# [PATCH] --walk-reflogs: actually find the right commit by date. 8 messages from 2007-01-20 to 2007-01-21. Participants: Johannes Schindelin, Junio C Hamano. Thread: https://gitlist.dev/t/6451 ## Johannes Schindelin, 2007-01-20 09:49 Subject: [PATCH] --walk-reflogs: actually find the right commit by date. Message-ID: URL: https://gitlist.dev/e/Pine.LNX.4.63.0701201048560.22628%40wbgn013.biozentrum.uni-wuerzburg.de ``` Embarassing thinko. Signed-off-by: Johannes Schindelin --- reflog-walk.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/reflog-walk.c b/reflog-walk.c index d4b49c7..d58ff9b 100644 --- a/reflog-walk.c +++ b/reflog-walk.c @@ -70,7 +70,7 @@ static int get_reflog_recno_by_time(struct complete_reflogs *array, unsigned long timestamp) { int i; - for (i = array->nr - 1; i >= 0; i++) + for (i = array->nr - 1; i >= 0; i--) if (timestamp >= array->items[i].timestamp) return i; return -1; -- 1.5.0.rc1.gd85c ``` ## Johannes Schindelin, 2007-01-20 09:52 Subject: Re: [PATCH] --walk-reflogs: actually find the right commit by date. Message-ID: URL: https://gitlist.dev/e/Pine.LNX.4.63.0701201051320.22628%40wbgn013.biozentrum.uni-wuerzburg.de In-Reply-To: ``` Hi, On Sat, 20 Jan 2007, Johannes Schindelin wrote: > Embarassing thinko. BTW this fixes your "git-log --walk-reflogs -1 --pretty=s master@{1.day}" bug. Ciao, Dscho ``` ## Junio C Hamano, 2007-01-20 10:11 Subject: Re: [PATCH] --walk-reflogs: actually find the right commit by date. Message-ID: <7vhcum80u4.fsf@assigned-by-dhcp.cox.net> URL: https://gitlist.dev/e/7vhcum80u4.fsf%40assigned-by-dhcp.cox.net In-Reply-To: ``` Johannes Schindelin writes: > Embarassing thinko. Aha, progress ;-). But it still seems to pick up entry that is less than specified number of hours if I do: $ git log --walk-reflogs -4 master@{1.hour} | head -2 commit 33dd19c2a9c0b114436f53d86fc1c2c5e00c26bb Reflog: master@{Sat, 20 Jan 2007 00:01:35 -0800} (Junio C Ham It is Jan 20th 02:10 now and the master was not pointing at that commit 1 hour ago. Anyway, I am done for the night -- I haven't finished testing the updated 'master' and 'next', so I'll push out the tip of 'pu' (which contains all of the tentitive 'master' and 'next' changes I haven't verified yet) and call it a day. ``` ## Johannes Schindelin, 2007-01-20 21:36 Subject: Re: [PATCH] --walk-reflogs: actually find the right commit by date. Message-ID: URL: https://gitlist.dev/e/Pine.LNX.4.63.0701202225020.22628%40wbgn013.biozentrum.uni-wuerzburg.de In-Reply-To: <7vhcum80u4.fsf@assigned-by-dhcp.cox.net> ``` Hi, On Sat, 20 Jan 2007, Junio C Hamano wrote: > Johannes Schindelin writes: > > > Embarassing thinko. > > Aha, progress ;-). But it still seems to pick up entry that is > less than specified number of hours if I do: > > $ git log --walk-reflogs -4 master@{1.hour} | head -2 > commit 33dd19c2a9c0b114436f53d86fc1c2c5e00c26bb > Reflog: master@{Sat, 20 Jan 2007 00:01:35 -0800} (Junio C Ham > > It is Jan 20th 02:10 now and the master was not pointing at that > commit 1 hour ago. I tried to reproduce this effect, but so far no luck. I even let the function get_reflog_recno_by_time() output what it does, and it picks exactly the right commit. Do you have a time skew? Ciao, Dscho ``` ## Junio C Hamano, 2007-01-21 07:09 Subject: Re: [PATCH] --walk-reflogs: actually find the right commit by date. Message-ID: <7vhcuk500j.fsf@assigned-by-dhcp.cox.net> URL: https://gitlist.dev/e/7vhcuk500j.fsf%40assigned-by-dhcp.cox.net In-Reply-To: ``` Johannes Schindelin writes: > Hi, > > On Sat, 20 Jan 2007, Junio C Hamano wrote: > >> Johannes Schindelin writes: >> >> > Embarassing thinko. >> >> Aha, progress ;-). But it still seems to pick up entry that is >> less than specified number of hours if I do: >> >> $ git log --walk-reflogs -4 master@{1.hour} | head -2 >> commit 33dd19c2a9c0b114436f53d86fc1c2c5e00c26bb >> Reflog: master@{Sat, 20 Jan 2007 00:01:35 -0800} (Junio C Ham >> >> It is Jan 20th 02:10 now and the master was not pointing at that >> commit 1 hour ago. > > I tried to reproduce this effect, but so far no luck. Sorry, I was reading the timestamp wrong. That is one minute thirty-five seconds past midnight, not thirty five minute past one o'clock. ``` ## Johannes Schindelin, 2007-01-21 10:53 Subject: Re: [PATCH] --walk-reflogs: actually find the right commit by date. Message-ID: URL: https://gitlist.dev/e/Pine.LNX.4.63.0701211152400.22628%40wbgn013.biozentrum.uni-wuerzburg.de In-Reply-To: <7vhcuk500j.fsf@assigned-by-dhcp.cox.net> ``` Hi, On Sat, 20 Jan 2007, Junio C Hamano wrote: > Sorry, I was reading the timestamp wrong. That is one minute > thirty-five seconds past midnight, not thirty five minute past one > o'clock. Wow! I missed that completely! Ciao, Dscho ``` ## Junio C Hamano, 2007-01-21 11:19 Subject: Re: [PATCH] --walk-reflogs: actually find the right commit by date. Message-ID: <7vac0c39uw.fsf@assigned-by-dhcp.cox.net> URL: https://gitlist.dev/e/7vac0c39uw.fsf%40assigned-by-dhcp.cox.net In-Reply-To: ``` Johannes Schindelin writes: > On Sat, 20 Jan 2007, Junio C Hamano wrote: > >> Sorry, I was reading the timestamp wrong. That is one minute >> thirty-five seconds past midnight, not thirty five minute past one >> o'clock. > > Wow! I missed that completely! Sorry for that. It was late and I was tired. By the way, "git log --walk-reflog master@{now}" seems to give me one non-reflog entry output for me at the end (which is I think the parent of the oldest reflog entry which itself does not have the record of being at the tip of the ref in question). Other than that I think we are Ok. Also doesn't --walk-reflog sound too long? ``` ## Johannes Schindelin, 2007-01-21 18:14 Subject: Re: [PATCH] --walk-reflogs: actually find the right commit by date. Message-ID: URL: https://gitlist.dev/e/Pine.LNX.4.63.0701211913060.22628%40wbgn013.biozentrum.uni-wuerzburg.de In-Reply-To: <7vac0c39uw.fsf@assigned-by-dhcp.cox.net> ``` Hi, On Sun, 21 Jan 2007, Junio C Hamano wrote: > By the way, "git log --walk-reflog master@{now}" seems to give me one > non-reflog entry output for me at the end (which is I think the parent > of the oldest reflog entry which itself does not have the record of > being at the tip of the ref in question). Other than that I think we > are Ok. Yes, I realized that. It might bear some information, but if you don't like it, I can change that. > Also doesn't --walk-reflog sound too long? Yes. Suggestions? (--reflog is already taken :-( ) Ciao, Dscho ```