Remove timestamp from non-watch test runs

[sotojuan]

Fixes #775.

WIP because I need to add some tests + need more review especially for testing. Would it be better to remove the time from all the modified tests if I have some specific tests that do test time on watch?

[novemberborn]

Would it be better to remove the time from all the modified tests if I have some specific tests that do test time on watch?

Yea. Having to add the time in all those tests should've been a clue that time was being output too much.

---

Passing the boolean is a bit awkward, but I think it's OK. Ideally the logger understands test runs, so we don't have to interact with it as precisely as we do in the watcher.

[sotojuan]

Fixed tests.

[novemberborn]

Can probably move these two lines into the new test now.

LGTMO!

[jamestalmage]

I think we should pass an options object to the reporter constructor with static details/config like isWatching.

It definitely shouldn't be a boolean param. {watching: true}

[sotojuan]

Good idea—just pushed some changes to make that work.

[jamestalmage]

isWatching => watching

[jamestalmage]

Just so we're all on the same page about reporter constructor options. IMO, I think the reporters should avoid maintaining any mutable state, especially as it relates to the test run status. We are violating that rule in a couple spots now (my fault, I was supposed to follow up after merging the PR that introduced RunStatus).

I do think it's fine to pass immutable config values that will never change during the test run (i.e. watching, basedir, pkgConf, debugging, etc). That should keep our reporters as state free and easily testable as possible, without necessitating passing around needless config data.

[sotojuan]

Pushed the changes (and thanks for the explanation, makes sense).

[jamestalmage]

@sotojuan

Move the two lines @novemberborn talked about here into the test that uses it.

Once you've done that (and tests pass), go ahead and merge.