Patrick Ohly

@thediveo: I prefer separate functions because it enables compile-time checking and simplifies linting.

@nunnatsa: if we go the route of extending `Be[True|False]`, would it be okay to extend ginkgolinter so that it optionally (and by default off) warns about plain calls without description?...

How about keeping `Be[True|False]` as-is (no description) and adding `func BeTrueF(format string, args....)`? This is not a pattern used in Gomega, but not unusual elsewhere. The advantage is that the...

I was about to comment that `go vet` won't like single-string parameters because it always expects printf-style, but you already found that! :grin: The commit in master looks good to...

See https://github.com/kubernetes/kubernetes/pull/112013 for an implementation outside of ginkgo and gomega.

> sounds like perhaps the timeout is too high? We have to have very high timeouts in Kubernetes E2E testing because sometimes an operation simply takes longer on a busy...

m_wildcardValue was just the tip of the iceberg. Several other global instances had similar issues. Some of that code is also used in cynarad, so this issue might go beyond...

> The second patch if I understand static variable initialization correctly makes it less likely to fail, > because there are less variables that we depend on. However, we still...

I've tried std::locale::classic() and that also works. I suggest that you just go ahead and commit a change created with sed. There's not much value in preparing a full patch.

> We still need separate common library repos and tooling since csi drivers are using them Those could be included in a single repo via individual go.mod files in sub-directories....