These instructions and sample code are for setting up OpenGL on Mac in Xcode as of MacOS Monterey and 2022. I followed these on my 2022 MacBook Air M2 as well as my M1 mac mini.
They use glew and glfw3 and also glm.
- glfw3
- glew
- glm
Install Homebrew if you don't have it already. There are different versions and paths they use for Intel binaries and Apple Silicon binaries.
For Apple silicon homebrew, it is typically underneath /opt/homebrew. For Intel it is under /usr/local.
These release notes at https://brew.sh/2021/02/05/homebrew-3.0.0/ describe how/why for the two paths.
Install glfw3
/opt/homebrew/bin/brew install glfw3
And install glew
/opt/homebrew/bin/brew install glew
And also glm
/opt/homebrew/bin/brew install glm
These instructions were based on Xcode Version 14.2.
Create an Xcode MacOS cli project. Choose C++ for the language.
Add the header search path in Build Settings tab for it to find the binaries.
Click the project at the top of the navigation pane on the left. Choose the Build Settings tab. Choose the project under the PROJECT heading. Use the Filter text input box at the top right to search for "paths". Find the ""Header Search Paths" setting in the "Search Paths" section. Double click the text value area to the right of "Header Search Paths". Click the + (plus) icon inside the popup to enter a new path.
Then add the paths /usr/local/include or /opt/homebrew/include or wherever they are based on Intel or Apple Silicon machine and homebrew install.
- Intel homebrew: /usr/local/include
- Apple Silicon Homebrew: /opt/homebrew/include
Add the binaries you need to link with. To do this,
- Go to project settings,
- Select your target
- Then click "Build Phases" tab,
- Expand "Link Binary with Libraries"
- Click +
Search for and add the OpenGL framework which comes with Xcode.
- Then click + again
- Select "Add Other..."
- Then select "Add files..." from the drop down.
- Hit Command + shift + G
- Type the path for your homebrew libraries: /opt/homebrew/Cellar for Apple silicon
- Go into the direcotory named for the specific binary(i.e. glew)
- Go into the version subdirectory, /opt/homebrew/Cellar/glfw/3.3.8
- Go into the "lib" subdirectory within that version.
- Find and choose the .dylib files that does NOT show as an alias.
For example, one of the files on my system is named /opt/homebrew/Cellar/glfw/3.3.8/lib/libglfw.3.3.dylib.
Repeat the above steps for the glew and glfw libraries. The glm library is a header only library, so it doesn't need this step.
If you get a bunch of warnings about signing and team mapping ID when you try to launch you will have to sort out the code signing settings.
I had to sign the library files and also double check some Xcode settings.
In particular, if CODE_SIGN_IDENTITY gets set to '-' in the pbproject file settings it will not launch even if I have signed the libraries with the instructions below.
Check how a library file is signed or not with this:
codesign -d -v /opt/homebrew/opt/glfw/lib/libglfw.3.dylib
find your identity with this:
security find-identity -v -p codesigning
Then sign it with your identity like this:
codesign -s "Apple Development: Your Name (blablah)" <lib_path>
If you get an "already signed" error then add "-f" flag in front of the library path and re-run.
In order to load the shader files using relative paths successfully you need to set the working directory in Xcode for running the program.
From the menu bar pick Product, then Scheme, then Edit Scheme. Click the "Use custom working directory" checkbox. Select the directory with your source code, which contains the "shaders/" subdirectory.
You can just enter the value "${SRCROOT}" here which should do the trick.