IRCloggy #git 2010-03-01

Provider of IRC logs since 2005.

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

	2010-03-01
	→ Azeroth-Working joined	00:00
	→ novas0x2a joined	00:00
	→ beeman_nl joined	00:00
beeman_nl	hello :)	00:00
helloworld_xyz	Aaaah, time for some coding!	00:00
beeman_nl	is there a way that i can make git checkout this specific tree: http://github.com/cakephp/cakephp1x/tree/1.2.6 or should i checkout the whole source and then switch to that tree?	00:00
	← Guest69622 left	00:00
	← Azeroth-Work left	00:00
beeman_nl	actually github calles it a tag i believe	00:01
terinjokes	bob2: is it normal for the web interface to be up, but the server be unreachable by client?	00:01
Ilari	beeman_nl: Clone the whole repo and checkout the right commit.	00:01
Fernandos	ahh this is exactly what I want :) http://mirror.averse.net/pub/software/scm/git/docs/v1.6.0.6/howto/using-merge-subtree.html	00:01
beeman_nl	Ilari: oke, thank you :)	00:01
Fernandos	jhelwig: thanks for the link!	00:01
helloworld_xyz	Meh, one more question: I have four other submodules of the project where I don't care about history nor about being able to push/pull from/to them. What is the command to add them in the ./lib folder without reading all of the history? Is it "git submodule add lib/other.sub1 --depth 3"?	00:01
	Fernandos time for backups	00:01
Ilari	terinjokes: "Host unreachable" should mean that no port works because there isn't network path to it.	00:01
bob2	terinjokes: depends on how it's setup I guess - ie if they're on the same host or not	00:02
	← thiago_home left	00:02
terinjokes	Ilari: I get "Network is unreachable"	00:02
	← tildeequals_ left	00:02
Ilari	terinjokes: Well, that would mean whole network segment can't be reached (instead just one host is unreachable).	00:02
	← Paraselene_ left	00:03
	→ rloaderro joined	00:03
impulze	how do i compare my current working dir with a specific revision?	00:04
	i.e. without it being added to the staging area	00:04
bob2	impulze: git diff takes a revid	00:04
impulze	yeah i tried git diff ..SOMEREV	00:04
	but it lists changes between HEAD and SOMEREV	00:04
	not between working dir and SOMEREV	00:04
Ilari	impulze: 'git diff SOMEREV'	00:05
impulze	hm nope	00:05
	still shows changes between HEAD and SOMEREV	00:05
	→ bambule joined	00:05
helloworld_xyz	about my own question: I can say that it was not the answer I sought, it downloads all of it	00:06
	→ vice-versa joined	00:06
impulze	oh nvm Ilari	00:07
	i was being stupid, sorry, thank you	00:07
	→ fool_ joined	00:07
	← fool_ left	00:07
	→ fool_ joined	00:07
terinjokes	Ilari, bob2: IDK what's up, but pingged the git server from another computer, and added the domain to my hosts, and worked just fine...	00:09
	→ Paraselene_ joined	00:09
	← Paraselene_ left	00:09
	← hyperair left	00:10
	→ sunil joined	00:10
helloworld_xyz	What if I check out a submodule and have all line endings set to CRLF, but one file of the submodule doesn't have that, so when I try to checkout the specific branch, it fails because I have so called modified that file?	00:10
	← plediii_ left	00:10
bob2	changing all your line endings is modifying the file	00:11
	← sunil left	00:11
helloworld_xyz	bob2: well I don't care about them	00:11
	← fool__ left	00:11
helloworld_xyz	bob2: so I would gladly change them for that file, especially so in all submodules, I don't care	00:11
	→ Paraselene_ joined	00:11
helloworld_xyz	bob2: is there a way to say "don't care" to git?	00:11
	← Paraselene_ left	00:11
	← sven^ left	00:12
	→ sven^ joined	00:12
	← rudi_s_ left	00:13
Fernandos	helloworld_xyz: .gitignore	00:13
	you put the path of file into that .gitignore file that you don't want in your repo	00:13
Mage	Fernandos: um	00:13
	Fernandos: he means for line endings; read up	00:14
Fernandos	whoops :-[	00:14
helloworld_xyz	;)	00:14
Mage	convert all your line endings to one format and dont ever touch them again	00:14
	:>	00:14
helloworld_xyz	Well, I have :p, but not all my dependencies have.	00:14
Fernandos	helloworld_xyz: I know that you can decide that upon installation of git itself	00:15
	helloworld_xyz: there are tree options: windows line endings, unix line endings, keep line endings as they are	00:15
helloworld_xyz	Fernandos: The problem isn't choosing line endings, the problem is when they are mixed and I have a dependency which I cannot change; then I don't want dependors on my fork + submodule to be bothered by it	00:15
	I don't even know if it's a problem, perhaps the easiest solution is to add the file to .gitignore?	00:16
	but there are sometimes a lot of these files	00:16
Fernandos	hmm.. dunno	00:16
	→ consolers joined	00:17
	← toxx_ left	00:18
	← hal left	00:23
	→ the7thbit joined	00:24
	← the7thbit left	00:25
	← sven^ left	00:25
	→ sven^ joined	00:25
	→ Paraselene__ joined	00:27
Fernandos	yeahh I had succcess mergint the subtree	00:27
	ehrm.. how do I update in future?	00:28
	do I need the subproject's folder anymore?	00:28
jhelwig	Fernandos: Do another subtree merge.	00:28
Fernandos	/path/to/B can be deleted now right?	00:28
	→ sunil joined	00:31
	← sunil left	00:31
	→ flaguy48 joined	00:32
Fernandos	jhelwig: so, I'm not sure, can I delete /path/to/B now? this is the folder that I merged into my main repo. I thought I'm no more dependant on this folder after doing a subtree merge	00:33
jhelwig	Fernandos: Did you want to merge the two repositories permanently, or just pull in what was in the other one?	00:33
Fernandos	permanantely	00:34
jhelwig	Ok, then you can delete the other one.	00:34
Fernandos	good :) jhelwig thanks again for helping git n00bs like me out	00:34
	I apprecite that	00:35
	appreciate^	00:35
jhelwig	No prob.	00:35
	← saste left	00:35
	← Sonderblade left	00:36
jhelwig	Fernandos: Thanks for RTFM, when someone points you at the FM. ;-) That doesn't happen nearly often enough. lol	00:36
Fernandos	hahahah lol	00:37
	:D	00:37
	I'm building a model based booking application ontop of zend integrated into concrete5 cms	00:37
	→ AkumaStreak joined	00:38
	→ andres joined	00:38
	← terinjokes left	00:38
Fernandos	non-commercial^ it's a university project	00:38
	→ hyperair joined	00:40
	→ basus joined	00:40
basus	I was wondering what would be the best way to have a repository that is essentially a collection of other repositories	00:41
Fernandos	maybe what I did? subtree merges..	00:41
	← d0ugal left	00:42
	→ Adjective joined	00:43
	← consolers left	00:43
	← ProperNoun left	00:46
basus	subtree merges looks like what I want	00:46
	I'm using the doc here: http://mirror.averse.net/pub/software/scm/git/docs/v1.6.0.6/howto/using-merge-subtree.html	00:46
	I'm not sure what line 5 in the code does	00:46
	← novas0x2a left	00:47
Fernandos	basus: this helped me http://mirror.averse.net/pub/software/scm/git/docs/v1.6.0.6/howto/using-merge-subtree.html	00:47
	not sure too, but I just did it xD	00:48
	→ squisher joined	00:48
basus	I'm not sure what I need to do when I want to pull up dates	00:48
	← cgkm left	00:50
	→ bdowning joined	00:56
	← sagsousuke left	00:58
	→ Kelly_zhou joined	01:01
helloworld_xyz	I need some more help, got stuck again! :(. Got ./myproject, ./myproject/lib/sub1 (submodule), ./myproject/lib/sub1/lib/subx (nested submodule)	01:01
	I want ./myproject to download all of sub1's submodules. How?	01:02
	PS C:\Users\xyz\dev\code\open source\castle\work\Castle.Facilities.AutomaticTransactionManagement> git submodule update	01:02
	--rebase --recursive	01:02
	Doesn't work	01:02
cbreak	if you want to update submodules	01:02
	do git submodule update	01:02
helloworld_xyz	i.e. --rebase --recursive	01:02
cbreak	but be aware	01:02
	this will not give you the newest version of the submodules	01:02
	it will give you the correct one	01:03
	(the one that matches your current commit)	01:03
helloworld_xyz	Yes, but how do I make it download the last one?	01:03
cbreak	you have to go into each of them and manually check it out	01:04
	maybe foreach works	01:04
helloworld_xyz	Ah, ok.	01:04
	→ Leemp joined	01:04
helloworld_xyz	Nothing recursive then?	01:04
	← DrNick left	01:05
cbreak	there's a flag for that	01:05
helloworld_xyz	something like git submodule init lib/sub1 --recursive?	01:05
	That didn't work	01:06
cbreak	no, of course not	01:06
	→ novas0x2a joined	01:06
cbreak	randomly selecting commands rarely works :)	01:06
	→ frogonwheels joined	01:06
cbreak	helloworld_xyz: man git-submodule	01:06
helloworld_xyz	lol	01:06
Gitbot	helloworld_xyz: the git-submodule manpage can be found at http://git.or.cz/man/git-submodule	01:06
cbreak	check out "foreach"	01:06
RandalSchwartz	foreach?	01:06
cbreak	and "git checkout master"	01:06
	yes	01:06
	it's in the man page	01:06
	→ sixvu joined	01:07
helloworld_xyz	I've been there a long time by now	01:07
sixvu	Hey	01:07
helloworld_xyz	First instance : google	01:07
	Second : more google	01:07
	Third : IRC	01:07
	Fourth : Crying	01:07
sixvu	How to I pull a branch from origin which I don't have yet? git clone just pull the master branch	01:07
cbreak	pull is a merge	01:08
sixvu	(that used to be different with 1.4.something, right?)	01:08
cbreak	do you really want to merge the branch into your master?	01:08
	you likely do not want	01:08
sixvu	I just want to have the branch available locally	01:08
cbreak	then do git branch -r	01:08
	it should show you the remote ones you can pick from	01:08
	then do git checkout -t origin/foobarz	01:08
helloworld_xyz	"Evaluates an arbitrary shell command in each checked out submodule." -> Seems like I can't init them through this as they aren't "checked out"?	01:08
cbreak	you should get a local foobarz branch	01:09
	helloworld_xyz: you have them already	01:09
	why would you want to init them?	01:09
	you want to check out	01:09
sixvu	cbreak: jep, that worked. so the actual data is already there? because git pull still says that everything is up to date	01:09
helloworld_xyz	I see, thx	01:09
	Then do a submodule update?	01:10
cbreak	sixvu: if you did a git pull, you MERGED it into your current branch	01:10
	the git fetch part probably fetched all branches	01:10
	helloworld_xyz: that will not give you the newest	01:10
	it will give you the correct	01:10
RandalSchwartz	pull = fetch + merge	01:10
	generally not useful	01:10
sixvu	cbreak: I just did a clone, then the commands you said (which switched me to the new 'remote tracking branch', then I did a pull to actually get the data	01:10
RandalSchwartz	you want fetch, then figure out what you want	01:10
cbreak	pull is very nice if you have set up tracking	01:11
RandalSchwartz	if you're smart	01:11
cbreak	and expect fast forward	01:11
RandalSchwartz	local tracking branches generally mean you're not clued yet	01:11
	so if you have a lot of those, oops :)	01:11
cbreak	sixvu: you should have had the data already. hmm.	01:11
RandalSchwartz	nearly always, there are two camps:	01:11
	people wanting to know what's upstream, and have it	01:12
	or people wanting to keep rebasing on an upstream	01:12
	neither of those are served by pull	01:12
	so almost always, pull is wrong	01:12
cbreak	most people just want to get what ever remote has	01:12
nevyn	local tracking is bad?	01:12
cbreak	pull is perfect for that	01:12
	not really	01:12
RandalSchwartz	if you have a local branch that is always just the same as the remote branch, that's dumb	01:13
cbreak	no	01:13
davr	there's also pull --rebase	01:13
RandalSchwartz	just checkout the remote branch	01:13
nevyn	oh duh.	01:13
RandalSchwartz	right pull -rebase makes sense	01:13
	but naive pull is bad	01:13
cbreak	rebase is problematic	01:13
	don't do it unless you know what you are doing	01:13
RandalSchwartz	only if you're pushing	01:13
	as I said, let's distinguish those categories	01:14
cbreak	... if you're not pushing, your work is probably not usefull anyway	01:14
RandalSchwartz	people just tracking remotes	01:14
	and people tracking local changes on remotes	01:14
	for both of those, fetch is better	01:14
cbreak	pulling IS fetching	01:14
	among others	01:14
RandalSchwartz	pull = fetch + merge	01:14
cbreak	exactly	01:15
RandalSchwartz	most people do not need the merge	01:15
	hence	01:15
cbreak	or fetch + rebase	01:15
RandalSchwartz	dude - see above	01:15
jaysern	do I undo a merge ?	01:15
	← pantsman left	01:15
RandalSchwartz	don't want to repeat for the fourth time	01:15
cbreak	don't know	01:15
	do you?	01:15
	← tvw left	01:15
jaysern	i just merged one branch to another, and want to undo it	01:15
cbreak	you can use git reset	01:15
RandalSchwartz	jaysern - just reset it	01:15
	git checkout THATBRANCH	01:15
	git reset --hard HEAD~1	01:16
helloworld_xyz	Can't you generalize what you two people are saying to that if you want some code but don't modify it, you can use rebase, but if you're working on it you also want the merge, so pull?	01:16
RandalSchwartz	nope	01:16
cbreak	if you have not made any changs	01:16
RandalSchwartz	I don't like merge, ever	01:16
cbreak	rebase isn't needed	01:16
RandalSchwartz	either you're tracking remote	01:16
cbreak	rebase is for when you have local changes which you want to put on top of what ever is upstream	01:16
RandalSchwartz	so fetch works	01:16
jaysern	I merged 'a' into 'b', so i go to 'b' and reset it from there ?	01:16
	← mike8901 left	01:16
RandalSchwartz	or I'm tracking local changes	01:16
cbreak	jayne: you want to reset it away from where it is	01:17
RandalSchwartz	so either pull --rebase, or fetch, and rebase	01:17
helloworld_xyz	So let's ask this: what can go wrong with rebase that's easier handled with merge?	01:17
RandalSchwartz	or if I'm developing with others, I follow rebase route generally	01:17
helloworld_xyz	According to cbreak	01:17
jaysern	oh, i forgot, so i have a private repo on github, which is just used by me. after merging, i pushed it	01:17
cbreak	both rebasing and merging can cause conflicts	01:17
jaysern	so i will have to undo the push ?	01:17
cbreak	you can force push afterwards	01:17
jaysern	so reset, then push	01:18
RandalSchwartz	yeah, force	01:18
jaysern	gotcha, thanks	01:18
cbreak	wether you rebase or merge depends on preference	01:18
	rebasing rewrites history	01:18
	for the purpose of making it linear	01:18
RandalSchwartz	but local history is heap	01:18
	cheap	01:18
cbreak	merging creates merge commits	01:18
RandalSchwartz	it depends on whether you've published that or not	01:18
	→ DWSR joined	01:18
RandalSchwartz	in general, rebasing local history is better for others	01:18
cbreak	which preserve the history, but you will get parallel strains of developnent	01:18
RandalSchwartz	rebasing pushed history is worse for others	01:19
	→ jinzhu joined	01:19
cbreak	if you pushed something, you should NEVER rebase	01:19
	unless everyone who pulls from there knows about it	01:19
RandalSchwartz	i think we violently agree, which is why I wonder why you're arguing with me	01:19
DWSR	Is there a way to a) remove all mention of select files from a repository or b) remove all mention of them from a push to a remote repository?	01:19
helloworld_xyz	hehe	01:19
	← Paraselene__ left	01:19
cbreak	no good questions to answer.	01:19
DWSR	Break you, cbreak.	01:19
helloworld_xyz	That's a good IRC-quote, Randal.	01:20
cbreak	DWSR: you can with man git-filter-branch	01:20
Gitbot	DWSR: the git-filter-branch manpage can be found at http://git.or.cz/man/git-filter-branch	01:20
cbreak	but be warned	01:20
	this will rewrite history	01:20
	and so, you will have to force the push to remote	01:20
DWSR	cbreak: that's fine.	01:20
RandalSchwartz	yup	01:20
DWSR	The remote is currently empty.	01:20
cbreak	and everyone who pulled from there will become very confused	01:20
	← Hackykid left	01:20
RandalSchwartz	history will be rewritten :)	01:20
	the nazis won!	01:21
DWSR	You mean like CANADA WON THE MENS HOCKEY GAME?	01:21
	Sorry.	01:21
RandalSchwartz	no - in this new universe, canada lost	01:21
jaysern	what is HEAD~1 ?	01:21
cbreak	git filter-branch --index-filter 'git rm --cached nazi.txt'	01:21
	← saccade_ left	01:21
jhelwig	DWSR: Yes, everyone was very confused by that. ;-)	01:21
RandalSchwartz	that's the consequence of your rewrite	01:21
cbreak	leftmost parent of HEAD	01:21
Fernandos	cbreak: hahaha funny command :D hahaah	01:21
	→ mike8901 joined	01:22
jhelwig	jaysern: HEAD~1 is the 1st generation 1st parent of HEAD.	01:22
Fernandos	rm -R /trash/nazi/	01:22
RandalSchwartz	mv /dev/nazi /dev/null	01:22
Fernandos	hahaha	01:22
	that's real geek humor	01:22
	I like that	01:22
RandalSchwartz	it happens	01:23
	hang out with me more. that's my humor level	01:23
	listen to FLOSS Weekly then	01:23
	twit.tv/floss	01:23
Fernandos	never heard	01:23
DWSR	cbreak: something like git-filter-branch --index-filter 'git rm <file>'?	01:23
RandalSchwartz	start. :)	01:23
cbreak	if you use --index-filter, you have to use --cached	01:23
	since it only changes the index, not the working dir	01:23
	(disclaimer: I've never used that command myself... make backups)	01:24
DWSR	This is a temp copy anyway	01:24
	If I use --tree-filter, will this completely obliterate the files, like a Frost DK with Killing Machine proc'ed?	01:24
cbreak	tree and index filter have the same result	01:25
	index is faster though	01:25
DWSR	So why use tree at all then?	01:25
cbreak	you can use non-git commands	01:25
DWSR	ah	01:26
cbreak	for example, run sed on the files	01:26
RandalSchwartz	are there non-git commands? :)	01:26
cbreak	index filter doesn't give you the files, they won't get checked out	01:26
DWSR	ugh, fail.	01:27
	git-filter-branch doesn't come with msysgit	01:27
bremner	orly?	01:28
rtyler	I'm blanking here, if I want to see a log of the changes that are unique in remote/branch, what's the right syntax?	01:28
cbreak	git filter-branch --tree-filter 'find . -name "*.cpp" sed -i bak \'s/foo/bar/m\' {} \; '	01:28
	maybe	01:28
bremner	DWSR: does msysgit have a problem with bash scripts?	01:28
cbreak	DWSR: it's git filter-branch	01:28
DWSR	cbreak: ah, thanks.	01:28
bremner	oops	01:28
rtyler	git log branch..remote/branch seems to be correct	01:28
DWSR	cbreak: How do I specify that I want to affect all repositories?	01:29
	err.	01:29
	s/repositories/revisions/	01:29
	← sixvu left	01:30
cbreak	hmm.	01:30
	no idea.	01:30
	it doesn't work when you specify master?	01:30
Fernandos	damn.. I've to push 17 changes now	01:30
	what is rebase? I merged all projects now	01:31
bremner	check the man page, there's an option for git filter-branch to filter all branches, if that is what you mean	01:31
Fernandos	looks good, except that the log tree looks freaky	01:31
cbreak	Fernandos: man git-rebase	01:31
Gitbot	Fernandos: the git-rebase manpage can be found at http://git.or.cz/man/git-rebase	01:31
cbreak	take a look at rebase -i, especially squash/edit	01:31
	although...	01:31
	no idea if you can squash merge commits	01:31
	← DavidKlein^ left	01:32
jhelwig	rtyler: git log master..origin/master # <-- That will show you what is in origin/master that you don't have in your master.	01:32
cbreak	if in doubt, make backups :)	01:32
Fernandos	cbreak: I already have a backup	01:32
DWSR	git filter-branch --index-filter'git rm --cached -r *.bat' HEAD is incorrect syntax?	01:32
	→ Hackykid joined	01:33
Fernandos	cbreak: what I want to do is, to remove all those branches and merge the logs into the base	01:33
cbreak	this should delete all bat files in the current directory	01:33
Fernandos	cbreak: I guess rebase is for that	01:33
bremner	DWSR: maybe space after index-filter	01:33
cbreak	yes, that too	01:33
	maybe you have to escape the *	01:34
	so that git gets it	01:34
	but I think it should work :)	01:34
DWSR	cbreak: Ah, yes.	01:35
Fernandos	I'm rebasing now ...	01:35
	→ loinclot_ joined	01:36
	← loincloth left	01:37
Fernandos	oh..	01:38
	rebase puts all files from all branches into the main dir?	01:38
cbreak	no	01:39
	rebase is powerful. history changing powerful	01:39
DWSR	cbreak: Ah.	01:39
cbreak	but it should not influence other branches	01:39
DWSR	cbreak: You need to add --ignore-unmatch in order to run through all of the revisions.	01:39
	← themill left	01:40
cbreak	maybe. as I said, I only read the docs :)	01:40
DWSR	cbreak: Else the filter-branch stops on the first commit that doesn't have match the filter.	01:40
Fernandos	rebase takes much time...	01:40
cbreak	depends on what you rebase	01:40
DWSR	I'm rewriting 1450 commits!	01:40
	→ SlickT10 joined	01:41
Fernandos	I have 19commits	01:41
	but each with thousands of files	01:41
cbreak	as far as I know, noninteractive rebase only takes the changeset of the commits, and puts it on top of whatever	01:41
Fernandos	I think	01:41
DWSR	Fernandos: Yeah, I probably have ~500.	01:42
Fernandos	then I was doing something wrong..	01:42
	← mjf left	01:44
	← kpreid_ left	01:45
	→ sobersabre joined	01:46
sobersabre	hi. is there a way to add into the source code special tags to be substituted by git to show author, last commit date, etc. from the file itself ?	01:46
	(like svn properties svn:keywords)	01:47
	← Nies left	01:48
cbreak	maybe you can hack together something with the hooks	01:48
	to replace them on checkout and commit	01:49
	→ cgkm joined	01:49
sobersabre	cbreak: it's something I'm just used to. It comes since the days of rcs, IIRC.	01:49
cbreak	or maybe filters	01:54
DWSR	sobersabre: Yes, it's definitely possible.	01:54
	You need to use hooks.	01:54
cbreak	sobersabre: man gitattributes, look for filter	01:54
Gitbot	sobersabre: the gitattributes manpage can be found at http://git.or.cz/man/gitattributes	01:54
cbreak	that's probably better than using hooks	01:55
	← rloaderro left	01:58
	← Hackykid left	02:00
bob2	(it does go back to rcs, but few modern systems implement it directly)	02:01
	← brettgoulder_ left	02:06
Fernandos	how do I find out if there is a commit that I have in a branche but not in the master?	02:06
	← ericpugh left	02:07
	→ ericpugh joined	02:07
	← ericpugh left	02:07
Fernandos	so commit diff between branch x and master	02:07
bob2	diff is easy	02:08
bremner	sobersabre: http://git.wiki.kernel.org/index.php/GitFaq#Does_git_have_keyword_expansion.3F	02:09
Fernandos	bob2: ok.. and how do I do that?	02:09
bob2	'git diff otherthign...'	02:09
	maybe log will do what you want	02:09
	→ saccade_ joined	02:09
Fernandos	I'm trying to see if I have a commit in a branch that I don't have in my master	02:09
	if not I'm goint to delete the branch	02:10
	→ brettgoulder joined	02:12
	← SlickT10 left	02:13
	→ plediii_ joined	02:14
SpookyET	Is Google going to add Git? Mercurial sucks.	02:18
sobersabre	bremner: thanks, READING this.	02:18
Fernandos	ok.. better idea.. how can I archive a branch?	02:21
	is that possible?	02:21
RandalSchwartz	what does "archive" mean?	02:21
	a branch is a point	02:21
	so you have a specific commit	02:21
	what do you want to do with it?	02:21
Fernandos	I just don't want it to be visible in branches, but let the history and files stay	02:21
	kinda archiving some branches instead of deleting them	02:22
cbreak	push it to some other bare repo, zip that repo, burn it on a CD and put it into a save	02:22
RandalSchwartz	just tag it	02:22
	and delete the branch	02:22
	tags are relatively permanent	02:22
Fernandos	1. switch and checkout branch B 2. tag branch B 3. delete branch B 4. switch and checkout master is this right?	02:23
	← Chillance left	02:27
	→ VVSiz_ joined	02:29
DWSR	http://wowuidev.pastey.net/133593-3axk <-- cbreak: I'm getting this message from WoWAce when I push. I rewrote history already. wtf?	02:29
cbreak	git log --stat cd6d82ca979850417cf4121b193f59f102c8d79d	02:30
	what does it show?	02:30
	also, git log, is cd6d82ca979850417cf4121b193f59f102c8d79d in the current branch?	02:30
DWSR	cbreak: I'm commiting master.	02:31
	How do I show which branch I'm on? git status?	02:31
cbreak	yes	02:31
	or git branch	02:31
DWSR	Oh, yeah, I'm in master.	02:31
	It's on the damned prompt line. >.<	02:31
	← VVSiz left	02:32
DWSR	Yes, cd6d82ca979850417cf4121b193f59f102c8d79d is in the current branch.	02:33
	← nullvariable left	02:33
	→ nullvariable1 joined	02:33
DWSR	And it doesn't really show much	02:33
	nullvariable1 → nullvariable	02:33
cbreak	git log --stat should show the changed files	02:33
Fernandos	RandalSchwartz: nice that worked	02:34
	my history is almost linear now	02:34
SpookyET	Does anyone have any idea what this means? http://pastie.org/private/wqffctlhfsfq3vpgnxog It failed to push.	02:35
	→ magn3ts joined	02:35
DWSR	cbreak: http://wowuidev.pastey.net/133595	02:35
	→ laj joined	02:35
cbreak	weird	02:36
	do a git checkout cd6d82ca979850417cf4121b193f59f102c8d79d, does it contain still any bat files?	02:36
	← magn3ts left	02:37
	→ magn3ts joined	02:37
DWSR	Yeah, it does. =\	02:37
	DWSR sighs	02:37
	→ jmcantrell joined	02:38
DWSR	I thought that git filter-branch would..you know, remove it.	02:39
cbreak	well... unless it's in a subdirectory :)	02:40
	I think I meantioned that	02:40
DWSR	I passed -r to git r.	02:40
	git rm*	02:40
cbreak	yeah. but -r means "remove folders including contents"	02:40
	it's like the -r to normal rm	02:40
DWSR	cbreak: Anyway, it removed the file from a bunch of commits.	02:41
	But not this specific one.	02:41
	Can I git rm from this checkout and then recommit?	02:41
cbreak	where is it?	02:41
DWSR	in the root	02:41
cbreak	you can, but you have to do git rebase -i	02:41
	it's probably in other commits too	02:42
	since it didn't show up in the log	02:42
	it has to be in at least the ancestor	02:42
	→ yml joined	02:42
cbreak	I'd try to do git filter-branch again, something must have gone wrong	02:42
DWSR	cbreak: How do I get back to head, git checkout HEAD?	02:42
cbreak	no	02:42
DWSR	(I'm used to working with a GUI).	02:42
cbreak	master	02:42
DWSR	ah k	02:43
cbreak	HEAD is where you are now	02:43
	← hyperair left	02:43
DWSR	ah	02:43
yml	I have clone a repo on github. Done some changes and now I want to push the modification onto the branch I have created here : http://github.com/yml/django-varnish	02:44
	I can't find the secret spell to make it push there	02:44
	← basus left	02:45
cbreak	why not?	02:46
yml	so far I am with this : git push --repo=git@github.com:yml/django-varnish.git origin master	02:46
cbreak	unm... what's that --repo for?	02:47
DWSR	cbreak: Once I do git filter-branch, do I have to recommit?	02:47
	cbreak: Because that's probably what I did wrong.	02:47
cbreak	I don't think so	02:47
	filter-branch should rewrite all your commits	02:47
	so the sha should change from all	02:47
	(you shouldn't have cd6d82ca979850417cf4121b193f59f102c8d79d anymore)	02:48
DWSR	Well I do, so obviously something's wrong. :P	02:48
yml	cbreak: I am trying to make git push to a different repo that I have clone from	02:49
jhelwig	DWSR: You'll still have that commit, but you shouldn't have it anymore in the branch you were running filter-branch on.	02:49
cbreak	then why do you specify origin if you don't want to push to origin?	02:49
	and why do you not add that other with git remote add first?	02:49
	→ kpreid_ joined	02:49
	→ f00li5h[HireMe] joined	02:49
yml	cbreak: In fact I am just trying to push to my newly created github branch	02:50
cbreak	don't write origin if you don't mean it	02:50
	→ scarabx joined	02:50
yml	cbreak: when I do not specify anything git tell me that everything is uptodate	02:51
cbreak	specifying something you don't want won't make it work...	02:52
yml	cbreak: I have started by this git push --repo=git@github.com:yml/django-varnish.git	02:52
	but it tell me that everything is uptodate	02:52
sergio_	yml, you must tell git you have a new remote which you would like to push to	02:53
cbreak	how about doing what I said two or three minutes ago	02:53
	and just add it as remote?	02:53
yml	where in my local repository there is 3 commits that are noton github	02:53
cbreak	if you don't want to read the man page	02:53
	at least read what I write :)	02:53
yml	cbreak: git remote add -m [email@hidden.address] <= does this looks ok ?	02:55
cbreak	why did you throw a random -m in there?	02:57
yml	cbreak: because I read the man page and it seems to be that I want to add a master remote repository	02:59
cbreak	... no	02:59
	that's not what -m does	02:59
	leave it away	02:59
yml	ok I also tried this : git remote add --mirror my_github [email@hidden.address]	03:00
cbreak	...	03:00
yml	git didn't complain	03:00
	so I thought it was a good sign	03:00
cbreak	why the hell did you decide to use --mirror?	03:00
	it is not what you want with 99% probability	03:01
yml	ok trying without	03:01
	cbreak: thanks for you guidance	03:01
cbreak	np.	03:02
	← locklace left	03:03
DWSR	cbreak: So I should be running filter-branch on master and not HEAD?	03:03
cbreak	depends	03:03
	← Leemp left	03:03
cbreak	didn't you say your HEAD was master?	03:03
DWSR	I guess, I don't really know.	03:04
cbreak	(you did git checkout master after all)	03:04
DWSR	yeah. :P	03:04
	So I guess so.	03:04
	Not really used to git cli	03:04
cbreak	from what I read, it should work that way	03:04
	since it doesn't you either did something wrong, or the way I thought it'd work doesn't as I thought	03:04
yml	cbreak: could you pleaaaseee give me the secret spell to switch to my github branch and push the modif there ?	03:05
cbreak	switch to your branch?	03:05
	you can switch branches with git checkout branchnameyouwanttogoto	03:06
	but I don't know why you'd want to	03:06
yml	cbreak: sorry don't know the git wording	03:06
	cbreak: I did : git clone git://github.com/justquick/django-varnish.git	03:06
cbreak	you can just push with git push remotenameyoujustchose branchnameyouwanttopush	03:06
yml	hack commit hack commit	03:07
	← SpookyET left	03:07
	→ locklace joined	03:07
cbreak	of course, you know you can't push to git://, right?	03:07
	→ j416 joined	03:07
cbreak	at least if my infos are correct, git:// is read only	03:07
yml	I now want to push my modification there: [email@hidden.address]	03:07
cbreak	then git remote add that...	03:08
	and git push	03:08
DWSR	cbreak: You don't have to add it as a remote.	03:08
yml	so that would be : git remote add [email@hidden.address]	03:08
	git push	03:08
	← helloworld_xyz left	03:08
cbreak	yeah... but I asume he wants to push repeatedly...	03:08
	no	03:08
	→ defectiv joined	03:08
yml	yes	03:08
cbreak	yml... read the docs	03:08
	yml: man git-remote-add	03:09
	ah, it's git-remote	03:09
yml	thanks	03:09
cbreak	you have to give the remotes you add a name	03:09
	git remote add nameyouwanttogiveit url	03:09
	anyway	03:09
	bed time.	03:09
	→ Tommy[D] joined	03:09
	← j416 left	03:10
	→ munichlinux joined	03:12
	→ hyperair joined	03:15
	→ sshc_ joined	03:16
	← jast left	03:17
	← sergio_ left	03:17
	← Bass10 left	03:18
	← squisher left	03:18
	→ reprore joined	03:18
Fernandos	what do I used "tag" for?	03:19
	eh I mean what is "tag" for	03:19
fryguy	provide a name to a fixed state of the repository (a released version, for example)	03:20
jhelwig	Fernandos: Generally people use tags to mark the various releases. If you get a copy of git.git, and do "git tag -l", you'll see all the various versions of Git.	03:20
Fernandos	so it's kind of a bookmark	03:21
fryguy	sort of	03:21
Fernandos	okay	03:21
	→ jast joined	03:22
	sshc_ → sshc	03:23
DWSR	jhelwig: Can't you do a lot more with tags?	03:23
	← cytrinox left	03:24
jhelwig	DWSR: You can tag any object, but tagging releases is the most common use.	03:24
	→ johnw joined	03:25
	→ cytrinox joined	03:25
Fernandos	can someone help me rebasing the concrete5 branch and refining_search branch into master?	03:26
	http://image-upload.de/image/Raph1N/4b28792ff2.png	03:26
DWSR	git rebase -i?	03:27
jhelwig	Fernandos: Git checkout concrete5 && git rebase master #<-- That doesn't do what you want?	03:28
Fernandos	ah I didn't know that I had to checkout	03:28
	I tried git rebase concrete5 master without success ^	03:29
	lol	03:29
	→ hakunin joined	03:29
	→ Blaine joined	03:29
	→ DrNick joined	03:29
Blaine	in a git diff, what would "index ba57016..9b46a3f 100644" mean?	03:30
Fernandos	jhelwig: and then: git checkout refining_search && git rebase master ?	03:31
jhelwig	Fernandos: Yeah.	03:31
Fernandos	error: pathspec 'refining_search' did not match any file(s) known to git.	03:31
jhelwig	Blaine: The two SHA1s are the hashes of the file in the diff, and the last bit is the permissions of the file.	03:32
Fernandos	jhelwig: aah it is git checkout refining_search/master && git rebase master..right?	03:32
jhelwig	Fernandos: No? Is refining_search a remote or a branch?	03:33
Fernandos	a branch	03:33
jhelwig	Then you shouldn't have the "/master".	03:33
Blaine	jhelwig: in `git show <commit>`, is the first SHA1 the file before the commit?	03:33
FFForever	Anyone know a good reliable git host? with 10-15 projects?	03:33
jhelwig	Blaine: Yeah, if you're still talking about the "index ..." line.	03:35
DWSR	FFForever: github.	03:35
Blaine	jhelwig: yup, same line	03:35
FFForever	DWSR, for private projects?	03:35
Blaine	I'm having trouble figuring out which version the -/+ lines are referring to	03:36
DWSR	FFForever: You can probably configure github projects to be private.	03:36
jhelwig	FFForever: If you have a paid account, you can have private repos on GitHub.	03:37
fryguy	or just get a server somewhere	03:37
	and do git via ssh	03:37
Blaine	jhelwig: same goes for version a and version b?	03:37
	e.g. a/somefile b/somefile	03:37
	← coolcat left	03:37
Blaine	a is before commit, b is after?	03:37
	→ RolaBlade joined	03:38
jhelwig	Blaine: You might want to check out man git-config and consider setting diff.mnemonicprefix	03:38
Gitbot	Blaine: the git-config manpage can be found at http://git.or.cz/man/git-config	03:38
Blaine	so if I have: --- a/somefile +++ b/somefile; + foobar; - foobaz; the commit adds foobar to somefile	03:38
RolaBlade	hello... I am trying to omit all object file when commiting but it does not work. I have *.o on my .gitignore file	03:39
	← Rolenun left	03:39
Blaine	oh, sweet	03:39
jhelwig	RolaBlade: Have you already committed the .o files at some point in the past?	03:39
RandalSchwartz	RolaBlade - do they show up in "git status"?	03:39
	← magn3ts left	03:39
RandalSchwartz	yeah - already added?	03:39
	→ Rolenun joined	03:39
RolaBlade	RandalSchwartz: yeah... they do show up in "git status"	03:40
	→ dys` joined	03:40
RolaBlade	jhelwig: I guess so	03:40
jhelwig	RolaBlade: If they show up in git status as anything other than "untracked files", then you need to "git rm <file>" and commit that before the .gitignore will keep them from showing up.	03:41
	If they show up as untracked files, then your .gitignore isn't quite right. ;-)	03:41
FFForever	does the server I am pushing to have to have git?	03:41
jhelwig	FFForever: As a general rule: Yes.	03:42
FFForever	as a requirement?	03:42
DWSR	No	03:42
	← jaysern left	03:42
bcoca	might just have a repo, but then your methods are limited	03:42
FFForever	how so?	03:42
	I don't need public reading	03:43
	← dys left	03:43
RandalSchwartz	you can't "push" unless you can push :)	03:43
bcoca	access methods, i think w/o git you only have local and ssh	03:43
FFForever	ahhh	03:43
	← alus left	03:43
RandalSchwartz	and push requires git remote :)	03:43
Ilari	Without git on remote side there's only dumb HTTP (aaa...) or rsync (aaa...)	03:43
jhelwig	bcoca: You need git on the remote end with ssh.	03:43
	→ jaysern joined	03:43
	← jaysern left	03:43
	bcoca stands corrected	03:44
	← parasti left	03:45
	→ AkumaStreak_ joined	03:46
	← AkumaStreak left	03:48
	← RobertMeta left	03:52
	→ multigrain_ joined	03:53
	→ magn3ts joined	03:55
	← multigrain left	03:57
	multigrain_ → multigrain	03:57
	→ consolers joined	04:00
	→ hipe joined	04:00
	→ eletuchy joined	04:05
	← stringo0 left	04:08
	← Twisted_Mentat left	04:10
	→ ArdRigh joined	04:14
	← aziz left	04:16
	→ MetaCosm_ joined	04:18
	→ RobertMeta joined	04:23
	← MetaCosm_ left	04:24
	→ Fernandos1 joined	04:26
	← Fernandos left	04:26
	Fernandos1 → Fernandos	04:26
	← RolaBlade left	04:26
	→ tedoc2000 joined	04:27
	dys` → dys	04:30
	← kukks left	04:30
	← defectiv left	04:31
	→ ilteris_ joined	04:32
	← Blaine left	04:32
	→ Blaine joined	04:32
	→ khmarbaise joined	04:34
	← Fernandos left	04:35
	→ ankit9 joined	04:38
	← dcampano left	04:38
	→ defn joined	04:39
	→ dcampano joined	04:39
defn	If i have a repo.git on my server and I scp it to another server, can I just change my local copy to point to the new server and continue where i left off without any issue?	04:39
	→ sunil joined	04:39
	← hipe left	04:39
	→ xinming_ joined	04:39
	→ hipe joined	04:40
consolers	defn yeah	04:40
	so far so good	04:40
	← sunil left	04:41
	→ Azeroth-Work joined	04:42
	← Azeroth-Working left	04:42
	← xinming left	04:43
DWSR	Is there a way that I can split a git repository into two repositories?	04:43
	← RobertMeta left	04:43
consolers	sideways? or across?	04:45
	→ susheel joined	04:47
	← tedoc2000 left	04:47
DWSR	http://stackoverflow.com/questions/359424/detach-subdirectory-into-separate-git-repository <-- Something like this is what I'm looking for.	04:48
	I'm assuming that those instructions are pretty accurate?	04:48
	← sven^ left	04:51
	→ sven^ joined	04:51
	→ greg8si joined	04:53
AAA_awright	DWSR: Yep	04:54
	← munichlinux left	04:57
DWSR	ok	04:57
	I just used git clone --no-hardlinks and somehow git's saying that a bunch of files are modified?	04:58
	← aresnick left	05:00
	susheel → susheelvarma	05:00
	← multigrain left	05:00
	susheelvarma → susheel	05:01
	→ MetaCosm_ joined	05:01
DWSR	http://xkcd.com/ win.	05:02
	← greg8si left	05:06
	→ greg8si joined	05:06
	→ Leemp joined	05:07
	← MetaCosm_ left	05:07
	→ MetaCosm_ joined	05:08
	← Rolenun left	05:08
	← khmarbaise left	05:08
	← magn3ts left	05:10
	← susheel left	05:12
	→ anathematic joined	05:13
	→ amerine joined	05:14
anathematic	how do I revert to an old commit? git checkout hash right?	05:14
bob2	depends what you mean by revert	05:14
	checkout will revert the working copy	05:14
	revert will "revert" a branch	05:14
	→ bugfux joined	05:15
ben_h	http://twitter.com/hyazvco	05:15
	anathematic waves to ben_h	05:15
ben_h	ergh wrong chan, sorry folks	05:15
	← eletuchy left	05:15
ben_h	hi anathematic :)	05:15
anathematic	bob2: I'd like to remove that commit for the moment	05:15
	← ankit9 left	05:16
ben_h	anathematic: revert makes a new commit that reverses an existing commit	05:16
anathematic	ben_h: that sounds like what I'm after	05:16
ben_h	you can remove previous commits with either reset —hard or rebase, but both of those change your history	05:16
anathematic	so git revert hash_number ?	05:16
ben_h	so you shouldn't use them unless you haven't pushed	05:16
	yep	05:16
anathematic	thanks <3	05:17
ben_h	in general you should only actually remove commits that were mistakes	05:17
	as opposed to code-level mistakes, which are legitimately part of your dev history	05:17
anathematic	ben_h: that's what I figured, I wasn't really sure what to do	05:17
	→ eletuchy joined	05:17
anathematic	I don't make mistakes so I odn't understand what you mean sorry ;-)	05:17
ben_h	lol.	05:19
	→ psankar joined	05:19
	← yml left	05:20
ben_h	there's another undo concept i like as well, which is 'undo the fact that i just made a commit'	05:20
	git reset —soft HEAD^	05:20
anathematic	that whole reset --soft --hard thing confuses me a bit	05:20
	← disappearedng left	05:20
ben_h	which means, reset HEAD to its parent, but don't change the working copy	05:20
	yeah it makes sense, just not in the way you might expect	05:20
	reset means "reset HEAD to a given commit"	05:21
	—soft means don't change the working copy, —hard means do change it	05:21
anathematic	ohh	05:21
ben_h	there's also —mixed, which i haven't looked into :)	05:22
anathematic	I think --soft and --hard are enough for me to deal with atm	05:22
ben_h	oh right, of course.	05:22
	—soft means don't change the index or the working copy	05:22
	—mixed means change the index, but not the working copy	05:22
	—hard means change both	05:22
	—mixed is the default	05:22
anathematic	kay will I'm going back to rails now because I know rails :-D	05:23
	thanks!	05:23
consolers	does gits getopt only accept -- (2 ascii char 45s) or does it accept unicode #x2014 em-dash	05:24
	← cgkm left	05:25
	← reprore left	05:25
consolers	ben_h your software is surreptitiously converting your -- to some unicode nonsense!	05:25
ben_h	consolers: i think it's —	05:28
	or whatever the char is	05:28
	irritating, let me see if i can turn it off	05:28
	→ ajonat joined	05:28
	← MetaCosm_ left	05:29
	→ MetaCosm_ joined	05:29
	→ disappearedng joined	05:30
	← anathematic left	05:31
	→ webus joined	05:32
	→ eut joined	05:33
eut	hello	05:33
	is there a way to temporarily revert a few commits? (i want to test performance between different patches)	05:33
bob2	git help revert -> -n	05:34
	← cloudhead left	05:34
	← mssm left	05:34
eut	once they have been reverted, how do i un-revert them?	05:38
offby1	"git reset --hard SHA1-FROM-BEFORE-YOU-REVERTED"	05:39
	→ defectiv joined	05:39
ben_h	eut: you should use a branch for that	05:39
	← ttvd left	05:39
bob2	git checkout -f	05:39
ben_h	git stash; git checkout -b test; git reset —hard OLD_ID	05:40
bob2	oh, no need for -f	05:40
	we already did the other bit	05:40
ben_h	then just 'git checkout branch_name' to switch between master and test	05:40
	→ ttvd joined	05:40
	→ ankit9 joined	05:41
	← jmcantrell left	05:41
	→ roop joined	05:43
	← MetaCosm_ left	05:44
	→ Matt_R joined	05:45
	← disappearedng left	05:46
	← consolers left	05:46
	→ MetaCosm_ joined	05:46
	← rgr left	05:49
	→ cgkm joined	05:49
	→ rgr joined	05:49
	← rgr left	05:51
	→ xxthink joined	05:51
	→ rgr joined	05:51
xxthink	How to cancel a commit by git?	05:51
bob2	cancel = ?	05:51
xxthink	I use git commit -a -m "xxx"	05:51
bob2	commit the inverse?	05:51
xxthink	yes	05:52
bob2	or remove it from history?	05:52
xxthink	I want to cancel this commit	05:52
	I only want to change the log message from "xxx" to "yyy"	05:52
bob2	commit --amend	05:53
xxthink	bob2: do you know how to do that?	05:53
	ok	05:53
bob2	assuming you never pushed it	05:53
	if you have, yhl, hand	05:53
xxthink	I still not push	05:53
	→ tildeequals joined	05:53
ben_h	if all you want to do is change the commit message, then you just want to recreate the commit	05:54
	which is commit —amend as bob2 said	05:54
	to "cancel" the commit, you can remove it from your history, using reset —hard, but that's probably not what you want.	05:55
	← nullvariable left	05:55
	→ nullvariable joined	05:55
	← project2501a left	05:56
	← nullvariable left	05:56
	→ moconnor joined	05:58
	→ disappearedng joined	06:00
	← scarabx left	06:00
	← defectiv left	06:01
	→ project2501a joined	06:02
	← brettgoulder left	06:04
	← roop left	06:05
	→ roop joined	06:07
	→ muri_one joined	06:09
	→ RobertMeta joined	06:09
	← MetaCosm_ left	06:10
xxthink	yes, thank you very much, ben_h and bob2	06:10
	I use --amend	06:10
	→ loincloth joined	06:11
muri_one	does git look at file modification times before it does any hashing on files?	06:11
	← Determinist left	06:11
	→ magn3ts joined	06:11
	→ taufiqkh joined	06:12
	← johnw left	06:12
ben_h	muri_one: afaik git ignores modification times altogether	06:13
	← loinclot_ left	06:14
	→ Ratler joined	06:15
	← amerine left	06:16
muri_one	it seems surprisingly fast to be relying soley on a hash of file contents	06:16
DrNick	it looks at both mtime and ctime	06:18
	← Textmode left	06:18
	→ amerine joined	06:19
	← webus left	06:20
ben_h	oh sorry i'm with you now	06:20
	i was thinking of what's included in the commits	06:20
defn	If i have a repo.git on my server and I scp it to another server, can I just change my local copy to point to the new server and continue where i left off without any issue?	06:21
bob2	should do	06:22
	scp might have trashed the perms though	06:22
eli	I'm getting lots of "Use of uninitialized value $git_dir in concatenation (.) or string at [...]" -- is there some known problem?	06:25
bob2	that's not from git	06:25
	that's from perl	06:25
eli	Yeah -- gitweb.	06:25
bob2	perhaps you didn't configure git_Dir	06:25
	spelt correctly even	06:25
eli	Looks like it's the repo directory, not a configuration option.	06:26
	(And otherwise things work fine.)	06:26
	← dutchie left	06:26
	→ MetaCosm_ joined	06:27
	→ Adaptee joined	06:27
	← rgr left	06:27
	← project2501a left	06:28
	→ rgr joined	06:28
	← ph^___ left	06:29
eli	...and it also happens only on the gitweb top page, not when viewing a repository.	06:29
	← amerine left	06:30
	← RobertMeta left	06:31
	← taufiqkh left	06:32
	→ ph^ joined	06:32
	→ reprore joined	06:32
	← EricInBNE left	06:36
	← magn3ts left	06:36
	→ reprore__ joined	06:39
	→ codeshepherd joined	06:41
codeshepherd	what is HEAD? I understand that HEAD points to the current branch.. but apart from that what is its purpose?	06:41
	← rgr left	06:42
	← reprore left	06:42
	→ rgr joined	06:44
	→ repro____ joined	06:44
	← bugfux left	06:44
	← reprore__ left	06:45
Mage	to point to the top of the current branch	06:45
	without having to know the exact branch head	06:45
	numer	06:45
	or hash, as it were	06:45
	← greg8si left	06:47
	← codeshepherd left	06:47
	→ codeshepherd joined	06:48
	← repro____ left	06:49
codeshepherd	thanks maattd\|away	06:49
	thanks Mage	06:49
	→ psoo joined	06:49
	← marc_kdab left	06:51
	→ magn3ts joined	06:51
	← mattcodes left	06:51
	→ Theravadan joined	06:52
	→ Ingmar joined	06:52
	← VVSiz_ left	06:55
	← codeshepherd left	06:55
	→ VVSiz joined	06:56
	→ codeshepherd joined	06:56
	→ RobertMeta joined	06:57
	← sven^ left	06:57
	→ sven^ joined	06:57
	→ cjs joined	06:58
	← codeshepherd left	06:58
	→ codeshepherd joined	06:59
	← rgr left	07:00
	← MetaCosm_ left	07:00
	→ tomekw joined	07:01
	← krawek left	07:02
	← wolog left	07:03
	← bambule left	07:03
	→ rgr joined	07:04
	← codeshepherd left	07:06
	→ voxpelli joined	07:06
	→ codeshepherd joined	07:06
	→ dpino joined	07:07
	← codeshepherd left	07:07
	→ rgr_ joined	07:08
	← cgkm left	07:08
	→ mattcodes joined	07:09
	→ thiago_home joined	07:11
	→ RdeWilde joined	07:17
	→ defectiv joined	07:19
	← NetHawk left	07:21
	→ ph^_ joined	07:22
	→ mastro joined	07:23
	→ sde_antrix1 joined	07:23
	← ajonat left	07:24
	→ maxthoursie joined	07:24
	← artagnon left	07:25
	→ froschi joined	07:27
	← Jeena left	07:28
	→ Azeroth-Working joined	07:28
	← Azeroth-Work left	07:29
	← spacebat left	07:31
	→ codeshepherd joined	07:31
codeshepherd	I did " git pull origin master" ==> git said the following ..	07:32
	From [email@hidden.address]	07:32
	* branch master -> FETCH_HEAD	07:32
	Already up-to-date.	07:32
	what exactly does *branch master -> FETCH_HEAD mean?	07:32
	→ spacebat joined	07:32
	← moconnor left	07:35
	→ taufiqkh joined	07:35
	→ priidu joined	07:35
	→ ProperNoun joined	07:36
	← Adjective left	07:38
	→ Jeena joined	07:40
	→ giallu joined	07:40
	← ph^ left	07:40
	→ ph^ joined	07:41
	→ munichlinux joined	07:41
	← voxpelli left	07:43
thiago_home	codeshepherd: it means git pull did: git fetch origin master	07:45
	which in turn downloads the master branch off origin and saves it in FETCH_HEAD	07:45
	FETCH_HEAD is a temporary ref that contains the result of the last fetch	07:45
defn	i need to learn git and quit doing crappy commits and learn how to merge properly	07:45
	i always get conflicted merges and then go through and edit files, commit stuff, get out of the conflicted merge but feel dirty afterwards	07:46
codeshepherd	thiago_home: where did it merge FETCH_HEAD with?	07:46
	→ DavidKlein joined	07:46
thiago_home	codeshepherd: master	07:46
	but the result was "Already up-to-date"	07:46
	→ voxpelli joined	07:47
codeshepherd	what confuses me is... which master does it refer to .. local master or origin master ... I am taking about the line * branch master -> FETCH_HEAD	07:47
	thiago_home: ^	07:48
bob2	local	07:48
	← eletuchy left	07:48
	→ skoop joined	07:48
	← skoop left	07:48
	→ skoop joined	07:48
	→ eletuchy joined	07:49
	← voxpelli left	07:50
	← Jeena left	07:50
	← Matt_R left	07:51
	← thiago_home left	07:51
taufiqkh	is there any easy way to get git to ignore cr/lf eol differences when merging or cherry-picking?	07:53
bob2	ultimately, fixing your developers will pay huge dividends	07:54
	← robinr left	07:54
bob2	or adding hooks that forbid it	07:54
taufiqkh	definitely, but in the meantime there's a fair bit of work to be done with an existing code base	07:54
	← Pupeno left	07:55
	← maxthoursie left	07:55
	→ Jeena joined	07:55
	→ maxthoursie joined	07:55
	→ Pupeno joined	07:56
	← Pupeno left	07:57
	→ wolog joined	07:58
	→ steffkes joined	07:58
	→ maxthoursie_ joined	07:59
	← maxthoursie left	07:59
	maxthoursie_ → maxthoursie	07:59
	maattd\|away → maattd	07:59
	→ maxthoursie_ joined	08:00
	← maxthoursie left	08:00
	maxthoursie_ → maxthoursie	08:00
	← taufiqkh left	08:00
	← defectiv left	08:00
	→ gnufied joined	08:02
	→ alreadygone joined	08:08
	← yhager left	08:08
	→ yhager joined	08:09
	← Jeena left	08:10
	→ EricInBNE joined	08:13
freedrull	i can't git pull on windows, it just sits there forever. however, i did a git clone fine...	08:13
	← maattd left	08:15
	→ janders joined	08:15
	→ FelipeMcMont joined	08:15
janders	Hi,	08:16
	→ project2501a joined	08:16
janders	I'm trying to create a new branch in a supermodule	08:16
	→ project2501b joined	08:17
	← FelipeMcMont left	08:17
janders	however, when I include some submodules with branches e.g. a,b and c, other submodules also get these branches	08:17
	Which they should not have.....	08:17
	Anyone familiar with these operations? I have more infomation if someone can help	08:18
	→ sis joined	08:19
	← Arafangion left	08:19
	← project2501a left	08:20
	← fmardini left	08:22
	→ ixti joined	08:25
	← loincloth left	08:26
	→ loincloth joined	08:27
	→ fmardini joined	08:28
	← fmardini left	08:32
	→ mbroeker joined	08:38
	→ fmardini joined	08:38
	→ sergio__ joined	08:39
	← eckzow left	08:40
	← ArdRigh left	08:40
	→ Jeena joined	08:40
kristian42_	Is there a global option for .gitignore ?	08:43
	← loincloth left	08:43
	→ loincloth joined	08:43
kristian42_	I.e. not just per-repo ?	08:43
codeshepherd	git config --global core.excludesfile ~/.gitignore kristian42_	08:45
	← psoo left	08:46
	→ d0ugal joined	08:51
	→ NetHawk joined	08:52
	→ psoo joined	08:55
	← hyperair left	08:55
	← froschi left	08:55
	← dreiss left	08:56
	→ mssm joined	08:57
	→ shruggar joined	08:57
	→ khmarbaise joined	08:58
	→ reprore joined	08:58
	← muri_one left	09:01
	→ twb joined	09:02
	← r11t left	09:02
	→ aziz joined	09:02
twb	Does git://git.sv.gnu.org/screen.git have any tags? A simple clone of it has an empty "git tag" output, but I'm not sure if I'm doing it wrong.	09:02
	ISTR needing some trick to make tags "visible" in the local checkout	09:03
Ilari	twb: It does not.	09:04
twb	OK, so obviously the OTHER repo, I've manually create a tag.	09:05
Ilari	twb: It has only three branches.	09:05
	(caption-never, cmd-alias and master)	09:05
	← ph^_ left	09:06
	← rgr left	09:06
	→ ph^_ joined	09:07
	← xxthink left	09:08
twb	How do I get "git tag" to tell me which hash the tag is pointing to?	09:09
	← EricInBNE left	09:09
	→ EricInBNE joined	09:10
twb	Never mind, I brute-forced it	09:10
Saur	git show <tag> should work	09:10
	← loincloth left	09:11
	→ loincloth joined	09:11
twb	I tried git log <tag>, which started from that tag	09:11
	→ Pupeno joined	09:12
	← EricInBNE left	09:12
Ilari	twb: 'git rev-parse <tag>^{}'.	09:12
twb	What's the ^{} for?	09:13
Saur	twb: To get the SHA1 for the commit the tag is pointing to, rather than the SHA1 for the tag	09:14
Ilari	twb: Peel tags operator.	09:14
	→ nicoulaj joined	09:14
twb	My head hurts	09:14
	→ nithin joined	09:15
	← Pupeno left	09:16
nithin	Hi ...is there any resource available that explains the git alternates mechanism?	09:16
	→ munichpython joined	09:17
Ilari	nithin: You are short on disk space or have some very large repositories?	09:19
nithin	no .. actually I have some code that I share b/w projects.. So was looking for the best way to set that up..	09:19
	← munichlinux left	09:20
Ilari	nithin: Alternates has nothing to do with that... Submodules or subtree merges maybe?	09:21
	← janders left	09:21
	→ janders joined	09:22
nithin	Ilari: would it be possible to share submodules with another repo?	09:24
	→ xxthink joined	09:24
	← xxthink left	09:25
	→ JoseLobato joined	09:27
Ilari	nithin: It is. Submodules are independent repos.	09:27
	← JoseLobato left	09:27
nithin	ILari: Thanks ..then I think thats what Im looking for..	09:28
	→ froschi joined	09:28
frogonwheels	nithin: look at it the other way round. you have your library as sperate project, then you import it as a submodule into the projects that require it.	09:29
	→ unreal_ joined	09:30
	→ Paraselene__ joined	09:30
	← mason-l left	09:30
eyzn	i use submodules to fetch a bunch of projects im interested in	09:30
nithin	ILari: Im confused on one point.. any changes made to a submodule via a project branch that includes it .. how does it reflect on the separate library project.. ?	09:30
eyzn	instead of writing a script, which fetchs each project - i do one git submodule update and its ok	09:31
	← nevyn left	09:31
eyzn	u have to push your changes in the submodule itself	09:31
	→ themill joined	09:31
frogonwheels	nithin: it's a seperate project from that point of view - you just push changes up.	09:31
eyzn	cd into it, change stuff commit and push	09:31
	← unreal left	09:31
	unreal_ → unreal	09:31
nithin	frogonwheels: oh ok ..makes sense.. ill just need to read up a bit on submodules.. any pointers?	09:34
	← tomekw left	09:35
	← psankar left	09:35
	← Theravadan left	09:35
	→ jkp joined	09:36
	← ph^_ left	09:36
	→ ph^_ joined	09:36
frogonwheels	nithin: this was suggested to me, and helped: http://progit.org/book/ch6-6.html	09:37
	→ tbf joined	09:38
frogonwheels	nithin: I'm only just ahead of you. gathering info about them before using them :)	09:38
	→ dj_tjerk joined	09:38
	→ tomekw joined	09:43
	→ unreal_ joined	09:44
	→ PointMan joined	09:44
	← reaVer left	09:44
	→ nizuri joined	09:44
	rgr_ → rgr	09:45
	→ tatsuya joined	09:45
	← unreal left	09:45
	unreal_ → unreal	09:45
	→ simNIX joined	09:47
	→ mason-l joined	09:47
	← NetHawk left	09:48
steffkes	if i have an f.e. symfony project .. with an _big_ lib/vendor folder .. a good idea to split this one in a submodule? or perhaps another idea for it?	09:50
	because lib/vendor inside the main project .. costs some seconds on an normal "git status" :/	09:50
	← froschi left	09:51
	→ bluenovember joined	09:53
Caelum	what's the command to see diff to previous commit (without using the md5)	09:54
	← twb left	09:54
Ilari	Caelum: 'git diff HEAD^'?	09:54
	Caelum: Or 'git diff HEAD'?	09:54
	Caelum: HEAD is last commit on current branch, HEAD^ is first parent of that.	09:55
Caelum	Ilari: thank you!	09:55
Ilari	Caelum: Also, MD5?	09:55
Caelum	from git log	09:55
bob2	man git-rev-parse	09:55
Gitbot	bob2: the git-rev-parse manpage can be found at http://git.or.cz/man/git-rev-parse	09:55
bob2	^ handy	09:55
Ilari	Caelum: Git uses SHA-1 as object name hash, not MD5.	09:56
Caelum	oh sorry	09:56
	← nizuri left	09:56
	→ nizuri joined	09:57
steffkes	nobody an suggestions for git / sf & submodule yes/no/do-it-the-other-way-round? :>	09:59
	← reprore left	09:59
	→ reprore_ joined	10:00
	→ aatanasov joined	10:00
	→ reaVer joined	10:01
	← reaVer left	10:01
	→ reaVer joined	10:01
	← alreadygone left	10:02
	→ anno^da joined	10:04
	→ sagsousuke joined	10:06
	→ willvarfar joined	10:06
	→ psankar joined	10:07
	willvarfar is terribly confused; I have a git server, and I have existing sourcecode, and I want to put it on the server	10:09
	← Kelly_zhou left	10:09
	← maxthoursie left	10:10
willvarfar	I have login working fine; on the git server there is a myproj.git already, but its empty; I clone it and I get an empty folder; I go into it, copy in my source, add, commit, push	10:10
	→ maxthoursie joined	10:10
willvarfar	it doesn't work: "You did not specify any refspecs to push" and "Perhaps you should specify a branch such as 'master'."	10:10
	→ froschi joined	10:11
codeshepherd	what is the difference between ^ and ~ ?	10:12
defn	git push origin master	10:12
	codeshepherd: in what context	10:12
charon	codeshepherd: man git-rev-parse	10:13
Gitbot	codeshepherd: the git-rev-parse manpage can be found at http://git.or.cz/man/git-rev-parse	10:13
codeshepherd	git diff HEAD^1 or git diff HEAD~1	10:13
willvarfar	defn, sweet! thx. so simple	10:13
defn	does anyone have a link to a good tutorial on git that uses branches, merges, etc. in a sort of real life context?	10:14
	kind of the git workflow for most users in an easy to follow tut.	10:14
nithin	frogonwheels: :)... thanks	10:14
PointMan	defn: check learn.github.com	10:14
	← psankar left	10:15
	→ dajero joined	10:15
defn	PointMan cool site -- thanks	10:15
	← RobertMeta left	10:16
	→ psankar joined	10:18
	→ consolers joined	10:19
	← reaVer left	10:22
	← loincloth left	10:23
	→ loincloth joined	10:23
	← nizuri left	10:23
	← eletuchy left	10:24
	← zjason``` left	10:24
	→ zjason``` joined	10:24
	← orafu left	10:26
	→ orafu joined	10:26
	← novas0x2a left	10:28
	→ novas0x2a joined	10:28
	→ rolando joined	10:29
	→ viniciusluiz joined	10:30
	← _Vi left	10:31
	← consolers left	10:31
	← robinsmidsrod left	10:32
	→ ijuma_ joined	10:33
	→ jetienne joined	10:33
jetienne	q. i got 7 repositories to pull, i currently do 7 git clone, is there a way to speed this up ?	10:33
_rane	pull?	10:34
	← TDJACR left	10:34
_rane	you could do shallow clones	10:35
	if that suits your needs	10:35
	→ tobiassjosten joined	10:35
Caelum	where do I put my ssh key for msys-git	10:38
	← Ingmar left	10:38
Caelum	if using the openssh it comes with	10:38
	→ TDJACR joined	10:38
	→ mstormo joined	10:39
	→ offby1` joined	10:40
jetienne	_rane: ok i do pull too. so i guess there is no magic bullet to speed up this case. Thanks	10:40
	← offby1 left	10:40
	→ reaVer joined	10:41
	← reaVer left	10:41
	→ reaVer joined	10:41
jast	jetienne: well, if they are completely different repositories, each clone has to get completely different data, so I don't think there could possibly any way to speed it up	10:42
jetienne	jast: thats cool. just wondered and ask to be sure	10:42
jast	sure :)	10:42
	→ tzafrir joined	10:42
	→ maxthoursie_ joined	10:45
	← maxthoursie left	10:45
	maxthoursie_ → maxthoursie	10:45
	→ _Vi joined	10:45
	← janders left	10:45
tzafrir	http://paste.debian.net/61951/ - my old git svn issue. Any idea what my next move can be?	10:46
	← yhager left	10:47
	→ helloworld_xyz joined	10:48
	→ yhager joined	10:48
	← helloworld_xyz left	10:49
	→ sylr joined	10:50
sylr	Hi	10:50
	← willvarfar left	10:51
	→ maxthoursie_ joined	10:51
	← sagsousuke left	10:51
tzafrir	sylr, hi	10:51
sylr	I have commited some changes on a wrong branch some times ago, what is the good behavior to have knowing this branch has been published and tagged ?	10:51
	← maxthoursie left	10:51
	maxthoursie_ → maxthoursie	10:51
	→ maxthoursie_ joined	10:53
	← maxthoursie left	10:53
	maxthoursie_ → maxthoursie	10:53
sylr	I was thinking : git diff 45baa94b~1 path/to/file/wrongly/commited \| patch -R	10:53
	but i'm not sure this is the best thing to do	10:53
jast	sylr: git revert does pretty much that, but automatically	10:54
	reverts a complete commit, though. I'm not sure whether you want that.	10:54
	← maxthoursie left	10:54
	→ maxthoursie_ joined	10:54
	maxthoursie_ → maxthoursie	10:54
	→ Ingmar joined	10:54
	← beeman_nl left	10:55
	← cYmen left	10:55
sylr	well, I don't want to revert entirely the commit, In fact I have switched branch but I forgot to commit in the previous branch	10:55
	← maxthoursie left	10:55
sylr	and the change has been commited in the checkedout branch	10:55
	→ maxthoursie joined	10:56
sylr	and I can not use git revert because the branch has been published and it will rewrite the branch history, right ?	10:57
teuf	git revert creates a new commit, does not mess with earlier history	10:58
	git rebase -i / git commit --amend would rewrite published history	10:58
sylr	ok but I have a lot of commit since the "wrong" commit	10:59
jast	revert just takes a single commit and commits the reverse of it on top of the current history	10:59
teuf	you can start with git revert (if it works cleanly) and then amend that commit before pushing it to get what you want ?	11:00
jast	it works exactly like cherry-pick but it reverses the patch, too	11:00
sylr	Ok I will try that	11:00
	thank you	11:01
	→ hyperair joined	11:02
	→ robinsmidsrod joined	11:03
	→ bentob0x joined	11:05
	← rolando left	11:07
tzafrir	specifically: how can I get rid of this wrongly-added commit 3504509917fab5479ba55ab7c008e92496463bd1 ?	11:09
	→ kang_ joined	11:10
tzafrir	(I'm not sure if this is the right question to ask, but it looks like a start)	11:10
	← maxthoursie left	11:13
	← jinzhu left	11:13
	→ maxthoursie joined	11:13
_rane	depends on your situation	11:19
	is it the latest commit, have you published it, etc.	11:19
	← reprore_ left	11:21
shruggar	and what you mean by "get rid of"	11:22
_rane	we need a "how to remove a commit" flow chart	11:24
shruggar	sounds like a good way to kill a lunch break :)	11:25
	→ darkskiez joined	11:26
darkskiez	im trying to use git svn, and it has created a bunch of remote branches, "trunk@rev", i'd like to rename these remote branches to the correct thing, any suggestions	11:27
shruggar	what do you think the "correct thing" would be?	11:27
darkskiez	the branch name that it came from	11:28
Rhonda	darkskiez: If you want to keep using git svn to commit back simply don't touch them.	11:29
darkskiez	right..hmm..	11:31
	← sven^ left	11:31
	← TDJACR left	11:31
darkskiez	we have a fairly convoluted and horrible svn repository, we dont have a 'trunk', just branches for each major release, in order to migrate to git, there would need to be a transitional period. I was hoping to educate the team in the way of the git, but its just so extra complicated when using an svn backend :[	11:31
	→ sven^ joined	11:32
shruggar	darkskiez, my recommendation tends to be "if your situation is even remotely complicated, do not track multiple branches using git-svn"	11:32
doener	mine is: If you have a choice, drop svn	11:33
Rhonda	darkskiez: blame that on git and the way "tags" work in there, and that branches don't need to have any clear branch point to start off with. :)	11:33
darkskiez	I want to drop svn entirely	11:33
tzafrir	darkskiez, http://gitorious.org/asterisk-tools/asterisk-tools/blobs/master/update_branches	11:33
Rhonda	bleah, s/on git/on svn/ obviously	11:33
darkskiez	but i need a migration path	11:33
tzafrir	(but then again, using git-svn with the asterisk repo and recent git will break your repo)	11:33
Rhonda	darkskiez: You are raising much more pain with wanting to have a transition period where both works, most of the times even with just basic svn usage.	11:34
darkskiez	at the moment we have a bunch of hook scripts to validate svn commits, will have to port those to git first, tho, because you can push multiple commits to the server in one got, the commit validation is all a bit different.	11:34
	s/in one got/in one go/	11:34
	→ d0k joined	11:36
	← dj_tjerk left	11:36
	→ TDJACR joined	11:37
darkskiez	I actually only wanted to checkout the current main branch as the trunk with git	11:38
	but it auto added all of these remote entries	11:38
	→ ivan joined	11:40
shruggar	what command did you use to clone the svn repos?	11:40
Caelum	the openssh in msysgit is giving me "referenced memory at "0x00000..."" probably because I have cygwin installed too...	11:41
darkskiez	git svn init --trunk=branches/build20 file://path/to/repo	11:41
shruggar	did you try "git svn init file://path/to/repo/branches/build20" ?	11:42
	(personally I tend to use git svn clone anyway)	11:43
darkskiez	shruggar: no, not tried that, will see what it does	11:48
tzafrir	Caelum, that is not really related to git	11:49
	← reaVer left	11:50
	tzafrir gives up on any chance of actually getting answers to his questions in this channel	11:50
darkskiez	shruggar: it did the same but named them git-svn instead of trunk	11:51
tzafrir	maybe it's time to look at hg-svn or bzr-svn :-(	11:51
defn	tzafrir: irc does not part like the red sea every single time you need a tooltip	11:51
	if i could help you i would	11:51
tzafrir	defn, I don't expect it to	11:51
	→ maxthoursie_ joined	11:52
	← maxthoursie left	11:52
	maxthoursie_ → maxthoursie	11:52
tzafrir	But right now I'm blocked by this issue, and lack the time to dig deeper into the technicalities (I already spent days on it)	11:52
	← anno^da left	11:53
tzafrir	git-svn is a useful svn browsing tool for me, but I can't afford to spend too much time maintaning it	11:53
	← Jeena left	11:53
tzafrir	I have already asked my question here and on the list several times, and hardly even got useful follow-ups	11:54
	← sven^ left	11:55
tzafrir	I really don't expect IRC / mailing list support to do my job. But I have some minimal expectations of a project that attempts to be thriving	11:55
	</rant>	11:55
	→ sven^ joined	11:55
locklace	is git-svn thriving?	11:57
	i see more problem reports than people capable of providing answers for it, at least here	11:57
tzafrir	locklace, so you suggest I check the alternative? :-(	11:57
locklace	i assumed the right forum for it was a dedicated mailing list	11:57
	i'm not suggesting anything, just mentioning what i've seen	11:58
shruggar	locklace, git-svn is still actively being developed	11:58
	locklace, that is, a quick skim of the git logs shows that there have been several commits for it in the past week. It's not dead or anything	11:58
tzafrir	locklace, I asked twice about problems I had with it. Once I got several useful followups, and the issue is still a "missing feature"	11:58
	(certificate authentication over https)	11:59
	Second time was this issue. And I only got one follow-up, suggesting something trivial.	11:59
locklace	the list must be the best place for hardcore git-svn questions then. best bet is probably just to repost there regularly until you get a useful response	12:00
	→ parasti joined	12:01
	← dajero left	12:02
	→ consolers joined	12:02
	→ catphish joined	12:06
	← hipe left	12:08
	→ abms1116 joined	12:09
	→ hipe joined	12:09
	→ dajero joined	12:10
	→ Adjective joined	12:13
	← ProperNoun left	12:15
	← munichpython left	12:19
	→ munichpython joined	12:19
	→ pheaver joined	12:21
	← catphish left	12:21
	← comps left	12:21
tzafrir	locklace, actually if you look at the paste I posted here, I'm not sure it's strictly a git-svn question:	12:25
	http://paste.debian.net/61951/	12:25
	→ snearch joined	12:29
	→ loinclot_ joined	12:31
	→ radarek joined	12:31
locklace	tzafrir: definitely looks like an issue for the git and/or git-svn devs. you might have more luck here at a different time of day, or else send mail to the git list as well as the git svn one	12:32
	← loincloth left	12:32
locklace	tzafrir: another pointed approach to take if you haven't done so already is file a bug report	12:33
	→ multigrain joined	12:34
	← neunon left	12:38
	← rduplain left	12:38
	→ JonathanT joined	12:38
	→ comps joined	12:40
	→ oliver3 joined	12:41
	→ Shapeshifter joined	12:41
	← giallu left	12:42
Shapeshifter	We have a terminal server with hundrets of users (they don't use it very often). Now amongst other things, we would like to provide some sort of git repository. People should be able to create a branch and allow others to make changes on it. any ideas how to do this? with multiple users having different branches on the same machine?	12:42
	→ spaceonline joined	12:42
	→ loincloth joined	12:43
	← consolers left	12:43
	← loinclot_ left	12:43
	→ maxthoursie_ joined	12:45
	← maxthoursie left	12:45
	maxthoursie_ → maxthoursie	12:45
tzafrir	Shapeshifter, sorry for the stupid question: how do users connect? (I'm not really sure what you mean by "terminal server)	12:46
	← loincloth left	12:47
	→ loinclot_ joined	12:47
Shapeshifter	tzafrir: mhh, I called it terminal server because people also have nx access. In any case, people have access over ssh, and we can open any ports we like if neccessary	12:47
	→ alienBOB joined	12:48
tzafrir	Do users have some sort of public_html ?	12:49
spb	if everyone has ssh access, you can just create a git repo in your home directory and use regular unix permissions to allow access	12:49
	← unijunction left	12:51
tzafrir	Shapeshifter, if they have, just consider the simple and lazy option: allow anybody to set up a repo on their home dir. No special write access	12:52
Shapeshifter	tzafrir: yes, we have public_html. what if we allo anybody to set up a repo in their home dir, can they allow others to access it nevertheless (even though other people do not have actual file access to other users folders?)	12:53
	← loinclot_ left	12:54
	→ loincloth joined	12:54
	← JEEB left	12:55
	→ JEEBcz joined	12:55
Ilari	Shapeshifter: Then if admins can get involved for setup, gitolite with "wildrepos" support enabled?	12:55
	→ dogmatic69 joined	12:56
dogmatic69	hi all	12:56
Ilari	...	12:56
	→ hced joined	12:56
	← aatanasov left	12:56
Shapeshifter	mhh, sounds interesting	12:56
dogmatic69	i have some new code/fies that are not yet commited to the repo, can i stash them and branch and then commit them to a new branch?	12:56
Ilari	Shapeshifter: The default mode works akin to MAC (Mandatory Access Control), but wildrepos mode is bit more like DAC (Discretory Access Control).	12:57
	Shapeshifter: You presumably need dedicated user account for gitolite... And also you need SSH keys from users (but CAP_DAC_READ_SEARCH/CAP_DAC_OVERRIDE is available, one could just read the from homedirs).	12:59
	← xinming_ left	12:59
	→ JackTheRabbit joined	12:59
charon	tzafrir: still looking for help on http://paste.debian.net/61951/ ? have you tried skipping the problematic revision with the -r option?	13:00
	tzafrir: also, the real issue is that a ref can never end in a dot, see git-check-ref-format(1)	13:01
	→ swordz joined	13:03
charon	i also suspect that ignore-paths only works for files in the repo, and not for paths that are branches/tags, but that's just a theory.	13:03
swordz	Hi. I'm fairly new to git, and am using it on a ubuntu machine (or 2). For the previous weekend I cloned some repositories onto my laptop to work on them as I went away. I've just pushed them back and all the commit messages are in the log. The files however, are not.	13:04
	What have I forgotten?	13:04
Ilari	swordz: $faq non-bare	13:04
	Oh darn, gitbot's borked.	13:04
Shapeshifter	Ilari: okay thanks	13:04
tzafrir	charon, thanks. The option -r for what? (I don't see it for 'git svn fetch)	13:05
	← loincloth left	13:05
swordz	so git push --non-bare ?	13:05
	It's not in the man pages	13:05
	→ loincloth joined	13:06
Ilari	swordz: Basically, don't push to checked out branch of repo with working copy.	13:06
charon	tzafrir: 'git svn fetch' takes a -r option that only lets it fetch the specified SVN revision range. so if revision N is problematic, you can try 'git svn init <options>; git svn fetch -r 1:N-1; git svn fetch -r N+1:HEAD'	13:06
	→ unijunction joined	13:06
tzafrir	charon, any idea what I can do to remove that broken commit?	13:07
Ilari	Shapeshifter: It seems even possible to only allow user 'foo' to create repos within '/users/foo/' (and then allow others to read and write to them).	13:07
charon	tzafrir: what do you mean "remove"?	13:07
tzafrir	not sure technically	13:07
	(git svn managed to put it in somehow already)	13:07
	← dogmatic69 left	13:07
charon	oh, it did?	13:07
	i.e., git show 3504509917fab5479ba55ab7c008e92496463bd1 works fine?	13:08
	← kpreid_ left	13:08
tzafrir	I also have a repo which worked fine with an older version of git, before that sanity check got in action	13:08
	charon, yes	13:08
	works fine	13:09
charon	hmm.	13:09
	i think you'll still have to use -r to skip it	13:09
	→ janders joined	13:09
charon	the fact that the commit object exists in your git repo doesn't seem to help much	13:09
	← janders left	13:09
charon	and/or devise a patch that filters git-svn's ref names through some sanitizer function that lets it be compatible with git's rules	13:10
	← abms1116 left	13:10
tzafrir	charon, -r seems to work . So far so good.	13:13
	Recreating this repo will probably take over a day	13:13
	→ f0rkit joined	13:13
tzafrir	So I'll only know for sure then	13:14
charon	tzafrir: you can keep that 'autotag_for_.' tag by using the update-ref provided, but adjusting the tag name so that git can cope with it	13:14
	tzafrir: also, if you have access to the server, i hear loading from file:// is much faster	13:15
	← fahadsadah left	13:15
tzafrir	no, I don't	13:15
	← pheaver left	13:16
tzafrir	at work I use svn over ssh, but connect through a control socket, which greatly reduces connection overhead	13:16
	→ auslander joined	13:17
	← auslander left	13:17
	→ auslander joined	13:17
	← RdeWilde left	13:18
	← defn left	13:19
	→ corni joined	13:19
	→ icwiener joined	13:20
	← darwin_ left	13:20
	→ darwin_ joined	13:21
	→ scarabx joined	13:22
	→ deuterium_ joined	13:22
deuterium_	re guys	13:22
	← psankar left	13:23
	→ FND joined	13:23
	← loincloth left	13:24
	→ loincloth joined	13:24
	→ Ryback_ joined	13:24
deuterium_	i've pushed my changes to origin/master, but now when my friend tries to pull these changes to his local repository, the tortoisegit client doesn't notice this change. what could be the problem?	13:24
	← sven^ left	13:24
	→ sven^ joined	13:25
FND	when defining a custom alias (e.g. foo = !echo $PWD), CWD is always the repo root - is this intentional, or could it be considered a bug?	13:25
charon	FND: hysterical raisins, can't be changed any more.	13:26
	← bentob0x left	13:26
FND	charon: is "hysterical raisins" code for backwards compatibility?	13:26
charon	it's a play on "historical reasons" that seems to be fairly common on the git list	13:27
FND	I don't suppose there's a workaround, as there's no way to pass in the spawning CWD?	13:27
	ah gotcha	13:27
	so it's a known wart	13:27
jetienne	old joke :)	13:27
charon	i'm afraid not. but you can place git-foo in your path and it will be invoked normally when you do 'git foo'	13:27
FND	that'll do, thanks	13:27
	(I currently have an alias gitsend, will make that git-send)	13:27
	→ reaVer joined	13:28
	→ bentob0x joined	13:28
	← saccade_ left	13:28
	← bentob0x left	13:30
	→ kpreid_ joined	13:31
	→ bentob0x joined	13:33
	← deuterium_ left	13:34
	← maxthoursie left	13:36
	→ maxthoursie joined	13:37
	→ deuterium joined	13:38
deuterium	re, sorry for the disconnect..	13:38
	→ mklappstuhl joined	13:38
	← Adaptee left	13:38
	→ GarethAdams\|Work joined	13:39
	→ style joined	13:40
deuterium	charon: do you have an idea what might have gone wrong if there are two "master" in my remote repository? might this be the reason why my collegue doesn't see my pushes when he pulls from the same remote repo (and me not seeing his)? i.e. that we both push to each our own remote masters instead of the same?	13:40
	← roop left	13:40
ben_h	that's normal, deuterium	13:41
style	any ideas, why git push says: "ssh: connect to host gitorius.org port 22: bad file number"?	13:41
	← sven^ left	13:41
	→ Yuuhi joined	13:41
	→ sven^ joined	13:41
	← uau left	13:41
wereHamster	style: proxy/firwall in the way?	13:41
	← multigrain left	13:42
ben_h	oh wait, sorry, i was thinking about the reverse, deuterium	13:42
style	firewall yes, but port 22 is open	13:42
	does it need some other ports too?	13:42
ben_h	i.e. fetching doesn't update your local master branch, only remotes/blah/master	13:42
wereHamster	style: no. try 'ssh -v [email@hidden.address] and see if anything stands out in the output	13:42
style	wereHamster: hmm says same	13:43
	→ uau joined	13:44
	← maxthoursie left	13:44
	→ LRN joined	13:44
wereHamster	style: I didn't expect any different outcome, but the output can give you a clue as to where the problem might be	13:44
	→ jaql joined	13:44
	→ maxthoursie joined	13:44
	← jaql left	13:44
wereHamster	style: use 'ssh -vvv ..' to get even more verbose output	13:45
style	mmh ok	13:45
Saur	style: Might wanna try strace as well...	13:45
	→ bremner joined	13:45
style	btw ssh connection to other servers works.. :(	13:45
deuterium	ben_h: so any idea what might be wrong with our remote repo? anybody ever experienced a similar situation (i.e. one doesn't see/get the other's pushes or that there are two masters in the remote repo)?	13:45
	← ben_alman left	13:46
Saur	deuterium: What do you mean by two masters?	13:46
style	it says "Attempt to connec timed out whitout establishing a connection"	13:46
	connect*	13:46
	← snearch left	13:47
	→ EXAAFZJ9 joined	13:47
wereHamster	deuterium: post the output of 'git for-each-ref' in the repo with the two masters	13:47
	→ defn joined	13:47
	→ nothingmuch joined	13:47
style	stupid me... its gitorious.org, not gitorius.org :D	13:48
Saur	lol	13:48
	→ eletuchy joined	13:49
	← deuterium left	13:49
	→ chuck joined	13:49
	→ deuteriunm joined	13:50
	→ SpookyET joined	13:50
	deuteriunm → deuterium	13:51
	→ yashi joined	13:51
	→ tychoish joined	13:51
SpookyET	Wondering if anyone uses git porcelains like eg, yap, etc.	13:51
deuterium	wereHamster, Saur: ok, mompl	13:51
	→ tstclair joined	13:51
	← swordz left	13:52
deuterium	Saur: in tortoisegit's pull dialog window there's a dropdown box to select/choose which remote branch to pull from. there are two 'master' entries there. there either seem to be two master branches in the remote repo or tortoisegit is buggy. which is more likely?	13:53
ben_h	deuterium: how do you mean, two masters?	13:53
	what does 'git branch -a' say?	13:53
	oh sorry, i lagged behind there	13:53
	→ hugoxrosa joined	13:53
deuterium	ben_h: two branches with the same name, i.e. two "master"	13:54
ben_h	deuterium: sounds like a tortoisegit bug	13:54
	you should use the commandline instead :)	13:54
deuterium	ben_h: oh ok	13:54
Saur	or it is not showing the full repfspec...	13:54
ben_h	yeah	13:54
	could be remotes/blah/master	13:54
	and it's just doing a basename	13:54
deuterium	let me check wereHamster's suggestion, too	13:54
	→ fahadsadah joined	13:55
ben_h	nice, wereHamster - didn't know that one	13:55
	→ Bass10 joined	13:57
	→ loinclot_ joined	13:58
deuterium	wereHamster: ok, the output is here: http://dpaste.com/166402/	13:58
wereHamster	deuterium: that repository has only one local branch 'master' (refs/heads/master)	13:59
	refs/remotes/origin/master is a remote tracking branch	13:59
	→ ben_alman joined	13:59
	← style left	13:59
deuterium	wereHamster: that's what's displayed when i do a 'git for-each-ref' on the same server that holds our remote repo	13:59
	wereHamster: so is this the expected output? or anything wrong?	14:00
Caelum	what's the difference between GIT_AUTHOR_NAME and GIT_COMMITTER_NAME	14:00
	← loincloth left	14:00
	→ remi joined	14:00
wereHamster	deuterium: is that repository non-bare?	14:00
	Caelum: author is the person who wrote the patch, committer who committed it to the repository	14:00
	→ mm_work joined	14:01
	→ giallu joined	14:01
deuterium	wereHamster: i.e. whether the one i issued 'git for-each-ref' contains outchecked files and dirs? yes	14:02
wereHamster	deuterium: it's not recommended to push into a non-bare repository	14:02
Caelum	wereHamster: thank you	14:02
deuterium	wereHamster: sorry, it's not the one we push to.. the one we push to should be bare (how can i check this?)	14:03
	should i do for-each-ref in the one we push to?	14:03
	← eletuchy left	14:04
	← FND left	14:04
	← munichpython left	14:05
wereHamster	yes	14:06
	well, do it in the repo which has two 'master' branches (which is what seems to confuse you)	14:06
	→ vmiklos joined	14:07
	→ mvanholstyn joined	14:07
deuterium	wereHamster: here's the output when i do 'for-each-ref' in the remote repo directly: http://dpaste.com/166403/	14:08
	← mvanholstyn left	14:08
deuterium	wereHamster: will also post the output when i do the same in the repo's clone on my client box (windows) here	14:09
	← LRN left	14:10
	← vmiklos left	14:12
	→ LRN joined	14:12
	← maxthoursie left	14:12
	→ bcardarella joined	14:12
	← rgr left	14:13
	→ vmiklos joined	14:13
	→ maxthoursie joined	14:14
	← scarabx left	14:14
	→ SLiMIT joined	14:16
	← deuterium left	14:19
	→ deuterium joined	14:19
	← yhager left	14:20
	← SLiMIT left	14:20
	→ yhager joined	14:20
deuterium	wereHamster: here's the output of 'git for-each-ref' when i execute it in the repo's clone on my local client box here: http://dpaste.com/166419/	14:25
	→ bauruine joined	14:25
	← spb left	14:25
deuterium	wereHamster: to me, that actually looks like everything's fine with the remote repo and the local clones?	14:25
	← jspiros left	14:29
	→ jspiros joined	14:30
	← jspiros left	14:30
	→ jspiros joined	14:30
	← bblfish left	14:30
	← vmiklos left	14:31
deuterium	does anybody else here use the latest stable version of tortoisegit? if so, experiencing similar problems?	14:32
	→ LiamH joined	14:32
	→ JJoop joined	14:32
	→ kibab joined	14:32
	→ vmiklos joined	14:33
kibab	git gc is only for working copies isn't it? (i.e. non-bare repositories)	14:34
	← JJoop left	14:34
	→ Mezner joined	14:37
shruggar	why wouldn't it be for bare repositories?	14:37
	→ Mike_MnD joined	14:37
Mike_MnD	i've added a file to the .gitignore but it's already tracked hot to make git stop tracking it from now on	14:38
kokx	git rm	14:38
	→ eletuchy joined	14:38
wereHamster	Mike_MnD: git rm --cached	14:38
kokx	wereHamster: thats only when its in the index	14:38
wereHamster	no, that's the one to use if you want to remove it from the index but not from the filesystem	14:39
kokx	Mike_MnD: is it already in the repo?	14:39
	wereHamster: and what if its in the repo?	14:39
wereHamster	he said it's already tracked.. (that means it's in the repo)	14:39
Mike_MnD	yes	14:39
kokx	does --cached work then?	14:39
	ok, then use --cached	14:39
kang_	ive a stupid question, i rsync'd a .git directory with a comitted merge on it, but i can't find the merge on the rsync'd version. is there anything obvious i've missed?	14:40
Mike_MnD	mmm how can check without making a pull :)	14:40
wereHamster	Mike_MnD: for what purpose?	14:40
kibab	shruggar: because it keeps saying it's not a repository (because no .git file exists in the parent-chain)... which makes sense for bare repositories	14:41
Mike_MnD	well i've did a git rm --cached	14:41
	so i'm going to make a pull and see what happeneds	14:41
	← PointMan left	14:41
Mike_MnD	but i'm curios if there is other way to check	14:41
wereHamster	you need to commit first	14:42
	→ ntoll joined	14:42
Mike_MnD	aham ok cause i was going to miss it since i commit jsut before removing the cached	14:43
	→ michuel joined	14:43
	→ Cascade- joined	14:44
Cascade-	MsnBot [Freezer] [MassAdder] [And More] Add 'bot@hackersrus.info' to your contact list for FREE access!	14:44
	← vmiklos left	14:44
	← Cascade- left	14:44
	← mm_work left	14:44
Mike_MnD	hmm :( after a pull it still takes and replace the file	14:45
	→ vmiklos joined	14:45
	← michuel left	14:45
Mike_MnD	i've got .gitignore in main folder and have webroot/index.php	14:46
	← radarek left	14:47
wereHamster	Mike_MnD: maybe you misunderstand how .gitignore works and expect a different outcome ...	14:48
	→ loincloth joined	14:49
	← loinclot_ left	14:49
	← plediii_ left	14:51
	← dcampano left	14:52
	← kpreid_ left	14:52
	→ cannonball joined	14:52
Mike_MnD	wereHamster: i don't misunderstand it since it's working for other files	14:53
	← Yuuhi left	14:53
Mike_MnD	but the case is i want to ignore and stop track	14:53
	→ Yuuhi joined	14:53
wereHamster	git rm --cached <file>; git commit	14:53
Mike_MnD	a file which is added and tracked in the moment	14:53
wereHamster	I don't see what that has to do with pull/checkout	14:53
Mike_MnD	well i've got some settings in index.php which are for dev and pub and	14:54
	when i make a pull for public for example it takes the dev settings	14:54
	and vice versa	14:54
	btw i think i found solution	14:54
	http://www.pagebakers.nl/tag/git-ignore-tracked-files/	14:54
jast	read the hints in the manual about --assume-unchanged. it has some pitfalls.	14:55
	→ bblfish joined	14:55
	→ kpreid_ joined	14:56
Mike_MnD	jast ok do you use/recomend other way for this ?	14:56
	→ nanoo joined	14:57
	← GarethAdams\|Work left	14:57
	← Mike_MnD left	14:59
	← tomekw left	15:00
	→ kukks joined	15:02
	← loincloth left	15:02
	→ rgr joined	15:03
	→ stepnem joined	15:04
	→ dcampano joined	15:04
	→ spb joined	15:04
	← Haarg left	15:05
	← hakunin left	15:06
	← rgr left	15:06
	← stepnem left	15:08
	→ jmcantrell joined	15:08
	→ stepnem joined	15:08
	→ enyovelcora joined	15:09
	← steveire left	15:10
	→ pheaver joined	15:11
	→ agile joined	15:11
	→ sagsousuke joined	15:11
	→ binjured joined	15:11
enyovelcora	I have a Webproject in a git repository. This Project depends on other JS Frameworks; some of them are written by me, some of them by other people in either SVN or git or whatever repos. I now want them in my project, in my git repository, so I don't have to maintain different repositories on my deployment server, and have to worry about them being up to date.	15:11
shruggar	I've got a merge conflict, and the conflict markers read "<<< HEAD ..[some data here].. >>> WIP" Is there any way to get the conflict markers to say something more useful than HEAD? HEAD sounds wrong to me. "HEAD" is, in my mind, where I was before I ran rebase. (aka: the commit with the message "WIP"). I don't have any commits with the message "HEAD".. it's all very head-slam-into-wall-ish	15:11
	← hugoxrosa left	15:12
	→ psankar joined	15:12
	← yairgo left	15:12
enyovelcora	I now thought of hard linking those files on my local drive, so any changes I do to other frameworks will be changed in my git folder for the webproject.. This way I have total control of when I want to checkin a new version of another framework.	15:12
	The problem with that, is that git seems to destroy those links from time to time... (probably when creating branches)	15:13
	← eMxyzptlk left	15:13
enyovelcora	Is there a good solution for that problem?	15:13
	→ eMxyzptlk joined	15:15
	→ marc_m joined	15:15
	→ khmarbaise_ joined	15:17
	→ alexpercsi joined	15:18
	→ smuf joined	15:18
shruggar	enyovelcora, submodules, perhaps?	15:18
alexpercsi	how do i create a branch from origin/master?	15:19
enyovelcora	shruggar: Is it possible to specify a specific version in submodules?	15:20
	→ marcusb joined	15:20
	← khmarbaise left	15:20
	khmarbaise_ → khmarbaise	15:20
	→ nizuri joined	15:21
shruggar	enyovelcora, as far as I know, it is not possible to NOT specify a specific version in submodules	15:22
	← marcusb left	15:22
	→ TeckniX joined	15:22
enyovelcora	shruggar: :) Ok... I guess that is what I'm looking for then... even it seems awfully complicated for what I need.	15:22
	← binjured left	15:22
shruggar	though I'm pretty sure that they act in a stupid and utterly broken manner when switching branches	15:23
	← marc_m left	15:24
enyovelcora	shruggar: arg.... that's no good then... I'm branching a lot!	15:24
	→ jaql joined	15:25
	→ marc_m joined	15:26
	← alienBOB left	15:26
	← maxthoursie left	15:26
	→ maxthoursie joined	15:27
	← kibab left	15:27
	→ rgr joined	15:28
shruggar	enyovelcora, they may have fixed it by now, but I think the way it works is: when you switch branches, submodules are not effected. In order to make the submodule switch to whatever it was pointed at for that branch, you need to manually say something like "git submodule update"	15:28
	← alexpercsi left	15:28
enyovelcora	shruggar: ... ok thanks! Sounds kinda horrible...	15:29
jast	you can write a hook to automatically run that. it takes about half a minute. ;)	15:29
shruggar	I don't know much about submodules because I gave up on them quickly	15:29
	→ angiez_way69 joined	15:29
	← marc_m left	15:31
	→ marc_m joined	15:32
	← spb left	15:35
	← nizuri left	15:35
	→ sonicrules1234 joined	15:36
	→ loincloth joined	15:36
	→ lxsameer joined	15:36
	→ spb joined	15:37
lxsameer	hi how can i make my website a repository i mean when i want to change some code of my site just push my local to it	15:37
	← skoop left	15:38
steffkes	lxsameer, what? :D	15:38
	→ kolokolo joined	15:38
	← themill left	15:38
	→ themill joined	15:39
	→ brodyberg joined	15:39
	← psoo left	15:39
lxsameer	steffkes: hmmmmmmm , hard to say . for example i work on example.com website , i want to add some code to it , so i add those code to my local clone and push them to server and example.com update with my code	15:39
kolokolo	quick question: I'm on a local branch (not yet present in the origin repo) that I have locally tagged. I want to push the tag (but ONLY the tag, not the branch). So I do "git push origin MYTAG". However, it won't work.	15:39
	Any idea ?	15:40
	← sonicrules1234 left	15:40
kolokolo	Can I only "push" branches ?	15:40
steffkes	lxsameer, try post-receive hook?	15:40
lxsameer	steffkes: my problem is setting up such repo	15:41
jast	kolokolo: try: git push origin tag MYTAG	15:42
steffkes	lxsameer, then .. what's your exact problem? create the repo on the server? clone it? push to the server? or what?	15:42
	→ nizuri joined	15:42
	→ cilly joined	15:43
	← pmarsh left	15:43
lxsameer	steffkes: if i put my repo in server then how can i access it , i mean with what url?	15:43
	→ nullvariable joined	15:43
	→ shd joined	15:43
	→ codeswing joined	15:43
kolokolo	jast: it says "everything is up to date".... but nothing was actually pushed	15:44
	← marc_m left	15:44
	→ marc_m joined	15:45
	→ jammi joined	15:47
kolokolo	ok... I found out what was going on... tags are actually pushed, however you ned to "git fetch -t" (-t !!!!!) to fetch those tags that are not reachable from remote branches	15:47
	← mattcodes left	15:48
kolokolo	Hmm... maybe "-t" shuld not be required for this... but anyway... :)	15:48
jammi	Hi, is it possible to clone a svn checkout (with uncommitted changes) using git svn somehow?	15:48
	The intention is to have a sub-repository using git and commit to svn now and then when things are in order.	15:49
steffkes	lxsameer, if you need to push to it .. ssh is your friend	15:50
	lxsameer, otherwise http or git:// or working for read-ony operations	15:51
parasti	jammi: that's what git-svn is for, except you would clone from the SVN repo and move your changes over manually	15:51
lxsameer	steffkes: thanks man	15:51
	→ aresnick joined	15:51
	→ brendon_work joined	15:51
jammi	parasti: Thanks, I was just figuring out if there would be an automated process for it	15:52
	→ Novelocrat joined	15:53
parasti	jammi: I mean your current uncommited changes, not every change you ever make	15:53
	→ spearce joined	15:53
	→ rduplain joined	15:53
jammi	parasti: Yes, I got that :)	15:54
	← pheaver left	15:56
	→ sumangurung joined	15:56
	→ webchick joined	15:57
	→ shbwvl joined	15:58
esc	hey, whats the syntax for using 'recursive' with 'theirs'?	15:59
	as a merge strategy	15:59
	← kolokolo left	16:00
	← kpreid_ left	16:00
	← mbroeker left	16:01
	→ smotchkkiss joined	16:01
	→ AAABeef joined	16:02
	→ binjured joined	16:03
	→ kpreid_ joined	16:04
	→ krawek joined	16:05
	← ixti left	16:07
	← maxthoursie left	16:07
	→ maxthoursie joined	16:07
	← bauruine left	16:08
	← gnufied left	16:08
	→ Matt_R joined	16:08
	← codeswing left	16:09
	← angiez_way69 left	16:09
	→ codeswing joined	16:09
	← codeshepherd left	16:10
	→ malumalu joined	16:10
	← f00li5h[HireMe] left	16:10
	← krawek left	16:10
	→ krawek joined	16:11
charon	-Xtheirs ?	16:11
	esc: ^	16:11
	← psankar left	16:12
	→ catphish joined	16:12
esc	charon: error: unknown switch `X'	16:15
	:-(	16:15
charon	update your git?	16:15
	esc: -X is only in 1.7	16:16
esc	awwww :-(	16:18
	← ph^ left	16:18
	→ ph^ joined	16:19
	→ kumbayo joined	16:19
	← sde_antrix1 left	16:19
	← ph^_ left	16:20
	→ munichlinux joined	16:23
	← jaql left	16:24
	← ab3 left	16:25
	→ f00li5h[HireMe] joined	16:25
	← deuterium left	16:28
	→ deuterium joined	16:28
lxsameer	steffkes: i build every thing but when i try git clone ssh://user@host.com/www/project/	16:33
	← deuterium left	16:33
	← Matt_R left	16:34
lxsameer	i get an error that say can not chdir to the address	16:34
	steffkes: what should i do	16:34
steffkes	lxsameer, not sure .. but insert an : bewteen host.com & the path	16:34
	→ alley_cat joined	16:34
steffkes	and use the absolute path on your box after the hostname	16:34
jast	no, that would be mixed syntax	16:34
	if you use the ssh:// URL style, use an absolute path	16:35
steffkes	mixed syntax jast?	16:35
	ah okay	16:35
jast	otherwise, use user@host:relative/path/to/home	16:35
	without URI scheme prefix	16:35
	← aresnick left	16:35
lxsameer	jast: so should i enter this user@host:www/project/?	16:36
	jast: should i append .git to path ?	16:36
	→ k23z__ joined	16:37
	→ SRabbelier joined	16:38
	→ lhz joined	16:39
lxsameer	jast: its works thanks man	16:39
	→ aresnick joined	16:39
	→ jrmuizel joined	16:39
	offby1` → offby1	16:40
	← offby1 left	16:40
	→ offby1 joined	16:40
	← sh1mmer left	16:40
steffkes	lxsameer, append .git or not is mainly depending on you've created an bare-repo or not .. i guess	16:42
	→ hugoxrosa joined	16:42
	← minaguib left	16:43
	→ inffcs00 joined	16:43
	← inffcs00 left	16:44
	→ ph^_ joined	16:44
	→ minaguib joined	16:46
	← mike8901 left	16:46
	→ GarethAdams\|Work joined	16:47
	→ Matt_R joined	16:49
	← catphish left	16:49
	← nanoo left	16:49
	→ xperience joined	16:49
	← marc_m left	16:51
	→ lresende joined	16:52
	← nizuri left	16:53
	→ nizuri joined	16:53
	← mklappstuhl left	16:53
	← f0rkit left	16:54
	← maxthoursie left	16:55
	→ maxthoursie joined	16:55
	← kpreid_ left	16:57
	→ abergal joined	16:57
	→ pheaver joined	16:59
	← kang_ left	17:00
	→ mike8901 joined	17:00
	← nizuri left	17:00
	← mike8901 left	17:02
	← JonathanT left	17:02
	→ Haarg joined	17:02
	→ muri_one joined	17:04
	← Yuuhi left	17:06
	→ dj_tjerk joined	17:08
	← sagsousuke left	17:09
	→ Yuuhi joined	17:09
	← fmardini left	17:10
	← hugoxrosa left	17:11
	→ psoo joined	17:12
	← maxthoursie left	17:12
	→ kipras joined	17:13
	→ tstclair_ joined	17:13
	→ maxthoursie joined	17:13
	← tstclair left	17:13
	→ priidu_ joined	17:13
	← pheaver left	17:14
	→ yann joined	17:15
	→ mike8901 joined	17:16
	← steffkes left	17:16
	← lhz left	17:18
	← themill left	17:18
	← SpookyET left	17:19
	← tobiassjosten left	17:19
	← Matt_R left	17:19
	← froschi left	17:21
	← bblfish left	17:21
	← hipe left	17:21
	→ defectiv joined	17:22
	← mcdebugger left	17:22
	→ NetHawk joined	17:24
	← chouquette left	17:25
	→ chouquette joined	17:25
shruggar	is there a quick way to see "what branch was I last on?", I know I can read through the reflog, looking for an entry that says "moving from foo/bar to foo/baz", but is there anything cleaner?	17:25
	→ hugoxrosa joined	17:25
	← novas0x2a left	17:27
tzafrir	shruggar, 'git branch' ?	17:27
	← dancor left	17:27
tzafrir	sorry, I missed the whole question	17:27
	→ troyt joined	17:27
	→ christophsturm joined	17:28
	→ bradhaydon joined	17:29
GarethAdams\|Work	what would happen if I put .gitignore in my .gitignore file?	17:29
	→ dancor joined	17:29
locklace	shruggar: i think the reflog exists pretty much to answer exactly that sort of question	17:29
GarethAdams\|Work	something would explode, right?	17:29
	like searching for Google in Google	17:29
	→ RdeWilde joined	17:31
shruggar	yeah, the reflog is a bit too human-centric for me :/	17:31
	"do-able, but uglier than I'd like"	17:32
	← parasti left	17:32
pasky	human-centric?	17:32
locklace	GarethAdams\|Work: i assume it would prevent the .gitignore file from being tracked	17:32
	← tetha left	17:32
	→ bdiego joined	17:33
	→ kpreid_ joined	17:33
shruggar	pasky, as in: it has text descriptions of what happened, designed to be read by a human. It tends not to have everything laid out in a nice clean layout which is simple for a script to parse	17:33
	→ gnufied joined	17:33
pasky	/checkout: moving from (.) to (.)/ ? :)	17:33
	but yes it does have ugly warts	17:34
	9b7e3fb8fd8af3d48a8ff847d81429b41e3f6973 0564a7fc7dddb48006c3f244ca34c9a7a205e2b2 Petr Baudis [email@hidden.address] 1265847046 +0100 HEAD^: updating HEAD	17:34
	especially this	17:34
	→ tetha joined	17:34
pasky	and the space after pull is not nice either :/	17:34
	← disappearedng left	17:35
locklace	shruggar: maybe what you actually want to do is read the files reflog reads	17:35
	→ mcdebugger joined	17:36
shruggar	GarethAdams\|Work: if you add .gitignore to .gitignore, exactly the same thing happens as would in any other situation you add something to .gitignore	17:37
	GarethAdams\|Work, ie: .gitignore (if not already tracked) will not show up in "git status" output, and (if not already tracked) recursive adds will not add it	17:38
pasky	locklace: but they are in the same format :)	17:38
locklace	oh well.	17:39
	→ nik0 joined	17:39
	← wolog left	17:39
	← shruggar left	17:40
	niko → Guest79188	17:43
	nik0 → niko	17:43
	→ stringo0 joined	17:43
	← Guest79188 left	17:44
	→ Vortex35 joined	17:45
	→ tedoc2000 joined	17:45
	→ MetaCosm_ joined	17:46
	← MetaCosm_ left	17:46
	→ mw joined	17:46
	→ MetaCosm_ joined	17:47
	→ mklappstuhl joined	17:47
	← Vortex34 left	17:47
	→ disappearedng joined	17:49
	← uau left	17:50
	← kpreid_ left	17:50
	← remi left	17:50
	→ uau joined	17:50
	→ Mike_lifeguard joined	17:51
	→ dreiss joined	17:51
	→ ixti joined	17:52
	→ remi joined	17:52
	→ nanoo joined	17:53
	→ PerlJam joined	17:54
	→ kpreid_ joined	17:54
	→ nizuri joined	17:54
	← munichlinux left	17:55
	→ munichlinux joined	17:55
	← bradhaydon left	17:57
	← przemoc left	17:57
	← d0ugal left	17:57
	→ przemoc joined	17:58
	← kpreid_ left	18:00
	→ greg8si joined	18:00
	→ bradhaydon joined	18:00
	← brodyberg left	18:01
	← greg8si left	18:01
	→ mbroeker joined	18:02
	→ ksn joined	18:03
	← disappearedng left	18:03
	→ sh1mmer joined	18:05
	← tildeequals left	18:08
	← f00li5h[HireMe] left	18:09
	→ Chillance joined	18:09
csgeek	is there an easy way of saying.. commit all these files that changed except these 3. I guess I could temporarily add to .gitignore.. but wonder if there's a cleaner way	18:10
jhelwig	csgeek: git add . && git reset the three files ?	18:10
csgeek	hmm.. that would work.. I suppose though the files are in different directories.. but general idea	18:11
	→ hipe joined	18:12
	→ brizly joined	18:12
	→ doug8429 joined	18:13
	← shbwvl left	18:13
	→ jaysern joined	18:14
	→ bauruine joined	18:15
	← smuf left	18:16
abergal	anyone have any idea why a 'git svn clone' might silently fail after the message "This may take a while on large repositories"?	18:16
	→ greg8si joined	18:16
	→ bblfish joined	18:16
abergal	doing a trace, the last msg is "trace: built-in: git 'gc' '--auto'"	18:17
	← psoo left	18:17
jhelwig	You sure it's silently failing, and not just done?	18:17
	→ pantsman joined	18:17
abergal	well the .git folder is created, so i guess it does the init, but not the fetch, since there's no other files in the directory	18:18
	→ f00li5h[HireMe] joined	18:19
	← lxsameer left	18:20
jhelwig	abergal: Is this a publicly available repo? What arguments are you giving to git clone? Can you pastie the full output you're getting, and what "echo $?" gives immediately after it silently fails?	18:21
	← dpino left	18:21
abergal	no, not publicly available...	18:21
	→ psoo joined	18:21
	→ disappearedng joined	18:21
thiago	$ git br --set-upstream staging/4.7	18:21
	Branch staging/4.7 set up to track local branch tmp/qtimestamp.	18:21
	hmm... I wanted the exact opposite. Aren't the semantics of git branch --set-upstream somehow inverted?	18:22
abergal	args: git svn clone --prefix svn/ -s svn://svnrepo/subdir/	18:22
	→ Theravadan joined	18:22
abergal	would you prefer the trace output for the pastie?	18:23
jhelwig	thiago: That does seem completely backwards to me. Post something to the mailing list about it?	18:24
	← sven^ left	18:25
	→ sven^ joined	18:25
Ilari	thiago: You hit the 'create branch' case?	18:25
thiago	Ilari: yes	18:25
jhelwig	abergal: I'm not sure what would help, honestly. If it was publicly available, I was just going to try cloning it myself using the same args as you.	18:25
abergal	ah ok, well "echo $?" returns 0 if that helps...	18:27
jhelwig	Well, that means that it thinks it completed successfully.	18:28
abergal	right	18:28
	← nicoulaj left	18:28
	→ zodiak joined	18:28
	← bluenovember left	18:28
	← sven^ left	18:29
abergal	fyi, i cloned the repo successfully using my Mac OS 10.5 machine, but having trouble on this Snow Leopard machine	18:29
jhelwig	abergal: Same version of git on both?	18:29
	→ sven^ joined	18:29
abergal	yep, 1.6.6	18:29
	→ blaines joined	18:30
	→ viniciusluiz1 joined	18:31
	← viniciusluiz left	18:31
jhelwig	abergal: Are there any differences in the normal output, or the trace output between the two boxes up until it fails on the one?	18:32
	→ _iron joined	18:32
	→ maxter joined	18:33
abergal	dont have the 10.5 machine on hand, but IIRC, they're the same	18:34
	← loincloth left	18:35
	→ loincloth joined	18:36
abergal	(the normal output that is, didnt do a trace on the 10.5 machine)	18:37
	→ ericpugh joined	18:37
	← tg left	18:37
	→ nadavoid joined	18:38
GarethAdams\|Work	I wrote my git remote add wrong: `git remote add [email@hidden.address] gareth`	18:38
	now when I try and remove it, I get `fatal: Invalid refspec 'refs/heads/:refs/remotes/git@github.com:gareth/jti.hand-i.net.git/'`	18:38
	→ tg joined	18:38
jhelwig	GarethAdams\|Work: You can just edit your .git/config file, and remove the remote that way.	18:39
	→ flazz joined	18:39
GarethAdams\|Work	jhelwig: is that the only place it's stored?	18:39
	→ JEEB__ joined	18:39
jhelwig	Yeah.	18:39
GarethAdams\|Work	cool	18:39
	← doug8429 left	18:40
flazz	say i want to split a project in two. can i get the history of only certain files from the old project and put them in the new one before i delete the files?	18:40
	→ tildeequals joined	18:40
jhelwig	flazz: You can by using filter-branch. man git-filter-branch	18:40
Gitbot	flazz: the git-filter-branch manpage can be found at http://git.or.cz/man/git-filter-branch	18:40
flazz	jhelwig: thanks	18:40
	← JEEBcz left	18:41
jhelwig	Basically, make a new branch, or copy the repo, and use filter-branch to delete everything _except_ what you want to split into a new repo.	18:41
	→ toxx_ joined	18:41
	→ d0ugal joined	18:42
	← project2501b left	18:43
	← GarethAdams\|Work left	18:44
	← oliver3 left	18:44
	→ godsflaw joined	18:45
	← blaines left	18:45
	← ribasushi left	18:47
	← sven^ left	18:55
	→ kpreid_ joined	18:55
	→ sven^ joined	18:55
	← psoo left	18:56
	← nizuri left	18:57
	→ jrosdahl_ joined	18:58
	← christophsturm left	18:59
	← jrosdahl_ left	19:00
	→ jrosdahl joined	19:00
	← tatsuya left	19:01
flazz	jhelwig: is a straight cp good enough to copy the repo?	19:01
	← f00li5h[HireMe] left	19:02
jhelwig	flazz: cp -r, but yeah, there's nothing special you need to do.	19:02
flazz	nice	19:02
	← sh1mmer left	19:02
	→ GodEater_ joined	19:04
	← GodEater_ left	19:04
	→ GodEater_ joined	19:04
	→ f00li5h[HireMe] joined	19:05
	→ JEEB joined	19:05
	GodEater_ → GodEater	19:05
	→ froschi joined	19:06
	→ rapha joined	19:06
rapha	Hi all!	19:06
	← JEEB__ left	19:07
rapha	A developer has asked me to check out a specific branch of his software, the URL is http://github.com/newrelic/rpm/tree/2.10.6_dev - how do I do that?	19:07
	→ justin-george joined	19:07
	→ dr_jesus joined	19:07
k23z__	is there a github client so that I can check my messages and send messages to other users	19:08
	← smotchkkiss left	19:08
	→ dmsuperman joined	19:08
dmsuperman	I'm having a strange behavior that makes me think I don't understand gitignore	19:08
	http://paste.cjohnson.me/252	19:08
	That says block all and then allow everything that begins with .vim right?	19:09
	Except .vim/swap and .viminfo	19:09
	← priidu left	19:10
	← priidu_ left	19:11
	← sergio__ left	19:11
doener	dmsuperman: the * isn't anchored, so it applies in every directory	19:12
	← magn3ts left	19:12
	→ sergio_ joined	19:12
doener	dmsuperman: you might want to use /* instead	19:12
dmsuperman	Ah	19:14
	Thank you :)	19:14
	→ berndj joined	19:15
	→ sh1m joined	19:16
	→ timcharper joined	19:17
	← sven^ left	19:18
	→ sven^ joined	19:18
	→ dutchie joined	19:19
	← dutchie left	19:19
	→ dutchie joined	19:19
	→ mjf joined	19:20
	→ tesseracter joined	19:20
	→ Tureba joined	19:21
	→ bambule joined	19:21
tesseracter	i want to revert the last... 17 commits. git revert HEAD-17?	19:21
dmsuperman	^17	19:21
zomg	Revert only reverts a single commit	19:22
tesseracter	zomg: so... how do i do git "HULK SMASH!"	19:22
zomg	I think you could git checkout head~17 and then just commit	19:22
	Not sure really	19:22
	=)	19:22
tesseracter	thats something i can try...	19:23
zomg	git help revert may provide more info	19:23
	← ikrabbe left	19:23
locklace	tesseracter: do you actually want to revert, or reset	19:24
	← dmsuperman left	19:25
	→ ehabkost joined	19:25
	→ sagsousuke joined	19:26
	→ DanielHolth joined	19:27
DanielHolth	Does "git tag" with no arguments do anything? I notice it fails on older versions of git, but runs without output on newer git.	19:27
	→ matthieu joined	19:27
doener	DanielHolth: it shows a list of existing tags	19:28
	→ robinr joined	19:28
	← mastro left	19:29
DanielHolth	thanks doener. on old git, I just have to say "git tag -l" I guess.	19:29
	← codeswing left	19:31
dr_jesus	hey, I'm trying to merge two repositories together and I'm not sure how to do it with git merge	19:31
	← rgr left	19:33
	→ brodyberg joined	19:33
dr_jesus	anyone?	19:34
	→ Textmode joined	19:34
	→ rgr joined	19:34
	→ kim-day joined	19:35
Weiss	dr_jesus: you merge branches not repositories (probably you understood correctly though)	19:35
dr_jesus	actually the two branches are in different repositories	19:35
Weiss	dr_jesus: best way is to have local copies of both branches, check out the one you're merging into, and do "git merge <otherbranch>" where <otherbranch> is the branch you want to merge in	19:35
	... or is it a more complicated scenario giving problems?	19:36
locklace	dr_jesus: something wrong with adding one as a remote of the other, fetching, and merging branches as usual?	19:36
	← ankit9 left	19:36
broonie	or git pull which isn't ideal but is a combined fetch+merge	19:36
dr_jesus	locklace: ah, I think that's what I want	19:36
	just set remote.otherbranch.url in the target branch?	19:37
davr	yeah you basically just add a second remote. your local git repository can have multiple remotes	19:38
	then you'd do something like 'git checkout remote1/branch -b new_local_branch_name' then 'git merge remote2/branch'	19:39
dr_jesus	hmm it looks like I can use git-remote to do that	19:39
	→ JonathanT joined	19:39
dr_jesus	hmm git branch -r doesn't show the branches in the other repository after I add it to the first one	19:41
davr	you have to do 'git fetch <remote>' first	19:41
dr_jesus	ah	19:41
	→ psoo joined	19:41
	← brizly left	19:43
	← aresnick left	19:43
	→ thiago_home joined	19:44
	← DanielHolth left	19:44
tesseracter	locklace: umm, whatever brings me back to before i tried the crazy thing that took 17 commits. is revert or reset appropriate? at the moment im in a new branch, because the changes were made on master.	19:44
	→ bdimcheff joined	19:45
	→ leitao joined	19:45
	→ aresnick joined	19:45
	dr_jesus winces as the merge creates a bunch of conflicts	19:46
locklace	tesseracter: revert creates a new commit that reverses the effect of an earlier one. reset (--hard) blows away the current state and replaces it with an earlier one	19:46
	→ agliodbs joined	19:47
	← kpreid_ left	19:47
	→ kpreid_ joined	19:47
agliodbs	hey, is there any way I can make git ignore emacs backup files?	19:47
	that is, files ending in ~	19:47
	→ shen joined	19:47
DrNick	stick *~ in .gitignore	19:47
locklace	agliodbs: man gitignore	19:47
Gitbot	agliodbs: the gitignore manpage can be found at http://git.or.cz/man/gitignore	19:47
agliodbs	thanks	19:47
Weiss	I thought they were ignored by default actually...?	19:47
agliodbs	Weiss: apparently not	19:47
DrNick	nothing is ignored by default	19:48
	← mike8901 left	19:49
leitao	Hi, after a lot of modification (without commit), I'd like to "reset" one specifically file to HEAD. How should I do that ?	19:49
agliodbs	hmmm, I think setting this in my git configm makes more sense	19:49
	man gitconfig	19:50
	← Mike_lifeguard left	19:50
agliodbs	hmmm	19:50
	Gitbot doesn't listen to me?	19:50
jast	agliodbs: man git-config	19:50
Gitbot	agliodbs: the git-config manpage can be found at http://git.or.cz/man/git-config	19:50
	→ blaines joined	19:50
jast	small but significant difference :)	19:50
	← blaines left	19:51
agliodbs	yeah	19:51
	anyone know where the gitconfig file lives on a mac?	19:52
	→ blaines joined	19:52
locklace	leitao: reset, not surprisingly. man git-reset	19:52
Gitbot	leitao: the git-reset manpage can be found at http://git.or.cz/man/git-reset	19:52
leitao	locklace: yea. I looked at it, but I want to keep the other files there. It says that I can't hard-reset just a file	19:53
locklace	you don't want to reset hard, just reset	19:54
agliodbs	core.excludesfile is a reference to a file of excludes, not a value, right?	19:54
locklace	checkout works too	19:54
dr_jesus	locklace: thanks!	19:55
	← munichlinux left	19:55
thiago_home	agliodbs: yes	19:56
agliodbs	why does this not work, then?	19:57
	git config --add core.excludesfile=~/.gitignore	19:57
thiago_home	because of the ~	19:57
agliodbs	but the docs specifically say to use ~/ for my home dir	19:57
	and apparently that's not the problem	19:58
	← flazz left	19:58
agliodbs	what I mean by "doesn't work" is "git config rejects the command"	19:58
	← shen left	19:58
thiago_home	you didn't paste an error, so I assumed the command succeeded	19:58
agliodbs	thiago_home: well, it just spits back the git command syntax	19:59
	that is, the git config command syntax	19:59
	no specific error	19:59
locklace	because the syntax is "variable value" not "variable=value"	19:59
agliodbs	ah, ok, thanks	19:59
	→ shen joined	20:00
agliodbs	since the format of git config -l was variable=value, I assumed the command matcvhed	20:00
locklace	what version of git are you using anyway	20:00
	← malumalu left	20:00
	← Adjective left	20:01
agliodbs	1.6.3	20:01
	→ Twisted_Mentat joined	20:01
agliodbs	yaaaay, that worked	20:01
	thanks for help	20:01
	→ ProperNoun joined	20:03
godsflaw	I need git log output on a particular file that I only know by sha1 hash. What is the quickest way to do this?	20:04
	in short, I need the log details of its last commit log.	20:04
	← zodiak left	20:05
davr	one way is "git rev-list --objects --all \| grep <HASH>", but it's a bit slow. might be a faster way	20:06
	← _Vi left	20:06
	→ _Vi joined	20:07
	← ProperNoun left	20:09
	→ Adjective joined	20:09
	→ magn3ts joined	20:18
	← abergal left	20:19
	← Boriskr left	20:20
	→ nizuri joined	20:21
	← d0ugal left	20:21
	→ MacGyverNL_ joined	20:22
	← MacGyverNL_ left	20:22
	→ pklingem1 joined	20:23
	← ttvd left	20:27
	→ Plouj joined	20:27
Plouj	hi	20:27
	is changing bare=true in config (in a bare repository) the right way to avoid messages about "warning: updating the current branch" when pushing to a bare repository?	20:27
	← Adjective left	20:28
	→ ProperNoun joined	20:28
jast	Plouj: yes	20:29
Plouj	jast: do you know if bare could have been set to false on actually bare repositories in and old version of git?	20:30
wereHamster	jast: no, the right way is not to push into a bare repo	20:30
	→ PointMan joined	20:30
wereHamster	s/bare/non-bare/	20:31
	→ juan_arandaalvar joined	20:31
	→ agib joined	20:31
Plouj	wereHamster: Hi. If you're talking about my scenario then I'm actually pushing to a bare repo where core.bare=false, for a _strange_ reason.	20:32
wereHamster	yeah, that was to you, sorry. Try to figure out why bare was set to false..	20:32
	→ salty-horse joined	20:32
Plouj	I'm not sure how. This repo was created long before I joined.	20:32
	I suspect it's due to a bug in an older git.	20:33
salty-horse	"time echo $(__git_ps1)" takes half a second on my machine. anyone else noticing how slow it is?	20:33
	← agib left	20:33
	→ johnw joined	20:33
	→ wolog joined	20:33
	→ neunon joined	20:34
Ilari	salty-horse: Cold caches? It takes 'real 0m0.639s' for fist time and subsequent ones are in class of: 'real 0m0.066s'.	20:34
salty-horse	Ilari, consecutive ones keep giving real ~0m0.520s	20:35
	I'm using the supplied ubuntu karmic bash completion script	20:35
	should I try the latest git?	20:35
	← imajes left	20:36
	← hugoxrosa left	20:36
	→ agib joined	20:36
	← nizuri left	20:36
salty-horse	others with ubuntu aren't seeing such a slowdown. maybe it's due to github's "hub" :/	20:36
	silly github and their silly ruby	20:37
wereHamster	heh? what does __git_ps1 have to do with github/ruby?	20:37
	← ehabkost left	20:37
salty-horse	yes, that's it. I was using http://github.com/defunkt/hub as a git alias	20:38
	→ ehabkost joined	20:38
	→ imajes joined	20:38
	← nadavoid left	20:39
jast	Plouj: I think it's more likely that a mistake was made when setting up the repository	20:39
	← Tureba left	20:40
	→ hakunin joined	20:40
	← viniciusluiz1 left	20:42
	→ gebi joined	20:45
	→ anjen joined	20:47
	← ericpugh left	20:48
	→ ttvd joined	20:49
	→ tillberg joined	20:49
	← tillberg left	20:50
	→ ericpugh joined	20:51
	→ saiiangs joined	20:52
	→ WebDragon joined	20:52
	← icwiener left	20:53
	← nanoo left	20:55
	→ grawity joined	20:56
	← grawity left	20:56
	→ hipe_ joined	20:56
	← saiiangs left	20:57
	→ nadavoid joined	21:00
	← kpreid_ left	21:00
	← hipe left	21:00
	→ kpreid_ joined	21:00
	→ DanielHolth joined	21:08
	→ mklappst1hl joined	21:10
	→ saste joined	21:13
	← pklingem1 left	21:13
	← mklappstuhl left	21:13
	→ tjc joined	21:14
	→ zodiak joined	21:15
	→ pklingem1 joined	21:15
	← kipras left	21:15
	→ kipras joined	21:18
	→ alcoheca joined	21:18
	← sven^ left	21:19
	→ tlb joined	21:20
	← minaguib left	21:24
	→ mike8901 joined	21:24
	→ minaguib joined	21:25
	← maxter left	21:26
alcoheca	evening. I'm getting loads of conflicts while merging a branch (fair enough I'm used to this from svn). But the strange thing is many of the supposed 'conflicts' are really simple changes. like <height>123</height> changing to <height>234</height>, or one line insertions. have tried other merge strategies but to no avail	21:27
	← corni left	21:27
	← kpreid_ left	21:28
grumpytoad	is git supposed to be stable over an nfs mount ?	21:30
	← stringo0 left	21:31
fr0sty	alcoheca: if the line was changed in both branches you are trying to merge it is not possible for git to resolve the conflict automatically.	21:31
	grumpytoad: It has worked fine for me...	21:31
	← greg8si left	21:32
grumpytoad	fr0sty: even using git stash, and git stash apply ?	21:32
fr0sty	yeah,	21:33
	What trouble are you having?	21:33
grumpytoad	hmm.. it must be my nfs mount options	21:33
	← salty-horse left	21:33
grumpytoad	on the nfs client after a stash apply, it keeps complaining about a dirty working tree	21:33
	though if I log onto the server it works normally	21:34
	← bdiego left	21:34
	← d0k left	21:38
fr0sty	usually your tree would be dirty after 'git stash apply'	21:38
	← JEEB left	21:39
fr0sty	btw, what command thinks your tree is dirty?	21:39
	← simNIX left	21:39
	← RdeWilde left	21:39
	← sergio_ left	21:40
	→ jaql joined	21:41
	← shen left	21:42
	← jaql left	21:42
	→ JEEB joined	21:42
	→ sergio_ joined	21:43
	→ shennyg joined	21:44
	→ project2501a joined	21:44
	→ jaql joined	21:45
	← jaql left	21:45
	→ robacarp joined	21:45
	→ maxter joined	21:46
	← JackTheRabbit left	21:46
robacarp	is there a special way to get to handle binary files? Its trying to make me merge versions of a binary file	21:46
	→ webus joined	21:47
	→ jaql joined	21:47
	→ tjc_ joined	21:49
	← csgeek left	21:49
	← DanielHolth left	21:49
	← tjc left	21:49
	→ emil_87 joined	21:49
	← maxter left	21:49
emil_87	what a sily question but what is git?!	21:50
	→ csgeek joined	21:50
	← tjc_ left	21:50
thiago_home	a distributed version control system	21:50
emil_87	thanks :)	21:50
	→ DaveIngram joined	21:50
robacarp	emil_87: source code management suite	21:50
	← froschi left	21:50
	← dj_tjerk left	21:50
	← jaql left	21:50
	← magn3ts left	21:51
	← DaveIngram left	21:52
rapha	A developer has asked me to check out a specific branch of his software, the URL is http://github.com/newrelic/rpm/tree/2.10.6_dev - how do I do that?	21:54
robacarp	rapha: git clone <url>	21:54
	← sis left	21:55
	← ehabkost left	21:55
	← mike8901 left	21:55
	→ sis joined	21:55
	→ DaveIngram joined	21:55
rapha	robacarp: when i do that it gives me lots of error messages and html	21:56
	← DaveIngram left	21:56
robacarp	I would surmise that the url is wrong then	21:56
	← rduplain left	21:57
rapha	robacarp: at the url there is a link to another url which works, but doesnt get me the branch im supposed to get ... i think you need to modify the command to get that branch	21:57
	← JonathanT left	21:57
	← emil_87 left	21:58
	← PointMan left	21:59
robacarp	rapha: far as I know you have to clone the repo first then switch to the branch	22:00
	→ PointMan joined	22:00
	← psoo left	22:00
robacarp	or rather, then pull the branch	22:00
rapha	but how do you do that?	22:01
robacarp	after you clone the repo, git pull origin <branch>	22:01
	← alcoheca left	22:02
rapha	oh that did something	22:02
	* branch 2.10.6_dev -> FETCH_HEAD ... i think that was it - thanks robacarp :-	22:02
	:-)	22:02
robacarp	yep	22:02
	← WebDragon left	22:02
	← uau left	22:02
	← spearce left	22:03
	→ uau joined	22:05
	← pklingem1 left	22:07
	→ kpreid_ joined	22:08
cj	error: unable to create temporary sha1 filename ./objects/68: File exists	22:08
	huh?	22:08
	this from C<git push origin master>	22:08
	← kumbayo left	22:08
	→ nevyn joined	22:11
	→ lxsameer joined	22:11
	← brodyberg left	22:12
	← uau left	22:12
	← Mezner left	22:15
	← ntoll left	22:17
	← ProperNoun left	22:17
	→ mike8901 joined	22:17
lxsameer	hi i try to pull but i get some conflict i delete that file but when i try to commit i get this error	22:18
	http://pastebin.com/kYeQk1Sc	22:18
cj	it was a permissions issue	22:18
	← jmcantrell left	22:19
lxsameer	hmm what should i do	22:19
	← PointMan left	22:21
thiago_home	lxsameer: have you fixed the merge?	22:21
	the merge conflict, that is	22:21
	→ ProperNoun joined	22:22
lxsameer	thiago_home: i delete that file , there is no need of that eny more	22:22
	← _iron left	22:22
thiago_home	how did you tell Git you deleted the file?	22:22
	→ kipras_ joined	22:22
lxsameer	thiago_home: how should i tell that ?	22:23
	← k23z__ left	22:23
	→ uau joined	22:23
	← kipras left	22:24
	← TeckniX left	22:24
thiago_home	lxsameer: how do you tell Git that a file should no longer be tracked?	22:24
	← webchick left	22:24
	→ TeckniX joined	22:24
	→ pklingem1 joined	22:24
lxsameer	thiago_home: i dont know , can you tell me	22:24
	→ mrlarner joined	22:25
_rane	thiago_home: git rm --cached <file>	22:25
	← Ryback_ left	22:25
thiago_home	lxsameer: git rm didn't occur to you?	22:26
lxsameer	let me try	22:26
	← kpreid_ left	22:27
lxsameer	thiago_home: i get this	22:28
	database: needs merge	22:28
	rm 'database'	22:28
	← nevyn left	22:28
thiago_home	lxsameer: what does git status now say?	22:28
	→ d0ugal joined	22:28
	→ nevyn joined	22:28
	→ kpreid_ joined	22:29
	← webus left	22:29
lxsameer	thiago_home: http://pastebin.com/HqcnsfMy	22:30
	← ben_alman left	22:31
thiago_home	lxsameer: your conflict is fixed. Commit.	22:32
	← tlb left	22:32
	→ Paraselene___ joined	22:32
	→ stringo0 joined	22:33
	← loincloth left	22:34
	← Paraselene__ left	22:34
	→ difekta joined	22:35
	← difekta left	22:38
	← LiamH left	22:39
	← lxsameer left	22:40
	← johnw left	22:40
	→ loincloth joined	22:40
	→ hugoxrosa joined	22:46
	→ Azeroth-Work joined	22:46
	← Azeroth-Working left	22:46
	← enyovelcora left	22:47
	← sis left	22:48
cj	is it possible to have a repo be updated when a commit is pushed to it? I've got a web page that I want to see update when I do a git push	22:48
	last time I pushed, I had to ssh to the host and do a 'git reset' in order to see the changes	22:49
tango_	cj: you probably want a post-receive hook that does the job	22:49
	ikiwiki for example does this	22:49
Plouj	jast: yeah, very possible, thanks	22:49
shennyg	so I have a problem with my workflow...	22:50
	← hced left	22:51
	→ Rolenun joined	22:51
shennyg	I have a svn repo I am connecting to, I am using git-svn to manage it. I make changes, then I do a git-svn rebase to get changes from the repo... and there are always conficts	22:51
	← bcardarella left	22:52
shennyg	so I don't think I am doing things in the righ torder	22:53
	I use git-svn rebase to update from the svn repo	22:53
	← dcampano left	22:53
shennyg	I edit the files, I do another git-svn rebase to get any changes that may have occured...	22:53
	← alley_cat left	22:54
shennyg	then I do a git-svn dcommit	22:54
	is that correct?	22:54
	← AAA_awright left	22:54
	→ AAA_awright joined	22:54
	→ dj_tjerk joined	22:55
	← stringo0 left	22:55
	← jotik left	22:56
	→ jotik joined	22:56
	→ Goosey joined	22:56
	← Theravadan left	22:57
	← v0n left	22:57
	← hakunin left	22:58
Goosey	I am using git-svn to interface with a centralized SVN repo. during "git svn fetch" i am repeatedly failing at the same place with the message "fatal: write error: Invalid argument". This seems to be happening during git's process of building it's revisions from the SVN revisions. The source repo has 554 revisions and includes many large binaries.	22:58
	← hugoxrosa left	22:58
	← jas4711 left	22:58
Goosey	Any advice for how I can furthur diagnose this error? "fatal: write error: Invalid argument" doesn't tell me much.	22:59
	→ freedrull joined	22:59
	← mike8901 left	22:59
	← sagsousuke left	22:59
	← tbf left	22:59
	→ Ramblurr joined	23:00
	→ jas4711 joined	23:01
	← neunon left	23:01
	→ trochala joined	23:03
	← dj_tjerk left	23:03
	→ neunon joined	23:04
	← ixti left	23:04
	← yatiohi left	23:04
	← Twisted_Mentat left	23:04
	← ph^_ left	23:04
	→ hgimenez joined	23:05
	→ Twisted_Mentat joined	23:05
	→ mike8901 joined	23:08
sh1m	this is, I'm sure, quite stupid of me. However, if I'm ahead of origin/master by 2 commits and I want to drop them how do I do it?	23:11
	something like `git revert master~2` ?	23:11
bob2	revert commits a new changeset undoing some previous changeset	23:12
	maybe you want reset?	23:12
sh1m	I did a reset, but they are commits so they weren't dropped	23:12
thiago_home	if you want to remove the latest commits from history, simply move your branch backwards to a past commit	23:12
	← dajero left	23:12
thiago_home	that's the git reset commit	23:12
locklace	sh1m: reset --hard	23:12
sh1m	I merged a branch and wanted to perge them	23:12
thiago_home	are they the top two commits in git log?	23:13
	← doener left	23:13
sh1m	thiago_home: yep	23:13
thiago_home	reset to the commit prior to both of them	23:13
	→ smuf joined	23:14
	→ doener joined	23:14
sh1m	oh I think I got it	23:14
	thanks :)	23:14
	← agile left	23:17
	← bdimcheff left	23:17
	→ JayM joined	23:18
	→ tbf joined	23:18
	→ bcardarella joined	23:20
cj	hurm... doesn't look like putting 'git commit --hard' in .git/hooks/post-receive and chmod +x .git/hooks/post-receive did what I expected	23:20
	← xperience left	23:20
cj	er, git reset --hard	23:20
	am I misunderstanding the docs?	23:21
bob2	maybe you just don't want to push to a non-bare repository	23:21
	and instead have a hook that pulls to another working copy	23:21
	→ davidfetter joined	23:22
davidfetter	hello	23:22
	→ Gullik joined	23:22
	← mklappst1hl left	23:23
davidfetter	i'd like to pull this project into git. this is a one-time operation http://pgfoundry.org/scm/?group_id=1000288	23:23
	i don't have commit bits on it	23:23
cj	bob2: hurm... so push to the bare repo and have a hook there that pushes to the site that's in the document root?	23:23
bob2	no	23:24
cj	or cd to the document root and then do the git pull?	23:24
bob2	pull	23:24
shennyg	i just merged two local branches (that had conflicts I had to manualy fix) and wanted to do the exact merge on the second branch... will I have to do the same manual conflict merges?	23:24
bob2	sounds amenable to experimentation!	23:24
	→ dj_tjerk joined	23:25
davidfetter	anybody?	23:25
cj	okay ;)	23:25
bob2	davidfetter: what was the question	23:26
davidfetter	i'd like to pull this project into git. this is a one-time operation http://pgfoundry.org/scm/?group_id=1000288	23:26
	→ ereslibre joined	23:26
	← ereslibre left	23:26
	→ ereslibre joined	23:26
bob2	that's not a question ;)	23:26
davidfetter	how do i pull it in?	23:26
Nugget	http://www.kernel.org/pub/software/scm/git/docs/git-cvsimport.html	23:28
	← lresende left	23:28
	← Goosey left	23:28
	← khmarbaise left	23:28
Nugget	or just cvs co it, delete all the CVS detritus, and do a "git init"	23:28
	if you don't care about historical activity	23:28
	→ redbaritone joined	23:29
davidfetter	that sounds like a way better idea, Nugget :)	23:29
	← redbaritone left	23:30
	← remi left	23:30
	← pklingem1 left	23:32
	→ Nafai joined	23:32
	← loincloth left	23:33
Nafai	I'm trying to use git filter-branch to remove commits, but the implementation of skip_commit on the man page is confusing (what's map? Is this perl code?)	23:33
	← TeckniX left	23:33
	→ remi joined	23:33
	niki_ → niki	23:35
	niki → Guest62982	23:35
	← Guest62982 left	23:36
	→ niki_ joined	23:36
	→ lresende joined	23:36
	← lresende left	23:36
	← tesseracter left	23:37
	→ mecromace1 joined	23:39
	niki_ → niki	23:40
shennyg	I am in the middle of a rebase and it is telling me to "did you forget to use 'git add'?"	23:40
	→ Theravadan joined	23:40
nevyn	shennyg: you've got to git add after fixing conflicts	23:40
shennyg	I did add the file in conflict, and doing status tells me I have .dotest/ not added	23:40
	I think the problem is the file doesn't show up as staged	23:41
	maybe b/c I fixed the confilct to what it was originally	23:41
	← thiago_home left	23:44
	← bcardarella left	23:44
	→ tiglionabbit joined	23:45
shennyg	I just have to edit the file slightly	23:45
	and it will recognize it as changed, then I can add it	23:45
	why does it do that?	23:45
	← Paraselene___ left	23:45
tiglionabbit	how can I look at the patch that was applied when a branch was merged? I look at the commit message, and all it says is "merge branch x into y" with no changed lines in it	23:45
jhelwig	tiglionabbit: Try this: git diff <merge-sha>^1 <merge-sha>^2	23:46
tiglionabbit	interesting	23:48
jhelwig	You should also check out the -c and --cc flags in man git-log	23:48
Gitbot	jhelwig: the git-log manpage can be found at http://git.or.cz/man/git-log	23:48
	← anjen left	23:49
	← Yuuhi left	23:52
	shennyg → shennyg\|away	23:54
cj	yep, cd && git pull worked	23:54
	bob2: thanks	23:54
	← Nafai left	23:54
	→ cousine joined	23:56
	← davidfetter left	23:56
Plouj	shennyg\|away: you get conflicts when doing dcommit?	23:57
	nvm	23:57
	← cannonball left	23:58
	→ plediii_ joined	23:59

Logs Search ←Prev date Next date→

Channels Documentation