IRCloggy #git 2008-02-07

Provider of IRC logs since 2005.

WARNING: As Freenode became unjoinable and lost all warnings in topics, we cannot log channels on Freenode anymore.

	2008-02-07
	← zapnap left	00:00
	→ robinr joined	00:02
	← cannonball left	00:02
	← ferdy left	00:03
	← Cerebon left	00:03
	← ReinH left	00:04
	→ enderson joined	00:04
	← devogon left	00:05
	← Ryback_ left	00:05
	→ _kukks_ joined	00:10
	← cnantais left	00:10
	← enderson left	00:10
	→ FunkeeMonk joined	00:16
	← HG` left	00:20
	← FunkeeMonk left	00:20
	← mithro left	00:24
	→ mithro joined	00:24
	→ semi_ joined	00:26
semi_	is there some simple way to do "git-push --tags ; git-push" in one go?	00:26
	← \|kukks\| left	00:27
	→ aazar joined	00:28
aazar	hey guys, any command symmetrical to git add -u to unstage all files currently staged?	00:28
	git reset...?	00:28
jeffpc	git reset --soft might do it	00:29
	not sure	00:29
	→ izittm joined	00:29
semi_	man git-reset \| grep -A3 -e " --mixed"	00:29
aazar	jeffpc: nah, according to the doc it wouldn't	00:29
wildfire	aazar, git reset --mixed	00:29
aazar	it won't have any other effect than to unstage staged files?	00:30
semi_	run that command to get explanation	00:30
wildfire	aazar, correct. mixed leave the working directory intact and reset the index	00:30
aazar	thansk a lot guys	00:30
	semi_: I'm on msysgit, with no man... :S	00:31
	semi_: although the doc on the web says the same thing. I was just afraid it might have another effect.	00:31
semi_	I usually create test repositories in /tmp to try things out if I'm unsure :)	00:32
aeruder	git reset HEAD will only touch the index	00:33
	not the working directory	00:33
	(or git reset --mixed HEAD for that matter)	00:33
	if you're really worried about it you can always commit it, git reset --hard HEAD^	00:33
	er	00:34
	git reset HEAD^	00:34
	and regardless you have your old commit saved for the time being	00:34
semi_	or just commit it, make changes, and commit again with git-commit --amend to make that old commit disappear	00:39
aeruder	yea that works too	00:39
	i use --amend a lot	00:39
	← halfline left	00:41
	← careo left	00:45
	← scroogey left	00:46
fujin	could someone take a quick look at this? http://rafb.net/p/QCxT2h65.html	00:47
	trying to work out what I'm doing wrong to cause this.	00:47
	→ halfline joined	00:49
	← mugwump left	00:49
	← semi_ left	00:50
	← timocratic left	00:51
	← tvachon left	00:54
loops	fujin, there may be more than one issue here.. but for a start you can't "git pull" into a bare repository, only "git fetch"	00:55
fujin	oh, hrm.	00:58
	i'm trying to work out how to do this properly.	00:58
	just re-created it as a bare repository with --shared, Ideally I'd just like to be able to push commits from both branches into this bare repo, and then pull all of those changes (pull? merge? fetch?) into that third /etc/puppet work directory	00:59
	← captnmark left	00:59
	→ captnmark joined	00:59
loops	it sounds like you're using Git in adventurous ways	01:01
	→ z3ro_ joined	01:01
fujin	um	01:01
	lol	01:01
	that's not ideal	01:01
	loops: suggest how I should do it correctly?	01:01
loops	but as one small tip.. you don't want to pull (which implies a merge), or merge "automatically".. both those operations sometimes need operator intervention	01:01
	so, i'd do your merging in aj / tony and push the result, which can then be fetched automatically to the other machine	01:02
fujin	eh?	01:03
	merging in aj?	01:03
	I want to merge the changes I've made in aj to master	01:03
gebi	how do i update a workingcopy of a remote git repo automatically to the last revision of a branch?	01:03
	fujin mind boggles	01:03
gebi	just after the git push to the repo	01:03
loops	well, i was speaking of the machines rather than the branch :)	01:03
fujin	oh..	01:03
	so, I should checkout master on aj/tony, then merge, then push?	01:03
	and then just pull master on the third one?	01:04
loops	gebi, there's nothing in Git to do that for you.	01:05
	fujin, that should work better, although i haven't really thought through your entire setup.. (and i'm just running out the door)	01:07
gebi	loops: which hook would be the right place	01:07
	post-update doesn't work	01:07
fujin	loops: ok, as you said, that is working better - thanks :)	01:07
	← mithro left	01:08
fujin	when a user connects over git:// protocol, do they need any permissions on the directory?	01:09
	or does it just have to be owned/grouped to the user/group of the git-daemon	01:10
	← z3ro left	01:12
	→ mithro joined	01:14
	← Yuuhi left	01:14
aeruder	fujin: the latter	01:15
duncanm	What do i do if i want to insert a new commit in the past?	01:18
	oooh	01:19
	hmm	01:19
	_kukks_ → kukks	01:20
gebi	loops: http://utsl.gen.nz/git/post-update	01:27
	timj → tjafk	01:29
	→ kanru joined	01:29
	tjafk → timj	01:29
	timj → tjafk	01:30
	tjafk → timj	01:30
robinr	duncanm: rebase	01:32
	rebase -i	01:33
	→ tarbo_ joined	01:33
duncanm	bah, i lost a commit again ;-(	01:34
	← Alan_ left	01:35
	← cnantais_ left	01:38
	← ecdpalma left	01:43
	→ kevwil joined	01:44
	← tarbo left	01:44
	→ vintrepid joined	01:49
	→ mpagano\|work joined	01:49
mpagano\|work	how can I identity the -gitX snapshot that a specific patch is contained within?	01:51
aeruder	that's a good question, but i'm not sure exactly where those gitX snapshots are available via git (i.e. it'd be nice if there were gitX tags made somewhere...)	01:52
gitster	That depends on how "-gitX snapshot" is created, doesn't it?	01:52
aeruder	but what I've done in the past is the crappy solution of just looking at commit dates	01:52
	vs. when the release date of the gitX snapshots	01:52
gitster	Is this about the kernel?	01:52
aeruder	yea	01:52
mpagano\|work	yes	01:52
gitster	what does a "gitX snapshot" come with?	01:53
	Just a tarball and nothing else?	01:53
aeruder	well afaik they are just a patch on kernel.org	01:53
	it may give a sha1 somewhere	01:53
	← ebzzry left	01:53
mpagano\|work	so I'm trying to help a user who needs a specific patch, and I want to tell him to test with 2.6.24-gitX	01:53
gitster	Is it created with "git archive", so that you can run "git get-tar-commit-id" on it?	01:53
mpagano\|work	but I can just grab the next gitX since the patche sin question were committed today	01:54
	→ bdiego joined	01:54
aeruder	gitster: well, maybe... but still, kinda a pain to automate the process of downloading 14 or whatever # of gitX snapshots	01:54
	mpagano\|work: ah	01:54
	in that case, I'd just manually do it	01:54
	I thought you were trying to narrow it down really closely	01:55
	just look at the time stamp	01:55
mpagano\|work	for instance, this one: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=1ec4a9398dc05061b6258061676fede733458893	01:55
	I was hoping there was an easy way to say. Ah its in git15 .	01:56
aeruder	ok so that one was	01:56
	← kevwil left	01:56
aeruder	Wed, 6 Feb 2008 18:41:18	01:56
	whereas the snapshot was made at 07:01	01:56
	so nope	01:56
	git16 :)	01:56
mpagano\|work	ok, great, that's a method I'll use for now on. Thanks!	01:57
	you guys are great	01:57
aeruder	mpagano\|work: it'd be really easy if there was a place in git that had those tag objects	01:57
	like someone actually went through and tagged 2.6.24-git15 etc..	01:57
	but I don't think anyone has	01:57
mpagano\|work	i would love that	01:57
	yeah, it doesn't appear so	01:57
aeruder	so if you're just doing a one off thing, timestamps are just as easy	01:57
mpagano\|work	aeruder: tks, will do	01:58
	← robinr left	01:58
aeruder	gitster: nah, they just release a snapshot patch-2.6.24-git15.bz2 thing from time to time	01:59
	→ agoode joined	02:00
aeruder	afaik there's no way to really get a gitX snapshot => sha1 without doing the hokey snapshot method	02:00
	(or downloading the patch, applying it, comparing tree sha1's or something)	02:01
	→ garrin joined	02:09
	← z3ro_ left	02:10
	→ kevwil joined	02:10
	→ z3ro joined	02:10
	← kevwil left	02:12
	← aroben left	02:14
	← agib left	02:17
	← p4tux left	02:17
	→ agib joined	02:18
	← vintrepid left	02:30
	→ tongueroo_ joined	02:37
	→ casual_keith joined	02:39
	← casual_keith left	02:43
	← keithpitty left	02:54
	→ stick joined	03:01
	→ brothers joined	03:04
	← tongueroo left	03:07
	← KirinDave left	03:09
	tarbo_ → tarbo	03:09
	← z3ro left	03:10
	← kukks left	03:12
	→ z3ro joined	03:14
	→ tongueroo joined	03:24
	→ tarbo_ joined	03:24
	← tongueroo left	03:25
	← tarbo left	03:34
	→ tongueroo joined	03:34
	← bobesponja left	03:39
	→ bobesponja joined	03:40
	← tongueroo_ left	03:40
	→ Bluefury joined	03:44
	← kmap left	03:46
	→ kmap joined	03:47
	← wildfire left	03:49
	← csc` left	03:50
	→ ironfroggy joined	03:51
	← ironfroggy left	03:51
	→ ironfroggy joined	03:52
	← tongueroo left	03:52
	→ albertito joined	03:53
fujin	just need a quick point in the right direction, I've created a bare repository (web.git) and I want to add submodule repositories to it, but, I can't clone it to add files to it	03:54
	→ FunkeeMonk joined	03:59
fujin	does the submodule have to have a seperate remote origin?	04:00
	blagh, nevermind submodules I'll just use seperate projects.	04:02
	← bdiego left	04:13
	→ Beket joined	04:20
cehteh	fujin: want to use this web.git as kindof container for unrelated projects?	04:21
	← mithro left	04:22
cehteh	if yes, sorry thats not really supported, submodules have to point to a specific commit/revision	04:22
	→ z3ro_ joined	04:25
	← FunkeeMonk left	04:26
	← brothers left	04:26
	→ mithro joined	04:27
	← mithro left	04:29
	← pigeon left	04:30
	→ pigeon joined	04:30
	← sgrimm left	04:30
fujin	cehteh: well, they're kind of related - they'll all be web projects	04:31
cehteh	fujin: submodule stores the 'url' of the remote repository and one exact commit at a time	04:32
	→ wildfire joined	04:33
	→ z3ro__ joined	04:34
fujin	cehteh: yeah, cool. Going to just use seperate projects	04:34
	← Beket left	04:38
	→ csc` joined	04:38
	→ vbgunz joined	04:39
	→ Cerebon joined	04:40
	← z3ro left	04:43
	← z3ro_ left	04:48
	← Bluefury left	04:49
	← csc` left	04:52
	← jwhitmire left	04:53
	→ FunkeeMonk joined	04:54
	← madewokherd left	04:56
	→ Alam_Debian joined	05:09
	← Cerebon left	05:09
	← izittm left	05:10
	→ z3ro joined	05:11
	→ mithro joined	05:13
	→ sgrimm joined	05:14
	← agib left	05:16
	← loops left	05:16
	→ rubydiamond joined	05:17
	→ loops joined	05:17
Alam_Debian	is there a HOWTO on getting git svn to update it's SVN URL?	05:18
	← z3ro left	05:19
	← rubydiamond left	05:26
	← z3ro__ left	05:29
	→ srid joined	05:29
	→ agib joined	05:37
	→ adante_ joined	05:40
	→ itodd joined	05:43
itodd	i'm trying to import my svn repo using git-svnimport. the command i'm using is: git-svnimport -v -C /var/git/books file:///var/svn/charged/infuse. It appears to read the repo as it prints out a line for every commit prepended with [revision #] Unrecognized path: /path/to/file/in/commit then fails: fatal: master: not a valid SHA1 and fatal: just how do you expect me to merge 0 trees?	05:46
	any ideas? i'm a newbie. this is my first attempt with git	05:46
	← aazar left	05:48
	→ captnmark_ joined	05:48
	← captnmark left	05:50
loops	itodd, git-svnimport is deprecated. Try using "git svn" instead.	05:51
itodd	ahh thanks	05:51
	→ binnyva joined	05:51
	← loops left	05:54
	→ loops joined	05:54
itodd	loops: awsome. working. thanks :)	05:55
	← FunkeeMonk left	05:56
	← adante left	05:56
	adante_ → adante	05:56
	← cwillu left	05:59
	→ kylem_ joined	05:59
	→ cwillu joined	05:59
	→ spearce joined	06:00
binnyva	what entry should i add to info/exclude in ignore ".svn" and "CVS" folders?	06:09
spearce	".svn" and "CVS"	06:09
	:)	06:09
	without quotes	06:09
binnyva	i tried that - its still being added	06:09
spearce	the ignore applies only at time git-add. did you run git-add and then try to put them into the ignore and expect it to change?	06:10
binnyva	no- wait - i tried .svn/*	06:10
	i did not try .svn	06:10
	1 sec	06:10
	yup - it work!	06:11
	thanks	06:11
	← amf left	06:16
	← kylem left	06:16
	← mithro left	06:17
	→ amf joined	06:17
	← stick left	06:19
	← srid left	06:19
	→ mithro joined	06:21
	← binnyva left	06:51
	→ devogon joined	06:55
	→ TuxCM- joined	07:00
	→ namenlos joined	07:01
	→ alb joined	07:02
	← bobesponja left	07:04
	← kmap left	07:04
	→ nkallen joined	07:05
	→ kmap joined	07:10
	← TuxCM left	07:16
	→ mediogre joined	07:21
	← albertito left	07:24
	← mithro left	07:33
	← vbgunz left	07:36
	→ bentob0x joined	07:39
	← cwillu left	07:42
	→ cwillu joined	07:42
duncanm	i'm trying to use git-filter-branch to fix my commits with the wrong e-mail address	07:46
	i don't really understand how the command works, i've read the man page	07:47
	dum de dum	07:52
Mikachu	it's complicated	07:52
	if that makes you feel better :)	07:52
	i think you want the --env-filter	07:53
duncanm	i do too, but i can't find any examples	07:53
Mikachu	and change GIT_AUTHOR_NAME and GIT_AUTHOR_EMAIL	07:53
	just make a shell script with a switch case or so	07:53
duncanm	it's simple	07:54
Mikachu	case $GIT_AUTHOR_NAME in	07:54
	myfriend)	07:54
duncanm	i just need everything set to one name and email	07:54
Mikachu	export GIT_AUTHOR_NAME=myfriendsname	07:54
	export GIT_AUTHOR_EMAIL=myfriendsemail	07:54
duncanm	i started using my desktop, and i didn't set the environment variable there properly	07:54
Mikachu	;;	07:54
	otherfriend)	07:54
	...	07:54
	esac	07:54
duncanm	no no	07:54
	it's just me, one name, one address	07:54
Mikachu	well, just pick out the two export lines then	07:54
	and make it unconditional	07:55
duncanm	i want to fix the commits that i made already	07:55
Mikachu	then you just say git-filter-branch --env-filter=yourscript HEAD	07:55
	iirc	07:55
duncanm	and the script just says export GIT_AUTHOR_EMAIL, etc?	07:55
Mikachu	hm maybe it's more like	07:55
	git filter-branch --env-filter 'export GIT_AUTHOR_EMAIL=foo; export GIT_AUTHOR_NAME=bar' HEAD	07:56
duncanm	ah	07:56
	← johan-s left	07:57
	→ mithro joined	07:58
	tarbo_ → tarbo	08:14
duncanm	i'm in the middle of a rebase, and now it says # Not currently on any branch.	08:15
	how does that work?	08:15
Mikachu	see the entry on detached heads in man git-commit	08:15
	git rebase detaches head to not spam the reflog of the branch you're rebasing	08:16
	→ nkallen_ joined	08:17
	→ nkallen__ joined	08:19
	← nkallen left	08:21
	← DrNick left	08:22
	→ ruphy joined	08:27
duncanm	Mikachu: i don't see anything about detached heads	08:28
	← spearce left	08:29
	→ DrNick joined	08:31
	← nkallen_ left	08:34
duncanm	hmm	08:36
	Mikachu: still here?	08:36
Mikachu	yeah	08:36
	what git version are you on?	08:36
duncanm	1.5.4	08:36
Mikachu	you should definitely have that section then	08:36
	maybe your manpages are old somehow	08:37
duncanm	it's in git checkout, not commit	08:37
	i found it	08:37
	i'm using tig, and i see that there's 'merge' commit	08:37
	is there a way to turn that into just a normal commit?	08:37
Mikachu	doh, did i say commit? i'm sorry	08:38
	that question is a bit too general to answer, i think	08:38
	except with "yes" :)	08:38
duncanm	commit 9e3d9330fafbb9f715e16966321e421e67d516f4	08:38
	Merge: 8fd14c8... 2a1e668...	08:38
	but i think that just encompasses one commit	08:38
Mikachu	is it the latest commit? are there commits on both the left and right branch?	08:38
duncanm	it's not the latest	08:39
Mikachu	do you want to rewind to the last commit on the left branch and rebase the ones on the right to the top?	08:39
	if there are other commits after it, it's trickier	08:39
	you would have to rebase them too	08:39
duncanm	yeah, that's what i want to do	08:39
Mikachu	with --onto	08:39
	step one is definitely to git checkout -b try-rebase	08:39
	:)	08:39
duncanm	okay	08:42
	i don't know how to use --onto	08:42
Mikachu	that way you can mess up all you want and recover easily	08:42
	no, --onto is very hard to use before you understand rebasing completely :)	08:42
	do you have this repo available for cloning somewhere?	08:42
	← bentob0x left	08:43
	→ bentob0x joined	08:43
duncanm	Mikachu: can i give you a tarball?	08:43
Mikachu	yeah	08:44
duncanm	Mikachu: http://www.a-chinaman.com/~duncan/scheme/edwin48.tar.bz2	08:46
	→ intinig joined	08:47
Mikachu	okay, so you want to put those five commits 'setup the command'..'port command and variable' in a straight line?	08:48
intinig	hi all, I have a small problem. If it's too newbiesh a question please tell me so :) -- I got a git repo from a friend, I did some work on it, I committed twice and now would like to send my friend 'something' to sync his repo with mine. Can't do that though, can anyone help me or point me to a document?	08:49
	← sverrej left	08:49
duncanm	Mikachu: i think so	08:49
	Mikachu: in tig, i just want to get rid of the Merge: 8fd14c8... 2a1e668...	08:50
	yeah	08:50
Mikachu	right	08:50
duncanm	i don't even know how it got that way	08:50
Mikachu	well, you merged at some point	08:50
duncanm	i just started learning to use branching this week	08:50
	can you show me what commands to run to straighten it out?	08:51
DrNick	intinig: you can either create a series of patches using git format-patch or a bundle using git bundle	08:51
Mikachu	duncanm: an outline first, you want to rebase one of the side branches on top of the others, then rebase --onto the commits that come after the merge to your new linear branch	08:51
	duncanm: so are you on your try-rebase branch?	08:52
duncanm	yeah, i'm on a fix-merge branch	08:52
Mikachu	okay, so first git reset --hard 2a1e668	08:53
	that will move your fix-merge branch to the right (as in not left) side of the merge	08:54
	then git rebase 8fd14c	08:54
	that will produce some conflicts, so resolve those, git add the files, and then git rebase --continue, until it is done	08:54
	→ johan-s joined	08:55
intinig	thank: DrNick	08:56
	I meant	08:56
	← CelticSoul left	08:56
	← Sho_ left	08:56
intinig	DrNick: thanks :D	08:56
Mikachu	hahah	08:56
duncanm	Mikachu: i got to a point where there are no more changes	08:58
	but my rebase is still active, and i'm now detached from a branch	08:58
Mikachu	duncanm: okay, so the "port command and variable" is the latest commit now on try-merge?	08:58
	duncanm: oh okay, then you aren't done	08:58
duncanm	so do i checkout back to the fix-merge branch?	08:58
Mikachu	duncanm: nono	08:59
	duncanm: what have you done yet, you got one conflict and resolved it in the editor?	08:59
duncanm	i fixed 3 conflicts	08:59
	my top commit is 'Add fixme'	08:59
	d5e3d4	08:59
Mikachu	duncanm: okay	09:00
	duncanm: so what you want now is git add -u . and then git rebase --continue	09:00
	(or you can git add each of the fixed files explicitly)	09:00
duncanm	i did	09:00
	i fixed them one by one and --continued	09:00
Mikachu	oh	09:01
	→ CelticSoul joined	09:01
Mikachu	somehow you got to a state where you already made the changes from the next commit	09:01
duncanm	oh	09:01
	i guess when i resolved the conflicts, i went the other way	09:01
	oh	09:02
Mikachu	it can be hard to remember what state you want to resolve to in a rebase :)	09:02
duncanm	yeah, it is	09:02
Mikachu	so you can either change the file to how you want it, git-add it, and git commit --amend, and then git rebase --continue	09:02
	or	09:02
	you can git rebase --skip	09:02
	← intinig left	09:03
duncanm	Mikachu: okay, i'm at 5915e8	09:03
	'Port command and variable'	09:04
Mikachu	okay	09:04
	so now you want to move over the rest of the commits that came after the merge	09:04
	the way i usually do it is	09:04
	git tag here	09:04
	git reset --hard origin/use-string-copy	09:04
	git rebase --onto here 9e3d933	09:04
duncanm	isn't origin/use-string-copy the latest?	09:05
Mikachu	if you wonder where that number came from, it's the name of the commit that comes before the first one you want to move over	09:05
	yes, you want to move the latest commit to your new linear branch	09:05
	if everything is right, this step shouldn't produce any conflicts	09:05
duncanm	fatal: Needed a single revision	09:07
Mikachu	who said that?	09:07
duncanm	rebase --onto	09:07
Mikachu	hrm	09:07
	did you do the git tag here?	09:07
	you can say 5915e8 instead of "here" since you noted the sha1	09:07
duncanm	i tagged before resetting	09:07
	→ intinig joined	09:08
Mikachu	maybe you just need more numbers, 9e3d9330faf ?	09:08
duncanm	nope	09:08
Mikachu	hrm	09:08
duncanm	i don't see 5915e8 either	09:08
	→ octopod joined	09:09
Mikachu	i am a bit confused as to what happened now	09:09
	i did try the steps here first and it worked	09:09
duncanm	hmm	09:09
	maybe i should try again	09:09
Mikachu	yeah, do that :)	09:09
	→ z3ro joined	09:12
duncanm	where should i be in the tree when i do the rebase --onto?	09:15
Mikachu	i don't know the exact sha1 since i don't know how you resolved the conflicts	09:15
duncanm	but it should be the absolute latest?	09:16
Mikachu	yeah	09:16
	→ Cerebon joined	09:16
Mikachu	at "Remove all references to *-"	09:16
	oh i do know the sha1 since it isn't rebased yet heh	09:16
	a778fd	09:17
	but i don't know the sha1 of the --onto argument	09:17
	← bentob0x left	09:17
Mikachu	it should obviously be the last place you were rebasing on though	09:17
duncanm	do i type 'here' literally?	09:17
Mikachu	if you did the git-tag again	09:17
duncanm	can i do everything by SHA1?	09:18
	i want to do git-rebase --onto <sha1 of the latest> <sha1 of the merge commit> ?	09:18
Mikachu	yeah	09:18
	that "latest" should be the new "Port command and variable macros"	09:18
	maybe it helps if you look at gitk --all or so	09:19
	but if you did a reset --hard to origin/use-string-copy, that won't be on any existing branch	09:19
	which is why i usually tag it, in case you forget the sha1 :)	09:19
duncanm	gee!	09:20
	i made a big mess	09:20
Mikachu	this is one of the trickier procedures, i think	09:24
duncanm	http://a-chinaman.com/~duncan/messy-git.png	09:26
	← intinig left	09:26
Mikachu	heh, that is sort of expected right now	09:26
	since you still have the old branches	09:26
	what you want to do after this is move all branches over to your new commit graph	09:27
	at least the ones that come after the rebase point	09:27
	that looks good so far	09:27
	→ robinr joined	09:27
	→ sverrej joined	09:28
duncanm	well	09:29
	i'm really lost	09:30
	i thought i had 'master', 'commands-variables' and then 'use-string-copy' (which is a new branch off the tip of c-v)	09:31
	instead, i have a big knot of commits	09:31
Mikachu	well, all the branches that come after the merge that you want to get rid off will cause that merge to still be shown until you move them over	09:32
	after you do that, it will look a lot better	09:32
duncanm	Mikachu: can you show me what you have as motivation? ;-)	09:33
Mikachu	well, i didn't move the branches over	09:33
duncanm	how do i move branches over?	09:33
Mikachu	check them out and rebase --hard to the right sha1	09:34
	doh	09:34
	reset --hard	09:34
duncanm	reset?	09:34
	i thought reset changes the tip	09:34
Mikachu	that's what you want to do	09:34
	→ aggieben_ joined	09:35
Mikachu	you have a tree which splits off, and all the branches are on the wrong half of the tree, so you want to reset --hard them over to the other half	09:35
	s/tree/graph/g is less confusing	09:35
	← pchrist left	09:35
Mikachu	see here, this is before you reset --hard	09:37
	http://mika.l3ib.org/before.png	09:37
	and this is after http://mika.l3ib.org/after.png	09:38
	what i did was	09:38
	git checkout commands-variables	09:38
	git reset --hard try-rebase^	09:38
	if you want, you can just (without checking it out first), git push -f . try-rebase^:commands-variables	09:40
	→ robfitz_ joined	09:46
	→ cedricv joined	09:47
duncanm	Mikachu: thanks so much for your help, but i'm too tired now	09:50
	it's nearly 5am	09:50
Mikachu	sure	09:50
duncanm	actually	09:51
	maybe the easiest thing would be for me to just take what you've got	09:51
	i'm pretty sure once the knots are out of the way, i can fix everything else linearly using rebase -i	09:51
Mikachu	i just did git add -u . on everything without resolving the conflicts :)	09:52
	but yeah, maybe you can just put edit on every line and fix them up	09:52
	you can probably catch me tomorrow	09:52
duncanm	perfect	09:52
Mikachu	let me make a tar first	09:53
	in case	09:53
	mika.l3ib.org/edwin48.mika.tar	09:53
duncanm	grazie	09:54
	← robfitz left	09:57
	robfitz_ → robfitz	10:03
	→ ebel joined	10:08
	→ janm joined	10:19
	→ bentob0x joined	10:25
	→ Yuuhi joined	10:51
	← garrin left	10:55
	→ njero joined	11:11
	→ chris2 joined	11:20
	→ HG` joined	11:26
	→ mpagano_\|work joined	11:27
	← Phoop left	11:30
	→ Phoop joined	11:31
	← mpagano\|work left	11:38
	→ will29d joined	11:42
	← kanru left	11:44
	→ lcapitulino joined	11:46
	← Catfish left	11:46
	→ Catfish joined	11:46
	→ kanru joined	11:47
	← octopod left	11:48
	→ Sho_ joined	11:49
will29d	'lo.	11:52
	I have a git-svn woe, anyone about to help?	11:52
	http://rafb.net/p/EwYJ5R87.html	11:52
	ahem http://rafb.net/p/zOTeBM33.html	11:53
	That would be the correct paste :)	11:54
Mikachu	so secretive	11:54
will29d	:)	11:54
	← DrNick left	12:10
	→ DrNick joined	12:10
	← fultilt left	12:16
thiago	with git-fetch, can't I fetch one specific object?	12:17
	i.e., do I need to fetch a refspec?	12:17
Mikachu	i think that is the case at least with git-daemon	12:19
	→ freshtonic joined	12:23
will29d	It looks like if I do a "git svn dcommit remotes/my-branch" git will commit to the branch not the trunk, which would solve my problem I guess. I would love to know why this happened though.	12:23
	thiago is working on a script to package Qt from any commit from any repository	12:24
thiago	I can do this with git archive --remote, but I can't with git-fetch	12:24
freshtonic	anyone know where the source for git-show is hidden or how it's built? there doesn't seem to be a corresponding .c or script for it, and yet I end up with a git-show binary when I make git. I must be missing something obvious...	12:24
Mikachu	thiago: well, git fetch would still get the whole history from that point down	12:25
thiago	Mikachu: that's ok	12:25
	Mikachu: my point is: I have a remote repository name and a ref-or-commit. I want to check that out.	12:26
Mikachu	i think the idea is that you don't want people fetching possibly unpublished sha1s	12:27
	and you don't want to walk the whole tree to see if it's published	12:27
thiago	my point is that I want repeatability	12:27
	imagine I create a package off HEAD when it is 532f8b4c845abb	12:28
Mikachu	you would have to clone it first i guess	12:28
thiago	then I realise I had a bug in the packaging script. I want to repackage that commit in specific.	12:28
	I have a local clone of the master tree	12:28
	← will29d left	12:28
Mikachu	freshtonic: cmd_show in builtin-log	12:29
thiago	but that's just to speed things up (avoiding transferring 300 MB with git-archive)	12:29
Mikachu	freshtonic: see handle_internal_command in git.c	12:29
freshtonic	mikachu: thanks for that :)	12:29
	Mikachu thinks the builtin- prefix on all source files is sort of redundant	12:30
	→ Dejan joined	12:30
freshtonic	mikachu: I was greping for git-show, my mistake. Thanks for the tip!	12:31
Dejan	hello guys, i have a quick question - is it possible to push changes to a repository which is on a FTP server?	12:31
	ie, to use ftp://	12:31
	i read manuals and nothing has been mentioned about this	12:32
	but, on Wikipedia i see clearly that Git supports FTP	12:32
Mikachu	thiago: but it is inconsistent if you can git-archive an arbitrary sha1 but not fetch it...	12:32
vmiklos	Dejan: not at the moment.	12:33
Mikachu	Dejan: no you don't	12:33
vmiklos	it could be done, as http push is already implemented, but nobody did it so far.	12:33
Mikachu	thiago: i guess upload-archive isn't enabled by default though	12:33
vmiklos	also git supports pulling from ftp://, but pushing	12:34
Dejan	ok, thank you fellows	12:34
	cheers	12:34
	← Dejan left	12:34
	→ csc` joined	12:34
freshtonic	another question: when I run 'git show --pretty=format:"%H%n%P" HEAD', I expect to see SHA1 of commit, followed by SHA1s or parent commits only. However, I also get the diff after it - is that a bug in git show? It doesn't seem to indicate that in the man page	12:37
	→ ddaa joined	12:37
freshtonic	i.e. when controlling output with --pretty=format:"foo"	12:37
Mikachu	freshtonic: give -s too	12:38
	freshtonic: if you read the description part in the man page, you'll see it also takes options from git-diff-tree	12:38
freshtonic	Mikachu: thanks again.	12:39
thiago	freshtonic: if you don't want to see the diff, why are you using git-show?	12:40
Mikachu	to see the commit and parent sha1s apparently :)	12:40
	freshtonic: btw, there's no need to quote %H%n%P	12:41
freshtonic	I'm looking to generate output from git containing commit and parents, and I wanted to generate it in an easy-to-parse manner. If there is an easier way, I'm all ears :)	12:41
	Mikachu: You may have missed my response as I didn't prefix it with you handle. Here it is again: I'm looking to generate output from git containing commit and parents, and I wanted to generate it in an easy-to-parse manner. If there is an easier way, I'm all ears :)	12:44
Mikachu	i did not	12:45
freshtonic	no probs	12:45
	→ leitao joined	12:51
vmiklos	what about git log -1 --pretty=format:"%H%n%P" HEAD?	12:53
freshtonic	vmiklos: that works too, thanks	12:55
Mikachu	saves a full letter	12:55
vmiklos	hm? his problem was that git show is like git log -p	12:56
	→ ferdy joined	12:57
Mikachu	git log -1 vs git show -s saves one letter :)	12:57
freshtonic	now I two options, decisions, decisions :)	12:58
	and a one char difference.	12:58
	→ destruct_ joined	13:00
vmiklos	ah	13:02
	i didn't know git show -s	13:03
	yes, because it's missing from git-show.txt	13:03
freshtonic	vmiklos: I missed that too	13:04
vmiklos	Mikachu: is -s inherited from a lowlevel command or is it simply undocumented?	13:04
	→ ebel_ joined	13:04
Mikachu	vmiklos: it is not undocumented	13:05
	← ebel left	13:05
	ebel_ → ebel	13:05
freshtonic	vmiklos: as Mikachu mentioned further up " if you read the description part in the man page, you'll see it also takes options from git-diff-tree"	13:05
vmiklos	ah, i see.	13:08
	i should send a patch to include the diff-tree options using asciidoc's include ;)	13:09
	← mpagano_\|work left	13:09
freshtonic	vmiklos: sounds like a plan coming together ;)	13:10
	← destruct left	13:11
vmiklos	it seems this "don't mention it takes options as git-foo, but include the options" will be my hobby-horse;)	13:12
freshtonic	vmiklos: I'd help you out, but I need to sleep. I'll keep a close watch on your progress ;)	13:15
	← cedricv left	13:24
	→ Fullmoon_ joined	13:26
	→ cedricv joined	13:29
	destruct_ → destruct	13:29
	→ cmarcelo joined	13:33
	ruphy → ruphy-lapdog	13:34
	← ironfroggy left	13:56
	→ Man-erg joined	14:02
	← mithro left	14:02
	midgar → bdiego	14:11
	→ kukks joined	14:14
	→ albertito joined	14:21
	→ benol joined	14:21
	→ EmilMedve joined	14:22
	← timj left	14:25
	→ tjafk1 joined	14:29
	← Cerebon left	14:33
	→ mikl joined	14:33
mikl	I'm trying to do "git cvsimport -d:pserver:anonymous:anonymous@cvs.drupal.org:/cvs/drupal DRUPAL-5", but i get "AuthReply: cvs [pserver aborted]: descramble: unknown scrambling method" as an error message :(	14:34
	does anyone know what that means?	14:35
	(I have git and git-cvs 1.5.2.5 and 1.12.13 and cvsps 2.1	14:36
robinr	see the end of this page: http://www.freebsddiary.org/cvs-remote.php	14:37
	first hit on google :)	14:37
	→ PosixJumbo11 joined	14:37
PosixJumbo11	I work on the netdev linux git tree ; suppose I wrongly tried to git pull from the drivers tree (it could be that I was careless).	14:38
	since the trees are somewhat symmetric - can it be that it updated some files from the wrong place ?	14:38
	is there a way I can compate my current git tree on my fs to daveM git tree ?	14:39
mikl	robinr: heh, wouldn't it be fun if you had this ancient and evil SCM that has the most arcane error messages :)	14:39
PosixJumbo11	?	14:39
robinr	PosixJumbo11: you can diff a ref in your tree agains an absolute path	14:40
Ilari	PosixJumbo11: See if the merge actually did complete, and if it did, undo it?	14:40
PosixJumbo11	robinr, how?	14:41
	→ Toad__ joined	14:41
robinr	mikl: maybe they changed from the infamous I HATE YOU error message	14:41
	PosixJumbo11: git diff HEAD /some/path	14:41
Ilari	robinr: That seems like it tries to auth using method the server does not support...	14:41
robinr	Ilari: I've never seen the message myself so I googled it at that's what I found	14:42
	→ tjafk2 joined	14:42
	→ cannonball joined	14:43
	← alb left	14:44
mikl	robinr: do you have any idea why I'm getting this error, though. I can do a normal CVS checkout just fine, but if I try to do it through git, I get this error :(	14:44
	if I leave out one of the anonymous, I get "AuthReply: I HATE YOU", so I must be doing something right	14:46
	→ ironfroggy joined	14:47
	← ironfroggy left	14:47
mikl	robinr: ah, I had to do a CVSROOT=:pserver:anonumous@cvs.drupal.org:/cvs/drupal cvs login	14:47
	:)	14:47
	that created a .cvspass, but it doesn't seem to work :(	14:48
	→ octopod joined	14:50
mikl	ARRRRRRRRRRRRgh	14:50
	if I only my hair was long enough, I'd pull it out	14:51
	← Toad left	14:52
	← mediogre left	14:52
	→ kristoffer joined	14:53
kristoffer	Trying to allow a developer to push to our git-repository. He's part of the developer group (which i use as "push-enabled" group). However he gets this when he tries to push	14:53
	error: unable to create temporary sha1 filename ./objects/tmp_obj_gm7Srf: Permission denied	14:53
	If I try to set --group= inside git-daemon script it says it requires --user, and I don't want all commits to be in my name	14:54
	← nottwo left	14:56
Mikachu	you don't want to enable pushing via git://	14:56
avar	Use the ssh young skywalker	14:57
kristoffer	it is through ssh	14:57
	← tjafk1 left	14:57
Mikachu	then git-daemon has nothing to do with it	14:57
kristoffer	he's got an ssh account and he's added to my dev usergroup	14:57
Ilari	kristoffer: Is the repository shared flag set?	14:57
kristoffer	[core]	14:58
	repositoryformatversion = 0	14:58
	filemode = true	14:58
	bare = true	14:58
	lol,	14:58
	→ ironfroggy joined	14:58
kristoffer	Oki , thx	14:58
Ilari	kristoffer: If not, set it, chgrp all files and directories to apporiate group, and set mode 2770 or 2775 on all directories in repo...	14:59
kristoffer	Ilari, oki, thx will do	15:00
	→ ceez joined	15:00
Ilari	(Incidentally, that is one task that my planned set of extra scripts for git would do). Not implemented yet though.	15:01
kristoffer	:) would have been helpful. I didn't notice the sharedrepository flags wasn't seet, since I was the only "pusher" until now	15:03
Ilari	Setting the shared repo flag isn't hard. There's still the task of executing the chmods/chgrps (especally if setting shared flag after repo creation).	15:05
kristoffer	agreed	15:06
	→ janm_ joined	15:09
	← janm left	15:22
	→ flazz joined	15:23
	← kanru left	15:24
	← benol left	15:31
	→ Ryback_ joined	15:31
	→ aroben joined	15:31
	← kristoffer left	15:32
	→ aroben_ joined	15:33
	← aroben left	15:33
	→ madewokherd joined	15:35
PosixJumbo11	for comparing against git path , I was advised to try:	15:36
	git diff HEAD git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6.git	15:36
	nut i get	15:37
	→ stick joined	15:37
PosixJumbo11	but I get the follwoing error	15:37
	fatal: ambiguous argument 'git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6.git': unknown revision or path not in the working tree.	15:37
	Use '--' to separate paths from revisions	15:37
	?	15:37
	why?	15:37
	(I simply to simply compare my current git tree on my fs to daveM git tree)	15:40
fonseca	PosixJumbo11: You need to fetch it down locally first.	15:40
PosixJumbo11	it is there	15:41
fonseca	What name did you give the remote?	15:41
PosixJumbo11	this is what I tried : git diff HEAD git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6.git	15:41
	and the git-clone was from this URL	15:42
fonseca	Sorry, I seem to misunderstand what you are trying to do.	15:44
	→ \|kukks\| joined	15:44
PosixJumbo11	fonseca, I will explain:	15:44
	I had git-clone this repo:	15:45
	git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6.git	15:45
	simply by git-clone	15:45
	now, it was a month ago	15:45
	I want to check if I tried wrongly somewhere in the last month to git-pull from a differnet URL	15:45
	becuase I work with 2 other repos, which are also linux kernel	15:46
	one is the net drivers and one is the wiereless net drivers	15:46
chrizz	what about git diff master..origin/master ?	15:47
fonseca	OK, the original clone should have created a "remote" for you. Try running: git remote show origin	15:47
PosixJumbo11	How can I check agains the original git repo to see if it is the same I have	15:47
	it gives:	15:47
	git remote show origin	15:47
	* remote origin	15:48
	URL: git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6.git	15:48
	Remote branch(es) merged with 'git pull' while on branch master	15:48
	master	15:48
	Tracked remote branches	15:48
	master	15:48
	does it say that I did not try to pull from a different git repo ***for sure** ???	15:48
Mikachu	it doesn't say anything about that	15:50
	i think the easiest is to just look at gitk and see if it looks correct	15:51
	if your master is in the same place as origin/master	15:51
	→ Elr joined	15:52
	← kukks left	15:53
	\|kukks\| → kukks	15:53
	← Elr left	15:54
	→ flazz_ joined	15:58
	→ ReinH joined	16:00
	← Ryback_ left	16:05
	→ \|kukks\| joined	16:05
	← PosixJumbo11 left	16:07
	← johan-s left	16:08
	→ njero_ joined	16:09
	→ devurandom joined	16:09
devurandom	Hi!	16:09
	How do I unrecord a commit?	16:09
	I.e. delete it.	16:10
numbah	git rebase -i <commit>^	16:11
	and then remove the line containing the commit	16:11
	→ _kukks_ joined	16:12
devurandom	And how do I revert my working copy to the state of HEAD? I.e. what "svn revert" does.	16:13
	← flazz left	16:13
	→ niki joined	16:16
	aroben_ → aroben	16:18
	→ _kukks_3 joined	16:18
	← sverrej left	16:19
	← kukks left	16:22
	trapni_ → trapni	16:27
	_kukks_3 → kukks	16:28
	← njero left	16:28
ReinH	devurandom: git reset --hard	16:28
	http://www.kernel.org/pub/software/scm/git/docs/git-reset.html	16:29
	← \|kukks\| left	16:29
	→ bobesponja joined	16:32
	Toad__ → Toad	16:34
	→ kyle__ joined	16:34
	← _kukks_ left	16:35
	← Fullmoon_ left	16:37
	← EmilMedve left	16:39
	kyle__ → kylem	16:40
	← Ademan left	16:45
jrockway	this article almost made me cry: http://texagon.blogspot.com/2008/02/use-mercurial-you-git.html	16:47
	because i felt so sorry for the author	16:47
	← kylem_ left	16:48
	→ \|kukks\| joined	16:49
	→ johan-s joined	16:49
	← kukks left	16:50
	\|kukks\| → kukks	16:50
	ruphy-lapdog → ruphy	16:50
	← namenlos left	16:52
	← flazz_ left	16:52
	→ \|kukks\| joined	16:53
loops	jrockway, The only point he makes that has any real merit is the lack of full support for Windows.	16:54
jrockway	yeah, the rest is just ranting about things that aren't true	16:55
	the next article in the series is even better	16:55
loops	all the other points are rather lame.. He even basically says you only need to learn 1 command to use mercurial (ie. "hg"), but you have to learn at least 10 commands for Git.	16:55
jrockway	he starts by complaining that he can't clone an empty repo (i guess git init is too hard)	16:55
loops	Stupidly ignoring the fact that you have to learn at least 10 sub-commands to use "hg"	16:55
Debolaz	Hey jrockway.	16:55
	→ _kukks_ joined	16:56
jrockway	http://texagon.blogspot.com/2008/02/let-play-git.html	16:56
	hi Debolaz	16:56
Debolaz	I admit, I tend to recommend mercurial over git if someone asks me.	16:56
jrockway	hg internals aren't as clean as git's	16:56
	worrying about the UI only is a great way to lose all your data	16:56
Debolaz	But externals are way cleaner.	16:56
jrockway	i already warped my brain to the ways of git	16:57
	anyway, i recommend git. most people ignore me.	16:57
Debolaz	There's no way you can use git without actually reading the documentation. With hg, you can for most basic tasks just go by feeling alone.	16:57
loops	Git has some horrible UI warts.. and the main developers tend to think of anything that makes it better as catering-to-idiots. which is unfortunate.	16:57
jrockway	loops: what do you think are the worst parts?	16:58
	from that article, he complains that "git merge HEAD" doesn't do anything	16:58
	uh... yeah...	16:59
	→ _kukks_3 joined	16:59
loops	Oh i dunno.. There's just a bunch of little things.. The fact that you have to preface branch names with refs/heads/ about _half_ the time... etc..	16:59
jrockway	that's a new "feature"	16:59
	you didn't have to do that in the past	16:59
	i don't understand the logic going on there	16:59
Debolaz	A good example is how most commands use --cached to refer to the index, instead of the more obvious choice --index.	16:59
jrockway	the index is a place, cached is a state	16:59
loops	jrockway, it's an internal implementation detail.. and you always had to do it... about half the time.	16:59
jrockway	now it's all the time, i guess	16:59
loops	Debolaz, yeah.. that one too	16:59
jrockway	i know for a fact i used to say git push origin foo:foo	17:00
	now it's git push origin foo:refs/heads/foo	17:00
loops	jrockway, that still works	17:00
	jrockway, about half the time ( ;o) )	17:00
jrockway	hahah	17:00
Debolaz	jrockway: Explicit use of refs/heads has a subtly different meaning.	17:00
jrockway	maybe it's trying to normalize to 50%	17:00
loops	it'll work if Git can find a refs/heads/foo already existing on the remote	17:00
jrockway	yeah	17:00
loops	but it will fail if the remote doesn't have a refs/heads/foo	17:00
	→ EmilMedve joined	17:00
jrockway	but to create the branch, you have to say refs/heads/foo	17:00
Debolaz	Which is of course completely unobvious, you need to read the documentation.	17:00
jrockway	yeah, i agree that that's not a particularly helpful feature	17:01
	the computer knows that i'm wrong, and there's nothing else i could possibly mean	17:01
	so it should fix it	17:01
loops	The thing is.. once you get past all these little annoying things.. Git is fucking awesome.	17:01
Debolaz	Yes, I love git for its fundamental ideas. But it's very hard to market it on that basis.	17:01
jrockway	yes, i can't believe i wasted so much time fucking with svk	17:02
	what a nightmare	17:02
	Debolaz: does the git community need users that don't care to understand their computer?	17:02
	→ _kukks_4 joined	17:02
jrockway	git-svn works well enough that i don't care what other people choose to use	17:02
	← CelticSoul left	17:03
Debolaz	git-svn tend to have some trouble with real life repositories unfortunately.	17:03
	Since most svn users don't actually follow the weakly dictated svn conventions.	17:03
jrockway	true, but it does much better now than it used to	17:03
PerlJam	git-svn++	17:03
jrockway	and everyone i work with uses svn correctly	17:03
	project/[trunk\|tags\|branches	17:04
PerlJam	jrockway: you mean "conventionally" :)	17:04
loops	Debolaz, i've had great luck with git svn on quite a few different repos	17:04
Debolaz	jrockway: Most likely, the branches directory will be a mixture of copies of trunk, copies of subdirectories of trunk, and non-svn-copies of trunk and subdirectories of trunk.	17:04
	jrockway: git-svn only handles one of those 4 cases properly.	17:04
jrockway	yeah, don't worry, i know	17:04
	trust me, i wasted a good portion of my life fucking around with retarded svn repositories	17:05
	→ _kukks_5 joined	17:05
jrockway	i just happen to be lucky that OSS projects i work on don't do that anymore, and work doesn't do that	17:05
	so while it's not a general solution, i'm happy :)(	17:05
PerlJam	Hmm. I wonder if this is another instance of the opinionated v. no-opinion. (i.e. svn has no opinion about repository layout but git does)	17:06
jrockway	no, this is different	17:06
Debolaz	I modified git-svn to deal correctly with svn copies of subdirectories of trunk.	17:06
jrockway	svn has no support for branching or tags	17:06
	while git does	17:06
	→ kumbayo joined	17:06
loops	PerlJam, that's an interesting point. It would be cool to write a Git-fs backend for svn.	17:06
ReinH	jrockway: svn has copy, what else do you need? ;)	17:06
jrockway	svn was a cute experiment	17:07
	aroben → aroben\|meeting	17:07
ReinH	loops: if you like exercizes in utter futility	17:07
jrockway	"can we make everything a file or directory, including branches and tags"	17:07
PerlJam	loops: what would be the point there?	17:07
jrockway	as it turns out, no	17:07
avar	loops: it would be cool to kill svn with FIAH	17:07
jrockway	can i have CVS back now?	17:07
	that's another "svn lie"	17:07
	"this is just bugfixes for CVS"	17:07
	no, it's TOTALLY different	17:07
ReinH	I'm in agreemend with Linus here	17:07
jrockway	a bugfix for CVS is git-cvsserver	17:07
loops	ReinH, it wouldn't be futile. It would be useful as a migration too.. just like the git cvs-server is.	17:07
ReinH	*agreement	17:08
loops	tool*	17:08
ReinH	"CVS done right" is a loosing proposition from the start	17:08
Debolaz	jrockway: That's kind of like saying perl has no support for classes, only modules. Like classes in perl, branches and tags is a mental construct in svn. It just fails because it requires people to actually follow conventions voluntarily instead of being forced to.	17:08
PerlJam	loops: migration tool? Only if you're migrating the wrong way	17:08
loops	PerlJam, you know git comes with a git-cvsserver right?	17:08
PerlJam	Debolaz: it also fails because of little things like ... oh, I don't know ... no merge tracking	17:09
jrockway	Debolaz: and indeed, that's a frequent criticism of perl	17:09
	→ _kukks_6 joined	17:09
Debolaz	PerlJam: That's another issue, but it'll be fixed in future subversion versions.	17:09
jrockway	but it's a little different	17:09
	← kukks left	17:09
loops	PerlJam, it's useful to let some people continue to access the repo as svn while others are using Git..	17:09
	There's actually talk (and a note in the svn source repo) about making a Mercurial backend.	17:10
PerlJam	Debolaz: no, it's a related issue. Tell people to do things by convention then provide minimal support for a useful convention and people get confused and cranky.	17:10
Debolaz	PerlJam: I think you misread me slightly.	17:10
	← Helios- left	17:10
ReinH	merge tracking would be a hack job on svn	17:11
loops	avar, FIAH ?	17:12
	→ _kukks_7 joined	17:12
PerlJam	Debolaz: no, voluntarily convention following works as long as there are "tools" to help you follow a useful convention. svn lacks the tools other than some documentation that says "these are best practices"	17:12
	← \|kukks\| left	17:12
Debolaz	Ok then.	17:12
PerlJam	In perl's case, at least it had CPAN where tools could grow.	17:13
	→ rubydiamond joined	17:13
loops	To be fair to svn, the convention is pretty damn easy to follow..	17:13
PerlJam	loops: How often have you done merging in svn?	17:13
loops	heh.. never.	17:13
rubydiamond	hey I dont have man git ... man pages of git installed on my mac	17:14
avar	loops: fire	17:14
PerlJam	Well, if you restrict yourself to very simple use, sure the convention may be just fine.	17:14
rubydiamond	but I have installed git from source	17:14
	now I want to install man pages..	17:14
	how do i do that?	17:14
Debolaz	It has occationally been tempting to reimplement a git-like scm based on the same ideas but with a sane ui.	17:14
Ilari	rubydiamond: Clone the git repository (if you don't already have it), and copy the stuff in man branch to /usr/share/man?	17:15
loops	rubydiamond, i think there is a separate package for man pages on mac.. git-doc git-man ??	17:15
	→ \|kukks\| joined	17:15
Ilari	rubydiamond: Or whatever is the manual pages directory.	17:15
	← _kukks_ left	17:16
rubydiamond	do u know how to clone git repository... I want the latest cache of it	17:16
	means.. I dont want it from previous versions	17:16
loops	rubydiamond, you want to know from where to clone the main git.git repo? git://git.kernel.org/pub/scm/git/git.git	17:17
	→ flazz joined	17:17
rubydiamond	loops: would it clone all the repository having all previous history	17:18
kergoth	hmm, can you stash a conflicted merge and have stash apply bring it back in its unmerged state?	17:18
	→ _kukks_ joined	17:18
loops	rubydiamond, yes.. that's the way Git works.. it's a pretty small repo though.. (you can do a shallow-clone, but i'd recommend against it)	17:18
	← _kukks_3 left	17:19
rubydiamond	loops: why do u recommend with all previous changes	17:19
loops	rubydiamond, because it's so cheap (small & fast) it's basically free, and because you sometimes will run into gotchyas with a partial clone.	17:20
rubydiamond	loops: how can I update my installed git.(by source)... ?	17:20
	loops: ok	17:20
	← _kukks_4 left	17:20
	→ _kukks_3 joined	17:21
	Thumper_ → BerntH	17:22
	← mithraic left	17:23
	→ cnantais joined	17:24
	→ tvachon joined	17:24
	→ _kukks_4 joined	17:24
	← flazz left	17:24
	→ jas4711 joined	17:24
kergoth	huh, thats weird	17:24
	using an offset for end in git blame -L acts oddly	17:25
	← _kukks_5 left	17:25
	← destruct left	17:25
kergoth	the number apparently includes the start line, and oddly, if you do +0, it acts the same as +2, not the same as +1	17:25
	→ Cerebon joined	17:25
	← _kukks_6 left	17:28
loops	rubydiamond, fwiw it appears that the macports package to install git man pages is git-manpages	17:29
rubydiamond	loops: I have git installed via source	17:29
	now should i install it with sudo port install git-manpages	17:30
	will it result in conflict	17:30
	← bentob0x left	17:30
	→ trapni_ joined	17:30
loops	rubydiamond, i don't really know enough about the mac to say.. but obviously you'd only want to install one way.. either from macports or source.	17:31
rubydiamond	loops: Error: Port git-manpages not found	17:31
	← _kukks_7 left	17:31
	BerntH → BerntH_	17:33
rubydiamond	loops: what is difference between git init , git init-db and git init .	17:33
	→ zapnap joined	17:34
	← \|kukks\| left	17:34
	BerntH_ → BerntH	17:34
	← njero_ left	17:36
kergoth	hmm	17:36
	Dodji → Dodj\|away	17:37
kergoth	i have 'git log' output and want to dump a log with a different format. how can i feed this to git rev-list or git diff-tree such that I can do that? Any ideas, other than grepping out 'commit 12345'?	17:37
	← _kukks_ left	17:37
	→ alley_cat joined	17:40
	← _kukks_3 left	17:40
	← trapni left	17:41
	← rubydiamond left	17:42
	← _kukks_4 left	17:43
	→ _kukks_4 joined	17:44
	→ proppy joined	17:44
proppy	Hi	17:44
	→ \|kukks\| joined	17:46
	→ asdx joined	17:47
	→ CelticSoul joined	17:48
	→ Helios- joined	17:50
	→ _kukks_ joined	17:50
	→ is_null joined	17:51
is_null	hello everybody, i'm trying to get gitweb to roll on http://gitweb.devangels.org:83/, but it show a 403/forbiden error. For testing purpose, the permissions are 777 on the gitweb folder recursively. This is the lighttpd config : http://nopaste.com/p/azlWUKFXb any tip or idea please ?	17:51
	note that i have the same problem with apache anyway, that is why i think that it's more of a gitweb mis-configuration	17:52
	← Catfish left	17:52
	→ _kukks_3 joined	17:53
	→ nkallen joined	17:55
	→ Catfish joined	17:55
	← nkallen left	17:55
	→ nkallen joined	17:56
	aroben\|meeting → aroben	17:57
	← _kukks_4 left	18:02
	→ pflanze joined	18:03
	← nkallen left	18:03
	← ebel left	18:04
pflanze	Hello. How can I prune my git-reflog history? I'm trying to run git-gc --prune to delete off much stuff which erroneously made it on top of my HEAD, but it seems git-reflog holds on to that storage.	18:04
	← \|kukks\| left	18:05
pflanze	Attempts at 'git-reflog expire --expire=0' or 'git-reflog expire --all' didn't help, even cloning doesn't help ;)	18:05
	→ careo joined	18:08
	← _kukks_ left	18:08
careo	quick question about the index and git-add. the docs say the index is a snapshot of the next commit, does that mean it's just the commit object and tree, or also any blobs that'll be added? so then does git-add append those blobs to the index? or, does git-add write the blobs to the repository and update the tree in the index?	18:09
	→ timocratic joined	18:10
Debolaz	careo: The latter.	18:10
careo	Debolaz: thanks! I'm giving a demo of git this evening and wanted to be sure I had it straight	18:11
	← nkallen__ left	18:11
	← _kukks_3 left	18:11
	→ drizzd joined	18:14
loops	pflanze, try git reflog --expire=0 --all	18:16
	← Yuuhi left	18:21
	→ pombreda joined	18:23
	← drizzd_ left	18:24
	→ kristoffer joined	18:30
kristoffer	Im exporting a tree (based on linux-2.6.git) where I have 2 branches. It works well. However I wish to update my *master with new updates from the linux-2.6.git. when I go into server folder and type git-pull, it says that it cannot do that without extracted userland	18:32
	is there any other command to handle 'pull' functions in an --bare environment?	18:32
	Seems silly to need to extract / pull / make bare	18:33
loops	kristoffer, you don't want pull, you want 'fetch'	18:33
kristoffer	so a simple git-fetch then?	18:33
loops	pull implies a possible merge, which must be done with a working tree..	18:34
	yes.	18:34
kristoffer	ah	18:34
	← Fullmoon left	18:34
	→ vbgunz joined	18:34
	← Catfish left	18:34
	→ d2dchat joined	18:46
d2dchat	How do I reset my repository to the latest on the main tree	18:46
	What happened was that I forked a tree	18:46
	made a bunch of changes	18:46
	and then the tree I forked from took my changes and made a bunch of changes to my code	18:46
	now I want the latest tree changes without my changes	18:46
	how do I do that?	18:46
	→ HobGoblin joined	18:48
	→ destruct joined	18:48
loops	d2dchat, was your repo originally a clone of the upstream repo ?	18:48
d2dchat	loops: it was a clone of the forked tree I believe yes	18:49
loops	d2dchat, a clone of the fork? or a clone of the upstream your'e trying to reset too?	18:50
d2dchat	hmm how do I verify?	18:50
loops	several ways.. what does git remote show origin say ?	18:50
	← janm_ left	18:51
	← wildfire left	18:51
d2dchat	loops: ok yes the original clone was the repository that I forked from	18:51
kristoffer	I've done git-fetch linux-2.6.git and it downloads, however doesn't seem to get merged into my master	18:52
loops	d2dchat, and now you want your local master branch to be exactly what is upstream ?	18:52
d2dchat	loops: yes	18:53
kristoffer	"5684 objects were added to complete this thin pack."	18:53
loops	d2dchat, are you worried about losing any uncommitted local changes?	18:53
Ilari	kristoffer: fetch does not merge.	18:53
d2dchat	loops: nope	18:53
	← is_null left	18:53
d2dchat	purge them all!	18:53
	git reset --hard head?	18:53
loops	d2dchat, git checkout master ; git reset --hard origin/master	18:53
d2dchat	what does checkout master do?	18:54
kristoffer	Ilari, oki, but how can I get linux-2.6.git upstream updates?.. my *master is identical to 2.6.24-x something	18:54
loops	d2dchat, was just making sure that was your current branch	18:54
kristoffer	Ilari, its a bare repository btw, thats shared	18:54
Ilari	kristoffer: 'git merge origin/master' probably?	18:54
loops	d2dchat, it checks out your local branch named master (your local branch will then track the upstream origin/master)	18:54
Ilari	kristoffer: You pretty much can't update shared repo except by pushing something into it.	18:55
kristoffer	Ilari, aah oki. I'll do that then	18:55
	→ bentob0x joined	18:57
	← proppy left	18:58
	← pombreda left	18:59
	← Helios- left	19:05
	← UukGoblin left	19:05
	→ Helios_ joined	19:06
	→ eetfunk joined	19:07
	Helios_ → Helios-	19:07
eetfunk	I want to pull changes made by somebody else from my origin server. However, GIT complains that I didn't specify which branch to merge when I do a "pull". I must be missing something, but basically, I want to merge all the branches and have the latest version in all of them. What am I doing wrong?	19:08
Randal	you're confusing pull with fetch. :)	19:08
	that's what you're doing wrong	19:08
	hint - "pull" is never what you want	19:08
	really	19:08
	it's a shiny object to confuse beginners	19:08
	→ Ademan joined	19:09
eetfunk	ok :)	19:10
	so just git fetch huh?	19:10
	← d2dchat left	19:15
	← chris2 left	19:15
	← eetfunk left	19:15
	→ eetfunk joined	19:16
	→ eetfunk_ joined	19:17
	← eetfunk left	19:17
	← eetfunk_ left	19:17
	→ eetfunk joined	19:18
	→ eetfunk_ joined	19:19
	← eetfunk left	19:19
Pieter	Heh	19:19
Mikachu	eetfunk_: are you done?	19:20
eetfunk_	Mikachu: done with what?	19:20
Pieter	TextMate breaks and does weird things when you have symlinks to non-existing directories in your ~/Lib/AppSupp/Txtmt/Bundles/	19:20
	oh	19:20
Mikachu	reconnecting 100 times	19:20
Pieter	wrong :)	19:20
eetfunk_	Mikachu: i think my connection is acting weird.	19:20
	← octopod left	19:20
careo	stupid git-show question, how come sometimes it will spew forth a large number of blank lines before actually printing stuff out?	19:25
	→ pombreda joined	19:26
loops	careo, do you have test case you can publish to show the problem ?	19:26
careo	loops: yeah. one sec	19:26
	→ Ryback_ joined	19:28
	→ Yuuhi joined	19:29
	→ benol joined	19:31
	← pombreda left	19:32
	← benol left	19:33
	→ mithraic joined	19:33
pflanze	loops: thanks, that finally purged the reflog; but git-gc --prune still doesn't free the space;	19:33
	I see references in .git/packed-refs	19:33
	and there's a .git/ORIG_HEAD	19:34
loops	pflanze, yeah you'll have to clean those up too before prune will werk	19:36
	→ octopod joined	19:37
	dsaxena_ → dsaxena	19:39
pflanze	Yep. Removing .git/ORIG_HEAD, putting the sha1 of the "master" entry from .git/packed-refs into .git/refs/heads/master and removing .git/packed-refs helped.	19:40
	← bentob0x left	19:46
	→ flazz joined	19:50
flazz	any common caveats for a .gitignore? i made one and git status still shows everything i think i'm ignoring	19:50
ReinH	flazz: files already tracked by git will continue to be tracked	19:51
flazz	i should delete them commit and then try?	19:51
ReinH	so you'll have to git rm it and then put it back	19:51
	unless anyone else has a better way?	19:52
flazz	what is the big difference between git-rm and a normal rm?	19:52
	that's a good enough way for me	19:52
ReinH	flazz: git-rm removes files from the tree and the index	19:52
	rm just removes them from the tree	19:52
	er	19:53
	s/tree/working copy	19:53
flazz	index?	19:53
	index == repository tree?	19:53
	the stuff in .git?	19:53
ReinH	index is the staging area that git uses to decide what gets committed	19:53
flazz	gotcha	19:54
ReinH	when you do git add you are adding changes to the index	19:54
careo	loops: I think it's related to paging. to get this to show in one screenfullI had to lengthen the window to show the. http://skitch.com/careo/f6ct/djensen-loki-desktop-87x37-ttys001-bash-68x55	19:54
flazz	its beyond an add but not beyond a commit?	19:54
ReinH	when you do git commit you commit what you've added to the index	19:54
kergoth	anyone know how i'd get to the version of a file in the index for git show or git cat-file, in a pre-commit hook?	19:55
flazz	thanks ReinH	19:55
loops	careo, does it do it everytime you show that commit? what happens if you export PAGER=cat does the problem go away?	19:55
ReinH	flazz: np	19:55
careo	loops: it does it every time if I've done anything like clear. and it does indeed go away if I set my pager to cat	19:56
	so it's probably less being stupid	19:56
loops	careo, yes, i think it's a less bug.	19:56
	careo, try export PAGER=more	19:57
careo	loops: same thing with more	19:57
loops	well they may actually be same program on your system.. still probably a pager bug.. i see the same thing when using "sudo" sometimes.	19:57
careo	oh no, sorry, if the pager is less it does it. with more it opens up more itself	19:58
	and then when I quit more, it's the same end result	19:58
	either way, I can live with just setting my pager to cat.	19:59
	thanks!	19:59
kergoth	ah, nm, found it	19:59
loops	kergoth, what did you end up with?	19:59
careo	gotta run for 30ish. bbiab	19:59
kergoth	i knwe you could look at the different staged versions of an unmerged file wtih :1:foo & the like, but the :# is optional, :foo is just the staged version of foo, it seems	20:00
	kergoth found it in specifying revisions in 'git help rev-parse'	20:00
kergoth	wanted a syntax checking pre-commit	20:00
	heh	20:00
	→ samgranieri joined	20:02
	← leitao left	20:03
	→ Voker57 joined	20:06
	← Cap_J_L_Picard left	20:07
	→ Cap_J_L_Picard joined	20:07
	← henr_k left	20:07
	→ henr_k joined	20:07
	← lcapitulino left	20:09
Illocution	git-magicians, is there a way to grep over a section of history looking for diffs that match x? I can do what I want in gitk with "Highlight: Commits adding/removing string" but there's got to be a command line equivalent... my weak google-fu is not finding anything	20:09
	→ CJUKAUS joined	20:10
loops	Illocution, i think you're looking for git log -S'pattern'	20:10
Illocution	yes indeed - fantastic, thanks	20:10
	→ avarab joined	20:10
Illocution	is there a way to get the output in more grep-like format, like commitid:filepath:matching line?	20:11
	→ zeenix joined	20:13
Illocution	can script it from git log --pretty=oneline I guess	20:13
	← kumbayo left	20:15
	← CJUKAUS left	20:16
	← amf left	20:18
	← Gitzilla left	20:18
	← avar left	20:21
	← Randal left	20:21
	→ Gitzilla joined	20:22
	← Voker57 left	20:22
	→ amf joined	20:23
Tv	Illocution: and --pretty=format is your bitch	20:25
robinr	odd, why is git reset --hard so slow now? it does three files per second only	20:26
Tv	robinr: something else is stressing the machine perhaps?	20:26
Illocution	Tv: right on, thanks	20:28
ReinH	Tv: generic but enthusiastic temporal greeting	20:28
robinr	Tv: seems I used too much memory, but those apps weren't much active.	20:28
Tv	ReinH: mumbled incoherent rambling about time being relative	20:28
robinr	still it must be suboptimal to do so many lstat calls	20:29
	firs one lstat, then unlink, lstat on all parent dirs, create and write and then lstat on the final file	20:31
aroben	is it a known bug that "git rebase -i HEAD branch" just checks out branch and does no rebasing?	20:32
loops	aroben, it's not a bug, if you specify upstream as HEAD, there is nothing to rebase.	20:35
aroben	loops: git rebase HEAD branch works as expected	20:36
Tv	loops: maybe his issue is HEAD before the checkout vs HEAD after	20:36
aroben	loops: (this is with a branch other than "branch" checked out)	20:36
Illocution	oh man and there's git log --diff-filter too... oh so much fun	20:37
loops	Tv, i don't think there is an issue :o)	20:37
	aroben, read the second paragraph of the git-rebase man page and see if it makes sense... ?	20:37
	" All changes made by commits in the current branch but that are not in <upstream> are saved to a temporary area. This is the same set of commits that would be shown by git log <upstream>..HEAD."	20:38
kergoth	Is there a way to have a user local version of /usr/share/git-core/templates/ ?	20:38
loops	aroben, so you are asking it to rebase all the changes between HEAD..HEAD	20:38
johan-s	is there something like a pre/post-fetch hook?	20:38
aroben	loops: no, I'm not	20:39
	loops: at the time I run git rebase, HEAD is pointing to 5aa51707	20:40
	loops: while "branch" is 2e133cc	20:40
	loops: git log 5aa51707..2e133cc shows the revisions I want to rebase	20:40
	loops: the first paragraph says that "If <branch> is specified, git-rebase will perform an automatic git checkout <branch> before doing anything else"	20:40
	loops: so it should checkout 2e133cc, then run git log 5aa51707..HEAD	20:41
loops	aroben, it sounds like you have are on a detached head.. which means you don't have _any_ branch checked out.	20:41
aroben	loops: no, I'm on a branch	20:41
	loops: sorry for not explaining this better	20:41
	loops: let me see if I can come up with a small testcase	20:41
robinr	... and when reset is done it lstats and opens every file again	20:43
loops	aroben, you're explaining it fine. When you invoke rebase you're passing HEAD as upstream. The manual explicitly states that what will be rebased is <upstream>..HEAD so you're trying to rebase HEAD..HEAD	20:43
aroben	loops: I guess I'd expect it to do `git rev-parse HEAD` before doing anything	20:44
	loops: which is what git-rebase seems to do, but not git-rebase -i	20:44
loops	aroben, you're seeing a difference when using -i ?	20:45
aroben	loops: exactly	20:45
	loops: sorry for not stating that earlier	20:45
	loops: git rebase -i HEAD branch is behaving differently from git rebase HEAD branch	20:45
	→ fultilt joined	20:45
aroben	loops: as far as I can tell, when running without -i "HEAD" is first passed to git-rev-parse	20:48
	loops: while with -i, "HEAD" is just taken as a string and used after the checkout	20:48
	loops: I guess I don't know which behavior is correct, but I definitely prefer the behavior without -i	20:48
	← kristoffer left	20:49
	→ kristoffer joined	20:50
robinr	aroben: your expectation seems more logical	20:51
loops	aroben, yeah you are right, i see the difference too and am surprised by it.	20:51
robinr	the HEAD at the time the command is issued	20:51
aroben	robinr: yeah	20:51
robinr	please -> git ml	20:51
aroben	robinr: I think of HEAD as just a synonym for whatever SHA1 happens to be the tip of the current branch	20:51
	robinr: will do	20:51
	← bobesponja left	20:54
	→ com4 joined	20:54
kergoth	how do i get a list of 'about to be committed' files?	20:56
Tv	kergoth: git status for humans, git ls-files with right options for computers	20:58
	kergoth is trying to figure out what the right options are :)	20:59
Tv	kergoth: i use this to autocommit all changes:	20:59
	git ls-files --deleted --modified --others -z \	20:59
	\| git update-index --add --remove -z --stdin	20:59
robinr	hard work	21:00
	git add -u	21:00
kergoth	right, but what i'm looking for is a list of changes that have already been updated in the index, but which havent yet been committed	21:00
	not working copy modifications	21:00
Tv	robinr: and old skool	21:00
	kergoth: git diff-tree HEAD ?	21:00
robinr	git ls-files --modified --deleted	21:00
kergoth	Tv: ah, good call. thanks	21:01
Tv	err not parents, that's not the right git diff-*	21:01
kergoth	hmm	21:01
Tv	kergoth: diff-index	21:01
kergoth	excellent, git diff-index --cached --name-only HEAD	21:02
	<3 thanks	21:02
	← Ingmar left	21:03
	← ferdy left	21:04
	→ Ingmar joined	21:04
	avarab → avar	21:08
	→ Ingmar_ joined	21:10
	← Ingmar left	21:14
	→ asdx_ joined	21:14
	bdiego → midgar	21:14
	← asdx left	21:17
	asdx_ → asdx	21:17
	← lamont left	21:17
	← freshtonic left	21:19
	← Ingmar_ left	21:19
	→ lamont joined	21:20
	→ sverrej joined	21:20
	→ Ingmar joined	21:23
	→ bobesponja joined	21:24
	← ironfroggy left	21:32
	← Ingmar left	21:35
	→ mnemoc joined	21:38
mnemoc	hi, how can i cherry-pick a range?	21:38
vmiklos	for + git rev-list + git cherry-pick :)	21:39
mnemoc	thanks for the hint, i'll experiment with git rev-list :)	21:40
	→ Ingmar joined	21:41
	← pflanze left	21:41
	← madewokherd left	21:42
	← dsaxena left	21:42
	→ madewokherd joined	21:45
	→ ferdy joined	21:47
	→ kukks joined	21:56
	← Ingmar left	22:05
	→ Ingmar joined	22:06
	← Ingmar left	22:13
	→ Ingmar joined	22:15
	← fujin left	22:15
	→ fujin joined	22:16
	← flazz left	22:16
	← kristoffer left	22:25
	→ aazar joined	22:27
	← mithraic left	22:27
	← eetfunk_ left	22:31
	→ mithraic joined	22:32
	→ dsaxena joined	22:37
	→ ironfroggy joined	22:40
	← octopod left	22:42
	← bobesponja left	22:44
	→ bdiego joined	22:45
	→ mithro joined	22:46
	→ tarbo_ joined	22:50
	← chrizz left	22:55
	← cmarcelo left	22:59
	← vbgunz left	23:01
	← tarbo left	23:01
Toad	hey	23:01
	is there some way to do a git-merge -s theirs?	23:02
	i.e. take all conflicts from the branch being merged in?	23:02
DrNick	do it in reverse using ours	23:03
Toad	hmm, that's a thought	23:03
	→ bobesponja joined	23:05
	avar → avarab	23:06
Toad	wait, does -s ours do a complete replace on all files?	23:06
	I want to only take theirs on conflicts	23:06
robinr	sound dangerous to do that automatically	23:07
	← samgranieri left	23:07
	→ toxx_ joined	23:07
	← devurandom left	23:08
Toad	robinr: since when does git try to prevent dangerous activity? :-P	23:09
	→ Debolaz2 joined	23:09
	← ceez left	23:09
	← Debolaz left	23:09
	← Man-erg left	23:10
	→ iXce joined	23:11
	← madewokherd left	23:12
	→ madewokherd joined	23:12
	→ b0rdslide joined	23:13
	← EmilMedve left	23:13
b0rdslide	hi guys. would someone be able to help out with a gitosis problem? I get the following error when trying to run gitosis-init... http://pastie.caboo.se/148968	23:14
	← alley_cat left	23:20
	← Ingmar left	23:23
	← Ryback_ left	23:25
	→ Ingmar joined	23:30
	← careo left	23:31
	← Cerebon left	23:33
	← stick left	23:34
	→ Randal joined	23:37
	← ddaa left	23:43
	← b0rdslide left	23:44
	→ sverrej_ joined	23:45
	← robinr left	23:46
	← sverrej left	23:52
	→ baggles joined	23:55
baggles	hi. is there a shorter way to do this: git-checkout mob && git-merge master && git-checkout master	23:56
	avarab → avar	23:57
	→ tjafk3 joined	23:58
	→ KirinDave joined	23:58
vmiklos	not really.	23:59
baggles	ok then.	23:59

Logs Search ←Prev date Next date→

Channels Documentation