Should we filter out known testing directories by default?

[PragTob]

By default SimpleCov also tracks "coverage" for the run tests (spec/test/features directories) - which has always confused me as that pretty much only tests "are we running all of our tests?" (which sometimes leads to surprising results but that's a different topic...).

We also do the same thing in our own test (aka we assert against the specs and all of them are 100% covered).

I find this to always artificially inflate coverage numbers as of course way more code is executed but that's not what you really want to check for - you care about your application code!

So one of the first things I do in any code base personally is add_filter "spec" and add_filter "features" or what not.

Of course if we did this, we'd also need a nicer way to remove filters/profiles again #803

What do you think? Would that be a good default? Yes/No/Maybe?

@colszowka @deivid-rodriguez @bf4

[deivid-rodriguez]

I personally prefer not to filter out testing directories, because it has helped me with tracking dead code inside my test suite. But yeah, it is true that the percentages are a bit inflated when you do that.

[bf4]

Seems reasonable to either have some sugar to opt in to the profile or otherwise document it in the readme

…

On Sat, Jan 11, 2020, 8:08 AM David Rodríguez ***@***.***> wrote: I personally prefer not to filter out testing directories, because it has helped me with tracking dead code inside my test suite. But yeah, it is true that the percentages are a bit inflated when you do that. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#816>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AABC4QT2SLTMW7IAERVSGZLQ5HHF7ANCNFSM4KFSEV3Q> .