Coworker and I are at an impasse and would appreciate the community’s vote to resolve it.

Elixir’s test-runner has a --trace option which does the following:

1. it discards all of Elixir’s concurrency abilities and (slowly) runs everything sequentially, one-by-one
2. it prints the test-names (here, here, etc.), even the passing ones. (The failed tests are, of course, always printed, even without --trace.)
3. sets the test-timeout to :infinity. It’s tracing, so this is expected/welcome (in dev/local environment). But not desired in a CI env as it can blow up the bill/costs.

Here’re coworker and my positions:

• Coworker’s argument for tracing is effectively the 2. above. In his words, it “gives a better visual understanding which tests are running”.
• My position against it are that:
  • Not only are downsides of 1. and 3. not worth it, but one can also
  • just look at the test file to see which tests are running. Moreover, one shouldn’t even care about the names of the passing tests, abiding by the Rule of Silence.