# [PATCH] silence git gc --auto --quiet output 4 messages from 2012-09-24 to 2012-09-24. Participants: Tobias Ulmer, Jeff King, Junio C Hamano. Thread: https://gitlist.dev/t/31631 ## Tobias Ulmer, 2012-09-24 02:40 Subject: [PATCH] silence git gc --auto --quiet output Message-ID: <20120924024024.GA12700@tin.tmux.org> URL: https://gitlist.dev/e/20120924024024.GA12700%40tin.tmux.org ``` When --quiet is requested, gc --auto should not display messages unless there is an error. Signed-off-by: Tobias Ulmer --- builtin/gc.c | 4 +--- 1 files changed, 1 insertions(+), 3 deletions(-) diff --git a/builtin/gc.c b/builtin/gc.c index 6d46608..6be6c8d 100644 --- a/builtin/gc.c +++ b/builtin/gc.c @@ -217,9 +217,7 @@ int cmd_gc(int argc, const char **argv, const char *prefix) */ if (!need_to_gc()) return 0; - if (quiet) - fprintf(stderr, _("Auto packing the repository for optimum performance.\n")); - else + if (!quiet) fprintf(stderr, _("Auto packing the repository for optimum performance. You may also\n" "run \"git gc\" manually. See " -- 1.7.6 ``` ## Jeff King, 2012-09-24 18:36 Subject: Re: [PATCH] silence git gc --auto --quiet output Message-ID: <20120924183641.GA15858@sigill.intra.peff.net> URL: https://gitlist.dev/e/20120924183641.GA15858%40sigill.intra.peff.net In-Reply-To: <20120924024024.GA12700@tin.tmux.org> ``` On Mon, Sep 24, 2012 at 04:40:24AM +0200, Tobias Ulmer wrote: > When --quiet is requested, gc --auto should not display messages unless > there is an error. > [...] > - if (quiet) > - fprintf(stderr, _("Auto packing the repository for optimum performance.\n")); That makes sense to me. I wonder if we should also be more careful about propagating quiet flags from the callers of "gc --auto". For example, "git am -q" will run "git gc --auto" without the quiet flag at the end, potentially producing unwanted output. -Peff ``` ## Junio C Hamano, 2012-09-24 22:42 Subject: Re: [PATCH] silence git gc --auto --quiet output Message-ID: <7v3927xcu1.fsf@alter.siamese.dyndns.org> URL: https://gitlist.dev/e/7v3927xcu1.fsf%40alter.siamese.dyndns.org In-Reply-To: <20120924024024.GA12700@tin.tmux.org> ``` Tobias Ulmer writes: > When --quiet is requested, gc --auto should not display messages unless > there is an error. > > Signed-off-by: Tobias Ulmer > --- > builtin/gc.c | 4 +--- > 1 files changed, 1 insertions(+), 3 deletions(-) > > diff --git a/builtin/gc.c b/builtin/gc.c > index 6d46608..6be6c8d 100644 > --- a/builtin/gc.c > +++ b/builtin/gc.c > @@ -217,9 +217,7 @@ int cmd_gc(int argc, const char **argv, const char *prefix) > */ > if (!need_to_gc()) > return 0; > - if (quiet) > - fprintf(stderr, _("Auto packing the repository for optimum performance.\n")); > - else > + if (!quiet) > fprintf(stderr, > _("Auto packing the repository for optimum performance. You may also\n" > "run \"git gc\" manually. See " This patch will break t5400; the test needs to be updated in the same patch to check auto-gc kicks in when it should in some other way. The test currently _relies_ on this message to see the gc is triggered. ``` ## Junio C Hamano, 2012-09-24 22:53 Subject: Re: [PATCH] silence git gc --auto --quiet output Message-ID: <7vy5jzvxq4.fsf@alter.siamese.dyndns.org> URL: https://gitlist.dev/e/7vy5jzvxq4.fsf%40alter.siamese.dyndns.org In-Reply-To: <7v3927xcu1.fsf@alter.siamese.dyndns.org> ``` Junio C Hamano writes: > Tobias Ulmer writes: > >> When --quiet is requested, gc --auto should not display messages unless >> there is an error. >> >> Signed-off-by: Tobias Ulmer >> --- >> builtin/gc.c | 4 +--- >> 1 files changed, 1 insertions(+), 3 deletions(-) >> >> diff --git a/builtin/gc.c b/builtin/gc.c >> index 6d46608..6be6c8d 100644 >> --- a/builtin/gc.c >> +++ b/builtin/gc.c >> @@ -217,9 +217,7 @@ int cmd_gc(int argc, const char **argv, const char *prefix) >> */ >> if (!need_to_gc()) >> return 0; >> - if (quiet) >> - fprintf(stderr, _("Auto packing the repository for optimum performance.\n")); >> - else >> + if (!quiet) >> fprintf(stderr, >> _("Auto packing the repository for optimum performance. You may also\n" >> "run \"git gc\" manually. See " > > This patch will break t5400; the test needs to be updated in the > same patch to check auto-gc kicks in when it should in some other > way. The test currently _relies_ on this message to see the gc is > triggered. I suspect that the following may be sufficient. The test prepares a stale garbage file in the repository an auto-gc is expected to happen, and makes sure the garbage file is removed after the operation that expects to trigger an auto-gc. The detection of the message is more or less superfluous. t/t5400-send-pack.sh | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git i/t/t5400-send-pack.sh w/t/t5400-send-pack.sh index 250c720..78ab177 100755 --- i/t/t5400-send-pack.sh +++ w/t/t5400-send-pack.sh @@ -174,8 +174,7 @@ test_expect_success 'receive-pack runs auto-gc in remote repo' ' cd parent && echo "Even more text" >>file.txt && git commit -a -m "Third commit" && - git send-pack ../child HEAD:refs/heads/test_auto_gc >output 2>&1 && - grep "Auto packing the repository for optimum performance." output + git send-pack ../child HEAD:refs/heads/test_auto_gc ) && test ! -e child/.git/objects/tmp_test_object ' ```