There are many C/C++ open source projects there. We don't re-invent wheels, but we use these wheels. As these third party projects are just consumed by our project, instead of modifying them (most cases), it might be a good practice to just use their "packages", i.e. header files, library files, instead of their source code. The generated header files + library files, is called artifacts.
How do we get third-party project's artifacts? Some open source project authors provide prebuilt ones, other doesn't. And you may also customize some options for building instead of the prebuilt one. So, let's build and install 3rd-party project artifacts manually.
In these directory, I put my daily used packages' building and installing steps(scripts). Hope they can insight you.
I suggest every C/C++ developer keep an local "artifacts" directory, to store 3rd-party packages. It's OK for one project to put artifacts under its own directory, but common packages like OpenCV, GoogleTest, if you just use it without customization, I prefer reusing artifacts, hence, a global artifacts storing directory is quite useful.
Say, on Linux / MacOS, I use:
export ARTIFACTS_DIR=~/artifactsAnd under $ARTIFACTS_DIR, each subdirectory is an package's name, then next level is its version, next level is platform/architecture.
And all the building and installing examples under this directory, will install artifacts to ${ARFIFACTS_DIR}.
By using ${ARTIFACTS_DIR} as prefix to packages, we have cross-machine same experience: for same package with same version and same platform, we use same path variables on different machines. Quite standard.
Before running these building and installing scripts, people should manually git clone reops, and manually choose branch / tags, then modify building scripts, specifing proper installing directory for correct version.