On Mon, Sep 20, 2021 at 8:18 AM Pádraig Brady <p...@draigbrady.com> wrote: > On 20/09/2021 15:19, Jim Meyering wrote: > > On Mon, Sep 20, 2021 at 3:14 AM Pádraig Brady <p...@draigbrady.com> wrote: > >> On 20/09/2021 06:46, Jim Meyering wrote: > >>> I noticed/fixed a spurious test failure on OS X: > >> > >> Oh interesting. > >> This seems like a more general issue. > >> Should we do this more centrally by adding to the list in > >> tests/envvar-check > > > > Adding that name to the list in tests/envvar-check would have no > > effect, since this case is not about inheriting from the original > > environment, but rather about something that's inserted upon each > > "env" invocation. > > I don't fully understand where that env var is being set, > but don't see a specific issue with the patch, so please push.
Done. > BTW that test should not be dependent on the host env(1) since: > https://git.sv.gnu.org/gitweb/?p=coreutils.git;a=commitdiff;h=c8539d36c Right. it's not. I confess I did not root-cause, but it seems to be something deeper. That envvar, __CF_USER_TEXT_ENCODING is referenced e.g., here https://developer.apple.com/library/archive/technotes/tn2228/_index.html and associated with a file, ~/.CFUserTextEncoding.
