Christopher Hiller

@Thaina I locked it because you were becoming belligerent. It's difficult to infer tone in written comments, but I'm sensing some of the same here. Please be polite when posting...

FWIW the use case here is *not* to run tests in parallel, but rather to *dynamically generate tests* from data which is retrieved via asynchronous I/O operations. `--delay` gets us...

If `before` or `beforeEach` fails, would you expect the tests to be _run_ or would you just expect them to be _reported_?

Some notes: - If you have a failing hook at or near the root of your test suites, the reporter output will repeat the same error, potentially _many, many_ times;...

I'm not really convinced this casts a wide enough net to be in core... feel free to make the case though. I see this as being specifically useful for certain...

I'm confused about this vs. #4223

@sgilroy Making any built-in reporter report this information is going to be semver-major, and doesn't need to be part of *this* PR, which would be to just *collect* these statistics...

(See the `json` reporter for an example of how to collect information about the tests and output it when the run is complete)

cc @Nokel81; you might want to consider this when looking at the stats gathering.

Thanks. This is a duplicate of...several issues. Can you run `mocha || true` instead? This is how I've seen people treat any result as "passing"