impl/src.hpp \
+	asio/impl/system_context.hpp \
+	asio/impl/system_context.ipp \
+	asio/impl/system_executor.hpp \
+	asio/impl/thread_pool.hpp \
+	asio/impl/thread_pool.ipp \
+	asio/impl/use_awaitable.hpp \
+	asio/impl/use_future.hpp \
+	asio/impl/write_at.hpp \
+	asio/impl/write.hpp \
+	asio/inline_executor.hpp \
+	asio/inline_or_executor.hpp \
+	asio/io_context.hpp \
+	asio/io_context_strand.hpp \
+	asio/ip/address.hpp \
+	asio/ip/address_v4.hpp \
+	asio/ip/address_v4_iterator.hpp \
+	asio/ip/address_v4_range.hpp \
+	asio/ip/address_v6.hpp \
+	asio/ip/address_v6_iterator.hpp \
+	asio/ip/address_v6_range.hpp \
+	asio/ip/bad_address_cast.hpp \
+	asio/ip/basic_endpoint.hpp \
+	asio/ip/basic_resolver_entry.hpp \
+	asio/ip/basic_resolver.hpp \
+	asio/ip/basic_resolver_iterator.hpp \
+	asio/ip/basic_resolver_query.hpp \
+	asio/ip/basic_resolver_results.hpp \
+	asio/ip/detail/endpoint.hpp \
+	asio/ip/detail/impl/endpoint.ipp \
+	asio/ip/detail/socket_option.hpp \
+	asio/ip/host_name.hpp \
+	asio/ip/icmp.hpp \
+	asio/ip/impl/address.hpp \
+	asio/ip/impl/address.ipp \
+	asio/ip/impl/address_v4.hpp \
+	asio/ip/impl/address_v4.ipp \
+	asio/ip/impl/address_v6.hpp \
+	asio/ip/impl/address_v6.ipp \
+	asio/ip/impl/basic_endpoint.hpp \
+	asio/ip/impl/host_name.ipp \
+	asio/ip/impl/network_v4.hpp \
+	asio/ip/impl/network_v4.ipp \
+	asio/ip/impl/network_v6.hpp \
+	asio/ip/impl/network_v6.ipp \
+	asio/ip/multicast.hpp \
+	asio/ip/network_v4.hpp \
+	asio/ip/network_v6.hpp \
+	asio/ip/resolver_base.hpp \
+	asio/ip/resolver_query_base.hpp \
+	asio/ip/tcp.hpp \
+	asio/ip/udp.hpp \
+	asio/ip/unicast.hpp \
+	asio/ip/v6_only.hpp \
+	asio/is_applicable_property.hpp \
+	asio/is_contiguous_iterator.hpp \
+	asio/is_executor.hpp \
+	asio/is_read_buffered.hpp \
+	asio/is_write_buffered.hpp \
+	asio/local/basic_endpoint.hpp \
+	asio/local/connect_pair.hpp \
+	asio/local/datagram_protocol.hpp \
+	asio/local/detail/endpoint.hpp \
+	asio/local/detail/impl/endpoint.ipp \
+	asio/local/seq_packet_protocol.hpp \
+	asio/local/stream_protocol.hpp \
+	asio/multiple_exceptions.hpp \
+	asio/packaged_task.hpp \
+	asio/placeholders.hpp \
+	asio/posix/basic_descriptor.hpp \
+	asio/posix/basic_stream_descriptor.hpp \
+	asio/posix/descriptor_base.hpp \
+	asio/posix/descriptor.hpp \
+	asio/posix/stream_descriptor.hpp \
+	asio/post.hpp \
+	asio/prefer.hpp \
+	asio/prepend.hpp \
+	asio/query.hpp \
+	asio/random_access_file.hpp \
+	asio/read_at.hpp \
+	asio/read.hpp \
+	asio/read_until.hpp \
+	asio/readable_pipe.hpp \
+	asio/recycling_allocator.hpp \
+	asio/redirect_disposition.hpp \
+	asio/redirect_error.hpp \
+	asio/registered_buffer.hpp \
+	asio/require.hpp \
+	asio/require_concept.hpp \
+	asio/serial_port_base.hpp \
+	asio/serial_port.hpp \
+	asio/signal_set_base.hpp \
+	asio/signal_set.hpp \
+	asio/socket_base.hpp \
+	asio/spawn.hpp \
+	asio/ssl/context_base.hpp \
+	asio/ssl/context.hpp \
+	asio/ssl/detail/buffered_handshake_op.hpp \
+	asio/ssl/detail/engine.hpp \
+	asio/ssl/detail/handshake_op.hpp \
+	asio/ssl/detail/impl/engine.ipp \
+	asio/ssl/detail/impl/openssl_init.ipp \
+	asio/ssl/detail/io.hpp \
+	asio/ssl/detail/openssl_init.hpp \
+	asio/ssl/detail/openssl_types.hpp \
+	asio/ssl/detail/password_callback.hpp \
+	asio/ssl/detail/read_op.hpp \
+	asio/ssl/detail/shutdown_op.hpp \
+	asio/ssl/detail/stream_core.hpp \
+	asio/ssl/detail/verify_callback.hpp \
+	asio/ssl/detail/write_op.hpp \
+	asio/ssl/error.hpp \
+	asio/ssl.hpp \
+	asio/ssl/host_name_verification.hpp \
+	asio/ssl/impl/context.hpp \
+	asio/ssl/impl/context.ipp \
+	asio/ssl/impl/error.ipp \
+	asio/ssl/impl/host_name_verification.ipp \
+	asio/ssl/impl/src.hpp \
+	asio/ssl/stream_base.hpp \
+	asio/ssl/stream.hpp \
+	asio/ssl/verify_context.hpp \
+	asio/ssl/verify_mode.hpp \
+	asio/static_thread_pool.hpp \
+	asio/steady_timer.hpp \
+	asio/strand.hpp \
+	asio/streambuf.hpp \
+	asio/stream_file.hpp \
+	asio/system_context.hpp \
+	asio/system_error.hpp \
+	asio/system_executor.hpp \
+	asio/system_timer.hpp \
+	asio/this_coro.hpp \
+	asio/thread.hpp \
+	asio/thread_pool.hpp \
+	asio/time_traits.hpp \
+	asio/traits/equality_comparable.hpp \
+	asio/traits/execute_member.hpp \
+	asio/traits/prefer_free.hpp \
+	asio/traits/prefer_member.hpp \
+	asio/traits/query_free.hpp \
+	asio/traits/query_member.hpp \
+	asio/traits/query_static_constexpr_member.hpp \
+	asio/traits/require_concept_free.hpp \
+	asio/traits/require_concept_member.hpp \
+	asio/traits/require_free.hpp \
+	asio/traits/require_member.hpp \
+	asio/traits/static_query.hpp \
+	asio/traits/static_require.hpp \
+	asio/traits/static_require_concept.hpp \
+	asio/ts/buffer.hpp \
+	asio/ts/executor.hpp \
+	asio/ts/internet.hpp \
+	asio/ts/io_context.hpp \
+	asio/ts/netfwd.hpp \
+	asio/ts/net.hpp \
+	asio/ts/socket.hpp \
+	asio/ts/timer.hpp \
+	asio/unyield.hpp \
+	asio/use_awaitable.hpp \
+	asio/use_future.hpp \
+	asio/uses_executor.hpp \
+	asio/version.hpp \
+	asio/wait_traits.hpp \
+	asio/windows/basic_object_handle.hpp \
+	asio/windows/basic_overlapped_handle.hpp \
+	asio/windows/basic_random_access_handle.hpp \
+	asio/windows/basic_stream_handle.hpp \
+	asio/windows/object_handle.hpp \
+	asio/windows/overlapped_handle.hpp \
+	asio/windows/overlapped_ptr.hpp \
+	asio/windows/random_access_handle.hpp \
+	asio/windows/stream_handle.hpp \
+	asio/writable_pipe.hpp \
+	asio/write_at.hpp \
+	asio/write.hpp \
+	asio/yield.hpp
+
+MAINTAINERCLEANFILES = \
+	$(srcdir)/Makefile.in

third_party/asio/include/asio.hpp

@@ -0,0 +1,205 @@
+//
+// asio.hpp
+// ~~~~~~~~
+//
+// Copyright (c) 2003-2026 Christopher M. Kohlhoff (chris at kohlhoff dot com)
+//
+// Distributed under the Boost Software License, Version 1.0. (See accompanying
+// file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
+//
+
+#ifndef ASIO_HPP
+#define ASIO_HPP
+
+#if defined(_MSC_VER) && (_MSC_VER >= 1200)
+# pragma once
+#endif // defined(_MSC_VER) && (_MSC_VER >= 1200)
+
+#include "asio/any_completion_executor.hpp"
+#include "asio/any_completion_handler.hpp"
+#include "asio/any_io_executor.hpp"
+#include "asio/append.hpp"
+#include "asio/as_tuple.hpp"
+#include "asio/associated_allocator.hpp"
+#include "asio/associated_cancellation_slot.hpp"
+#include "asio/associated_executor.hpp"
+#include "asio/associated_immediate_executor.hpp"
+#include "asio/associator.hpp"
+#include "asio/async_result.hpp"
+#include "asio/awaitable.hpp"
+#include "asio/basic_datagram_socket.hpp"
+#include "asio/basic_file.hpp"
+#include "asio/basic_io_object.hpp"
+#include "asio/basic_random_access_file.hpp"
+#include "asio/basic_raw_socket.hpp"
+#include "asio/basic_readable_pipe.hpp"
+#include "asio/basic_seq_packet_socket.hpp"
+#include "asio/basic_serial_port.hpp"
+#include "asio/basic_signal_set.hpp"
+#include "asio/basic_socket.hpp"
+#include "asio/basic_socket_acceptor.hpp"
+#include "asio/basic_socket_iostream.hpp"
+#include "asio/basic_socket_streambuf.hpp"
+#include "asio/basic_stream_file.hpp"
+#include "asio/basic_stream_socket.hpp"
+#include "asio/basic_streambuf.hpp"
+#include "asio/basic_waitable_timer.hpp"
+#include "asio/basic_writable_pipe.hpp"
+#include "asio/bind_allocator.hpp"
+#include "asio/bind_cancellation_slot.hpp"
+#include "asio/bind_executor.hpp"
+#include "asio/bind_immediate_executor.hpp"
+#include "asio/buffer.hpp"
+#include "asio/buffer_registration.hpp"
+#include "asio/buffered_read_stream_fwd.hpp"
+#include "asio/buffered_read_stream.hpp"
+#include "asio/buffered_stream_fwd.hpp"
+#include "asio/buffered_stream.hpp"
+#include "asio/buffered_write_stream_fwd.hpp"
+#include "asio/buffered_write_stream.hpp"
+#include "asio/buffers_iterator.hpp"
+#include "asio/cancel_after.hpp"
+#include "asio/cancel_at.hpp"
+#include "asio/cancellation_signal.hpp"
+#include "asio/cancellation_state.hpp"
+#include "asio/cancellation_type.hpp"
+#include "asio/co_composed.hpp"
+#include "asio/co_spawn.hpp"
+#include "asio/completion_condition.hpp"
+#include "asio/compose.hpp"
+#include "asio/composed.hpp"
+#include "asio/config.hpp"
+#include "asio/connect.hpp"
+#include "asio/connect_pipe.hpp"
+#include "asio/consign.hpp"
+#include "asio/coroutine.hpp"
+#include "asio/defer.hpp"
+#include "asio/deferred.hpp"
+#include "asio/default_completion_token.hpp"
+#include "asio/detached.hpp"
+#include "asio/dispatch.hpp"
+#include "asio/disposition.hpp"
+#include "asio/error.hpp"
+#include "asio/error_code.hpp"
+#include "asio/execution.hpp"
+#include "asio/execution/allocator.hpp"
+#include "asio/execution/any_executor.hpp"
+#include "asio/execution/blocking.hpp"
+#include "asio/execution/blocking_adaptation.hpp"
+#include "asio/execution/context.hpp"
+#include "asio/execution/context_as.hpp"
+#include "asio/execution/executor.hpp"
+#include "asio/execution/invocable_archetype.hpp"
+#include "asio/execution/mapping.hpp"
+#include "asio/execution/occupancy.hpp"
+#include "asio/execution/outstanding_work.hpp"
+#include "asio/execution/prefer_only.hpp"
+#include "asio/execution/relationship.hpp"
+#include "asio/executor.hpp"
+#include "asio/executor_work_guard.hpp"
+#include "asio/file_base.hpp"
+#include "asio/generic/basic_endpoint.hpp"
+#include "asio/generic/datagram_protocol.hpp"
+#include "asio/generic/raw_protocol.hpp"
+#include "asio/generic/seq_packet_protocol.hpp"
+#include "asio/generic/stream_protocol.hpp"
+#include "asio/handler_continuation_hook.hpp"
+#include "asio/high_resolution_timer.hpp"
+#include "asio/immediate.hpp"
+#include "asio/inline_executor.hpp"
+#include "asio/inline_or_executor.hpp"
+#include "asio/io_context.hpp"
+#include "asio/io_context_strand.hpp"
+#include "asio/ip/address.hpp"
+#include "asio/ip/address_v4.hpp"
+#include "asio/ip/address_v4_iterator.hpp"
+#include "asio/ip/address_v4_range.hpp"
+#include "asio/ip/address_v6.hpp"
+#include "asio/ip/address_v6_iterator.hpp"
+#include "asio/ip/address_v6_range.hpp"
+#include "asio/ip/network_v4.hpp"
+#include "asio/ip/network_v6.hpp"
+#include "asio/ip/bad_address_cast.hpp"
+#include "asio/ip/basic_endpoint.hpp"
+#include "asio/ip/basic_resolver.hpp"
+#include "asio/ip/basic_resolver_entry.hpp"
+#include "asio/ip/basic_resolver_iterator.hpp"
+#include "asio/ip/basic_resolver_query.hpp"
+#include "asio/ip/host_name.hpp"
+#include "asio/ip/icmp.hpp"
+#include "asio/ip/multicast.hpp"
+#include "asio/ip/resolver_base.hpp"
+#include "asio/ip/resolver_query_base.hpp"
+#include "asio/ip/tcp.hpp"
+#include "asio/ip/udp.hpp"
+#include "asio/ip/unicast.hpp"
+#include "asio/ip/v6_only.hpp"
+#include "asio/is_applicable_property.hpp"
+#include "asio/is_contiguous_iterator.hpp"
+#include "asio/is_executor.hpp"
+#include "asio/is_read_buffered.hpp"
+#include "asio/is_write_buffered.hpp"
+#include "asio/local/basic_endpoint.hpp"
+#include "asio/local/connect_pair.hpp"
+#include "asio/local/datagram_protocol.hpp"
+#include "asio/local/seq_packet_protocol.hpp"
+#include "asio/local/stream_protocol.hpp"
+#include "asio/multiple_exceptions.hpp"
+#include "asio/packaged_task.hpp"
+#include "asio/placeholders.hpp"
+#include "asio/posix/basic_descriptor.hpp"
+#include "asio/posix/basic_stream_descriptor.hpp"
+#include "asio/posix/descriptor.hpp"
+#include "asio/posix/descriptor_base.hpp"
+#include "asio/posix/stream_descriptor.hpp"
+#include "asio/post.hpp"
+#include "asio/prefer.hpp"
+#include "asio/prepend.hpp"
+#include "asio/query.hpp"
+#include "asio/random_access_file.hpp"
+#include "asio/read.hpp"
+#include "asio/read_at.hpp"
+#include "asio/read_until.hpp"
+#include "asio/readable_pipe.hpp"
+#include "asio/recycling_allocator.hpp"
+#include "asio/redirect_disposition.hpp"
+#include "asio/redirect_error.hpp"
+#include "asio/registered_buffer.hpp"
+#include "asio/require.hpp"
+#include "asio/require_concept.hpp"
+#include "asio/serial_port.hpp"
+#include "asio/serial_port_base.hpp"
+#include "asio/signal_set.hpp"
+#include "asio/signal_set_base.hpp"
+#include "asio/socket_base.hpp"
+#include "asio/static_thread_pool.hpp"
+#include "asio/steady_timer.hpp"
+#include "asio/strand.hpp"
+#include "asio/stream_file.hpp"
+#include "asio/streambuf.hpp"
+#include "asio/system_context.hpp"
+#include "asio/system_error.hpp"
+#include "asio/system_executor.hpp"
+#include "asio/system_timer.hpp"
+#include "asio/this_coro.hpp"
+#include "asio/thread.hpp"
+#include "asio/thread_pool.hpp"
+#include "asio/use_awaitable.hpp"
+#include "asio/use_future.hpp"
+#include "asio/uses_executor.hpp"
+#include "asio/version.hpp"
+#include "asio/wait_traits.hpp"
+#include "asio/windows/basic_object_handle.hpp"
+#include "asio/windows/basic_overlapped_handle.hpp"
+#include "asio/windows/basic_random_access_handle.hpp"
+#include "asio/windows/basic_stream_handle.hpp"
+#include "asio/windows/object_handle.hpp"
+#include "asio/windows/overlapped_handle.hpp"
+#include "asio/windows/overlapped_ptr.hpp"
+#include "asio/windows/random_access_handle.hpp"
+#include "asio/windows/stream_handle.hpp"
+#include "asio/writable_pipe.hpp"
+#include "asio/write.hpp"
+#include "asio/write_at.hpp"
+
+#endif // ASIO_HPP

third_party/asio/include/asio/any_completion_executor.hpp

@@ -0,0 +1,338 @@
+//
+// any_completion_executor.hpp
+// ~~~~~~~~~~~~~~~~~~~~~~~~~~~
+//
+// Copyright (c) 2003-2026 Christopher M. Kohlhoff (chris at kohlhoff dot com)
+//
+// Distributed under the Boost Software License, Version 1.0. (See accompanying
+// file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
+//
+
+#ifndef ASIO_ANY_COMPLETION_EXECUTOR_HPP
+#define ASIO_ANY_COMPLETION_EXECUTOR_HPP
+
+#if defined(_MSC_VER) && (_MSC_VER >= 1200)
+# pragma once
+#endif // defined(_MSC_VER) && (_MSC_VER >= 1200)
+
+#include "asio/detail/config.hpp"
+#if defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+# include "asio/executor.hpp"
+#else // defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+# include "asio/execution.hpp"
+#endif // defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+
+#include "asio/detail/push_options.hpp"
+
+namespace asio {
+ASIO_INLINE_NAMESPACE_BEGIN
+
+#if defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+
+typedef executor any_completion_executor;
+
+#else // defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+
+/// Polymorphic executor type for use with I/O objects.
+/**
+ * The @c any_completion_executor type is a polymorphic executor that supports
+ * the set of properties required for the execution of completion handlers. It
+ * is defined as the execution::any_executor class template parameterised as
+ * follows:
+ * @code execution::any_executor<
+ *   execution::prefer_only<execution::outstanding_work_t::tracked_t>,
+ *   execution::prefer_only<execution::outstanding_work_t::untracked_t>
+ *   execution::prefer_only<execution::relationship_t::fork_t>,
+ *   execution::prefer_only<execution::relationship_t::continuation_t>
+ * > @endcode
+ */
+class any_completion_executor :
+#if defined(GENERATING_DOCUMENTATION)
+  public execution::any_executor<...>
+#else // defined(GENERATING_DOCUMENTATION)
+  public execution::any_executor<
+      execution::prefer_only<execution::outstanding_work_t::tracked_t>,
+      execution::prefer_only<execution::outstanding_work_t::untracked_t>,
+      execution::prefer_only<execution::relationship_t::fork_t>,
+      execution::prefer_only<execution::relationship_t::continuation_t>
+    >
+#endif // defined(GENERATING_DOCUMENTATION)
+{
+public:
+#if !defined(GENERATING_DOCUMENTATION)
+  typedef execution::any_executor<
+      execution::prefer_only<execution::outstanding_work_t::tracked_t>,
+      execution::prefer_only<execution::outstanding_work_t::untracked_t>,
+      execution::prefer_only<execution::relationship_t::fork_t>,
+      execution::prefer_only<execution::relationship_t::continuation_t>
+    > base_type;
+
+  typedef void supportable_properties_type(
+      execution::prefer_only<execution::outstanding_work_t::tracked_t>,
+      execution::prefer_only<execution::outstanding_work_t::untracked_t>,
+      execution::prefer_only<execution::relationship_t::fork_t>,
+      execution::prefer_only<execution::relationship_t::continuation_t>
+    );
+#endif // !defined(GENERATING_DOCUMENTATION)
+
+  /// Default constructor.
+  ASIO_DECL any_completion_executor() noexcept;
+
+  /// Construct in an empty state. Equivalent effects to default constructor.
+  ASIO_DECL any_completion_executor(nullptr_t) noexcept;
+
+  /// Copy constructor.
+  ASIO_DECL any_completion_executor(
+      const any_completion_executor& e) noexcept;
+
+  /// Move constructor.
+  ASIO_DECL any_completion_executor(
+      any_completion_executor&& e) noexcept;
+
+  /// Construct to point to the same target as another any_executor.
+#if defined(GENERATING_DOCUMENTATION)
+  template <class... OtherSupportableProperties>
+    any_completion_executor(
+        execution::any_executor<OtherSupportableProperties...> e);
+#else // defined(GENERATING_DOCUMENTATION)
+  template <typename OtherAnyExecutor>
+  any_completion_executor(OtherAnyExecutor e,
+      constraint_t<
+        conditional<
+          !is_same<OtherAnyExecutor, any_completion_executor>::value
+            && is_base_of<execution::detail::any_executor_base,
+              OtherAnyExecutor>::value,
+          typename execution::detail::supportable_properties<
+            0, supportable_properties_type>::template
+              is_valid_target<OtherAnyExecutor>,
+          false_type
+        >::type::value
+      > = 0)
+    : base_type(static_cast<OtherAnyExecutor&&>(e))
+  {
+  }
+#endif // defined(GENERATING_DOCUMENTATION)
+
+  /// Construct to point to the same target as another any_executor.
+#if defined(GENERATING_DOCUMENTATION)
+  template <class... OtherSupportableProperties>
+    any_completion_executor(std::nothrow_t,
+      execution::any_executor<OtherSupportableProperties...> e);
+#else // defined(GENERATING_DOCUMENTATION)
+  template <typename OtherAnyExecutor>
+  any_completion_executor(std::nothrow_t, OtherAnyExecutor e,
+      constraint_t<
+        conditional<
+          !is_same<OtherAnyExecutor, any_completion_executor>::value
+            && is_base_of<execution::detail::any_executor_base,
+              OtherAnyExecutor>::value,
+          typename execution::detail::supportable_properties<
+            0, supportable_properties_type>::template
+              is_valid_target<OtherAnyExecutor>,
+          false_type
+        >::type::value
+      > = 0) noexcept
+    : base_type(std::nothrow, static_cast<OtherAnyExecutor&&>(e))
+  {
+  }
+#endif // defined(GENERATING_DOCUMENTATION)
+
+  /// Construct to point to the same target as another any_executor.
+  ASIO_DECL any_completion_executor(std::nothrow_t,
+      const any_completion_executor& e) noexcept;
+
+  /// Construct to point to the same target as another any_executor.
+  ASIO_DECL any_completion_executor(std::nothrow_t,
+      any_completion_executor&& e) noexcept;
+
+  /// Construct a polymorphic wrapper for the specified executor.
+#if defined(GENERATING_DOCUMENTATION)
+  template <ASIO_EXECUTION_EXECUTOR Executor>
+  any_completion_executor(Executor e);
+#else // defined(GENERATING_DOCUMENTATION)
+  template <ASIO_EXECUTION_EXECUTOR Executor>
+  any_completion_executor(Executor e,
+      constraint_t<
+        conditional<
+          !is_same<Executor, any_completion_executor>::value
+            && !is_base_of<execution::detail::any_executor_base,
+              Executor>::value,
+          execution::detail::is_valid_target_executor<
+            Executor, supportable_properties_type>,
+          false_type
+        >::type::value
+      > = 0)
+    : base_type(static_cast<Executor&&>(e))
+  {
+  }
+#endif // defined(GENERATING_DOCUMENTATION)
+
+  /// Construct a polymorphic wrapper for the specified executor.
+#if defined(GENERATING_DOCUMENTATION)
+  template <ASIO_EXECUTION_EXECUTOR Executor>
+  any_completion_executor(std::nothrow_t, Executor e);
+#else // defined(GENERATING_DOCUMENTATION)
+  template <ASIO_EXECUTION_EXECUTOR Executor>
+  any_completion_executor(std::nothrow_t, Executor e,
+      constraint_t<
+        conditional<
+          !is_same<Executor, any_completion_executor>::value
+            && !is_base_of<execution::detail::any_executor_base,
+              Executor>::value,
+          execution::detail::is_valid_target_executor<
+            Executor, supportable_properties_type>,
+          false_type
+        >::type::value
+      > = 0) noexcept
+    : base_type(std::nothrow, static_cast<Executor&&>(e))
+  {
+  }
+#endif // defined(GENERATING_DOCUMENTATION)
+
+  /// Assignment operator.
+  ASIO_DECL any_completion_executor& operator=(
+      const any_completion_executor& e) noexcept;
+
+  /// Move assignment operator.
+  ASIO_DECL any_completion_executor& operator=(
+      any_completion_executor&& e) noexcept;
+
+  /// Assignment operator that sets the polymorphic wrapper to the empty state.
+  ASIO_DECL any_completion_executor& operator=(nullptr_t);
+
+  /// Destructor.
+  ASIO_DECL ~any_completion_executor();
+
+  /// Swap targets with another polymorphic wrapper.
+  ASIO_DECL void swap(any_completion_executor& other) noexcept;
+
+  /// Obtain a polymorphic wrapper with the specified property.
+  /**
+   * Do not call this function directly. It is intended for use with the
+   * asio::require and asio::prefer customisation points.
+   *
+   * For example:
+   * @code any_completion_executor ex = ...;
+   * auto ex2 = asio::require(ex, execution::relationship.fork); @endcode
+   */
+  template <typename Property>
+  any_completion_executor require(const Property& p,
+      constraint_t<
+        traits::require_member<const base_type&, const Property&>::is_valid
+      > = 0) const
+  {
+    return static_cast<const base_type&>(*this).require(p);
+  }
+
+  /// Obtain a polymorphic wrapper with the specified property.
+  /**
+   * Do not call this function directly. It is intended for use with the
+   * asio::prefer customisation point.
+   *
+   * For example:
+   * @code any_completion_executor ex = ...;
+   * auto ex2 = asio::prefer(ex, execution::relationship.fork); @endcode
+   */
+  template <typename Property>
+  any_completion_executor prefer(const Property& p,
+      constraint_t<
+        traits::prefer_member<const base_type&, const Property&>::is_valid
+      > = 0) const
+  {
+    return static_cast<const base_type&>(*this).prefer(p);
+  }
+};
+
+#if !defined(GENERATING_DOCUMENTATION)
+
+template <>
+ASIO_DECL any_completion_executor any_completion_executor::prefer(
+    const execution::outstanding_work_t::tracked_t&, int) const;
+
+template <>
+ASIO_DECL any_completion_executor any_completion_executor::prefer(
+    const execution::outstanding_work_t::untracked_t&, int) const;
+
+template <>
+ASIO_DECL any_completion_executor any_completion_executor::prefer(
+    const execution::relationship_t::fork_t&, int) const;
+
+template <>
+ASIO_DECL any_completion_executor any_completion_executor::prefer(
+    const execution::relationship_t::continuation_t&, int) const;
+
+namespace traits {
+
+#if !defined(ASIO_HAS_DEDUCED_EQUALITY_COMPARABLE_TRAIT)
+
+template <>
+struct equality_comparable<any_completion_executor>
+{
+  static const bool is_valid = true;
+  static const bool is_noexcept = true;
+};
+
+#endif // !defined(ASIO_HAS_DEDUCED_EQUALITY_COMPARABLE_TRAIT)
+
+#if !defined(ASIO_HAS_DEDUCED_EXECUTE_MEMBER_TRAIT)
+
+template <typename F>
+struct execute_member<any_completion_executor, F>
+{
+  static const bool is_valid = true;
+  static const bool is_noexcept = false;
+  typedef void result_type;
+};
+
+#endif // !defined(ASIO_HAS_DEDUCED_EXECUTE_MEMBER_TRAIT)
+
+#if !defined(ASIO_HAS_DEDUCED_QUERY_MEMBER_TRAIT)
+
+template <typename Prop>
+struct query_member<any_completion_executor, Prop> :
+  query_member<any_completion_executor::base_type, Prop>
+{
+};
+
+#endif // !defined(ASIO_HAS_DEDUCED_QUERY_MEMBER_TRAIT)
+
+#if !defined(ASIO_HAS_DEDUCED_REQUIRE_MEMBER_TRAIT)
+
+template <typename Prop>
+struct require_member<any_completion_executor, Prop> :
+  require_member<any_completion_executor::base_type, Prop>
+{
+  typedef any_completion_executor result_type;
+};
+
+#endif // !defined(ASIO_HAS_DEDUCED_REQUIRE_MEMBER_TRAIT)
+
+#if !defined(ASIO_HAS_DEDUCED_PREFER_MEMBER_TRAIT)
+
+template <typename Prop>
+struct prefer_member<any_completion_executor, Prop> :
+  prefer_member<any_completion_executor::base_type, Prop>
+{
+  typedef any_completion_executor result_type;
+};
+
+#endif // !defined(ASIO_HAS_DEDUCED_PREFER_MEMBER_TRAIT)
+
+} // namespace traits
+
+#endif // !defined(GENERATING_DOCUMENTATION)
+
+#endif // defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+
+ASIO_INLINE_NAMESPACE_END
+} // namespace asio
+
+#include "asio/detail/pop_options.hpp"
+
+#if defined(ASIO_HEADER_ONLY) \
+  && !defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+# include "asio/impl/any_completion_executor.ipp"
+#endif // defined(ASIO_HEADER_ONLY)
+       //   && !defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+
+#endif // ASIO_ANY_COMPLETION_EXECUTOR_HPP

third_party/asio/include/asio/any_completion_handler.hpp

@@ -0,0 +1,824 @@
+//
+// any_completion_handler.hpp
+// ~~~~~~~~~~~~~~~~~~~~~~~~~~
+//
+// Copyright (c) 2003-2026 Christopher M. Kohlhoff (chris at kohlhoff dot com)
+//
+// Distributed under the Boost Software License, Version 1.0. (See accompanying
+// file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
+//
+
+#ifndef ASIO_ANY_COMPLETION_HANDLER_HPP
+#define ASIO_ANY_COMPLETION_HANDLER_HPP
+
+#include "asio/detail/config.hpp"
+#include <cstring>
+#include <functional>
+#include <memory>
+#include <utility>
+#include "asio/any_completion_executor.hpp"
+#include "asio/any_io_executor.hpp"
+#include "asio/associated_allocator.hpp"
+#include "asio/associated_cancellation_slot.hpp"
+#include "asio/associated_executor.hpp"
+#include "asio/associated_immediate_executor.hpp"
+#include "asio/cancellation_state.hpp"
+#include "asio/recycling_allocator.hpp"
+
+#include "asio/detail/push_options.hpp"
+
+namespace asio {
+ASIO_INLINE_NAMESPACE_BEGIN
+namespace detail {
+
+class any_completion_handler_impl_base
+{
+public:
+  template <typename S>
+  explicit any_completion_handler_impl_base(S&& slot)
+    : cancel_state_(static_cast<S&&>(slot), enable_total_cancellation())
+  {
+  }
+
+  cancellation_slot get_cancellation_slot() const noexcept
+  {
+    return cancel_state_.slot();
+  }
+
+private:
+  cancellation_state cancel_state_;
+};
+
+template <typename Handler>
+class any_completion_handler_impl :
+  public any_completion_handler_impl_base
+{
+public:
+  template <typename S, typename H>
+  any_completion_handler_impl(S&& slot, H&& h)
+    : any_completion_handler_impl_base(static_cast<S&&>(slot)),
+      handler_(static_cast<H&&>(h))
+  {
+  }
+
+  struct uninit_deleter
+  {
+    typename std::allocator_traits<
+      associated_allocator_t<Handler,
+        asio::recycling_allocator<void>>>::template
+          rebind_alloc<any_completion_handler_impl> alloc;
+
+    void operator()(any_completion_handler_impl* ptr)
+    {
+      std::allocator_traits<decltype(alloc)>::deallocate(alloc, ptr, 1);
+    }
+  };
+
+  struct deleter
+  {
+    typename std::allocator_traits<
+      associated_allocator_t<Handler,
+        asio::recycling_allocator<void>>>::template
+          rebind_alloc<any_completion_handler_impl> alloc;
+
+    void operator()(any_completion_handler_impl* ptr)
+    {
+      std::allocator_traits<decltype(alloc)>::destroy(alloc, ptr);
+      std::allocator_traits<decltype(alloc)>::deallocate(alloc, ptr, 1);
+    }
+  };
+
+  template <typename S, typename H>
+  static any_completion_handler_impl* create(S&& slot, H&& h)
+  {
+    uninit_deleter d{
+        (get_associated_allocator)(h,
+          asio::recycling_allocator<void>())};
+
+    std::unique_ptr<any_completion_handler_impl, uninit_deleter> uninit_ptr(
+        std::allocator_traits<decltype(d.alloc)>::allocate(d.alloc, 1), d);
+
+    any_completion_handler_impl* ptr =
+      new (uninit_ptr.get()) any_completion_handler_impl(
+        static_cast<S&&>(slot), static_cast<H&&>(h));
+
+    uninit_ptr.release();
+    return ptr;
+  }
+
+  void destroy()
+  {
+    deleter d{
+        (get_associated_allocator)(handler_,
+          asio::recycling_allocator<void>())};
+
+    d(this);
+  }
+
+  any_completion_executor executor(
+      const any_completion_executor& candidate) const noexcept
+  {
+    return any_completion_executor(std::nothrow,
+        (get_associated_executor)(handler_, candidate));
+  }
+
+  any_completion_executor immediate_executor(
+      const any_io_executor& candidate) const noexcept
+  {
+    return any_completion_executor(std::nothrow,
+        (get_associated_immediate_executor)(handler_, candidate));
+  }
+
+  void* allocate(std::size_t size, std::size_t align_size) const
+  {
+    typename std::allocator_traits<
+      associated_allocator_t<Handler,
+        asio::recycling_allocator<void>>>::template
+          rebind_alloc<unsigned char> alloc(
+            (get_associated_allocator)(handler_,
+              asio::recycling_allocator<void>()));
+
+    std::size_t space = size + align_size - 1;
+    unsigned char* base =
+      std::allocator_traits<decltype(alloc)>::allocate(
+        alloc, space + sizeof(std::ptrdiff_t));
+
+    void* p = base;
+    if (detail::align(align_size, size, p, space))
+    {
+      std::ptrdiff_t off = static_cast<unsigned char*>(p) - base;
+      std::memcpy(static_cast<unsigned char*>(p) + size, &off, sizeof(off));
+      return p;
+    }
+
+    std::bad_alloc ex;
+    asio::detail::throw_exception(ex);
+    return nullptr;
+  }
+
+  void deallocate(void* p, std::size_t size, std::size_t align) const
+  {
+    if (p)
+    {
+      typename std::allocator_traits<
+        associated_allocator_t<Handler,
+          asio::recycling_allocator<void>>>::template
+            rebind_alloc<unsigned char> alloc(
+              (get_associated_allocator)(handler_,
+                asio::recycling_allocator<void>()));
+
+      std::ptrdiff_t off;
+      std::memcpy(&off, static_cast<unsigned char*>(p) + size, sizeof(off));
+      unsigned char* base = static_cast<unsigned char*>(p) - off;
+
+      std::allocator_traits<decltype(alloc)>::deallocate(
+          alloc, base, size + align -1 + sizeof(std::ptrdiff_t));
+    }
+  }
+
+  template <typename... Args>
+  void call(Args&&... args)
+  {
+    deleter d{
+        (get_associated_allocator)(handler_,
+          asio::recycling_allocator<void>())};
+
+    std::unique_ptr<any_completion_handler_impl, deleter> ptr(this, d);
+    Handler handler(static_cast<Handler&&>(handler_));
+    ptr.reset();
+
+    static_cast<Handler&&>(handler)(
+        static_cast<Args&&>(args)...);
+  }
+
+private:
+  Handler handler_;
+};
+
+template <typename Signature>
+class any_completion_handler_call_fn;
+
+template <typename R, typename... Args>
+class any_completion_handler_call_fn<R(Args...)>
+{
+public:
+  using type = void(*)(any_completion_handler_impl_base*, Args...);
+
+  constexpr any_completion_handler_call_fn(type fn)
+    : call_fn_(fn)
+  {
+  }
+
+  void call(any_completion_handler_impl_base* impl, Args... args) const
+  {
+    call_fn_(impl, static_cast<Args&&>(args)...);
+  }
+
+  template <typename Handler>
+  static void impl(any_completion_handler_impl_base* impl, Args... args)
+  {
+    static_cast<any_completion_handler_impl<Handler>*>(impl)->call(
+        static_cast<Args&&>(args)...);
+  }
+
+private:
+  type call_fn_;
+};
+
+template <typename... Signatures>
+class any_completion_handler_call_fns;
+
+template <typename Signature>
+class any_completion_handler_call_fns<Signature> :
+  public any_completion_handler_call_fn<Signature>
+{
+public:
+  using any_completion_handler_call_fn<
+    Signature>::any_completion_handler_call_fn;
+  using any_completion_handler_call_fn<Signature>::call;
+};
+
+template <typename Signature, typename... Signatures>
+class any_completion_handler_call_fns<Signature, Signatures...> :
+  public any_completion_handler_call_fn<Signature>,
+  public any_completion_handler_call_fns<Signatures...>
+{
+public:
+  template <typename CallFn, typename... CallFns>
+  constexpr any_completion_handler_call_fns(CallFn fn, CallFns... fns)
+    : any_completion_handler_call_fn<Signature>(fn),
+      any_completion_handler_call_fns<Signatures...>(fns...)
+  {
+  }
+
+  using any_completion_handler_call_fn<Signature>::call;
+  using any_completion_handler_call_fns<Signatures...>::call;
+};
+
+class any_completion_handler_destroy_fn
+{
+public:
+  using type = void(*)(any_completion_handler_impl_base*);
+
+  constexpr any_completion_handler_destroy_fn(type fn)
+    : destroy_fn_(fn)
+  {
+  }
+
+  void destroy(any_completion_handler_impl_base* impl) const
+  {
+    destroy_fn_(impl);
+  }
+
+  template <typename Handler>
+  static void impl(any_completion_handler_impl_base* impl)
+  {
+    static_cast<any_completion_handler_impl<Handler>*>(impl)->destroy();
+  }
+
+private:
+  type destroy_fn_;
+};
+
+class any_completion_handler_executor_fn
+{
+public:
+  using type = any_completion_executor(*)(
+      any_completion_handler_impl_base*, const any_completion_executor&);
+
+  constexpr any_completion_handler_executor_fn(type fn)
+    : executor_fn_(fn)
+  {
+  }
+
+  any_completion_executor executor(any_completion_handler_impl_base* impl,
+      const any_completion_executor& candidate) const
+  {
+    return executor_fn_(impl, candidate);
+  }
+
+  template <typename Handler>
+  static any_completion_executor impl(any_completion_handler_impl_base* impl,
+      const any_completion_executor& candidate)
+  {
+    return static_cast<any_completion_handler_impl<Handler>*>(impl)->executor(
+        candidate);
+  }
+
+private:
+  type executor_fn_;
+};
+
+class any_completion_handler_immediate_executor_fn
+{
+public:
+  using type = any_completion_executor(*)(
+      any_completion_handler_impl_base*, const any_io_executor&);
+
+  constexpr any_completion_handler_immediate_executor_fn(type fn)
+    : immediate_executor_fn_(fn)
+  {
+  }
+
+  any_completion_executor immediate_executor(
+      any_completion_handler_impl_base* impl,
+      const any_io_executor& candidate) const
+  {
+    return immediate_executor_fn_(impl, candidate);
+  }
+
+  template <typename Handler>
+  static any_completion_executor impl(any_completion_handler_impl_base* impl,
+      const any_io_executor& candidate)
+  {
+    return static_cast<any_completion_handler_impl<Handler>*>(
+        impl)->immediate_executor(candidate);
+  }
+
+private:
+  type immediate_executor_fn_;
+};
+
+class any_completion_handler_allocate_fn
+{
+public:
+  using type = void*(*)(any_completion_handler_impl_base*,
+      std::size_t, std::size_t);
+
+  constexpr any_completion_handler_allocate_fn(type fn)
+    : allocate_fn_(fn)
+  {
+  }
+
+  void* allocate(any_completion_handler_impl_base* impl,
+      std::size_t size, std::size_t align) const
+  {
+    return allocate_fn_(impl, size, align);
+  }
+
+  template <typename Handler>
+  static void* impl(any_completion_handler_impl_base* impl,
+      std::size_t size, std::size_t align)
+  {
+    return static_cast<any_completion_handler_impl<Handler>*>(impl)->allocate(
+        size, align);
+  }
+
+private:
+  type allocate_fn_;
+};
+
+class any_completion_handler_deallocate_fn
+{
+public:
+  using type = void(*)(any_completion_handler_impl_base*,
+      void*, std::size_t, std::size_t);
+
+  constexpr any_completion_handler_deallocate_fn(type fn)
+    : deallocate_fn_(fn)
+  {
+  }
+
+  void deallocate(any_completion_handler_impl_base* impl,
+      void* p, std::size_t size, std::size_t align) const
+  {
+    deallocate_fn_(impl, p, size, align);
+  }
+
+  template <typename Handler>
+  static void impl(any_completion_handler_impl_base* impl,
+      void* p, std::size_t size, std::size_t align)
+  {
+    static_cast<any_completion_handler_impl<Handler>*>(impl)->deallocate(
+        p, size, align);
+  }
+
+private:
+  type deallocate_fn_;
+};
+
+template <typename... Signatures>
+class any_completion_handler_fn_table
+  : private any_completion_handler_destroy_fn,
+    private any_completion_handler_executor_fn,
+    private any_completion_handler_immediate_executor_fn,
+    private any_completion_handler_allocate_fn,
+    private any_completion_handler_deallocate_fn,
+    private any_completion_handler_call_fns<Signatures...>
+{
+public:
+  template <typename... CallFns>
+  constexpr any_completion_handler_fn_table(
+      any_completion_handler_destroy_fn::type destroy_fn,
+      any_completion_handler_executor_fn::type executor_fn,
+      any_completion_handler_immediate_executor_fn::type immediate_executor_fn,
+      any_completion_handler_allocate_fn::type allocate_fn,
+      any_completion_handler_deallocate_fn::type deallocate_fn,
+      CallFns... call_fns)
+    : any_completion_handler_destroy_fn(destroy_fn),
+      any_completion_handler_executor_fn(executor_fn),
+      any_completion_handler_immediate_executor_fn(immediate_executor_fn),
+      any_completion_handler_allocate_fn(allocate_fn),
+      any_completion_handler_deallocate_fn(deallocate_fn),
+      any_completion_handler_call_fns<Signatures...>(call_fns...)
+  {
+  }
+
+  using any_completion_handler_destroy_fn::destroy;
+  using any_completion_handler_executor_fn::executor;
+  using any_completion_handler_immediate_executor_fn::immediate_executor;
+  using any_completion_handler_allocate_fn::allocate;
+  using any_completion_handler_deallocate_fn::deallocate;
+  using any_completion_handler_call_fns<Signatures...>::call;
+};
+
+template <typename Handler, typename... Signatures>
+struct any_completion_handler_fn_table_instance
+{
+  static constexpr any_completion_handler_fn_table<Signatures...>
+    value = any_completion_handler_fn_table<Signatures...>(
+        &any_completion_handler_destroy_fn::impl<Handler>,
+        &any_completion_handler_executor_fn::impl<Handler>,
+        &any_completion_handler_immediate_executor_fn::impl<Handler>,
+        &any_completion_handler_allocate_fn::impl<Handler>,
+        &any_completion_handler_deallocate_fn::impl<Handler>,
+        &any_completion_handler_call_fn<Signatures>::template impl<Handler>...);
+};
+
+template <typename Handler, typename... Signatures>
+constexpr any_completion_handler_fn_table<Signatures...>
+any_completion_handler_fn_table_instance<Handler, Signatures...>::value;
+
+} // namespace detail
+
+template <typename... Signatures>
+class any_completion_handler;
+
+/// An allocator type that forwards memory allocation operations through an
+/// instance of @c any_completion_handler.
+template <typename T, typename... Signatures>
+class any_completion_handler_allocator
+{
+private:
+  template <typename...>
+  friend class any_completion_handler;
+
+  template <typename, typename...>
+  friend class any_completion_handler_allocator;
+
+  const detail::any_completion_handler_fn_table<Signatures...>* fn_table_;
+  detail::any_completion_handler_impl_base* impl_;
+
+  constexpr any_completion_handler_allocator(int,
+      const any_completion_handler<Signatures...>& h) noexcept
+    : fn_table_(h.fn_table_),
+      impl_(h.impl_)
+  {
+  }
+
+public:
+  /// The type of objects that may be allocated by the allocator.
+  typedef T value_type;
+
+  /// Rebinds an allocator to another value type.
+  template <typename U>
+  struct rebind
+  {
+    /// Specifies the type of the rebound allocator.
+    typedef any_completion_handler_allocator<U, Signatures...> other;
+  };
+
+  /// Construct from another @c any_completion_handler_allocator.
+  template <typename U>
+  constexpr any_completion_handler_allocator(
+      const any_completion_handler_allocator<U, Signatures...>& a)
+    noexcept
+    : fn_table_(a.fn_table_),
+      impl_(a.impl_)
+  {
+  }
+
+  /// Equality operator.
+  constexpr bool operator==(
+      const any_completion_handler_allocator& other) const noexcept
+  {
+    return fn_table_ == other.fn_table_ && impl_ == other.impl_;
+  }
+
+  /// Inequality operator.
+  constexpr bool operator!=(
+      const any_completion_handler_allocator& other) const noexcept
+  {
+    return fn_table_ != other.fn_table_ || impl_ != other.impl_;
+  }
+
+  /// Allocate space for @c n objects of the allocator's value type.
+  T* allocate(std::size_t n) const
+  {
+    if (fn_table_)
+    {
+      return static_cast<T*>(
+          fn_table_->allocate(
+            impl_, sizeof(T) * n, alignof(T)));
+    }
+    std::bad_alloc ex;
+    asio::detail::throw_exception(ex);
+    return nullptr;
+  }
+
+  /// Deallocate space for @c n objects of the allocator's value type.
+  void deallocate(T* p, std::size_t n) const
+  {
+    fn_table_->deallocate(impl_, p, sizeof(T) * n, alignof(T));
+  }
+};
+
+/// A protoco-allocator type that may be rebound to obtain an allocator that
+/// forwards memory allocation operations through an instance of
+/// @c any_completion_handler.
+template <typename... Signatures>
+class any_completion_handler_allocator<void, Signatures...>
+{
+private:
+  template <typename...>
+  friend class any_completion_handler;
+
+  template <typename, typename...>
+  friend class any_completion_handler_allocator;
+
+  const detail::any_completion_handler_fn_table<Signatures...>* fn_table_;
+  detail::any_completion_handler_impl_base* impl_;
+
+  constexpr any_completion_handler_allocator(int,
+      const any_completion_handler<Signatures...>& h) noexcept
+    : fn_table_(h.fn_table_),
+      impl_(h.impl_)
+  {
+  }
+
+public:
+  /// @c void as no objects can be allocated through a proto-allocator.
+  typedef void value_type;
+
+  /// Rebinds an allocator to another value type.
+  template <typename U>
+  struct rebind
+  {
+    /// Specifies the type of the rebound allocator.
+    typedef any_completion_handler_allocator<U, Signatures...> other;
+  };
+
+  /// Construct from another @c any_completion_handler_allocator.
+  template <typename U>
+  constexpr any_completion_handler_allocator(
+      const any_completion_handler_allocator<U, Signatures...>& a)
+    noexcept
+    : fn_table_(a.fn_table_),
+      impl_(a.impl_)
+  {
+  }
+
+  /// Equality operator.
+  constexpr bool operator==(
+      const any_completion_handler_allocator& other) const noexcept
+  {
+    return fn_table_ == other.fn_table_ && impl_ == other.impl_;
+  }
+
+  /// Inequality operator.
+  constexpr bool operator!=(
+      const any_completion_handler_allocator& other) const noexcept
+  {
+    return fn_table_ != other.fn_table_ || impl_ != other.impl_;
+  }
+};
+
+/// Polymorphic wrapper for completion handlers.
+/**
+ * The @c any_completion_handler class template is a polymorphic wrapper for
+ * completion handlers that propagates the associated executor, associated
+ * allocator, and associated cancellation slot through a type-erasing interface.
+ *
+ * When using @c any_completion_handler, specify one or more completion
+ * signatures as template parameters. These will dictate the arguments that may
+ * be passed to the handler through the polymorphic interface.
+ *
+ * Typical uses for @c any_completion_handler include:
+ *
+ * @li Separate compilation of asynchronous operation implementations.
+ *
+ * @li Enabling interoperability between asynchronous operations and virtual
+ *     functions.
+ */
+template <typename... Signatures>
+class any_completion_handler
+{
+#if !defined(GENERATING_DOCUMENTATION)
+private:
+  template <typename, typename...>
+  friend class any_completion_handler_allocator;
+
+  template <typename, typename>
+  friend struct associated_executor;
+
+  template <typename, typename>
+  friend struct associated_immediate_executor;
+
+  const detail::any_completion_handler_fn_table<Signatures...>* fn_table_;
+  detail::any_completion_handler_impl_base* impl_;
+#endif // !defined(GENERATING_DOCUMENTATION)
+
+public:
+  /// The associated allocator type.
+  using allocator_type = any_completion_handler_allocator<void, Signatures...>;
+
+  /// The associated cancellation slot type.
+  using cancellation_slot_type = cancellation_slot;
+
+  /// Construct an @c any_completion_handler in an empty state, without a target
+  /// object.
+  constexpr any_completion_handler()
+    : fn_table_(nullptr),
+      impl_(nullptr)
+  {
+  }
+
+  /// Construct an @c any_completion_handler in an empty state, without a target
+  /// object.
+  constexpr any_completion_handler(nullptr_t)
+    : fn_table_(nullptr),
+      impl_(nullptr)
+  {
+  }
+
+  /// Construct an @c any_completion_handler to contain the specified target.
+  template <typename H, typename Handler = decay_t<H>>
+  any_completion_handler(H&& h,
+      constraint_t<
+        !is_same<decay_t<H>, any_completion_handler>::value
+      > = 0)
+    : fn_table_(
+        &detail::any_completion_handler_fn_table_instance<
+          Handler, Signatures...>::value),
+      impl_(detail::any_completion_handler_impl<Handler>::create(
+            (get_associated_cancellation_slot)(h), static_cast<H&&>(h)))
+  {
+  }
+
+  /// Move-construct an @c any_completion_handler from another.
+  /**
+   * After the operation, the moved-from object @c other has no target.
+   */
+  any_completion_handler(any_completion_handler&& other) noexcept
+    : fn_table_(other.fn_table_),
+      impl_(other.impl_)
+  {
+    other.fn_table_ = nullptr;
+    other.impl_ = nullptr;
+  }
+
+  /// Move-assign an @c any_completion_handler from another.
+  /**
+   * After the operation, the moved-from object @c other has no target.
+   */
+  any_completion_handler& operator=(
+      any_completion_handler&& other) noexcept
+  {
+    any_completion_handler(
+        static_cast<any_completion_handler&&>(other)).swap(*this);
+    return *this;
+  }
+
+  /// Assignment operator that sets the polymorphic wrapper to the empty state.
+  any_completion_handler& operator=(nullptr_t) noexcept
+  {
+    any_completion_handler().swap(*this);
+    return *this;
+  }
+
+  /// Destructor.
+  ~any_completion_handler()
+  {
+    if (impl_)
+      fn_table_->destroy(impl_);
+  }
+
+  /// Test if the polymorphic wrapper is empty.
+  constexpr explicit operator bool() const noexcept
+  {
+    return impl_ != nullptr;
+  }
+
+  /// Test if the polymorphic wrapper is non-empty.
+  constexpr bool operator!() const noexcept
+  {
+    return impl_ == nullptr;
+  }
+
+  /// Swap the content of an @c any_completion_handler with another.
+  void swap(any_completion_handler& other) noexcept
+  {
+    std::swap(fn_table_, other.fn_table_);
+    std::swap(impl_, other.impl_);
+  }
+
+  /// Get the associated allocator.
+  allocator_type get_allocator() const noexcept
+  {
+    return allocator_type(0, *this);
+  }
+
+  /// Get the associated cancellation slot.
+  cancellation_slot_type get_cancellation_slot() const noexcept
+  {
+    return impl_ ? impl_->get_cancellation_slot() : cancellation_slot_type();
+  }
+
+  /// Function call operator.
+  /**
+   * Invokes target completion handler with the supplied arguments.
+   *
+   * This function may only be called once, as the target handler is moved from.
+   * The polymorphic wrapper is left in an empty state.
+   *
+   * Throws @c std::bad_function_call if the polymorphic wrapper is empty.
+   */
+  template <typename... Args>
+  auto operator()(Args&&... args)
+    -> decltype(fn_table_->call(impl_, static_cast<Args&&>(args)...))
+  {
+    if (detail::any_completion_handler_impl_base* impl = impl_)
+    {
+      impl_ = nullptr;
+      return fn_table_->call(impl, static_cast<Args&&>(args)...);
+    }
+    std::bad_function_call ex;
+    asio::detail::throw_exception(ex);
+  }
+
+  /// Equality operator.
+  friend constexpr bool operator==(
+      const any_completion_handler& a, nullptr_t) noexcept
+  {
+    return a.impl_ == nullptr;
+  }
+
+  /// Equality operator.
+  friend constexpr bool operator==(
+      nullptr_t, const any_completion_handler& b) noexcept
+  {
+    return nullptr == b.impl_;
+  }
+
+  /// Inequality operator.
+  friend constexpr bool operator!=(
+      const any_completion_handler& a, nullptr_t) noexcept
+  {
+    return a.impl_ != nullptr;
+  }
+
+  /// Inequality operator.
+  friend constexpr bool operator!=(
+      nullptr_t, const any_completion_handler& b) noexcept
+  {
+    return nullptr != b.impl_;
+  }
+};
+
+template <typename... Signatures, typename Candidate>
+struct associated_executor<any_completion_handler<Signatures...>, Candidate>
+{
+  using type = any_completion_executor;
+
+  static type get(const any_completion_handler<Signatures...>& handler,
+      const Candidate& candidate = Candidate()) noexcept
+  {
+    any_completion_executor any_candidate(std::nothrow, candidate);
+    return handler.fn_table_
+      ? handler.fn_table_->executor(handler.impl_, any_candidate)
+      : any_candidate;
+  }
+};
+
+template <typename... Signatures, typename Candidate>
+struct associated_immediate_executor<
+    any_completion_handler<Signatures...>, Candidate>
+{
+  using type = any_completion_executor;
+
+  static type get(const any_completion_handler<Signatures...>& handler,
+      const Candidate& candidate = Candidate()) noexcept
+  {
+    any_io_executor any_candidate(std::nothrow, candidate);
+    return handler.fn_table_
+      ? handler.fn_table_->immediate_executor(handler.impl_, any_candidate)
+      : any_candidate;
+  }
+};
+
+ASIO_INLINE_NAMESPACE_END
+} // namespace asio
+
+#include "asio/detail/pop_options.hpp"
+
+#endif // ASIO_ANY_COMPLETION_HANDLER_HPP

third_party/asio/include/asio/any_io_executor.hpp

@@ -0,0 +1,353 @@
+//
+// any_io_executor.hpp
+// ~~~~~~~~~~~~~~~~~~~
+//
+// Copyright (c) 2003-2026 Christopher M. Kohlhoff (chris at kohlhoff dot com)
+//
+// Distributed under the Boost Software License, Version 1.0. (See accompanying
+// file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
+//
+
+#ifndef ASIO_ANY_IO_EXECUTOR_HPP
+#define ASIO_ANY_IO_EXECUTOR_HPP
+
+#if defined(_MSC_VER) && (_MSC_VER >= 1200)
+# pragma once
+#endif // defined(_MSC_VER) && (_MSC_VER >= 1200)
+
+#include "asio/detail/config.hpp"
+#if defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+# include "asio/executor.hpp"
+#else // defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+# include "asio/execution.hpp"
+# include "asio/execution_context.hpp"
+#endif // defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+
+#include "asio/detail/push_options.hpp"
+
+namespace asio {
+ASIO_INLINE_NAMESPACE_BEGIN
+
+#if defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+
+typedef executor any_io_executor;
+
+#else // defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+
+/// Polymorphic executor type for use with I/O objects.
+/**
+ * The @c any_io_executor type is a polymorphic executor that supports the set
+ * of properties required by I/O objects. It is defined as the
+ * execution::any_executor class template parameterised as follows:
+ * @code execution::any_executor<
+ *   execution::context_as_t<execution_context&>,
+ *   execution::blocking_t::never_t,
+ *   execution::prefer_only<execution::blocking_t::possibly_t>,
+ *   execution::prefer_only<execution::outstanding_work_t::tracked_t>,
+ *   execution::prefer_only<execution::outstanding_work_t::untracked_t>,
+ *   execution::prefer_only<execution::relationship_t::fork_t>,
+ *   execution::prefer_only<execution::relationship_t::continuation_t>
+ * > @endcode
+ */
+class any_io_executor :
+#if defined(GENERATING_DOCUMENTATION)
+  public execution::any_executor<...>
+#else // defined(GENERATING_DOCUMENTATION)
+  public execution::any_executor<
+      execution::context_as_t<execution_context&>,
+      execution::blocking_t::never_t,
+      execution::prefer_only<execution::blocking_t::possibly_t>,
+      execution::prefer_only<execution::outstanding_work_t::tracked_t>,
+      execution::prefer_only<execution::outstanding_work_t::untracked_t>,
+      execution::prefer_only<execution::relationship_t::fork_t>,
+      execution::prefer_only<execution::relationship_t::continuation_t>
+    >
+#endif // defined(GENERATING_DOCUMENTATION)
+{
+public:
+#if !defined(GENERATING_DOCUMENTATION)
+  typedef execution::any_executor<
+      execution::context_as_t<execution_context&>,
+      execution::blocking_t::never_t,
+      execution::prefer_only<execution::blocking_t::possibly_t>,
+      execution::prefer_only<execution::outstanding_work_t::tracked_t>,
+      execution::prefer_only<execution::outstanding_work_t::untracked_t>,
+      execution::prefer_only<execution::relationship_t::fork_t>,
+      execution::prefer_only<execution::relationship_t::continuation_t>
+    > base_type;
+
+  typedef void supportable_properties_type(
+      execution::context_as_t<execution_context&>,
+      execution::blocking_t::never_t,
+      execution::prefer_only<execution::blocking_t::possibly_t>,
+      execution::prefer_only<execution::outstanding_work_t::tracked_t>,
+      execution::prefer_only<execution::outstanding_work_t::untracked_t>,
+      execution::prefer_only<execution::relationship_t::fork_t>,
+      execution::prefer_only<execution::relationship_t::continuation_t>
+    );
+#endif // !defined(GENERATING_DOCUMENTATION)
+
+  /// Default constructor.
+  ASIO_DECL any_io_executor() noexcept;
+
+  /// Construct in an empty state. Equivalent effects to default constructor.
+  ASIO_DECL any_io_executor(nullptr_t) noexcept;
+
+  /// Copy constructor.
+  ASIO_DECL any_io_executor(const any_io_executor& e) noexcept;
+
+  /// Move constructor.
+  ASIO_DECL any_io_executor(any_io_executor&& e) noexcept;
+
+  /// Construct to point to the same target as another any_executor.
+#if defined(GENERATING_DOCUMENTATION)
+  template <class... OtherSupportableProperties>
+    any_io_executor(execution::any_executor<OtherSupportableProperties...> e);
+#else // defined(GENERATING_DOCUMENTATION)
+  template <typename OtherAnyExecutor>
+  any_io_executor(OtherAnyExecutor e,
+      constraint_t<
+        conditional_t<
+          !is_same<OtherAnyExecutor, any_io_executor>::value
+            && is_base_of<execution::detail::any_executor_base,
+              OtherAnyExecutor>::value,
+          typename execution::detail::supportable_properties<
+            0, supportable_properties_type>::template
+              is_valid_target<OtherAnyExecutor>,
+          false_type
+        >::value
+      > = 0)
+    : base_type(static_cast<OtherAnyExecutor&&>(e))
+  {
+  }
+#endif // defined(GENERATING_DOCUMENTATION)
+
+  /// Construct to point to the same target as another any_executor.
+#if defined(GENERATING_DOCUMENTATION)
+  template <class... OtherSupportableProperties>
+    any_io_executor(std::nothrow_t,
+      execution::any_executor<OtherSupportableProperties...> e);
+#else // defined(GENERATING_DOCUMENTATION)
+  template <typename OtherAnyExecutor>
+  any_io_executor(std::nothrow_t, OtherAnyExecutor e,
+      constraint_t<
+        conditional_t<
+          !is_same<OtherAnyExecutor, any_io_executor>::value
+            && is_base_of<execution::detail::any_executor_base,
+              OtherAnyExecutor>::value,
+          typename execution::detail::supportable_properties<
+            0, supportable_properties_type>::template
+              is_valid_target<OtherAnyExecutor>,
+          false_type
+        >::value
+      > = 0) noexcept
+    : base_type(std::nothrow, static_cast<OtherAnyExecutor&&>(e))
+  {
+  }
+#endif // defined(GENERATING_DOCUMENTATION)
+
+  /// Construct to point to the same target as another any_executor.
+  ASIO_DECL any_io_executor(std::nothrow_t,
+      const any_io_executor& e) noexcept;
+
+  /// Construct to point to the same target as another any_executor.
+  ASIO_DECL any_io_executor(std::nothrow_t, any_io_executor&& e) noexcept;
+
+  /// Construct a polymorphic wrapper for the specified executor.
+#if defined(GENERATING_DOCUMENTATION)
+  template <ASIO_EXECUTION_EXECUTOR Executor>
+  any_io_executor(Executor e);
+#else // defined(GENERATING_DOCUMENTATION)
+  template <ASIO_EXECUTION_EXECUTOR Executor>
+  any_io_executor(Executor e,
+      constraint_t<
+        conditional_t<
+          !is_same<Executor, any_io_executor>::value
+            && !is_base_of<execution::detail::any_executor_base,
+              Executor>::value,
+          execution::detail::is_valid_target_executor<
+            Executor, supportable_properties_type>,
+          false_type
+        >::value
+      > = 0)
+    : base_type(static_cast<Executor&&>(e))
+  {
+  }
+#endif // defined(GENERATING_DOCUMENTATION)
+
+  /// Construct a polymorphic wrapper for the specified executor.
+#if defined(GENERATING_DOCUMENTATION)
+  template <ASIO_EXECUTION_EXECUTOR Executor>
+  any_io_executor(std::nothrow_t, Executor e);
+#else // defined(GENERATING_DOCUMENTATION)
+  template <ASIO_EXECUTION_EXECUTOR Executor>
+  any_io_executor(std::nothrow_t, Executor e,
+      constraint_t<
+        conditional_t<
+          !is_same<Executor, any_io_executor>::value
+            && !is_base_of<execution::detail::any_executor_base,
+              Executor>::value,
+          execution::detail::is_valid_target_executor<
+            Executor, supportable_properties_type>,
+          false_type
+        >::value
+      > = 0) noexcept
+    : base_type(std::nothrow, static_cast<Executor&&>(e))
+  {
+  }
+#endif // defined(GENERATING_DOCUMENTATION)
+
+  /// Assignment operator.
+  ASIO_DECL any_io_executor& operator=(
+      const any_io_executor& e) noexcept;
+
+  /// Move assignment operator.
+  ASIO_DECL any_io_executor& operator=(any_io_executor&& e) noexcept;
+
+  /// Assignment operator that sets the polymorphic wrapper to the empty state.
+  ASIO_DECL any_io_executor& operator=(nullptr_t);
+
+  /// Destructor.
+  ASIO_DECL ~any_io_executor();
+
+  /// Swap targets with another polymorphic wrapper.
+  ASIO_DECL void swap(any_io_executor& other) noexcept;
+
+  /// Obtain a polymorphic wrapper with the specified property.
+  /**
+   * Do not call this function directly. It is intended for use with the
+   * asio::require and asio::prefer customisation points.
+   *
+   * For example:
+   * @code any_io_executor ex = ...;
+   * auto ex2 = asio::require(ex, execution::blocking.possibly); @endcode
+   */
+  template <typename Property>
+  any_io_executor require(const Property& p,
+      constraint_t<
+        traits::require_member<const base_type&, const Property&>::is_valid
+      > = 0) const
+  {
+    return static_cast<const base_type&>(*this).require(p);
+  }
+
+  /// Obtain a polymorphic wrapper with the specified property.
+  /**
+   * Do not call this function directly. It is intended for use with the
+   * asio::prefer customisation point.
+   *
+   * For example:
+   * @code any_io_executor ex = ...;
+   * auto ex2 = asio::prefer(ex, execution::blocking.possibly); @endcode
+   */
+  template <typename Property>
+  any_io_executor prefer(const Property& p,
+      constraint_t<
+        traits::prefer_member<const base_type&, const Property&>::is_valid
+      > = 0) const
+  {
+    return static_cast<const base_type&>(*this).prefer(p);
+  }
+};
+
+#if !defined(GENERATING_DOCUMENTATION)
+
+template <>
+ASIO_DECL any_io_executor any_io_executor::require(
+    const execution::blocking_t::never_t&, int) const;
+
+template <>
+ASIO_DECL any_io_executor any_io_executor::prefer(
+    const execution::blocking_t::possibly_t&, int) const;
+
+template <>
+ASIO_DECL any_io_executor any_io_executor::prefer(
+    const execution::outstanding_work_t::tracked_t&, int) const;
+
+template <>
+ASIO_DECL any_io_executor any_io_executor::prefer(
+    const execution::outstanding_work_t::untracked_t&, int) const;
+
+template <>
+ASIO_DECL any_io_executor any_io_executor::prefer(
+    const execution::relationship_t::fork_t&, int) const;
+
+template <>
+ASIO_DECL any_io_executor any_io_executor::prefer(
+    const execution::relationship_t::continuation_t&, int) const;
+
+namespace traits {
+
+#if !defined(ASIO_HAS_DEDUCED_EQUALITY_COMPARABLE_TRAIT)
+
+template <>
+struct equality_comparable<any_io_executor>
+{
+  static const bool is_valid = true;
+  static const bool is_noexcept = true;
+};
+
+#endif // !defined(ASIO_HAS_DEDUCED_EQUALITY_COMPARABLE_TRAIT)
+
+#if !defined(ASIO_HAS_DEDUCED_EXECUTE_MEMBER_TRAIT)
+
+template <typename F>
+struct execute_member<any_io_executor, F>
+{
+  static const bool is_valid = true;
+  static const bool is_noexcept = false;
+  typedef void result_type;
+};
+
+#endif // !defined(ASIO_HAS_DEDUCED_EXECUTE_MEMBER_TRAIT)
+
+#if !defined(ASIO_HAS_DEDUCED_QUERY_MEMBER_TRAIT)
+
+template <typename Prop>
+struct query_member<any_io_executor, Prop> :
+  query_member<any_io_executor::base_type, Prop>
+{
+};
+
+#endif // !defined(ASIO_HAS_DEDUCED_QUERY_MEMBER_TRAIT)
+
+#if !defined(ASIO_HAS_DEDUCED_REQUIRE_MEMBER_TRAIT)
+
+template <typename Prop>
+struct require_member<any_io_executor, Prop> :
+  require_member<any_io_executor::base_type, Prop>
+{
+  typedef any_io_executor result_type;
+};
+
+#endif // !defined(ASIO_HAS_DEDUCED_REQUIRE_MEMBER_TRAIT)
+
+#if !defined(ASIO_HAS_DEDUCED_PREFER_MEMBER_TRAIT)
+
+template <typename Prop>
+struct prefer_member<any_io_executor, Prop> :
+  prefer_member<any_io_executor::base_type, Prop>
+{
+  typedef any_io_executor result_type;
+};
+
+#endif // !defined(ASIO_HAS_DEDUCED_PREFER_MEMBER_TRAIT)
+
+} // namespace traits
+
+#endif // !defined(GENERATING_DOCUMENTATION)
+
+#endif // defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+
+ASIO_INLINE_NAMESPACE_END
+} // namespace asio
+
+#include "asio/detail/pop_options.hpp"
+
+#if defined(ASIO_HEADER_ONLY) \
+  && !defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+# include "asio/impl/any_io_executor.ipp"
+#endif // defined(ASIO_HEADER_ONLY)
+       //   && !defined(ASIO_USE_TS_EXECUTOR_AS_DEFAULT)
+
+#endif // ASIO_ANY_IO_EXECUTOR_HPP

third_party/asio/include/asio/append.hpp

@@ -0,0 +1,67 @@
+//
+// append.hpp
+// ~~~~~~~~~~
+//
+// Copyright (c) 2003-2026 Christopher M. Kohlhoff (chris at kohlhoff dot com)
+//
+// Distributed under the Boost Software License, Version 1.0. (See accompanying
+// file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
+//
+
+#ifndef ASIO_APPEND_HPP
+#define ASIO_APPEND_HPP
+
+#if defined(_MSC_VER) && (_MSC_VER >= 1200)
+# pragma once
+#endif // defined(_MSC_VER) && (_MSC_VER >= 1200)
+
+#include "asio/detail/config.hpp"
+#include <tuple>
+#include "asio/detail/type_traits.hpp"
+
+#include "asio/detail/push_options.hpp"
+
+namespace asio {
+ASIO_INLINE_NAMESPACE_BEGIN
+
+/// Completion token type used to specify that the completion handler
+/// arguments should be passed additional values after the results of the
+/// operation.
+template <typename CompletionToken, typename... Values>
+class append_t
+{
+public:
+  /// Constructor.
+  template <typename T, typename... V>
+  constexpr explicit append_t(T&& completion_token, V&&... values)
+    : token_(static_cast<T&&>(completion_token)),
+      values_(static_cast<V&&>(values)...)
+  {
+  }
+
+//private:
+  CompletionToken token_;
+  std::tuple<Values...> values_;
+};
+
+/// Completion token type used to specify that the completion handler
+/// arguments should be passed additional values after the results of the
+/// operation.
+template <typename CompletionToken, typename... Values>
+ASIO_NODISCARD inline constexpr
+append_t<decay_t<CompletionToken>, decay_t<Values>...>
+append(CompletionToken&& completion_token, Values&&... values)
+{
+  return append_t<decay_t<CompletionToken>, decay_t<Values>...>(
+      static_cast<CompletionToken&&>(completion_token),
+      static_cast<Values&&>(values)...);
+}
+
+ASIO_INLINE_NAMESPACE_END
+} // namespace asio
+
+#include "asio/detail/pop_options.hpp"
+
+#include "asio/impl/append.hpp"
+
+#endif // ASIO_APPEND_HPP

third_party/asio/include/asio/as_tuple.hpp

@@ -0,0 +1,154 @@
+//
+// as_tuple.hpp
+// ~~~~~~~~~~~~
+//
+// Copyright (c) 2003-2026 Christopher M. Kohlhoff (chris at kohlhoff dot com)
+//
+// Distributed under the Boost Software License, Version 1.0. (See accompanying
+// file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
+//
+
+#ifndef ASIO_AS_TUPLE_HPP
+#define ASIO_AS_TUPLE_HPP
+
+#if defined(_MSC_VER) && (_MSC_VER >= 1200)
+# pragma once
+#endif // defined(_MSC_VER) && (_MSC_VER >= 1200)
+
+#include "asio/detail/config.hpp"
+#include "asio/detail/type_traits.hpp"
+
+#include "asio/detail/push_options.hpp"
+
+namespace asio {
+ASIO_INLINE_NAMESPACE_BEGIN
+
+/// A @ref completion_token adapter used to specify that the completion handler
+/// arguments should be combined into a single tuple argument.
+/**
+ * The as_tuple_t class is used to indicate that any arguments to the
+ * completion handler should be combined and passed as a single tuple argument.
+ * The arguments are first moved into a @c std::tuple and that tuple is then
+ * passed to the completion handler.
+ */
+template <typename CompletionToken>
+class as_tuple_t
+{
+public:
+  /// Tag type used to prevent the "default" constructor from being used for
+  /// conversions.
+  struct default_constructor_tag {};
+
+  /// Default constructor.
+  /**
+   * This constructor is only valid if the underlying completion token is
+   * default constructible and move constructible. The underlying completion
+   * token is itself defaulted as an argument to allow it to capture a source
+   * location.
+   */
+  constexpr as_tuple_t(
+      default_constructor_tag = default_constructor_tag(),
+      CompletionToken token = CompletionToken())
+    : token_(static_cast<CompletionToken&&>(token))
+  {
+  }
+
+  /// Constructor.
+  template <typename T>
+  constexpr explicit as_tuple_t(
+      T&& completion_token)
+    : token_(static_cast<T&&>(completion_token))
+  {
+  }
+
+  /// Adapts an executor to add the @c as_tuple_t completion token as the
+  /// default.
+  template <typename InnerExecutor>
+  struct executor_with_default : InnerExecutor
+  {
+    /// Specify @c as_tuple_t as the default completion token type.
+    typedef as_tuple_t default_completion_token_type;
+
+    /// Construct the adapted executor from the inner executor type.
+    template <typename InnerExecutor1>
+    executor_with_default(const InnerExecutor1& ex,
+        constraint_t<
+          conditional_t<
+            !is_same<InnerExecutor1, executor_with_default>::value,
+            is_convertible<InnerExecutor1, InnerExecutor>,
+            false_type
+          >::value
+        > = 0) noexcept
+      : InnerExecutor(ex)
+    {
+    }
+  };
+
+  /// Type alias to adapt an I/O object to use @c as_tuple_t as its
+  /// default completion token type.
+  template <typename T>
+  using as_default_on_t = typename T::template rebind_executor<
+      executor_with_default<typename T::executor_type>>::other;
+
+  /// Function helper to adapt an I/O object to use @c as_tuple_t as its
+  /// default completion token type.
+  template <typename T>
+  static typename decay_t<T>::template rebind_executor<
+      executor_with_default<typename decay_t<T>::executor_type>
+    >::other
+  as_default_on(T&& object)
+  {
+    return typename decay_t<T>::template rebind_executor<
+        executor_with_default<typename decay_t<T>::executor_type>
+      >::other(static_cast<T&&>(object));
+  }
+
+//private:
+  CompletionToken token_;
+};
+
+/// A function object type that adapts a @ref completion_token to specify that
+/// the completion handler arguments should be combined into a single tuple
+/// argument.
+/**
+ * May also be used directly as a completion token, in which case it adapts the
+ * asynchronous operation's default completion token (or asio::deferred
+ * if no default is available).
+ */
+struct partial_as_tuple
+{
+  /// Default constructor.
+  constexpr partial_as_tuple()
+  {
+  }
+
+  /// Adapt a @ref completion_token to specify that the completion handler
+  /// arguments should be combined into a single tuple argument.
+  template <typename CompletionToken>
+  ASIO_NODISCARD inline
+  constexpr as_tuple_t<decay_t<CompletionToken>>
+  operator()(CompletionToken&& completion_token) const
+  {
+    return as_tuple_t<decay_t<CompletionToken>>(
+        static_cast<CompletionToken&&>(completion_token));
+  }
+};
+
+/// A function object that adapts a @ref completion_token to specify that the
+/// completion handler arguments should be combined into a single tuple
+/// argument.
+/**
+ * May also be used directly as a completion token, in which case it adapts the
+ * asynchronous operation's default completion token (or asio::deferred
+ * if no default is available).
+ */
+ASIO_INLINE_VARIABLE constexpr partial_as_tuple as_tuple;
+
+ASIO_INLINE_NAMESPACE_END
+} // namespace asio
+
+#include "asio/detail/pop_options.hpp"
+
+#include "asio/impl/as_tuple.hpp"
+
+#endif // ASIO_AS_TUPLE_HPP

third_party/asio/include/asio/associated_allocator.hpp

@@ -0,0 +1,216 @@
+//
+// associated_allocator.hpp
+// ~~~~~~~~~~~~~~~~~~~~~~~~
+//
+// Copyright (c) 2003-2026 Christopher M. Kohlhoff (chris at kohlhoff dot com)
+//
+// Distributed under the Boost Software License, Version 1.0. (See accompanying
+// file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
+//
+
+#ifndef ASIO_ASSOCIATED_ALLOCATOR_HPP
+#define ASIO_ASSOCIATED_ALLOCATOR_HPP
+
+#if defined(_MSC_VER) && (_MSC_VER >= 1200)
+# pragma once
+#endif // defined(_MSC_VER) && (_MSC_VER >= 1200)
+
+#include "asio/detail/config.hpp"
+#include <memory>
+#include "asio/associator.hpp"
+#include "asio/detail/functional.hpp"
+#include "asio/detail/type_traits.hpp"
+
+#include "asio/detail/push_options.hpp"
+
+namespace asio {
+ASIO_INLINE_NAMESPACE_BEGIN
+
+template <typename T, typename Allocator>
+struct associated_allocator;
+
+namespace detail {
+
+template <typename T, typename = void>
+struct has_allocator_type : false_type
+{
+};
+
+template <typename T>
+struct has_allocator_type<T, void_t<typename T::allocator_type>> : true_type
+{
+};
+
+template <typename T, typename A, typename = void, typename = void>
+struct associated_allocator_impl
+{
+  typedef void asio_associated_allocator_is_unspecialised;
+
+  typedef A type;
+
+  static type get(const T&) noexcept
+  {
+    return type();
+  }
+
+  static const type& get(const T&, const A& a) noexcept
+  {
+    return a;
+  }
+};
+
+template <typename T, typename A>
+struct associated_allocator_impl<T, A, void_t<typename T::allocator_type>>
+{
+  typedef typename T::allocator_type type;
+
+  static auto get(const T& t) noexcept
+    -> decltype(t.get_allocator())
+  {
+    return t.get_allocator();
+  }
+
+  static auto get(const T& t, const A&) noexcept
+    -> decltype(t.get_allocator())
+  {
+    return t.get_allocator();
+  }
+};
+
+template <typename T, typename A>
+struct associated_allocator_impl<T, A,
+  enable_if_t<
+    !has_allocator_type<T>::value
+  >,
+  void_t<
+    typename associator<associated_allocator, T, A>::type
+  >> : associator<associated_allocator, T, A>
+{
+};
+
+} // namespace detail
+
+/// Traits type used to obtain the allocator associated with an object.
+/**
+ * A program may specialise this traits type if the @c T template parameter in
+ * the specialisation is a user-defined type. The template parameter @c
+ * Allocator shall be a type meeting the Allocator requirements.
+ *
+ * Specialisations shall meet the following requirements, where @c t is a const
+ * reference to an object of type @c T, and @c a is an object of type @c
+ * Allocator.
+ *
+ * @li Provide a nested typedef @c type that identifies a type meeting the
+ * Allocator requirements.
+ *
+ * @li Provide a noexcept static member function named @c get, callable as @c
+ * get(t) and with return type @c type or a (possibly const) reference to @c
+ * type.
+ *
+ * @li Provide a noexcept static member function named @c get, callable as @c
+ * get(t,a) and with return type @c type or a (possibly const) reference to @c
+ * type.
+ */
+template <typename T, typename Allocator = std::allocator<void>>
+struct associated_allocator
+#if !defined(GENERATING_DOCUMENTATION)
+  : detail::associated_allocator_impl<T, Allocator>
+#endif // !defined(GENERATING_DOCUMENTATION)
+{
+#if defined(GENERATING_DOCUMENTATION)
+  /// If @c T has a nested type @c allocator_type, <tt>T::allocator_type</tt>.
+  /// Otherwise @c Allocator.
+  typedef see_below type;
+
+  /// If @c T has a nested type @c allocator_type, returns
+  /// <tt>t.get_allocator()</tt>. Otherwise returns @c type().
+  static decltype(auto) get(const T& t) noexcept;
+
+  /// If @c T has a nested type @c allocator_type, returns
+  /// <tt>t.get_allocator()</tt>. Otherwise returns @c a.
+  static decltype(auto) get(const T& t, const Allocator& a) noexcept;
+#endif // defined(GENERATING_DOCUMENTATION)
+};
+
+/// Helper function to obtain an object's associated allocator.
+/**
+ * @returns <tt>associated_allocator<T>::get(t)</tt>
+ */
+template <typename T>
+ASIO_NODISCARD inline typename associated_allocator<T>::type
+get_associated_allocator(const T& t) noexcept
+{
+  return associated_allocator<T>::get(t);
+}
+
+/// Helper function to obtain an object's associated allocator.
+/**
+ * @returns <tt>associated_allocator<T, Allocator>::get(t, a)</tt>
+ */
+template <typename T, typename Allocator>
+ASIO_NODISCARD inline auto get_associated_allocator(
+    const T& t, const Allocator& a) noexcept
+  -> decltype(associated_allocator<T, Allocator>::get(t, a))
+{
+  return associated_allocator<T, Allocator>::get(t, a);
+}
+
+template <typename T, typename Allocator = std::allocator<void>>
+using associated_allocator_t
+  = typename associated_allocator<T, Allocator>::type;
+
+namespace detail {
+
+template <typename T, typename A, typename = void>
+struct associated_allocator_forwarding_base
+{
+};
+
+template <typename T, typename A>
+struct associated_allocator_forwarding_base<T, A,
+    enable_if_t<
+      is_same<
+        typename associated_allocator<T,
+          A>::asio_associated_allocator_is_unspecialised,
+        void
+      >::value
+    >>
+{
+  typedef void asio_associated_allocator_is_unspecialised;
+};
+
+} // namespace detail
+
+/// Specialisation of associated_allocator for @c std::reference_wrapper.
+template <typename T, typename Allocator>
+struct associated_allocator<reference_wrapper<T>, Allocator>
+#if !defined(GENERATING_DOCUMENTATION)
+  : detail::associated_allocator_forwarding_base<T, Allocator>
+#endif // !defined(GENERATING_DOCUMENTATION)
+{
+  /// Forwards @c type to the associator specialisation for the unwrapped type
+  /// @c T.
+  typedef typename associated_allocator<T, Allocator>::type type;
+
+  /// Forwards the request to get the allocator to the associator specialisation
+  /// for the unwrapped type @c T.
+  static type get(reference_wrapper<T> t) noexcept
+  {
+    return associated_allocator<T, Allocator>::get(t.get());
+  }
+
+  /// Forwards the request to get the allocator to the associator specialisation
+  /// for the unwrapped type @c T.
+  static auto get(reference_wrapper<T> t, const Allocator& a) noexcept
+    -> decltype(associated_allocator<T, Allocator>::get(t.get(), a))
+  {
+    return associated_allocator<T, Allocator>::get(t.get(), a);
+  }
+};
+
+ASIO_INLINE_NAMESPACE_END
+} // namespace asio
+
+#include "asio/detail/pop_options.hpp"
+
+#endif // ASIO_ASSOCIATED_ALLOCATOR_HPP

third_party/asio/include/asio/associated_cancellation_slot.hpp

@@ -0,0 +1,223 @@
+//
+// associated_cancellation_slot.hpp
+// ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+//
+// Copyright (c) 2003-2026 Christopher M. Kohlhoff (chris at kohlhoff dot com)
+//
+// Distributed under the Boost Software License, Version 1.0. (See accompanying
+// file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
+//
+
+#ifndef ASIO_ASSOCIATED_CANCELLATION_SLOT_HPP
+#define ASIO_ASSOCIATED_CANCELLATION_SLOT_HPP
+
+#if defined(_MSC_VER) && (_MSC_VER >= 1200)
+# pragma once
+#endif // defined(_MSC_VER) && (_MSC_VER >= 1200)
+
+#include "asio/detail/config.hpp"
+#include "asio/associator.hpp"
+#include "asio/cancellation_signal.hpp"
+#include "asio/detail/functional.hpp"
+#include "asio/detail/type_traits.hpp"
+
+#include "asio/detail/push_options.hpp"
+
+namespace asio {
+ASIO_INLINE_NAMESPACE_BEGIN
+
+template <typename T, typename CancellationSlot>
+struct associated_cancellation_slot;
+
+namespace detail {
+
+template <typename T, typename = void>
+struct has_cancellation_slot_type : false_type
+{
+};
+
+template <typename T>
+struct has_cancellation_slot_type<T, void_t<typename T::cancellation_slot_type>>
+  : true_type
+{
+};
+
+template <typename T, typename S, typename = void, typename = void>
+struct associated_cancellation_slot_impl
+{
+  typedef void asio_associated_cancellation_slot_is_unspecialised;
+
+  typedef S type;
+
+  static type get(const T&) noexcept
+  {
+    return type();
+  }
+
+  static const type& get(const T&, const S& s) noexcept
+  {
+    return s;
+  }
+};
+
+template <typename T, typename S>
+struct associated_cancellation_slot_impl<T, S,
+  void_t<typename T::cancellation_slot_type>>
+{
+  typedef typename T::cancellation_slot_type type;
+
+  static auto get(const T& t) noexcept
+    -> decltype(t.get_cancellation_slot())
+  {
+    return t.get_cancellation_slot();
+  }
+
+  static auto get(const T& t, const S&) noexcept
+    -> decltype(t.get_cancellation_slot())
+  {
+    return t.get_cancellation_slot();
+  }
+};
+
+template <typename T, typename S>
+struct associated_cancellation_slot_impl<T, S,
+  enable_if_t<
+    !has_cancellation_slot_type<T>::value
+  >,
+  void_t<
+    typename associator<associated_cancellation_slot, T, S>::type
+  >> : associator<associated_cancellation_slot, T, S>
+{
+};
+
+} // namespace detail
+
+/// Traits type used to obtain the cancellation_slot associated with an object.
+/**
+ * A program may specialise this traits type if the @c T template parameter in
+ * the specialisation is a user-defined type. The template parameter @c
+ * CancellationSlot shall be a type meeting the CancellationSlot requirements.
+ *
+ * Specialisations shall meet the following requirements, where @c t is a const
+ * reference to an object of type @c T, and @c s is an object of type @c
+ * CancellationSlot.
+ *
+ * @li Provide a nested typedef @c type that identifies a type meeting the
+ * CancellationSlot requirements.
+ *
+ * @li Provide a noexcept static member function named @c get, callable as @c
+ * get(t) and with return type @c type or a (possibly const) reference to @c
+ * type.
+ *
+ * @li Provide a noexcept static member function named @c get, callable as @c
+ * get(t,s) and with return type @c type or a (possibly const) reference to @c
+ * type.
+ */
+template <typename T, typename CancellationSlot = cancellation_slot>
+struct associated_cancellation_slot
+#if !defined(GENERATING_DOCUMENTATION)
+  : detail::associated_cancellation_slot_impl<T, CancellationSlot>
+#endif // !defined(GENERATING_DOCUMENTATION)
+{
+#if defined(GENERATING_DOCUMENTATION)
+  /// If @c T has a nested type @c cancellation_slot_type,
+  /// <tt>T::cancellation_slot_type</tt>. Otherwise
+  /// @c CancellationSlot.
+  typedef see_below type;
+
+  /// If @c T has a nested type @c cancellation_slot_type, returns
+  /// <tt>t.get_cancellation_slot()</tt>. Otherwise returns @c type().
+  static decltype(auto) get(const T& t) noexcept;
+
+  /// If @c T has a nested type @c cancellation_slot_type, returns
+  /// <tt>t.get_cancellation_slot()</tt>. Otherwise returns @c s.
+  static decltype(auto) get(const T& t,
+      const CancellationSlot& s) noexcept;
+#endif // defined(GENERATING_DOCUMENTATION)
+};
+
+/// Helper function to obtain an object's associated cancellation_slot.
+/**
+ * @returns <tt>associated_cancellation_slot<T>::get(t)</tt>
+ */
+template <typename T>
+ASIO_NODISCARD inline typename associated_cancellation_slot<T>::type
+get_associated_cancellation_slot(const T& t) noexcept
+{
+  return associated_cancellation_slot<T>::get(t);
+}
+
+/// Helper function to obtain an object's associated cancellation_slot.
+/**
+ * @returns <tt>associated_cancellation_slot<T,
+ * CancellationSlot>::get(t, st)</tt>
+ */
+template <typename T, typename CancellationSlot>
+ASIO_NODISCARD inline auto get_associated_cancellation_slot(
+    const T& t, const CancellationSlot& st) noexcept
+  -> decltype(associated_cancellation_slot<T, CancellationSlot>::get(t, st))
+{
+  return associated_cancellation_slot<T, CancellationSlot>::get(t, st);
+}
+
+template <typename T, typename CancellationSlot = cancellation_slot>
+using associated_cancellation_slot_t =
+  typename associated_cancellation_slot<T, CancellationSlot>::type;
+
+namespace detail {
+
+template <typename T, typename S, typename = void>
+struct associated_cancellation_slot_forwarding_base
+{
+};
+
+template <typename T, typename S>
+struct associated_cancellation_slot_forwarding_base<T, S,
+    enable_if_t<
+      is_same<
+        typename associated_cancellation_slot<T,
+          S>::asio_associated_cancellation_slot_is_unspecialised,
+        void
+      >::value
+    >>
+{
+  typedef void asio_associated_cancellation_slot_is_unspecialised;
+};
+
+} // namespace detail
+
+/// Specialisation of associated_cancellation_slot for @c
+/// std::reference_wrapper.
+template <typename T, typename CancellationSlot>
+struct associated_cancellation_slot<reference_wrapper<T>, CancellationSlot>
+#if !defined(GENERATING_DOCUMENTATION)
+  : detail::associated_cancellation_slot_forwarding_base<T, CancellationSlot>
+#endif // !defined(GENERATING_DOCUMENTATION)
+{
+  /// Forwards @c type to the associator specialisation for the unwrapped type
+  /// @c T.
+  typedef typename associated_cancellation_slot<T, CancellationSlot>::type type;
+
+  /// Forwards the request to get the cancellation slot to the associator
+  /// specialisation for the unwrapped type @c T.
+  static type get(reference_wrapper<T> t) noexcept
+  {
+    return associated_cancellation_slot<T, CancellationSlot>::get(t.get());
+  }
+
+  /// Forwards the request to get the cancellation slot to the associator
+  /// specialisation for the unwrapped type @c T.
+  static auto get(reference_wrapper<T> t, const CancellationSlot& s) noexcept
+    -> decltype(
+      associated_cancellation_slot<T, CancellationSlot>::get(t.get(), s))
+  {
+    return associated_cancellation_slot<T, CancellationSlot>::get(t.get(), s);
+  }
+};
+
+ASIO_INLINE_NAMESPACE_END
+} // namespace asio
+
+#include "asio/detail/pop_options.hpp"
+
+#endif // ASIO_ASSOCIATED_CANCELLATION_SLOT_HPP