Re: [PATCH v2 00/12] fsmonitor: Implement fsmonitor for Linux

> Eric Sunshine <sunshine@sunshineco.com> writes:
> > On Thu, Oct 20, 2022 at 12:13 PM Junio C Hamano <gitster@pobox.com> wrote:
> >> All other jobs were using the default DC_SHA1 (which is a
> >> recommended practice), but the default for macOS jobs being Apple's
> >> common crypt, we didn't catch recent breakage soon enough.
> >
> > "recent breakage" is quite vague and probably won't help future
> > readers understand what this is actually fixing. Possible
> > improvements: (1) a prose description of the breakage; (2) the actual
> > compiler error message; (3) a pointer[1] to the email reporting the
> > problem. One or more of the above improvements to the commit message
> > would help future readers.
>
> I do not think (2) or (3) would help all that much.  A finger that
> points at the exact commit that broke the build (with the condition
> under which the build breaks) would probably be the most useful to
> those who read "git log" output.
>
> ----- >8 --------- >8 --------- >8 --------- >8 --------- >8 -----
> Subject: [PATCH] ci: use DC_SHA1=YesPlease on osx-clang job for CI
>
> 7b8cfe34 (Merge branch 'ed/fsmonitor-on-networked-macos',
> 2022-10-17) broke the build on macOS with sha1dc by bypassing our
> hash abstraction (git_SHA_CTX etc.), but it wasn't caught before the
> problematic topic was merged down to the 'master' branch.  Nobody
> was even compile testing with DC_SHA1 set, although it is the
> recommended choice in these days for folks when they use SHA-1.
>
> This was because the default for macOS uses Apple Common Crypto, and
> both of the two CI jobs did not override the default.  Tweak one of
> them to use DC_SHA1 to improve the coverage.