What impact does Coenobita have on build and runtime? How many lines must be changed for a crate to be capability safe? Those are the questions we intend to answer with these case studies.
You can measure Coenobita's impact on build and runtime by running the benchmark.py script. For example, the command below will measure the build time of every crate in this directory using Coenobita.
python3 benchmark.py build --wrapper coenobita --iters 5
python3 benchmark.py [build|test] [OPTION]
-w, --wrapper | Which `rustc` wrapper to use
-i, --iters | How many iterations to perform (default is 10)
-r, --root | The root directory (default is '.')
-l, --log | Which log level to use (default is none)