cmc

TODO

• Optimize create_chunk_and_light_data_packet
• Optimize chunk_section_update_mesh

Project setup

Debian Linux

If you are on debian linux run ./dev-setup.sh and ./import-data.sh.

To build run

• mkdir build && cd build
• cmake ..
• cd ..
• cmake --build build -j8

To run

• Create a Minecraft profile and set a custom Java executable to the path of custommc.sh
• Run that Minecraft profile which creates a file called ./run-with-token.sh
• Edit that file to change the server hostname and port.
• Run ./run-with-token.sh

To run with a local server online mode turned off (the final two args are unused auth placeholders)

• ./build/cmc <any-username> <host> <port> _ _

Generic instructions

• mkdir lib
• Download the prebuilt WGPU libraries from https://github.com/gfx-rs/wgpu-native/releases and put the extracted zip in the lib/wgpu directory
• Download the GLFW sources into lib/glfw from the version pointed to from https://github.com/gfx-rs/wgpu-native/tree/trunk/examples/vendor
• From the main directory, run git clone https://github.com/recp/cglm.git -o ./lib/cglm to get cglm. It works as a header-only library.
• Retrieve blocks.json from the server jar with the following and place it in /data.

java -DbundlerMainClass=net.minecraft.data.Main -jar paper-1.20.6-145.jar --reports

• To load default client resources, copy your client jar from ~/.minecraft/versions to /data and extract it with gunzip 1.20.6.jar. To clean things up you can rm *.class as these files are not needed.
• To compile and run follow the instructions above

Features

• Offline and online mode support
• Encryption support
• Basic rendering
• Building and breaking blocks
• Player movement

Debugging

If there is a core dump:

coredumpctl dump -r `pwd`/a.out > core && gdb a.out -c core

Notes on how I got started with a WebGPU example in C

• framework.h and framework.c are from https://github.com/gfx-rs/wgpu-native/tree/trunk/examples/framework
• main.c and shader.wsgl are from https://github.com/gfx-rs/wgpu-native/tree/trunk/examples/triangle
• In main.c I updated the includes to their local locations:

#include "wgpu/webgpu.h"
#include "wgpu/wgpu.h"
// ...
#include "GLFW/include/GLFW/glfw3.h"
#include "GLFW/include/GLFW/glfw3native.h"

• I added

#define GLFW_EXPOSE_NATIVE_WAYLAND
#define GLFW_EXPOSE_NATIVE_X11

before

#include "GLFW/include/GLFW/glfw3.h"

• I commented out this line that causes an error (probably a mismatched wgpu version)

//  .depthSlice = WGPU_DEPTH_SLICE_UNDEFINED,

To figure out which texture keys there are in model files:

jq -r '.textures | keys[]' block/*.json | sort | uniq -c | sort -nr