Hi, Le 30/07/2018 à 20:25, SZEDER Gábor a écrit : >> diff --git a/sequencer.c b/sequencer.c >> index 1c035ceec7..d257903db0 100644 >> --- a/sequencer.c >> +++ b/sequencer.c > >> +int write_basic_state(struct replay_opts *opts, const char *head_name, >> + const char *onto, const char *orig_head) >> +{ >> + const char *quiet = getenv("GIT_QUIET"); >> + >> + if (head_name) >> + write_file(rebase_path_head_name(), "%s\n", head_name); >> + if (onto) >> + write_file(rebase_path_onto(), "%s\n", onto); >> + if (orig_head) >> + write_file(rebase_path_orig_head(), "%s\n", orig_head); >> + >> + if (quiet) >> + write_file(rebase_path_quiet(), "%s\n", quiet); >> + else >> + write_file(rebase_path_quiet(), ""); > > This is not a faithful conversion of the original. git-rebase.sh writes > this 'quiet' file with: > > echo "$GIT_QUIET" > "$state_dir"/quiet > > which means that a single newline character was written even when > $GIT_QUIET was unset/empty. > > I seem to recall a case in the past, when a shell-to-C conversion > accidentally dropped a newline from a similar state-file, which then > caused some issues later on. But I don't remember the specifics and a > quick search didn't turn up anything relevant either... > I don’t think it’s a problem here, but we’re better safe than sorry. I’ll send a fix soon. Cheers, Alban