On Sun, Mar 10 2019, Luis Ressel wrote: > POSIX doesn't specify the flushing behaviour of the STDOUT stream, so > it's invalid to assume a particular order between the stdout and stderr > output. The current test breaks on musl due to this. > --- > test/T030-config.sh | 7 +++++-- > 1 file changed, 5 insertions(+), 2 deletions(-) > > diff --git a/test/T030-config.sh b/test/T030-config.sh > index f36695c6..9404390b 100755 > --- a/test/T030-config.sh > +++ b/test/T030-config.sh > @@ -43,7 +43,9 @@ notmuch config set foo.nonexistent > test_expect_equal "$(notmuch config get foo.nonexistent)" "" > > test_begin_subtest "List all items" > -notmuch config list 2>&1 | notmuch_config_sanitize > OUTPUT > +notmuch config list 2>OUTPUT-ERR | notmuch_config_sanitize > OUTPUT > +echo "====" >> OUTPUT > +notmuch_config_sanitize < OUTPUT-ERR >> OUTPUT IMO the above would be better (model/pattern for further use) as: notmuch config list >STDOUT 2>STDERR { cat STDOUT; echo "===="; cat STDERR; } | notmuch_config_sanitize > OUTPUT (I would like to put that cat-echo-cat into function but naming is hard...) alternatives: 1) drop ====, then one cat STDOUT STDERR ... would be enough 2) printf '%s====\n%s' "$(< STDOUT)" "$(< STDERR)" 3) head -1000 STDOUT STDERR -- in case of (3) are we sure that all head(1) implementations print the filenames (and formats (possible) extra newlines) the same way. -- ( (2) brings $(< ...) (faster replacement for $(cat ...) which we haven't used in notmuch test suite so far -- but we could!) Tomi _______________________________________________ notmuch mailing list notmuch@notmuchmail.org https://notmuchmail.org/mailman/listinfo/notmuch