Building from source

Quick links:

  1. Setup Your Environment

  2. Install Numpy & Pillow

  3. Pre-Build Instructions

  4. Build Instructions

Windows

  1. Install Any 1 of the 2 Toolchains:

    1. Install MinGW via chocolatey
      • choco install mingw
    2. Install Visual Studio or Visual Studio: Build Tools.
      • In the 'Visual Studio Installer' ensure the following components are selected to installed, everything else can be completely disabled to save bandwidth:
        • Desktop development with C++
          • C++ AddressSanitizer
          • Clang Compiler For Windows
          • One Of These SDKs:
            • Windows 10 SDK (XX.X.XXXXX.X)
            • Windows 11 SDK (XX.X.XXXXX.X)

  2. Install python & git

    • Can be done manually:

    • or Use chocolatey

      • choco install python git

    • or Use winget

      • winget install -e --id Python.Python.3.9
      • winget install -e --id Git.Git

Linux

  1. Install any 1 of the 2 compiler toolchain listed below:

    • GCC (gcc & g++)

      • Debian & Children: apt install gcc
      • Arch & Children: pacman -S gcc

    • Clang (clang & clang++)

      • Debian & Children: apt install clang
      • Arch & Children: pacman -S clang

  2. Install any 1 of the 2 build system listed below:

    • Ninja Build

      • Debian & Children: apt install ninja-build
      • Arch & Children: pacman -S ninja

    • GNU Make

      • Debian & Children: apt install make
      • Arch & Children: pacman -S make

  3. Install required dependencies:

    • Arch & Children: pacman -S git cmake libxrandr libxinerama libx11 libxcursor libxi python3
    • Debian & Children: apt install git cmake libxrandr-dev libxrandr2 libxinerama-dev libxinerama1 libx11-dev libx11-6 libxcursor-dev libxcursor1 libxi-dev libxi6 python3 python3-pip

macOS

  1. Install 1 of any 3 compiler toolchain listed below:

    • Xcode: https://developer.apple.com/xcode/ (Technically not a toolchain, but can still be used if already installed but other alternatives are much more smaller in size)
    • GCC: brew install gcc
    • Clang: as far as i know clang is pre-installed on MacOS

  2. Install required deps with homebrew:

    • brew install git cmake python@3.9

Install Numpy & Pillow

Pre-Build Instructions

  1. Clone csprite Repository: git clone https://github.com/csprite/csprite --recursive
  2. Move Into Cloned Repository Folder: cd csprite
  3. Generate Icons & Assets:
    • Icons: python tools/create_icons.py
    • Assets: python tools/create_assets.py
  4. Build Csprite using Build Instructions & you should have your final binary/package in a folder name on of the following: Release, RelWithDebInfo or Debug.

Build Instructions

these instructions are used for build SDL2 & Csprite.

  1. Generate Build Files: cmake -S ./ -B ./build/ <other-options>

    • -DCMAKE_BUILD_TYPE=Release, all possible values:
      • Debug
      • Release
      • RelWithDebInfo
    • -G option specifies the generator to use.
      • On Linux, you can use -G "Ninja" or -G "Unix Makefiles".
      • On MacOS, you can use: -G "Xcode" or -G "Unix Makefiles".
      • On Windows, you can use: -G "MinGW Makefiles" or If you installed Visual Studio you can use the following generators:
        • you can specify the Visual Studio generator depending on the version of Visual Studio you installed, possible values are:
          • Visual Studio 17 2022
          • Visual Studio 16 2019
          • Visual Studio 15 2017
          • Visual Studio 14 2015
    • -T "ClangCL" flag specifies the toolset to use, this is meant for Windows (Visual Studio only) & is used to Compile using Clang since MSVC is dumb.
    • -A flag specifies the platform you are building the library for & this flag is only used on Window.
      • use -A "x64" for x64/x86_64 platform.
      • use -A "Win32" for x86/i686 platform.
    • -DCMAKE_OSX_ARCHITECTURES=x86_64 flag specifies the Mac CPU Architecture to build for.
      • use arm64 for Apple Silicon M1 or M2 (or new CPUs apple might introduce with arm64 architecture).
      • use x86_64 for Apple CPUs with x64 or x86_64 architecture, this can be used with Apple's M1 & M2 chips since they can run x64 binaries too.
    • -DBUILD_APPLE_BUNDLE=ON flag specifies wether to build a Apple's .app bundle or not, by default the value is OFF.

  2. Build The Code With Generated Build Files: cmake --build ./build/ <other-options>

    • --config Release flag specifies the build configuration to build for, the value should be same as the value of -DCMAKE_BUILD_TYPE flag above.
    • --parallel 4 flag specifies the number of jobs while building, the value should be set to your number of CPU Cores & if affects your systems performance set the value to something lower.