On Jul 5, 2016, at 05:18, Adi Roiban <adi@roiban.ro> wrote:
On 3 July 2016 at 20:32, Craig Rodrigues <rodrigc@crodrigues.org <mailto:rodrigc@crodrigues.org>> wrote:
On Sun, Jul 3, 2016 at 9:10 AM, Glyph Lefkowitz <glyph@twistedmatrix.com <mailto:glyph@twistedmatrix.com>> wrote:
For now, let's just bite the bullet and require 100% patch coverage from here on out. If we hit a really nasty case where it really is a significant investment of effort, then maybe we can revisit this discussion and explore a better way to express this exception without losing information about test coverage completely.
Requiring 100% patch coverage sounds reasonable. However, what if the infrastructure for running coverage and uploading reports to codecov.io <http://codecov.io/> isn't working? Running coverage under Pypy is apparently not working ( https://github.com/twisted/twisted/pull/223#issuecomment-228626722 <https://github.com/twisted/twisted/pull/223#issuecomment-228626722> ). This is blocking forward progress on patches to fix the Pypy tests.
I have disabled codecov patch coverage for now as I think that codecov.io <http://codecov.io/> reporting is buggy.
For coverage merge protection please see See https://github.com/twisted-infra/braid/issues/213 <https://github.com/twisted-infra/braid/issues/213> Thanks for documenting this.
This PR was recently rejected due to lack of test coverage (with no other feedback): https://twistedmatrix.com/trac/ticket/5705#comment:15. If we believe codecov is buggy, are we sure that this is actually the case, and it wasn't just a codecov bug? -glyph