> I don't think research papers normally come with a simple portable way for others to rerun the calculations.
...which, for situations where a readable/narrated test suite is entirely possible, is awful.
...which, for situations where a readable/narrated test suite is entirely possible, is awful.