Skip to content

Latest commit

 

History

History
113 lines (67 loc) · 5.23 KB

README.md

File metadata and controls

113 lines (67 loc) · 5.23 KB

ABCg

build workflow GitHub release (latest by date)

Development framework accompanying the course MCTA008-17 Computer Graphics at UFABC.

Documentation | Release notes

ABCg is a lightweight C++ framework that simplifies the development of 3D graphics applications based on OpenGL, OpenGL ES, WebGL, and Vulkan. It is designed for the tutorials and assignments of the course "MCTA008-17 Computer Graphics" taught at Federal University of ABC (UFABC).


Main features

  • Supported platforms: Linux, macOS, Windows, WebAssembly.
  • Supported backends: OpenGL 3.3+, OpenGL ES 3.0+, WebGL 2.0 (via Emscripten), Vulkan 1.3.
  • Applications that use the common subset of functions between OpenGL 3.3 and OpenGL ES 3.0 can be built for WebGL 2.0 using the same source code.
  • OpenGL functions can be qualified with the abcg:: namespace to enable throwing exceptions with descriptive GL error messages that include the source code location.
  • Includes helper classes and functions for loading textures (using SDL_image), loading OBJ 3D models (using tinyobjloader), and compiling GLSL shaders to SPIR-V with glslang.

Requirements

The following minimum requirements are shared among all platforms:

  • CMake 3.21.
  • A C++ compiler with at least partial support for C++20 (tested with GCC 12, Clang 16, MSVC 17, and emcc 3.1.42).
  • A system with support for OpenGL 3.3 (OpenGL backend) or Vulkan 1.3 (Vulkan backend). Conformant software rasterizers such as Mesa's Gallium llvmpipe and lavapipe (post Jun 2022) are supported. Mesa's D3D12 backend on WSL 2.0 is supported as well.

For WebAssembly:

  • Emscripten.
  • A browser with support for WebGL 2.0.

For building desktop applications:

  • SDL 2.0.
  • SDL_image 2.0.
  • GLEW 2.2.0 (required for OpenGL-based applications).
  • Vulkan 1.3 (required for Vulkan-based applications).

Desktop dependencies can be resolved automatically with Conan, but it is disabled by default. To use Conan, install Conan 1.47 or a later 1.* version (ABCg is not compatible with Conan 2.0!) and then configure CMake with -DENABLE_CONAN=ON.

The default renderer backend is OpenGL (CMake option GRAPHICS_API=OpenGL). To use the Vulkan backend, configure CMake with -DGRAPHICS_API=Vulkan.


Installation and usage

Start by cloning the repository:

# Get abcg repo
git clone https://github.com/hbatagelo/abcg.git

# Enter the directory
cd abcg

Follow the instructions below to build the "Hello, World!" sample located in abcg/examples/helloworld.

Windows

  • Run build-vs.bat for building with the Visual Studio 2022 toolchain.
  • Run build.bat for building with GCC (MinGW-w64).

build-vs.bat and build.bat accept two optional arguments: (1) the build type, which is Release by default, and (2) an extra CMake option. For example, for a Debug build with -DENABLE_CONAN=ON using VS 2022, run

build-vs.bat Debug -DENABLE_CONAN=ON

Linux and macOS

Run ./build.sh.

The script accepts two optional arguments: (1) the build type, which is Release by default, and (2) an extra CMake option. For example, for a Debug build with -DENABLE_CONAN=ON, run

./build.sh Debug -DENABLE_CONAN=ON

WebAssembly

  1. Run build-wasm.bat (Windows) or ./build-wasm.sh (Linux/macOS).
  2. Run runweb.bat (Windows) or ./runweb.sh (Linux/macOS) for setting up a local web server.
  3. Open http://localhost:8080/helloworld.html.

Docker setup

ABCg can be built in a Docker container. The Dockerfile provided is based on Ubuntu 22.04 and includes Emscripten.

  1. Create the Docker image (abcg):

    sudo docker build -t abcg .
    
  2. Create the container (abcg_container):

    sudo docker create -it \
      -p 8080:8080 \
      -v /tmp/.X11-unix:/tmp/.X11-unix:rw \
      -e DISPLAY \
      --name abcg_container abcg
    
  3. Start the container:

    sudo docker start -ai abcg_container
    

    On NVIDIA GPUs, install the NVIDIA Container Toolkit to allow the container to use the host's NVIDIA driver and X server. Expose the X server with sudo xhost +local:root before starting the container.


License

ABCg is licensed under the MIT License. See LICENSE for more information.