With Jest, is it ok to consider each `describe()` as a linear scenario in which each `it()` relies on the state (of the tested system) created by previous `it()` test cases?

Example: describe('jobs-runner', () => { it('can accept a new job', () => { system.add({ job: DUMMY }); expect(system.getJobs()).toHaveLength(1); }); it('can delete a job', () => { system.delete({ job: DUMMY }); expect(system.ge...