Andre Brisco

The goal isn't to solidify on an old release, it's to be able to assess what code I'm picking up when I update. It would be very helpful to have...

> Why not just use `coverage run` directly instead of pytest-cov? `pytest-cov` integrates with [pytest-dist](https://github.com/pytest-dev/pytest-xdist) better. I was not able to easily collect coverage running `coverage run`.

@ionelmc friendly ping, does something like this sound reasonable to support?

> Having coverage always on? Absolutely not. Most users do not desire something like that. No, a flag to force coverage to turn on in cases where a coveragerc has...

> If pytest-cov is misbehaving for you, try using coverage.py without it: > > ``` > coverage run -m pytest > ``` I had originally tried to run coverage directly...

I revisited my need for this feature and was able to get the plugin to behave as I had wanted. I'm not sure what I'd done before to cause it...

I'm running into this trying to build a release artifact for [14.0.6](https://github.com/llvm/llvm-project/tree/llvmorg-14.0.6) on a [Redhat-8.4](https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/8/html/8.4_release_notes/index) machine. Anyone have a suggested work around? ``` ./llvm/utils/release/test-release.sh -no-test-suite -release 14.0.6 -triple x86_64-linux-gnu-rhel-8.4 -final...

I don't know how or why I was able to build the artifact but by only installing the following from a clean RHEL-8.4 VM I was able to build an...

> We already run mypy in strict mode and it did not catch any of this. Uh-oh. Does mypy run on the tests? Also, wouldn't mypy in strict mode disallow...

> No, and the test suite itself is untyped. This might be why this was missed, Unless there's a caller trying to use a function with a different type I...