this would allow easier detection of what's failing and follows the principle of one assertion per test.