From: Junio C Hamano Date: Tue, 17 Nov 2026 01:21:32 GMT Subject: Re: [PATCH v4 10/12] builtin/rerere: turn "gc" into a proper subcommand Message-ID: In-Reply-To: <20261111-pks-rerere-subcommands-v4-10-df0c8bf79e16@pks.im> Patrick Steinhardt writes: > As explained in a preceding commit, a subsequent commit will introduce > proper subcommands as supported by our parse-options subsystem. Refactor > the "gc" subcommand to use parse-options as a preparatory step. Even though the cover letter may have mentioned it, this should also describe the deliberate UI regression to help future readers of "git log". When they see a bug report on the mailing list and bisect leads to this commit, it makes a huge difference if you tell them that the change in behavior was deliberate, with a clear reason explained in the log message. Otherwise they would say "the UI regressed with this commit without explaining why it was thought to be a good idea. The commit unceremoniously drops a test that points out this exact regression". > Signed-off-by: Patrick Steinhardt > --- > builtin/rerere.c | 43 ++++++++++++++++++++++++++++++------------- > t/t4200-rerere.sh | 5 ----- > 2 files changed, 30 insertions(+), 18 deletions(-) > diff --git a/t/t4200-rerere.sh b/t/t4200-rerere.sh > index ab412ecc6e8..0d4cbf5114a 100755 > --- a/t/t4200-rerere.sh > +++ b/t/t4200-rerere.sh > @@ -263,11 +263,6 @@ test_expect_success 'gc --skip-locked prunes while MERGE_RR is not locked' ' > test_path_is_missing $rr2/preimage > ' > > -test_expect_success '--skip-locked is only accepted by gc' ' > - test_must_fail git rerere --skip-locked clear 2>err && > - test_grep "option .--skip-locked. requires .gc." err > -' > - > test_expect_success 'a held lock is waited out within rerere.lockTimeout' ' > git reset --hard && > rm -rf $rr &&