100% Test Coverage

[Uzlopak]

Obligatory issue.

Also: If we have 100% test coverage, we could drop codecov.

[mertcanaltin]

Wouldn't it be good to keep the codecov for future testing?

[Uzlopak]

We can use c8 and configure it so that it return non 0 return value when running in the CI/CD Pipeline. So if we have a PR without a proper test, we could nag about the missing test. Codecov is only necessary if we have below 100% coverage.

[mertcanaltin]

We can use c8 and configure it so that it return non 0 return value when running in the CI/CD Pipeline. So if we have a PR without a proper test, we could nag about the missing test. Codecov is only necessary if we have below 100% coverage.

I got it, thank you very much

[KhafraDev]

I don't see the benefit in having 100% code coverage.

[Uzlopak]

I am not saying, that you personally should start implementing tests.

E.g. I am still writing tests for cookies and I constantly find bugs. And I can not fix them, because upstreaming in deno is basically impossible, because I see that they cant even invest time to review the PR to fix the maxAge bug. I will open for the cookies problem a separate issue. But yeah, it is annoying.

We need to investigate with the tests if the underlying code is actually working or not. So it is not only achieve 100% test coverage but also investigate if the logic is valid or not.

[KhafraDev]

100% test coverage and validating logic are mutually exclusive issues, the latter one I have no problems with. I see it as not only a false sense of security, but also a waste of time.

[Uzlopak]

I disagree. Code which is not covered by any test has imho the inherent assumption to be erroneous or being obsolete.

[KhafraDev]

I consider tests that exist for the sake of code coverage erroneous and obsolete 😉

[Uzlopak]

We agree to disagree