Somebody should compile everything (make sure the build system works), and then run the tests (and make sure the test script works). Any algorithms that are broken should have issues made.
We also need to go through all algorithms manually to make sure an edge case hasn't been missed in both the implementation and the test. Or if an edge case exists that isn't tested, but does work, that should be added too to be more thorough.
Somebody should compile everything (make sure the build system works), and then run the tests (and make sure the test script works). Any algorithms that are broken should have issues made.
We also need to go through all algorithms manually to make sure an edge case hasn't been missed in both the implementation and the test. Or if an edge case exists that isn't tested, but does work, that should be added too to be more thorough.