https://pantsbuild.org/ logo
#general
Title
# general
h

happy-kitchen-89482

08/23/2019, 8:08 PM
I've been looking at our approach to running coverage, and it can be confusing. Basically if you have B_lib that depends on A_lib, and you run B_tests, we'll take all the sources in B_lib and A_lib for coverage measurement, but since we're not running A_tests, we'll get skewed numbers - there's no reason to expect B's tests to provide good coverage on some dependency...