"Youse, Charles" <charles.youse@...> writes:
We have a process in place. It failed here due to lack of coverage,
which unfortunate oversight and I'm very sorry for it, especially for
those who ran into it. However, it was lack of coverage, and not lack of
process, that was to blame.
West has a test suite that works at both API and functional levels. The
repository CI runs it on each commit (on Linux, via shippable). We also
run it on supported platforms before cutting a release. In this case,
that's Windows 10 and macOS. (This is in addition to informal testing
that people are doing on the code in the master branch, which isn't
relevant to this point since you are asking about process.)
More testers and attention to testing are always welcome, so feel free
to make any concrete suggestions for how we can improve our QA.