There is no good visual distinction between different types of failure states

This is true for both workers and jobs.

Whether something fails, is aborted, times out (other conditions?) you get a red X. You have to dig into the YAML to see the true status.

We should either come up with distinct visual indicators for each state or keep the red X and add a textual description of the state to the summary tab.

In this example, the `test-unit-go` job was aborted, but it just _looks_ like a failure.