IRCloggy #git 2010-08-03

Provider of IRC logs since 2005.

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

	2010-08-03
	← sammcd left	00:00
	→ spaceonline joined	00:00
	← infinity0 left	00:00
	← Yuuhi left	00:01
	← sivy left	00:01
	→ MattDiPa_ joined	00:01
	← sivy_ left	00:03
	← Paraselene_ left	00:03
	→ sivy joined	00:04
	← mw left	00:04
	→ hotwings joined	00:04
hotwings	is it possible to download just a single dir from a git repository or are you forced to download the whole thing? i only need the latest v4l drivers but so far it seems im forced to download the whole kernel just to get the dir	00:05
	← Jferro left	00:06
	→ Magicmat joined	00:06
	← bentob0x left	00:07
	← MattDiPa_ left	00:08
rane_	it's not	00:09
hotwings	in that case it seems idiotic that v4l doesnt at least offer daily snapshots of drivers like alsa does. very disappointed to hear that but thanks for the answer rane_	00:10
	→ Paraselene_ joined	00:10
	← tedoc2000 left	00:12
	→ zimbres joined	00:15
	← jim0203 left	00:15
	→ infinity0 joined	00:16
	← Magicmat left	00:19
	← tstclair left	00:20
airborn	hotwings, or it should be avaible as a submodule IMO	00:21
	← shiba_yu36 left	00:22
hotwings	yes, something better then forcing an entire kernel download.	00:22
	need about 17MB worth of files, forced to download 872MB just to get it is absurd	00:23
	→ samer joined	00:23
	→ sako joined	00:25
DrNick	add it as a remote to your existing kernel repository	00:26
	→ fbrunel joined	00:26
	← masterkorp left	00:27
hotwings	i dont want any kernel repository	00:29
airborn	hotwings, which repo did you use? git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git	00:30
	→ toni joined	00:31
	→ MartinKonicek joined	00:31
hotwings	git://linuxtv.org/v4l-dvb.git	00:31
		00:31
avar	you may be interested in the patch series on-list that's working towards partial checkouts	00:32
hotwings	that would be great. im surprised such a feature doesnt already exist to be honest	00:33
	← jrmuizel left	00:34
airborn	ok, time for me	00:34
	night	00:34
avar	well, because mostly people complain and don't implement	00:34
	And there are still a lot of issues with it, it probably won't get in.	00:35
	← airborn left	00:35
	→ jrmuizel joined	00:35
	← fbrunel left	00:36
	← sh1mmer left	00:36
	→ JoeyA joined	00:37
	← rettub_ left	00:37
JoeyA	Does git perform deduplication when you commit files into it?	00:38
avar	Yes, inherently.	00:38
	It's a content addressable filesystem	00:38
JoeyA	(not semantically, but in terms of memory usage)	00:38
	(disk space, I should say)	00:38
hotwings	avar - was the feature intentionally left out or just not thought of @ design time?	00:39
JoeyA	I have about a million files taking up 200GB or so of space, and I want to put them in a deduplicative archive.	00:39
	← ignatp left	00:39
	→ rettub_ joined	00:39
Arrowmaster	well now i cant find it but i thought there was an article named 'git from the ground up'	00:39
JoeyA	That's sort of an abuse of git, though. I'm not sure that's a well-trodden path.	00:40
avar	hotwings: Do you know how Git tracks content?	00:40
	hotwings: It was considered, but usually you want entire projects, not some subset.	00:40
	→ MartinKonicek__ joined	00:40
avar	hotwings: And it's inherently hard in Git, due to how it does tracking	00:40
hotwings	that seems like a pretty big assumption to make that you'd want an entire project. the kernel itself is a perfect example as you may be working on one portion thats completely unaffected by everything else.	00:41
	→ ignatp joined	00:41
avar	How are you going to build the kernel to test out your changes if you don't have the whole tree?	00:42
hotwings	well, doesnt that assume you need to build a kernel every time instead of just a module?	00:42
	← Paraselene_ left	00:43
	→ scarabx joined	00:43
	← MartinKonicek left	00:43
hotwings	or rather, changes dont automatically mean a new kernel needs to be built when compiling modules	00:44
avar	Yeah, I don't build the kernel a lot.	00:44
	← intripoon left	00:44
	← ExtraSpice left	00:44
	← hobodave left	00:45
hotwings	i build the kernel pretty much only once, but i build alsa and dvb driver modules often for testing purposes	00:45
	→ shiba_yu36 joined	00:46
Arrowmaster	there was a really good article i read that basically walked through the creation of a dvcs exactly like git by using usage scenario examples for each design decision	00:47
	← shiba_yu36 left	00:48
avar	"meh"	00:49
	Git is pretty much the biggest git thingy I've ever cloned, but it's still only 1GB or so, that's nothing.	00:49
	← MartinKonicek__ left	00:49
	→ kpreid joined	00:49
	← JoeyA left	00:50
	→ shiba_yu36 joined	00:52
	→ _numbers joined	00:53
hotwings	v4l lost scores of testers when they switched to git. for that matter, a bunch of the devs wont even use it either because they dont want to be forced into using a testing kernel rather then stable	00:53
	← _numbers left	00:53
	→ Theravadan joined	00:54
	← valeo left	00:54
avar	That doesn't make sense, they lost testers because they switched to Git? Did they make some code changes that made it impossible to run v4l on stable kernels while doing that?	00:54
	If those testers really mind the big linux-2.6.git checkout then setting up a repo that uses git-filter-branch to provide only v4l is only 10 line cronjob away.	00:55
hotwings	its not git they hate, its being forced to download the kernel constantly	00:56
Arrowmaster	that still makes no sense	00:56
hotwings	people have complained about it all from metered bandwidth to just disliking it and everything in between	00:56
	time, diskspace, etc	00:57
	take your pick, its been a complaint	00:57
avar	< Arrowmaster> that still makes no sense	00:57
Arrowmaster	are they doing a full clone or something everytime?	00:58
avar	My linux git is ~400MB once I've downloaded it once doing git pull is only a few hundred K	00:58
Arrowmaster	avar: git.git?	00:58
avar	So if anything they're probably using less bandwidth in the long term	00:58
	no, linux-2.6.git	00:58
	← gigamonkey left	00:58
Arrowmaster	oh, not sure i have linux-2.6.git	00:58
avar	"to just disliking it and everything in between" is also a bit unclear	00:58
Arrowmaster	git.git is tiny though	00:58
hotwings	yup, full clone. end-users dont seem very keen on the idea of maintaining a git repository	00:58
Arrowmaster	then they arent using the system properly so they have no right to complain	00:59
avar	Anyway, like I said it's trivial for someone to maintain a shallow subdir export of linux-2.6.git if those testers really want that.	00:59
	→ guilhermeblanco joined	01:00
	← hasen_ left	01:00
hotwings	you dont need a right to complain about something and considering all those users are a pretty important piece of the puzzle, i dont think its wise to just brush them off like they dont matter	01:00
	keep in mind, ive seem tons of commits to v4l that wernt even tested, and written by people that dont even have the hardware. that in itself is a big mistake in my opinion -- committing patches hoping they work, and if not then hoping someone will report problems quickly.. thats just asking for broken code to wind up in the kernel	01:02
	← linusoleander left	01:02
	slyphon → slyphon_away	01:02
	← adxp-2`` left	01:03
	→ adxp-2`` joined	01:03
	← SOCiETii left	01:03
avar	That applies to 99.9% of all Linux commits. Most of the codepaths are going to be run on hardware that the committer has never tested.	01:03
hotwings	i dont find that to be the case regarding dvb drivers	01:04
avar	Anyway, if this really is a problem for v4l folks as you claim then you could contribute usefully by offering shallow + partial linux clones to people that want them.	01:04
	→ gigamonkey joined	01:04
	→ nuoHep joined	01:04
gebi	are the v4l trees now in git? thought they used mercurial for development	01:06
	→ linusoleander_ joined	01:06
hotwings	yup gebi	01:07
	there are several git trees now	01:07
	the mercurial one still exists but is lagged far behind. its not even up-to-date with the newest stable kernel	01:07
	→ flaguy48 joined	01:08
gebi	ah nice :)	01:08
	← sako left	01:10
	← nuoHep left	01:11
	→ Vampire0 joined	01:12
	← rettub_ left	01:12
	← linusoleander_ left	01:13
	← parasti left	01:13
	→ rettub_ joined	01:13
	← andrewl left	01:13
	← mithro left	01:14
	← Vampire0_ left	01:14
	← SoCo_cpp_ left	01:17
	→ sako joined	01:17
	→ SoCo_cpp_ joined	01:18
	→ rayners joined	01:19
	← unreal left	01:19
	→ unreal joined	01:19
	→ masonkramer joined	01:20
	← Aaaarg left	01:21
	← dreiss left	01:24
	← jspiros left	01:24
	→ jspiros joined	01:24
	← SoCo_cpp_ left	01:25
	→ SoCo_cpp_ joined	01:25
	→ SOCiETii joined	01:26
	← Bombe left	01:28
	→ Bombe joined	01:28
	→ hobodave joined	01:28
	← frogonwheels left	01:29
	→ frogonwheels joined	01:29
	← andrewwatts left	01:30
	→ andrewwatts joined	01:30
	← sagsousuke left	01:32
	→ nuoHep joined	01:34
	← andrewwatts left	01:34
	← frogonwheels left	01:36
	→ frogonwheels_ joined	01:36
	← RobertLaptop left	01:37
	← flazz left	01:38
	→ Zyclops joined	01:41
	← nuoHep left	01:42
	← mar77i left	01:46
	← whaley left	01:46
	← MattDiPasquale left	01:46
	← SoCo_cpp_ left	01:47
	→ SoCo_cpp_ joined	01:47
	← kpreid left	01:48
	→ whaley joined	01:48
	← orafu left	01:48
	→ orafu joined	01:48
	← aziz_ left	01:49
	← hagabaka left	01:49
	→ nuoHep joined	01:54
	← dumb_ass left	01:59
	← redondos left	02:01
	→ redondos joined	02:02
	← nuoHep left	02:02
	← flaguy48 left	02:05
	← masonkramer left	02:07
	→ sh1mmer joined	02:08
	← adxp-2`` left	02:08
	→ adxp-2`` joined	02:08
	→ Aaaarg joined	02:09
	← Oxf13 left	02:12
	→ coderdad joined	02:12
	← kenyon left	02:17
	← scarabx left	02:17
	← kukks left	02:18
	→ nuoHep joined	02:19
	→ portablejim_ joined	02:19
	→ EvanR joined	02:20
	← stringoO left	02:20
EvanR	gah. git diff shows me the diff in color, but it shows control codes instead of colors	02:20
	vim works in color	02:20
	→ cytrinox_ joined	02:21
EvanR	is there some term config option for git?	02:21
	→ kenyon joined	02:22
frogonwheels_	EvanR: it should pick up the reported term .. what terminal?	02:22
EvanR	xterm	02:22
portablejim_	How do I pull a new branch from a remote?	02:22
frogonwheels_	EvanR: perhaps it is your pager then?	02:22
	← beatak left	02:23
frogonwheels_	portablejim_: just fetch/pull	02:23
EvanR	hmm	02:23
frogonwheels_	(that's what termcap/terminfo is for, after all)	02:23
EvanR	man pages show with bold on some words	02:23
frogonwheels_	portablejim_: then you should be able to checkout <branch>	02:23
avar	export LESS="--IGNORE-CASE --LONG-PROMPT --QUIET --chop-long-lines --RAW-CONTROL-CHARS"	02:23
frogonwheels_	avar: thanks :)	02:24
	← cytrinox left	02:24
	cytrinox_ → cytrinox	02:24
EvanR	right now LESS is -M ;)	02:24
	it worked	02:25
frogonwheels_	EvanR: .. well I was right .. avar possibly a touch more actually useful :)	02:25
EvanR	where did you learn these magic words	02:25
	→ rchavik joined	02:27
	← nuoHep left	02:27
	→ hagabaka joined	02:28
	← zimbres left	02:29
portablejim_	frogonwheels_: I did a git pull origin, and it made origin/new-branch . I am guessing I need to do a 'git merge' to fix it up, am I right?	02:31
frogonwheels_	portablejim_: is there a 'new-branch' branch there?	02:32
portablejim_	only origin/new-branch	02:32
frogonwheels_	programble: ok I think git checkout --track origin/new-branch should create and switch you to a tracking branch called new-branch	02:33
	.. after that git checkout new-branch willdo.	02:33
Mage_Dude_	How can I check what ssh protocol is being used from git bash?	02:33
	It's echo $GIT_SSH?	02:34
frogonwheels_	Mage_Dude_: what does that give you?	02:34
Mage_Dude_	Blank line	02:35
frogonwheels_	Mage_Dude_: os?	02:35
Mage_Dude_	Win 7	02:35
frogonwheels_	yeah GIT_SSH should be the one.	02:36
	← Elfe left	02:36
	→ Elfe_ joined	02:36
Mage_Dude_	Weird. That works on WinXP	02:37
portablejim_	frogonwheels_: is there a way to track new branches on a pull?	02:37
Mage_Dude_	I had the keys setup on WinXP, but trying to get it going on Win7 isn't going as well	02:37
	frogonwheels_ shrugs	02:38
frogonwheels_	portablejim_: you creating that many new branches?	02:38
	Mage_Dude_: you want to use plink or ssh?	02:38
portablejim_	more wanting automation.	02:39
Mage_Dude_	frogonwheels_: SSH I think.	02:39
frogonwheels_	well that would be the default.	02:39
	Mage_Dude_: GIT_SSH is just a pointer to the SSH program AFAIK	02:39
Mage_Dude_	Hmmm. Just typing 'ssh' hints to the fact that there is an ssh program installed. But I might need to install a different one too?	02:40
	← SoCo_cpp_ left	02:41
frogonwheels_	Mage_Dude_: nah, should be good. just need to set up your keys in the right dir.	02:41
	.. I think git gui has some nice helpers even for that.	02:41
	→ SoCo_cpp_ joined	02:42
Mage_Dude_	The weird thing is that pulling works fine, but pushing doesn't	02:42
	But I'm pushing from Eclipse and pulling from git-bash	02:43
	→ cirwin joined	02:44
cirwin	I often look through diff, and see a file I need to fix. Is there a better way of choosing which file to edit in this situation than parsing the output of git status through dmenu and then to vim?	02:46
	(the main problem I have with the dmenu approach is that dmenu appears a long way from the current terminal window)	02:47
	→ tstclair joined	02:48
offby1	dunno what dmenu is.	02:48
	← gebi left	02:48
cirwin	it's just a way of selecting one item from a list using the keyboard	02:48
	I'd like a nice way to do that in the current terminal	02:49
offby1	really don't understand your question.	02:49
	If you see a file name in the output of "git status", why not just run vim on that file?	02:50
cirwin	because I see multiple file names	02:50
	and I want to select only one of them	02:50
offby1	I'm still failing to understand	02:50
	Are you trying to avoid retyping the name?	02:50
cirwin	yes	02:50
offby1	ah	02:50
	I just use the mouse -- select the name, then type "vim" and paste	02:50
cirwin	as there are only 3 or 4 files, I shouldn't have to type 15 -> 20 chars	02:51
Arrowmaster	i use tab complete	02:51
offby1	(actually I live inside Emacs, and use a function called "ffap")	02:51
cirwin	unfortunately I am the victim of a wierd X bug which means my mouse is confined to one of my virtual desktops	02:51
	← SoCo_cpp_ left	02:51
offby1	huh	02:51
cirwin	tab complete is a bit rubbish for files within directories	02:51
	offby1: I did say wierd :P	02:51
Arrowmaster	well i use zsh and not bash so my tabcomplete is awesome	02:51
cirwin	Arrowmaster: oh?	02:52
	← timj__ left	02:52
	→ timj__ joined	02:52
	→ SoCo_cpp_ joined	02:53
offby1	I have this dim recollection that "git status" isn't intended to be machine-readable ... which implies there's some other command which is.	02:53
cirwin	git status --porcelain	02:53
	→ archis joined	02:53
	← spearce left	02:53
offby1	ha, you're ahead of me	02:53
	← archis left	02:54
cirwin	ah well, I'll stick with dmenu, it's not as though 10 cm is `that` far to move focus, it just feels slow	02:54
offby1	here's an insane idea -- git status --porcelain \| cut -c 4- \| xargs vim	02:56
	:)	02:56
	You'd have to :q across all the ones you don't care about	02:56
	→ hasenj joined	02:56
cirwin	yeah	02:56
	and I don't want to open any icons	02:56
offby1	but vim is like über-programmable. Isn't there a git interface for it?	02:57
cirwin	sure is	02:57
	but you first need to be in vim :p	02:57
	maybe I should stay there more	02:57
offby1	git status --porcelain \| fgrep -v '.xmp' \| cut -c 4- \| xargs vim	02:57
cirwin	but the commandline interface isn't very good	02:57
offby1	s/xmp/xbm/	02:57
	you get the idea	02:57
cirwin	yes	02:57
	it's a reasonable idea	02:57
	but I think dmenu is faster than my buffer navigation skills	02:58
	maybe i can get vim to present me with a choice	02:58
	hrm, too much hardness	02:58
offby1	The hardness, I hates it	02:59
	← SoCo_cpp_ left	02:59
	→ SoCo_cpp_ joined	02:59
	→ v0n joined	03:03
	← SoCo_cpp_ left	03:03
	← m41n1 left	03:04
	→ hilem joined	03:04
	→ spearce joined	03:05
	← spearce left	03:05
	← portablejim_ left	03:05
	← hotwings left	03:06
	→ ari-__-e joined	03:07
cirwin	bashes builtin "select" function is what I'm using now - if anyone cares	03:08
	bash's even	03:08
	← ComaWhite left	03:10
	← ari-_-e left	03:11
	→ mdel joined	03:11
	← cirwin left	03:15
	← Zyclops left	03:15
	→ nadavoid joined	03:16
	→ frakturfreak_ joined	03:17
	→ alt-dot-net-geek joined	03:18
	→ Zyclops joined	03:20
	← frakturfreak left	03:20
	← ProperNoun left	03:21
	← alt-dot-net-geek left	03:23
	← __name__ left	03:24
	→ unclown joined	03:24
	→ dreiss joined	03:24
	← Zyclops left	03:25
	← slonopotamus left	03:28
skeeziks	Is it possible for me to track a branch in a remote that is itself a remote tracking branch?	03:32
	(i.e. when the server with the remote tracking branch is a bare repo?)	03:32
	→ dumb_ass joined	03:35
	→ Heimidal joined	03:35
	→ ProperNoun joined	03:36
nevyn	sure	03:37
skeeziks	Can you give me an example of what the remote's fetch line would be for my local tracking branch?	03:38
	Er, sorry	03:38
	s/local tracking branch/local repo/	03:38
	← gbacon left	03:39
skeeziks	The default `fetch = +refs/heads/:refs/remotes/origin/' just ain't cutting it.	03:39
	← Blackguard left	03:43
nevyn	hrm wait..	03:48
	the branch has to exist on the origin.	03:48
	nevyn is confused now.	03:49
skeeziks	X is my local, non-bare repo, Y is the intermediate, bare repo, and it routinely does a fetch from Z for all its branches. I'd like to track in X one of Z's branches via Y's remotes.	03:50
doener	skeeziks: add a second line like: fetch = +refs/remotes/:refs/remotes/origin/ (adjust to your needs)	03:50
	← friskd left	03:51
skeeziks	doener: OK, trying...	03:51
	→ webchick joined	03:51
	→ friskd joined	03:51
doener	skeeziks: e.g. instead of using globs, you could specify the full name for the one ref you actually want	03:52
	→ hohoho joined	03:52
webchick	Hey, there. I'm with the Drupal project and we're in the process of migrating our crappy lame CVS repo to shiny awesome Git. In CVS we had some "pre-commit hooks" that would do things like prevent you from uploading blacklisted files (*.tar.gz, LICENSE,.txt, etc.). We're discussing our options in Git and it sounds like there's not much, since commits happen locally, and we could only check at push-time.	03:53
	Could someone check over this comment http://drupal.org/node/869746#comment-3281340 and just verify that that is indeed the case?	03:53
skeeziks	doener: Success! I added `fetch = +refs/remotes/:refs/remotes/vendor/' and it pulled them all down. Thanks!	03:53
	← jfkw left	03:53
doener	skeeziks: you're welcome	03:53
webchick	Cos if that's the case, it sounds like our recourse is either abandoning this plan (and then how do we stop people from committing warez) or else shifting the burden of clean-up from the developers to the infrastructure maintainers.	03:53
skeeziks	Is there some magic that happens when the lhs of the fetch spec specifes refs/remotes rather than refs/heads?	03:54
webchick	of which we have something like a 1000:1 ratio	03:54
	anyway, thoughts would be awesome. :)	03:54
doener	skeeziks: no, you're just selecting a different set of refs to be fetched	03:54
	skeeziks: branch heads live in the refs/heads/ namespace, remote tracking branches live in the refs/remotes/ namespace	03:55
	skeeziks: the left side of the refspec tells git what to fetch, the right side tells it where to store it	03:55
	→ RobertLaptop joined	03:56
doener	skeeziks: so "refs/remotes/:refs/remotes/vendor/" means: Fetch everything from refs/remotes/ and store it locally in refs/remotes/vendor/	03:56
skeeziks	doener: OK, I was thinking about it in reverse then, and I can simplify the second fetch line (the one I posted ended up creating `remotes/vendor/vendor/...' locally)	03:56
doener	skeeziks: no magic involved there	03:56
	→ dumb_ass1 joined	03:56
doener	skeeziks: right	03:57
	skeeziks: by adding the "vendor" thing to the left: +refs/remotes/vendor/:refs/remotes/vendor/	03:57
skeeziks	dbatbold: I feel like I figured this out once before, heh. Thanks a lot!	03:58
	← dumb_ass left	03:59
	→ MacHat joined	04:00
doener	webchick: provide a pre-commit hook so everyone can run the checks locally, but also and a pre-receive hook to your main repo	04:01
	webchick: if a push gets rejected by the pre-receive hook, it should be because the pusher refused to use the pre-commit hook, his fault, so he'll have to deal with the consequences	04:02
webchick	doener: but then we'd have to tell 5,000 developers to change their local .gitconfig x each of their repositories?	04:02
	hrm.	04:02
	I wish I could say they were that good at reading documentation, but… ;)	04:02
doener	webchick: .gitconfig isn't involved, just "cp pre-commit .git/hooks"	04:02
webchick	doener: and thats global, or per-repo?	04:03
	per-repo, looks like.	04:03
doener	webchick: per repo, but you could have some kind of "bootstrapping" script and does such setup stuff	04:03
	s/and does/that does/	04:03
webchick	I don't think that'll work. :\ We literally have at least 5,000 active developers. Each one of which might have between 5 and 300 repositories.	04:03
	We do have a set of shell scripts called "drush" (drupal shell) where we were thinking of adding some of this integration, since probably at least 30% of them use this.	04:04
	but.. it's a sticky problem.	04:04
doener	300 repositories? Didn't know that drupal has so many subprojects	04:04
webchick	So there's no way for us to enforce this server-side without the clients each doing manual config, via script or whatever.	04:04
	doener: Oh. Drupal has 6,000 sub-projects.	04:04
	doener: But each developer won't have a checkout of all 6,000. Just the ones they maintain/work on.	04:05
doener	heh, call me clueless ;-)	04:05
webchick	np. :) There's not really a big reason for a git person to know what drupal even is. :D	04:05
	but it makes for lots of "fun" trying to wrap our heads around how to do this migration.	04:06
	we're trying to do a sort of "middle-ground" phase where we keep our same centralized workflow for most things, but swap out CVS to Git, to help ease the training burden.	04:06
	But the more we dig into it, the less likely it looks like that's going to be workable.	04:06
	← aresnick left	04:06
	← phlowx left	04:07
	← unclown left	04:07
webchick	doener: it's also fun because we have a non-trivial number of those 5,000 devs who are Windows/Mac designers/themers and work in Dreamweaver and have no effing clue what a version control system is, working entirely in GUI. So asking them to make modifications to a config file, or run a shell script is… challenging.	04:07
	anyway, my problems are not your problems. ;)	04:07
doener	hm, I wonder how the eclipse guys handled that, AFAIK they have some strict policies as well	04:08
webchick	But it sounds like our git lead's well-informed and there's not some obvious option that we missed.	04:08
	i just had to try. ;)	04:08
	doener: Oh, interesting. I could try asking around.	04:08
	doener: anyway, thanks a lot for your help!	04:09
	← apsysoev left	04:09
doener	no problem	04:09
	← zenazn left	04:09
	→ d-snp joined	04:10
	← paradoxx left	04:14
	← luis left	04:21
	← electrogeek left	04:21
	→ electrogeek joined	04:21
	→ luis joined	04:21
	→ WebDragon joined	04:22
	→ terinjokes joined	04:22
terinjokes	is it possible to setup git to use a specified ssh key?	04:23
	← thiago left	04:23
avar	no, but you can set up ssh to do that	04:25
terinjokes	avar: we looked at that on kernel.org's wiki	04:25
	the probablem is, the use that needs to run git doesn't have a home directory	04:25
	s/probablem/problem/g (it's been a long day)	04:26
avar	give him a homedir?	04:26
terinjokes	him is apache	04:26
	can i just define $HOME?	04:27
	← adxp-2`` left	04:27
	→ adxp-2`` joined	04:28
blueshift	webchick: the fix for you #9 comment failure is that the dev needs to rewrite their commit rather than do a new one	04:28
	that is not a big deal when you get into git	04:29
webchick	yeah?	04:29
	← dreiss left	04:29
	→ comcor_ joined	04:29
webchick	blueshift: now imagine you're talking to someone who's never seen a command line before and works primarily in Photoshop and Dreamweaver on Windows. Is it still not a big deal? :D	04:29
blueshift	not sure how the push errors work but I assume you can send them some text?	04:29
	← v0n left	04:29
webchick	Yeah, I guess there's something called a … well it says in the issue somewhere. "pre-merge" or something	04:29
	hook	04:29
	← comcor left	04:30
blueshift	how are they going to use git w/o a command line?	04:30
webchick	blueshift: misys and tortoisegit	04:30
	apparently	04:30
	or ide integration	04:30
	lots of ways	04:30
	they're web developers. they're very, very crafty at getting around having to learn stuff. ;)	04:30
	it's an always interesting time seeing how they can stuff up our CVS repo. ;)	04:31
terinjokes	webchick: some of us love learning new stuff	04:32
webchick	me too. :D	04:32
blueshift	I'm not sure git is really compatible with people who aren't prepared to learn it to a moderate extent. although it is getting friendlier as time goes on	04:32
webchick	but for a lot of people, it's pulling teetht o get them to contribute their code back, so we need to make it as easy on them as possible.	04:32
	blueshift: yeah, there will definitely be a learning curve. we've acepted that. but re-writing commit history doesn't have an analog in anything else I can think of, so it's a tricky thing to explain, IMO.	04:33
	I can see it being very useful	04:33
	but I can see it also confusing the living crap out of a good 20-30% of our contrbutor base	04:33
blueshift	webchick: if you want to make it really easy, you could let them push to a staging repo which sanitises their commit and rewrites it if needed, by scripts	04:33
webchick	(if not more)	04:33
	blueshift: So re-write the commit on the server side..?	04:34
blueshift	their commit will arrive but be out of sync with what they pushed, the individual dev will have problems later on	04:34
	but, you get their work and.. hmm.. the more I think about it the more problems that would probably cause	04:34
webchick	Right.	04:34
blueshift	it makes it easy once. :D	04:34
webchick	AFAIK our git lead has ruled out that option for the reasons you state	04:34
	Yep. :D	04:34
	← SandCube left	04:34
	← guilhermeblanco left	04:37
blueshift	you could bundle a script which rewrites history to remove offending files and put that in the tree	04:38
	then the push error could say "you messed up, run xyz.sh and try again"	04:38
	← mdel left	04:38
blueshift	or history-fixup.bat for the windows people. they can even run it from explorer.	04:39
	← nadavoid left	04:43
	→ dumb_ass joined	04:44
	← dumb_ass1 left	04:47
	← batok left	04:47
	← tstclair left	04:51
	→ mdel joined	04:51
	→ alester joined	04:58
	→ bemson joined	04:59
	→ freimatz joined	04:59
bemson	does git auto-generate docs from commits?	04:59
	I'm creating js projects, and currently employ jsdoc comments...	05:00
	→ girishr joined	05:01
	← Mage left	05:01
	→ bambule joined	05:01
bemson	newb here - does git auto-generate docs from commits?	05:01
	← jonmasters left	05:02
bemson	I'm working on some js projects and use jsdoc comments for documentation	05:02
	How does this work with wikis, read-mes, etc?	05:02
	There seems to be no documentation on how to integrate/trigger documentation	05:02
	← hilem left	05:03
	→ zenazn joined	05:04
	← jspiros left	05:04
	→ jspiros_ joined	05:04
frogonwheels_	bemson: hmm.. the git-cola I know has some kind of auto-generate rlease notes thing happening'	05:04
	git-cola project I mean.	05:04
bemson	frogonwheels_: fair enough and I expected/heard as much - I'm just trying to get a handle on documenation/tutorials to this end	05:05
	Or is it all, just wiki and what I manually upload, regarding docs?	05:05
frogonwheels_	bemson: I wrote a _batch_ file to generate the developer release notes for our meetings :)	05:05
	← bauruine left	05:05
blueshift	bemson: when do you want to generate docs? locally on the dev machine when they commit, or on the server when they push?	05:06
frogonwheels_	bemson: Git-cola has auto-generated release notes collated by developer.	05:06
bemson	but what about API docs? YUI is on git, and they use jsdoc comments - does git auto-generate these?	05:06
	frogonwheels_: I guess when I commit	05:06
	I'm definitely not talking about release notes	05:06
	I'm learning git, but it seems it would not auto-update/generate API docs for a project	05:07
frogonwheels_	bemson: ooh ... API docs.. just put a post-receive hook on the server.. or a ??post-commit?? hook that could, I guess regeenrate api docs for that file??	05:07
blueshift	well, have a look at the hooks	05:07
frogonwheels_	bemson: you have hooks.	05:07
bemson	frogonwheels_: ok, so I can't do this all online? I have to use the terminal?	05:07
blueshift	you can write or adapt a hook to do what you want (whatever that is)	05:07
bemson	@all, didn't know about "hooks"! :-)	05:07
	→ hilem joined	05:08
bemson	I'll look into that, and see how to set this up on my terminal - I was hoping to just cut/paste and save and git would do the rest	05:08
	(yes, I'm that lazy)	05:08
	← hasenj left	05:08
bemson	many thanks for the tips - I'll look into this in the morrow :-)	05:08
blueshift	I'd think you want it to be server side.. a commit hook will slow down commits and be annoying	05:08
bemson	blueshift: duly noted	05:08
	thanks again all!	05:08
	lata	05:08
	← bemson left	05:09
blueshift	maybe a push hook if you want to enforce some doc build sanity checks	05:09
	hum	05:09
madduck	how do I test whether git-pull would update the checkout?	05:13
frogonwheels_	madduck: you can always git-fetch and manually merge it.	05:13
madduck	one trivial way might be to fetch, and then to compare	05:13
	git show --pretty=format:%H HEAD origin/master	05:13
	but I cannot stop git-show from printing the diffs	05:14
	oh!	05:14
	→ jonmasters joined	05:14
madduck	if 'git log --pretty=format:%H HEAD..origin/master' outputs no lines…	05:14
	→ Guest30979 joined	05:15
frogonwheels_	madduck: you probably also want to know if you're going to merge	05:15
blueshift	madduck: you could compare merge-base to remote/branch	05:16
frogonwheels_	madduck: whcih I guses is if the common ancestor isn't the same as the tracking branch	05:16
madduck	frogonwheels_: nah, not in this case.	05:16
	in this case i just want to know if a git-pull would do anything at all	05:16
frogonwheels_	madduck: which is what blueshift said :)	05:16
blueshift	madduck: just looking at the output from git pull isn't good enough I supose..?	05:17
	er, git fetch	05:17
	those names are pretty shitty	05:17
	fetch/pull ambiguity	05:17
madduck	blueshift: frogonwheels_ well, the log check is cleaere I think	05:18
	← adxp-2`` left	05:19
	→ adxp-2`` joined	05:19
	← doener left	05:20
	→ doener joined	05:20
	← dumb_ass left	05:20
	→ hasenj joined	05:21
	→ whatdoesitwant joined	05:21
	→ maleknet joined	05:21
	← friskd left	05:22
webchick	headed off. thanks for the feedback, folks!	05:23
	← webchick left	05:23
	→ bambule___ joined	05:26
	← bambule left	05:26
	← MacHat left	05:27
	→ MacHat joined	05:27
	→ mattikus joined	05:27
	→ mar77i joined	05:29
	→ brothers joined	05:30
	← gigamonkey left	05:31
	→ slonopotamus joined	05:31
	→ hakunin joined	05:32
	← Guest30979 left	05:33
	← hakunin left	05:35
	→ hakunin joined	05:35
	← brothers left	05:36
	← girishr left	05:36
	→ nataraj joined	05:37
nataraj	Hi	05:37
	→ JonathanT joined	05:37
nataraj	how to get a datewise sorted list of branches?	05:38
	→ choey joined	05:38
	← Modius left	05:39
	← BryanWB left	05:39
blueshift	nataraj: date the branch pointer changed, or the date of the commit it points to?	05:41
	→ johnw joined	05:41
blueshift	if you want the former, ls -t .git/refs/heads	05:41
	← hilem left	05:42
nataraj	date of creation	05:42
	→ amerine joined	05:42
nataraj	or else a tree like display of branches	05:43
blueshift	date of the first commit on that branch? git doesn't record that	05:43
	← johnw left	05:43
blueshift	oh that you can do. gitk.	05:43
	→ johnw joined	05:43
	→ apsysoev joined	05:43
	← ignatp left	05:44
nataraj	blueshift, need to find out which branch derives from which	05:44
blueshift	I don't know a simple way to do that. someone may	05:45
	← ToxicFrog left	05:45
blueshift	have you tried git show-branch?	05:45
	← mattikus left	05:46
blueshift	or maybe	05:47
	git log --graph --decorate --pretty=oneline --abbrev-commit	05:47
	→ hasen joined	05:49
	← whatdoesitwant left	05:49
	→ hasen_ joined	05:50
	← jrmuizel left	05:50
	→ girishr joined	05:51
terinjokes	it looks like I really should have been using GIT_SSH	05:51
	→ jrmuizel joined	05:51
	← hasenj left	05:52
	← hasen left	05:53
blueshift	nataraj: the problem is that git only thinks of the tip of a branch, it doesn't know what branches earlier commits were	05:54
	→ Titosemi joined	05:57
	← hasen_ left	05:57
	→ poseid joined	05:58
	→ nuoHep joined	05:58
nataraj	blueshift: need to search for a string in the logs , spanning all the branches	05:58
	i meant commits	05:58
	← Eaven left	05:59
	→ bauruine joined	06:00
	→ giallu joined	06:01
doener	nataraj: git log --grep=whatever --all ?	06:02
	→ MacHat_ joined	06:02
	← MacHat left	06:03
	MacHat_ → MacHat	06:03
	← nuoHep left	06:04
nataraj	doener, grep works	06:06
	→ Magicmat joined	06:06
	← slonopotamus left	06:06
	→ mastroDani joined	06:07
	→ slonopotamus joined	06:07
nataraj	would like to get last commit dates of all branches, (getting old you know ;)	06:11
	→ nuoHep joined	06:11
	→ ph^ joined	06:12
	← jrmuizel left	06:13
	← madewokherd left	06:13
	→ elitexray joined	06:13
	→ lanthan_ joined	06:13
	← nuoHep left	06:14
	→ Nemurenai joined	06:14
doener	nataraj: git for-each-ref --format='%(refname:short) %(committerdate)' refs/heads	06:15
	→ cbreak_work joined	06:15
	← lanthan left	06:16
	→ priidu_ joined	06:17
	→ priidu joined	06:17
	→ apsysoev1 joined	06:17
	ari-__-e → ari-_-e	06:17
	← slonopotamus left	06:17
	← d1zzy left	06:18
	← uau left	06:18
	← adxp-2`` left	06:19
	→ adxp-2`` joined	06:19
	→ roop joined	06:19
	← bauruine left	06:19
nataraj	doener:how to specify a date format? %(committerdate:"dd-mm-yy")?	06:19
	← apsysoev left	06:20
	← jamescarr left	06:20
	→ Schmallon joined	06:22
	← jspiros_ left	06:22
	→ jspiros joined	06:22
	← bambule___ left	06:22
	← MacHat left	06:23
	→ \ask_ joined	06:23
	→ paradoxx joined	06:24
	← \ask_ left	06:25
	→ \ask_ joined	06:25
	← \ask_ left	06:25
	← trodrigues left	06:25
	→ d1zzy joined	06:25
	→ trodrigues joined	06:26
	→ hagebake joined	06:26
	← hagebake left	06:26
	→ hagebake joined	06:26
	← hagabaka left	06:27
	← eletuchy left	06:29
	→ eletuchy joined	06:30
	→ vu3rdd joined	06:32
	→ opalka joined	06:34
	← dunny left	06:36
	← rretzbach left	06:37
	← hagebake left	06:37
	← giallu left	06:38
	→ jamescarr joined	06:38
	← johnw left	06:39
	→ hasenj joined	06:40
	← zenazn left	06:41
	← Amorphous left	06:42
	→ urkud joined	06:42
	← rayners left	06:44
	← alester left	06:45
	← Magicmat left	06:45
	→ airborn joined	06:46
	→ Magicmat joined	06:47
	→ ChrisBuchholz joined	06:48
	← eletuchy left	06:52
	→ [Po]lentino joined	06:52
	→ marc_kdab joined	06:55
	← joevano left	06:55
	→ Rene joined	06:55
Rene	hi all!	06:55
	→ jason237 joined	06:55
Rene	Is it possible with a git command to push to a server a local git repo, and that this repo would be initialized and added under eg. /var/lib/git? I could not find any notes about this..	06:56
Sieben	KIKOOOOOOOOOOOO RENE !!!!!!!!!!!!! :o)))))))))))	06:56
Rene	:-)	06:57
	what i found was to log in with ssh, create that directory, and do a bare-init..	06:57
	→ Amorphous joined	07:00
	→ hebz0rl joined	07:00
Mage_Dude_	Rene: Are you not using a repository manager?	07:00
Rene	Mage_Dude_, what's a "repository manager"?	07:01
	→ xiong joined	07:01
	→ leo2007 joined	07:01
Mage_Dude_	Rene: It's a system that helps you manage all your repositories including giving access rights to specific users.	07:01
cbreak_work	Rene: you can only push to git repositories, so if there's nothing inited, you can't push	07:01
Rene	ahaa..	07:01
	ok, so to my question i have to log on with eg ssh, do a eg --bare init and start from there?	07:02
cbreak_work	you could use a program that creates empty repositories in places where you try to push to I guess.	07:02
	for example	07:02
	yes	07:02
Mage_Dude_	An exception to that... is using something like 'gitolite' which allows you to declare a repository exists...then push to it	07:02
Rene	ok, that's then ok	07:02
	→ enherit joined	07:03
cbreak_work	no exceptions :)	07:03
Titosemi	hey guys, any good doc about integrating git and kdiff3 as the merge & diff tool ?	07:04
	← Tommy[D] left	07:05
Rene	cbreak_work, this is really fine with me :-) I just wanted to check that i did not any unnessesary steps :-)	07:05
	← enherit left	07:05
	→ Tommy[D] joined	07:06
	→ enherit joined	07:06
	→ nuoHep joined	07:07
	→ skoop joined	07:08
	→ joevano joined	07:09
sitaram	Titosemi: why do you need docs? "git config diff.tool kdiff3" and similarly merge.tool	07:09
	→ joeytwiddle joined	07:09
Titosemi	sitaram: if i make git diff, it just shows like always and if i make git difftool i think t doesnt find it	07:13
	i guess it´s not in path	07:13
	← amerine left	07:13
Rene	thanks Mage_Dude_ and cbreak_work for your help! cu!	07:13
	← Rene left	07:13
	← nuoHep left	07:13
Mage_Dude_	Oh... I was thinking of emailing you sitaram.	07:13
	← Nemurenai left	07:14
	→ ethy joined	07:15
	← adxp-2`` left	07:15
	→ adxp-2`` joined	07:15
Mage_Dude_	sitaram: I was searching through GitHub and Google Groups trying to see if the issue comes up but I can't seem to find any talk of the issue. That means it's my mistake. It's fixed now but didn't know if you'd like teh solution or whether it's important or not.	07:15
	jexmex → jexmex_	07:16
	← opalka left	07:19
	← vu3rdd left	07:20
	← poseid left	07:20
	← sivy left	07:20
	→ opalka joined	07:21
	→ giallu joined	07:22
	← airborn left	07:22
	→ airborn joined	07:23
	→ eletuchy joined	07:26
	→ workmad3 joined	07:27
	→ chittoor joined	07:29
	← Schmallon left	07:30
	→ galderz joined	07:32
	→ napster_123 joined	07:32
	← Heimidal left	07:32
napster_123	hi, i see many changes in my file's diff that are just due to line endings "^M"	07:32
	is there any way to get rid of it	07:33
	i'm using giv-svn on windows	07:33
sgronblo	Is it possible to stash untracked files?	07:34
	→ nuoHep joined	07:36
	← opalka left	07:36
	→ opalka joined	07:37
	→ valeo joined	07:39
	→ mtkd joined	07:39
	→ mtkd_ joined	07:40
	← nuoHep left	07:40
	→ Paraselene_ joined	07:40
	← mtkd left	07:41
	← ChrisBuchholz left	07:42
	← hasenj left	07:43
	→ ntoll joined	07:43
cbreak_work	sgronblo: you can start tracking them	07:44
	← Aaaarg left	07:45
napster_123	cbreak_work, any thing about "^M" stuff?	07:45
	cbreak_work, i see a lot of fake changes which is really annoying.	07:46
	→ trivol joined	07:46
	← mtkd_ left	07:46
blueshift	napster_123: maybe not much use to you but I fixed all the windows line endings in my history yesterday with this command, under unix:	07:48
	git filter-branch --tree-filter 'find . -path ./.git -prune -o -type f -print0 \| xargs -0 sed -i "s/\x0d//;s/[ \t]*$//"' HEAD	07:48
	← FauxFaux left	07:48
blueshift	that also strips trailing whitespace.. and may damage any binary files you have in git	07:48
	← programble left	07:48
	← choey left	07:49
napster_123	blueshift, have you tried this command? did it successfully work for you?	07:49
blueshift	napster_123: you can either store files with windows line endings under git, or better may be to store unix line endings and configure git to auto convert to windows style when checking out on windows	07:49
napster_123	blueshift, but then i'll also need to configure it to change line-endings back to unix style while committing changes right?	07:50
blueshift	yes it worked for me but may not be what you want. you should understand that it will rewrite history and what that means	07:50
	→ FauxFaux joined	07:50
blueshift	correct. the one config switch does both	07:50
napster_123	blueshift, yeh that's why i re-asked you about that command	07:50
	blueshift, ok and what's that nice config switch?	07:51
	→ v0n joined	07:51
charon	napster_123: which side has the CRs? is the diff you're seeing that the repo has LFs and your editor saves CRLF, or is it the other way around?	07:51
blueshift	can't remember, sorry :) there are several ways to do it though. there's a global auto eol thing, and then per wildcard or per file attributes	07:51
sitaram	Mage_Dude_: I didn't understand the context. What were you searching for?	07:52
napster_123	charon, i'm not exactly sure about what to call this but i see "^M" append to eol as mine made changes to the file	07:53
	→ shiba_yu_ joined	07:53
blueshift	napster_123: related, your editor sucks.	07:53
	← workmad3 left	07:54
napster_123	blueshift, i think platforms' differences suck	07:54
charon	napster_123: in that case it should be sufficient to set core.autocrlf = input, see man git-config. note that i have never actually used this feature because i use a real editor on a real OS, so YMMV, but rewriting history won't do any good.	07:54
Gitbot	napster_123: the git-config manpage can be found at http://git.or.cz/man/git-config	07:54
charon	(also, running g-f-b on windows is quite silly anyway because it's so slow)	07:55
blueshift	napster_123: they do, but you can't change them. you can change to a better editor, though	07:55
	← shiba_yu36 left	07:55
	← FauxFaux left	07:55
	→ hasenj joined	07:55
	← freshtonic left	07:55
blueshift	charon: rewriting history won't stop him from producing messed-up line endings, but may clean up history if some have been committed already	07:56
	→ pantsman joined	07:56
	→ Fullmoon joined	07:56
	→ LiHong joined	07:58
	← asmega left	07:59
	← LiHong left	07:59
	→ FauxFaux joined	07:59
	→ mtkd joined	08:00
	← stepnem left	08:00
	charon shrugs	08:00
charon	blueshift: we'll just direct all questions about suddenly disconnected history and such things to you then :)	08:01
	→ Jaxan joined	08:01
	← hasenj left	08:01
charon	it's true, if he already committed the changes it needs some rewriting, but the wording of the first line seems to indicate it's only in 'git diff' so far	08:01
	→ justin-george joined	08:01
blueshift	I shall go mysteriously silent when history rewrite problems arise :)	08:03
	← _Vi left	08:03
	→ froschi1 joined	08:03
blueshift	though refs/original, the reflog and reset --hard are useful get out of jail free cards	08:04
drizzd_	can I fool git diff into thinking that output is to the terminal, even though it's to a pipe?	08:08
	← adxp-2`` left	08:08
	→ adxp-2`` joined	08:08
	→ zidoh joined	08:08
napster_123	not sure if should feel good or not... this thing has happened to only one file of mine	08:08
drizzd_	I am doing something like (git diff a; git diff b) \| less -SFRX	08:09
napster_123	i can go through the pain to saving my good changes to a temp file, delete the repo file and restore my changes after i checkout a fresh copy of this file	08:09
drizzd_	so git turns off color, but if pass --color to diff I force it on even if it's off by default	08:10
blueshift	napster_123: you just need to fix the line endings on one file?	08:10
	most editors can do that with "save as"	08:11
napster_123	blueshift, yeh.. i'm surprised none other file has been affected	08:11
blueshift	or document properties or somesuch	08:11
	or you can do it with sed or tr if you have them	08:11
napster_123	blueshift, can i remove these "^M" line ending with sed/awk?	08:11
	anybody knows the command for that?	08:11
	← AAA_awright left	08:12
drizzd_	napster_123: note that git diff shows ^M only at the end of new changes, even if the file had CR line endings already	08:13
	← sako left	08:13
drizzd_	napster_123: sed -i 's/\r//g'	08:13
napster_123	drizzd_, my problematic file has ^M appended to it even if there are no new changes.. and you know there's like 200 lines like this	08:14
drizzd_	or dos2unix	08:14
	→ asmega joined	08:14
drizzd_	in vi it's :set ff=unix	08:15
blueshift	napster_123: or, sed -i "s/\x0d//"	08:15
	→ AAA_awright joined	08:15
blueshift	(and then your filename)	08:15
	→ davidw joined	08:16
	← davidw left	08:16
	→ davidw joined	08:16
	→ nuoHep joined	08:17
napster_123	blueshift, drizzd_ these both command do same thing? removing the ^M from any line ending if there's any	08:17
	← jason237 left	08:18
blueshift	should do. I don't know if sed understands \r but \x0d should work.	08:19
	make a copy of your file and try it	08:19
	→ stepnem joined	08:19
napster_123	blueshift, ok	08:19
	→ Aaaarg joined	08:21
	← nuoHep left	08:21
	→ ExtraSpice joined	08:23
	→ poseid joined	08:24
	→ hyperair joined	08:24
	← pantsman left	08:27
	→ lhz joined	08:27
	→ ProperN[out] joined	08:27
	← ProperN[out] left	08:27
	→ ProperN[out] joined	08:27
knittl	is it possible with gitk to display all branches of a remote?	08:28
	zz_rubydiamond → rubydiamond	08:29
charon	knittl: given a sufficiently new git you can 'gitk --remotes=trast' etc, see the description of --remotes in man git-log	08:30
Gitbot	knittl: the git-log manpage can be found at http://git.or.cz/man/git-log	08:30
knittl	charon: yeah, i build git frequently from source	08:30
	← ProperNoun left	08:30
	ProperN[out] → ProperNoun	08:30
knittl	seems to work. thanks!	08:31
	→ mulp joined	08:31
mulp	hello everyone	08:32
	← enherit left	08:33
	← justin-george left	08:33
knittl	can i configure a remote to only pull a single branch?	08:34
charon	you mean fetch? edit the remote.<nick>.fetch line	08:35
knittl	i only need a single branch	08:35
	→ zrbecker joined	08:35
	→ ojii joined	08:39
ojii	hi everyone	08:39
	→ jason237 joined	08:39
ojii	I have problems setting the default push location for a branch, can someone help? for some reasons it tries to push master->master instead of currentbranch->currentbranch	08:39
cbreak_work	by default, it will push all matching	08:42
	→ Schmallon joined	08:44
	dwmw2_gone → dwmw2	08:44
cbreak_work	does currentbranch exist on the remote?	08:44
	→ psankar joined	08:44
	→ courtain joined	08:45
	← jspiros left	08:45
	→ jspiros_ joined	08:45
	→ HG` joined	08:46
	→ gebi joined	08:47
	← nevyn left	08:47
	→ orangemeetyou joined	08:51
	→ tizbac joined	08:51
ojii	cbreak_work, yes it does	08:51
	← eletuchy left	08:53
cbreak_work	but git push remotename just pushes master?	08:53
	what does git push remotename : do? (the : means to push all matching)	08:53
	← Theravadan left	08:54
	→ ixti joined	08:54
	← ixti left	08:54
	→ ixti joined	08:54
cbreak_work	ojii: also check out man git-push, the examples section	08:55
Gitbot	ojii: the git-push manpage can be found at http://git.or.cz/man/git-push	08:55
	→ bluenovember joined	08:56
	→ aspotashev joined	08:58
	← candybar left	08:59
	→ nevyn joined	08:59
	← Sieben left	09:00
	→ workmad3 joined	09:00
	← orangemeetyou left	09:02
	→ engla joined	09:02
	→ rindolf joined	09:02
rindolf	Hi all.	09:02
	← engla left	09:02
	→ engla joined	09:02
rindolf	How do I tell which branch the current branch was forked from?	09:02
	→ zaytsev joined	09:03
zaytsev	Hi! Anybody using git for latex? I'm looking for a way to make diff wrap lines	09:03
cbreak_work	rindolf: not really possible	09:04
charon	rindolf: you can't, in general. if you're lucky your branch has an "upstream" configured, which you can see in 'git branch -vv'	09:04
zaytsev	I have already found --word-diff, which is of HUGE help, but... I don't want to wrap and rewrap the paragraphs manually all the time	09:04
charon	zaytsev: take -S out of the core.pager for that repo	09:04
	or manually type -S<ENTER> in less when you need it	09:04
	→ fivetwentysix joined	09:05
ndim	Is there a way to get a hash for all all the filesystem content a certain branch contains, regardless of what commit history and commit comments lead to that content?	09:06
	→ jtdowney joined	09:06
ndim	I want to find branches that are identical on the filesystem, but differ in the history.	09:06
cbreak_work	ndim: commits have a tree object, which is basically a hash of the content only	09:06
ndim	cbreak_work: That sounds just like what I need.	09:07
charon	ndim: you can read the tree sha1 with 'git rev-parse commit_sha1^{tree}'	09:07
	→ phryk joined	09:07
phryk	Hi there.	09:07
	← adxp-2`` left	09:07
	→ adxp-2`` joined	09:07
ndim	charon: THanks a lot	09:07
phryk	I am pretty new to git, and a merge failed. What do I do? I edited the file it failed at. Now what?	09:07
ndim	(THough I wonder why ^{tree}, not @{tree})	09:08
zaytsev	charon, just added pager = "" to [core] in .git/config. Looks utterly cool, MANY thanks. I was actually under impression that it was using vim for paging because of ~ on empty lines. Are you aware of any page that summarizes nice latex tips? I'm googling them one by one... maybe there's more to it?	09:08
	← girishr left	09:08
cbreak_work	ndim: I don't get the logic behind that notation either	09:09
	rubydiamond → zz_rubydiamond	09:09
workmad3	ndim: ^ is used in several places to get parts of a commit object	09:09
	→ uau joined	09:09
charon	zaytsev: umm, an empty pager?	09:09
workmad3	don't know why ^ instead of a different char, but that's the way it is	09:09
charon	ndim: @{} is the reflog access	09:09
zaytsev	charon, interestingly it worked :)	09:09
charon	zaytsev: anyway, you may also want '*.tex diff=tex' in .gitattributes	09:09
cbreak_work	there's @{upstream}	09:09
zaytsev	charon, is there a way to set this globally?	09:10
	From what I know .gitattributes is per repo	09:10
ndim	ah, reflog. well, upstream is reflog, parts of commit object are not reflog.	09:10
	makes sense.	09:10
charon	don't think so, but a core.attributesfile patch would be welcome	09:10
ndim	thank you all.	09:10
cbreak_work	upstream is not reflog :)	09:10
charon	ndim, cbreak_work, workmad3: and btw it's not just for commit object's "parts". you can also say foo^{commit} which guarantees that you will get a commit pointed to by foo, even if foo is an annotated tag	09:11
	← gebi left	09:11
workmad3	charon: right :) still seems to be referring to 'parts' (I don't know a better word... maybe sub-objects?) of a composite object... just not exclusively commits	09:12
	(and I didn't know about that notation for annotated tags, so thanks :) )	09:12
	← ntoll left	09:12
zaytsev	charon, actually pager = less -FXR (w/o S) still does not wrap lines.	09:12
	→ raichoo joined	09:12
	→ vital joined	09:15
	→ project2501b joined	09:15
	← maleknet left	09:16
	→ maleknet joined	09:17
	→ candybar joined	09:17
charon	zaytsev: maybe your $LESS includes -S. try adding -+S	09:19
avar	hrm, is there a printf() format for unsigned int?	09:19
	avar tries to compile git with clang	09:19
jast	%u	09:19
	actually that's not in the standard implementation	09:19
	oh yes, it is	09:19
	just missed it because the manpage is formatted in an annoying way	09:20
avar	hrm, odd, it'll still warn	09:20
	log-tree.c:297:21: warning: field width should have type 'int', but argument has type 'unsigned int' [-Wformat]	09:20
	"Subject: [%s %0*u/%d] ",	09:20
	← poseid left	09:21
charon	avar: sounds like it's complaining about the type for the *, not the u	09:22
jast	for some reason the field width is supposed to be signed	09:22
zaytsev	charon, YESSS! Thanks!	09:22
	← raichoo left	09:22
avar	ah, it's too earlie. Guess the right thing is just an (int) cast there	09:23
	it'll still die on this though: http://paste.scsys.co.uk/47535	09:23
	"usage.c:56:1: error: function declared 'noreturn' should not return"	09:23
	silly	09:23
	→ sagsousuke joined	09:24
	← courtain left	09:24
	→ girishr joined	09:24
	zz_rubydiamond → rubydiamond	09:24
	← phryk left	09:24
jast	well, what part of that is line 56?	09:25
avar	just the } that closes the function	09:25
charon	i'd just assert(0, "usage_routine must not return") there	09:26
	disclaimer: i made up the assert() syntax	09:26
	→ cilly joined	09:28
frerich	The common pattern for such assertions is assert( !"Unreachable" );	09:29
	→ courtain joined	09:29
avar	It doesn't like that either, I think it just hates everything.	09:30
johto	abort();?	09:30
jast	what is the declaration of usagef?	09:32
	← krawek left	09:32
avar	void usagef(const char *err, ...), it's a vararg	09:32
charon	i think the point is that it's also NORETURN	09:32
jast	yeah, that's what I was after	09:33
	you can easily satisfy the noreturn constraint by calling exit(1) at the end of the function	09:34
FauxFaux	Apparently I'm in the middle of a rebase from aaages ago, but git rebase --abort reports "error: Ref refs/heads/imp is at bd.. but expected 5a..; fatal: Cannot lock the ref 'refs/heads/imp'. Could not move back to refs/heads/imp", probably because I've fiddled with that branch in other ways. How else can I get out of a rebase?	09:34
jast	FauxFaux: remove the .git/rebase-apply directory (and manually recover anything that the partial rebase might have lost back then, if possible and if you actually want to recover stuff)	09:35
FauxFaux	Deleting the branch has unsuprisingly made it more angry. ¬_¬	09:36
	→ ereslibre joined	09:36
FauxFaux	That gets it, cool, thanks.	09:36
	← engla left	09:37
	← maleknet left	09:38
	← toni left	09:38
avar	Ah, I see the problem. It's defining NORETURN on __GNUC__	09:38
	← workmad3 left	09:40
	← cbreak_work left	09:41
charon	avar: but then usage() wouldn't be noreturn either	09:42
avar	it's just hating the void __attribute__((__noreturn__)) prototype, it needs to just be void	09:43
	← ojii left	09:44
avar	This works: http://paste.scsys.co.uk/47538	09:46
	← v0n left	09:46
	→ cbreak_work joined	09:47
charon	which is even wackier since now it apparently lets you a pointer to a function not declared __noreturn__ into a pointer field which is declared __noreturn__	09:48
	→ comawhite joined	09:48
charon	anyway my point was that if it properly ignored the __GNUC__ bit, then there shouldn't have been a problem to begin with	09:48
	since then none of the usage() and friends would actually be __noreturn__ so there's nothing to complain about	09:48
	→ Textmode joined	09:49
charon	well, unless it pretends to be MSC of course.	09:49
avar	nope, it's pretending to be __GNUC__, I checked manually	09:49
	← leo2007 left	09:51
	→ toni joined	09:51
charon	at second reading i just think it completely ignores __noreturn__ on pointers	09:52
	that should be consistent with what you're seeing	09:53
	→ maleknet joined	09:53
avar	otoh I can't build the perl stuff because I confusingly get this:	09:56
	v perl (clang) $ /usr/bin/perl Makefile.PL PREFIX='/home/avar'	09:56
	Only one of PREFIX or INSTALL_BASE can be given. Not both.	09:56
	on master, maybe our Makefile.PL doesn't work with some new EU::MM	09:57
	→ tatsuya joined	09:57
napster_123	i think git config core.whitespace cr-at-eol was something i was looking for	09:58
	→ tomoyuki28jp joined	09:58
tomoyuki28jp	How can I get a list of remote branches?	09:58
schweizer	git branch -a	09:58
tomoyuki28jp	schweizer: thanks!	10:00
	← rchavik left	10:02
	adamholt → adamholt_away	10:03
	adamholt_away → adamholt	10:03
	← adxp-2`` left	10:04
	→ adxp-2`` joined	10:04
	← jason237 left	10:05
	← froschi1 left	10:05
	→ yngress joined	10:05
	→ whatdoesitwant joined	10:06
	→ rousing joined	10:06
	→ nuoHep joined	10:10
	→ scarabx joined	10:11
	→ bauruine joined	10:12
	← fivetwentysix left	10:12
	→ hipe joined	10:13
	→ kipras joined	10:13
	→ intripoon joined	10:13
	← Magicmat left	10:14
	→ happypolla joined	10:14
	← steph021 left	10:15
	← Guest31675 left	10:16
	← andyj left	10:21
	→ zorzar joined	10:24
	← shiba_yu_ left	10:24
	← jtdowney left	10:24
	→ asic joined	10:27
	← zorzar_ left	10:27
	← asic left	10:27
	rubydiamond → zz_rubydiamond	10:30
avar	/w 4	10:30
	sent a patch for the perl thing, odd that nobody saw it before	10:30
	→ beni- joined	10:31
	courtain → curtana	10:34
	← intripoon left	10:35
	← priidu left	10:35
	← happypolla left	10:36
	← hipe left	10:36
	→ priidu joined	10:36
	← priidu_ left	10:38
	→ parasti joined	10:40
	→ priidu_ joined	10:40
	→ lemmy joined	10:43
	→ marcob joined	10:43
lemmy	hi, is it possible to selectively check out a subfolder from a git tree? E.g. smallSubFolder/ from hugeGitRepo/	10:44
maxb	I'm fairly sure it isn't	10:47
	→ Yuuhi joined	10:47
	← napster_123 left	10:50
avar	no	10:51
	→ formale joined	10:52
	← aspotashev left	10:54
	→ bentob0x joined	10:54
	← sagsousuke left	10:54
	← kipras left	10:54
	→ jetienne_ joined	10:55
	← bauruine left	10:56
	→ tvw joined	10:58
	→ herdingcat joined	10:58
	← herdingcat left	10:59
	→ kumbayo joined	10:59
	← fr0sty left	10:59
jetienne_	q. i got submodules in a repo, and i would like to move the folder location inside the repo, how can i do that ?	11:00
rindolf	jetienne_: git mv	11:00
	→ gebi joined	11:00
	→ fr0sty joined	11:00
jetienne_	rindolf: ?	11:01
rindolf	jetienne_: that's the answer to your question.	11:01
	jetienne_: use git mv	11:01
jetienne_	rindolf: it will modify .submodule and .git/config ?	11:02
rindolf	jetienne_: what?	11:02
	jetienne_: what is .submodule?	11:02
FauxFaux	Good to see that neither ditz nor ticgit work at all (due entirely to being written in Ruby), time for a project in a serious language, like cmd.	11:02
	→ froschi1 joined	11:03
jetienne_	$ git mv contrib/vendor/underscore .	11:04
	fatal: source directory is empty, source=contrib/vendor/underscore, destination=underscore	11:04
	rindolf: git mv doesnt seem to understand submodule... or im missing something	11:04
	→ Elfe joined	11:05
	← Elfe_ left	11:06
	→ bloopletech joined	11:08
	→ aziz_ joined	11:08
	← adxp-2`` left	11:12
	→ adxp-2`` joined	11:12
	→ picardo12 joined	11:13
	← Schmallon left	11:13
	→ MattDiPasquale joined	11:13
	→ Aikawa joined	11:13
	→ intripoon joined	11:14
	← Liberum left	11:16
	← MattDiPasquale left	11:16
	→ tcurdt joined	11:17
	→ aspotashev joined	11:18
	→ jim0203 joined	11:20
	→ _Vi joined	11:21
	← daethorian left	11:27
	→ t0rc joined	11:30
	← ixti left	11:31
	→ ixti joined	11:32
	← mar77i left	11:33
	→ d0k joined	11:35
	→ jason237 joined	11:35
	→ jfb joined	11:37
	→ Schmallon joined	11:37
	alin → alin-food	11:38
	jfb → DjiDji	11:38
bloopletech	Is there a way via the git:// protocol to get the size of the repo without doing a full clone?	11:39
	I.e. the "xxMB to send" when you do a clone?	11:39
	→ interferon joined	11:42
	← _Vi left	11:42
interferon	i have a branch B originally created from A. I'm in the B branch and i have a number of unstaged changes. i want to apply just two of these changed files back to the parent branch A. how can i do this? it won't let me switch branches with local changes and git stash stashes all unstaged files but i want to have my two relevant changes left alone.	11:43
FauxFaux	At worst, just commit them, switch, and cherry-pick or git checkout B path/to/file	11:44
crab	interferon: commit those two changes, and cherry-pick the commits on A	11:45
frerich	You can switch branches with unstaged changes, no? I do that all the time.	11:46
	← beni- left	11:46
frerich	I suppose it might be a problem in case switching a branch involves checking out a file which was modified.	11:46
	→ engla joined	11:48
	← engla left	11:48
	→ engla joined	11:48
	← psankar left	11:49
	← whatdoesitwant left	11:53
	→ whatdoesitwant joined	11:54
	← hakunin left	11:54
	→ Liberum joined	11:54
	→ Dave^\| joined	11:57
	→ saidi joined	11:57
	→ snitko joined	11:57
	← t0rc left	11:58
	frakturfreak_ → frakturfreak	11:58
	→ timotei joined	11:59
timotei	is any way I can FORCE git to make eol style to be unix?	11:59
	automatically on commit?	11:59
	← coderdad left	11:59
jetienne_	q. i got submodules in a repo, and i would like to move the folder location inside the repo, how can i do that ?	12:00
	→ masonkramer joined	12:00
	→ workmad3 joined	12:00
	← rindolf left	12:00
	← bremner left	12:02
	← airborn left	12:03
luis	timotei: I think you want core.autocrlf=input	12:03
timotei	hmm, ok, I'll checkit	12:03
	→ jetienne joined	12:04
	→ khelll joined	12:09
	→ ToxicFrog joined	12:10
	← valeo left	12:11
	→ becomingGuru joined	12:11
saidi	jetienne_, maybe git mv	12:11
	alin-food → alin	12:13
	→ blaines joined	12:14
	← blaines left	12:14
jetienne	saidi: several people told me that: but git mv submodule1 submodule2 deosnt work. and git mv doesnt talk at all about submodule	12:15
	← ne0futur left	12:16
	← zrbecker left	12:16
	← scarabx left	12:17
	← NetHawk left	12:17
	→ cubuanic joined	12:18
	← adxp-2`` left	12:18
	→ adxp-2`` joined	12:18
charon	jetienne: mv A B && git rm --cached A && git add B	12:18
	and send a report to [email@hidden.address] so we hopefully won't forget to fix	12:18
	(the above is just spelling out what 'git mv A B' should do anyway)	12:18
	→ NetHawk joined	12:19
jetienne	charon: and this will handle the .gitmodules and .git/config ?	12:19
charon	ah, no	12:19
jetienne	git submodule mv A B <- would love this	12:19
	← timotei left	12:19
	← ajpiano left	12:20
	← jetienne_ left	12:20
charon	i'm also doubting my advice now, i tested and for some reason i can't understand yet the 'git add B' doesn't do anything for me	12:20
jetienne	B is an empty dir ?	12:21
charon	ah, other mistake.	12:21
	← maleknet left	12:21
charon	no, i don't actually use submodules so i dug up a testing repo i had from a dir->submodule conversion test, and for some reason the submodule wasn't checked out	12:21
	→ bloopletech_ joined	12:21
charon	the above only works if you have it cloned	12:21
	→ nchase joined	12:22
	← bloopletech left	12:23
masonkramer	which parent is which in a 2-parent merge commit?	12:23
	→ blaines joined	12:23
	← DjiDji left	12:23
	→ Rearden joined	12:24
	← blaines left	12:24
masonkramer	I merge feature into master, and HEAD is now pointing at the resultant merge commit	12:24
	HEAD^1 is the old master or the old feature?	12:24
	→ blaines joined	12:24
	← picardo12 left	12:25
	→ charnel joined	12:25
charnel	what is svn update in git ?	12:25
	← blaines left	12:25
codejunky	git pull	12:25
	→ TheNewAndy joined	12:26
	← Woody2143 left	12:26
	→ blaines joined	12:27
	→ Woody2143 joined	12:27
charnel	codejunky Thanks.	12:27
	← hohoho left	12:28
wereHamster	masonkramer: first parent = the branch/commit which was HEAD before you did the merge	12:29
masonkramer	wereHamster: thanks, I couldn't find that in the docs anywhere	12:29
saidi	i have committed a big binary file long time a go, and now i want to remove it from the object database, i deleted that commit by re-basing my repo, what should i do next?	12:29
	→ q0rban joined	12:30
frerich	Run 'git gc'; that will remove any unreachable objects from the harddisk.	12:31
	However, note that if you have any commits around which you might want to ressurrect later, but there is no branch which can be used to access them - those commits will be gone, too.	12:31
	→ voxpelli joined	12:32
fr0sty	frerich: assuming they are not referenced in the reflog	12:33
	→ Ryback_ joined	12:34
interferon	FauxFaux, crab: thanks, i just did git checkout B file	12:34
	→ zrbecker joined	12:35
	← Rearden left	12:35
	→ _Vi joined	12:38
	← interferon left	12:38
	→ IslandUsurper joined	12:39
	→ thiago joined	12:39
	← roop left	12:39
saidi	frerich, that's all?	12:39
frerich	saidi: That's it, yes.	12:39
saidi	okay, thank you	12:40
	→ jrmuizel joined	12:42
	→ name joined	12:47
frerich	Does anybody know of a script which lets me merge git branches without explicitely having a checkout of the branch I want to merge into?	12:48
teuf	what would it do when there is a merge conflict ?	12:48
frerich	It would complain, just like normal git merge.	12:49
	→ coderdad joined	12:49
teuf	and fail without doing anything ?	12:49
	→ tstclair joined	12:49
frerich	Yes, asking me to resolve, or to run 'mymergescript --abort' (a bit like git rebase --abort) in case I want to drop my work (and the temporary checkout)	12:50
	→ jacobwg joined	12:50
	→ poseid joined	12:50
teuf	to resolve ? but how would you do that since the working dir isn't where the merge happens ?	12:50
frerich	Well it is during the merge, but not after the merge.	12:50
	So in case there's a conflict, the script bails out and I'm left in a detached HEAD	12:51
	in case everything works out, I get my working tree reset to whatever it was before I did the merge.	12:51
charon	frerich: you can probably build your own from clever usage of read-tree	12:51
	name → __name__	12:52
jacobwg	Is it possible to prevent a file from being committed using a pre-receive hook?	12:53
charon	jacobwg: you cannot stop the user from committing anything (they could override the hooks) but you can prevent the server from accepting commits containing such files	12:54
	(since git-commit is local, but git-push asks the remote side to accept something new)	12:54
jacobwg	charon: right... makes sense, but is it possible to rewrite the commit to not include a file?	12:55
	In a hook?	12:55
charon	bad idea	12:55
	just reject	12:55
	it's most likely possible but it would violate all sorts of expectations, so don't do that	12:55
jacobwg	I see. I'm thinking of a repo for a CMS where it basically enforces a gitignore on the config file that contains the database passwords.	12:56
	Sort of like an automatic .gitignore	12:56
frerich	jacobwg: If you don't want people to commit a certain file, I suggest to just reject the commit. To make it less inconvenient from people who push to you, they could add the file in question to .gitignore so that a 'git add .' doesn't add it every time.	12:56
jacobwg	Right... probably the way I'll have to go.	12:57
frerich	jacobwg: I think you can adjust commits in a hook, but I'd refrain from doing so since I tend to forget about such silent things.	12:57
	So eventually I wonder why I can't adjust a particular file without realizing that some clever script fixes things up on the server side.	12:58
	In fact, wouldn't that bring your local repository and the upstream repository out of sync? You local commit includes an evil file, but the server rewrites the commit before accepting so that the file is not included. That means the server gets a different commit in his repository than you have.	12:58
jacobwg	Good to know... I'm trying to setup a CMS-smart repo server that would auto-detect what to commit and what not to commit.	12:58
rane_	wise words	12:58
jacobwg	frerich: right, I hadn't thought of that...	12:59
	Although, you might want the config file in your local repo but not the remote one since it has a different setup... probably easier to have the remote server do a checkout and then rewrite the config file.	13:00
	→ unclown joined	13:05
charon	jacobwg: maybe this helps: http://thomasrast.ch/git/local-config.html	13:05
jacobwg	charon: Thanks! Looks like it will be helpful...	13:06
	→ flazz joined	13:07
jacobwg	I'm thinking now that I probably don't want the repository to have the remote config file, either, so it's probably better to just keep whatever's in the repo, and then modify the config file after an export on the app server.	13:08
	→ Nemurenai joined	13:09
	← urkud left	13:09
	→ urkud joined	13:10
jacobwg	Trying to build heroku for PHP (for my projects), if that makes sense	13:11
	← EvanR left	13:11
	→ thurd joined	13:14
	← zrbecker left	13:14
	→ rgr joined	13:15
	→ shvedsky joined	13:15
	← mquin left	13:16
shvedsky	hi. Is there any version of git, which has an option to fetch subprojects at clone stage?	13:16
	Like, git clone --with-subprojects git://repo/repo.git	13:16
thiago	git clone --recursive	13:17
	→ andrewwatts joined	13:17
	→ mquin joined	13:17
saidi	frerich, my repository size didn't change after doing that	13:17
	← coderdad left	13:20
	→ coderdad joined	13:20
saidi	removing an big file from history by re-basing the repo should reduce that repository size?	13:22
	→ redbaritone joined	13:22
saidi	an/a	13:22
	→ andreaa joined	13:22
thiago	no	13:23
	you need to expire the reflog first	13:23
saidi	git reflog expire?	13:26
thiago	and set the limit to 0	13:27
	to expire everything now	13:27
	then repack + prune	13:27
	→ roop joined	13:27
	→ beatak joined	13:28
	← nataraj left	13:29
	← jacobwg left	13:31
saidi	how can i set the limit to 0? i didn't find that in the doc	13:32
	→ jaql joined	13:33
	← ph^ left	13:33
	← giallu left	13:33
	← urkud left	13:35
	← formale left	13:36
	→ hoijui joined	13:37
	→ stefl joined	13:38
	→ TeckniX joined	13:39
	→ hilem joined	13:39
stefl	Hi all - wondered if a kind person could help me with a strange git / chef deployment problem..?	13:40
Ilari_antrcomp	stefl: Error messages?	13:41
stefl	@Ilari_antrcomp Chef checks out the current head from a repo, does various capistrano-like things and brings the new app up. I'm getting this though, which I've never seen before (repo works and I can clone it manually on the server):	13:44
	git ls-remote [email@hidden.address] HEAD	13:44
	STDOUT: 7f6029a5db06a2e22549832dcb528b997475ac62 HEAD	13:44
	git fetch origin && git fetch origin --tags && git reset --hard 7f6029a5db06a2e22549832dcb528b997475ac62	13:44
	STDERR: fatal: Could not parse object '7f6029a5db06a2e22549832dcb528b997475ac62'.	13:44
	→ b0fh_ua joined	13:45
Ilari_antrcomp	stefl: Could not parse object indicates it can't read that object. What does 'git fsck --full \| egrep -v "^dangling"' say?	13:46
b0fh_ua	Hello there! Can somebody please advice - if I do have SVN and branches A and B, then those branches were "merged" in SVN with trunk using "reintegrate" feature - will it still be possible to correctly merge them with GIT?	13:46
	or there's no way of merging things correctly in git with SVN?	13:47
stefl	@Ilari_antrcomp thanks - it gives me no output. In fact git fsck —full gives me no output too. Hmm...	13:48
PerlJam	b0fh_ua: if you've already merged with svn, why do you need to merge with git?	13:48
Ilari_antrcomp	stefl: That giving no output means it didn't find any problems from that repo.	13:49
b0fh_ua	PerlJam: in case if I did some changes and committed them into my local GIT repository	13:50
fr0sty	stefl: is 'origin' == github.com:steflewandowski... ?	13:50
	'git ls-remote origin HEAD' ?	13:50
	← darkskiez_ left	13:50
	→ darkskiez_ joined	13:51
stefl	@fr0sty yep I've just checked .git/config and all is in order...	13:52
	← dwmw2 left	13:52
	→ urkud joined	13:52
stefl	@fr0sty and it gives me back the same via ls-remote	13:52
	→ dwmw2_gone joined	13:52
	← marc_kdab left	13:53
	→ xuxc000 joined	13:53
stefl	Thanks guys - pretty stumped.	13:54
	→ tty1 joined	13:54
	→ Theravadan joined	13:54
tty1	can someone tell me why i am getting this error: http://pastebin.com/LwBQgZK0 also what are the reprocutions and how do i fix it?	13:55
	→ psankar joined	13:56
	← adxp-2`` left	13:56
	→ mar77i joined	13:56
	→ adxp-2`` joined	13:56
	→ MattDiPasquale joined	13:56
	→ ajpiano joined	13:56
	← Schmallon left	13:57
	→ bilboa joined	13:58
	→ tech13 joined	13:58
bilboa	hi, that's me again ;)	13:58
	→ Schmallon joined	13:58
	← Ilari left	13:58
	Ilari_antrcomp → Ilari	13:58
	← snitko left	13:58
	offby1 ducks	13:58
tech13	is there a way to get a portion of an other git repo as a submodule? http://pastie.org/private/jey0m7whzzwinkezqb3lng	13:58
fr0sty	stefl: and 'git fetch origin' claims everything up-to-date?	13:58
	→ ivan-kanis joined	13:59
	← jrmuizel left	13:59
	jexmex_ → jexmex	13:59
	→ mkramer joined	14:00
	→ smuf joined	14:00
offby1	tech13: I doubt it	14:01
	→ steph021 joined	14:01
	← freimatz left	14:01
bilboa	is there any way to tell git that some locally modified and committed files are pristine files from upstream (so that they get auto-merged)? (some attribute or something)	14:01
stefl	@fr0sty it does on my local machine. I'm using chef-solo to do the deployment on the server. It checks the repository out into a local cache, then moves the files into the deployment location. It doesn't get that far...	14:02
	← ryanakca left	14:02
stefl	@fr0sty only thing I can think to do is rsync the repo to the server, then ask chef to pull from that repo instead of checking it out. This has only failed for one repo out of five… odd.	14:03
	→ ehabkost joined	14:03
tty1	anyone?	14:04
	← blaines left	14:05
	← opalka left	14:05
fr0sty	tty1: faq non-bare	14:06
Gitbot	tty1: Pushing to non-bare repositories is discouraged. See https://git.wiki.kernel.org/index.php/GitFaq#non-bare	14:06
	← girishr left	14:06
tty1	Gitbot: it isnt a non-bare repository though, ive been using this same repo for ages and the warning just recently started showing up	14:07
	Gitbot: so im not sure how it could have just randomly turned itsel non-bare	14:07
codejunky	lol	14:08
	← hobodave left	14:08
	→ hobodave joined	14:08
tty1	err i meant fr0sty	14:08
shvedsky	tty1: perhaps, it was always non-bare? And new verion of git, to which you updated, started showing the warning?	14:09
tty1	let me go double check jsut to make sure	14:09
	→ Ilari_ joined	14:09
tty1	shvedsky: its def bare by the looks of it, i went to the repo ont he server and did an ls, and it looks liek the content i usually see in .git .. which as far as i know means its bare	14:10
	← b0fh_ua left	14:10
	← bloopletech_ left	14:10
	← hobodave left	14:13
	← JonathanT left	14:14
	← bilboa left	14:15
	→ hohoho joined	14:15
	→ sivy joined	14:15
	← HG` left	14:16
	→ elmob joined	14:17
tty1	any ideas?	14:18
	am i misunderstanding what bare means?	14:18
	or maybe i am checking if its bare the wrong way?	14:18
	← spaceonline left	14:18
Ilari_	tty1: Well, bare is just that.	14:19
	→ leo2007 joined	14:19
tty1	other than doing a ls ont he directory how can i check it is bare?	14:19
	→ bilboa joined	14:19
Ilari_	tty1: See if it has .git directory in it?	14:19
	→ loincloth joined	14:19
ivan-kanis	i have a problem with git an a large repo, it takes all the memory on the server when compressing objects	14:19
Ilari_	tty1: And if it hasn't, check that it has HEAD, objects and refs.	14:20
	ivan-kanis: Large files do that...	14:20
	Ilari → Ilari_antrcomp	14:20
	Ilari_ → Ilari	14:20
tty1	Ilari: other than doing ls	14:21
	see i think what happened is that i was non bare	14:21
bilboa	when i clone a.git, modify file.c inside, then upstream modify file.c the same way as i did, then I change again file.c i another place, when pulling it's going to conflict. if i didn't modify file.c the same way as upstream first, it wouldn't conflict. Since git uses "heuristic algorithms" to find the ancestor i would think it would make no difference. Any way to override this?	14:21
tty1	but i had just copied the .git contents out to make it look bare	14:21
	but i think it still thinks its non-bare as a result	14:21
	→ ph^ joined	14:21
	← saidi left	14:21
	← yngress left	14:22
wereHamster	bilboa: which of the two changes conflicts? Both? Or only the second change?	14:22
	→ \|REM\| joined	14:22
ivan-kanis	Ilari: thanks	14:22
	← TheNewAndy left	14:22
	← mtkd left	14:22
	← cubuanic left	14:22
	→ mtkd joined	14:22
bilboa	wereHamster, it's my yesterday's problem, only the first one conflict actually	14:23
Ilari	ivan-kanis: Basically, RAM usage is 2 * largest file + some quasi-constant.	14:23
	← vital left	14:23
tty1	assuming this is what happened, how do i make a repo that thinks its non-bare think it is bare?	14:23
wereHamster	bilboa: and are you sure you edited it the exact same way upstream did? Do the two conflict hunks look the same?	14:23
	tty1: git config core.bare true	14:24
bilboa	wereHamster, maybe i didn't explain precisely enough in fact:	14:24
	wereHamster, on linux-2.6.git, there was a commit made locally with a newer linux version, then new updates have been pulled from upstream and these updates conflict because there's been more changes in between (upstream changes)	14:24
	that is, upstream made the same changes, then in later commits modified them again	14:24
tty1	wereHamster: do i have to do that from the server itself, or is it done ont he client side? and if its done on the client side will it be applied tot he server when i push?	14:24
wereHamster	tty1: in the repo you want to amke bare	14:25
tty1	thanks	14:25
	i think that did it...	14:25
bilboa	wereHamster, at this point i pull upstream, and the files conflict - but the changes have been integrated upstream already (and modified again upstream)	14:25
	← charnel left	14:25
tty1	wereHamster: how could i confirm it worked? is there some way i can double check from the client side?	14:25
ivan-kanis	Ilari: thanks again that really helps	14:26
	→ darkskiez__ joined	14:26
	← rousing left	14:26
	← apsysoev1 left	14:26
	← darkskiez_ left	14:26
	← Aaaarg left	14:26
	→ spearce joined	14:27
	← WebDragon left	14:28
wereHamster	bilboa: so.. how do the conflict hunks look like? Do they differ?	14:28
	if yes than it means it's a genuine conflict	14:28
bilboa	the hunk differs; but i'm not sure if you understood me	14:29
	→ frakturfreak_ joined	14:29
bilboa	file.c has some piece of code inside, it gets updated and comitted locally, the exact same update is made upstream and committed	14:30
	→ jrmuizel joined	14:30
ivan-kanis	the largest object is 100M however git takes 4G ...	14:30
bilboa	then upstream changes this piece of code again, and commit it	14:30
	then it is pulled and conflicts - the hunk has to look different since upstream has changed it further	14:30
	→ ari-__-e joined	14:30
wereHamster	bilboa: and you don't undersdtand why it causes a conflict?	14:30
bilboa	in a 3 way merge, if the file was not modified locally, it would be automatically merged with fast forward	14:31
	(even if theres a different hunk)	14:31
Ilari	ivan-kanis: Maybe some large file has gotten committed and then "removed"?	14:31
	← jspiros_ left	14:31
	→ jspiros joined	14:31
	← frakturfreak left	14:31
wereHamster	bilboa: there would be no 3-way merge if the file wasn't modified locally	14:31
ivan-kanis	Ilari: i don't think so, it's a repo that's been converted from StarTeam	14:32
	→ CoinOpeBoy joined	14:32
bilboa	wereHamster, well the thing is i want the file to be "fast forwarded" and the local commit to be ignored, or the merge to be replayed properly in order from upstream so that it does not conflict	14:32
Ilari	ivan-kanis: The types and sizes of objects can be obtained from: "git rev-list --branches --tags --objects \| cut -b 1-40 \| git cat-file --batch-check". The third column is size.	14:32
	ivan-kanis: What's the biggest value there?	14:33
	→ munichlinux joined	14:33
bilboa	it seems like it merges HEAD from upstream without looking at commits in between (this is probably how git works since it doesnt use any attribute to store last pull)	14:33
	→ mspitsc joined	14:33
wereHamster	bilboa: you want to ignore all local changes? Or only those to that particular file?	14:33
bilboa	i want to ignore only the changes made by this commit (it affects 300 files ><)	14:33
	← mspitsc left	14:33
wereHamster	bilboa: revert that commit, then merge	14:33
	← ari-_-e left	14:34
bilboa	wereHamster, as far as i understand, git-revert creates a new commit that reverse the effect of the earlier commit, wouldn't i get the same problem next pull?	14:35
	that is, the very next pull will work, then upstream will change a few times and it will conflict again, won't it?	14:35
wereHamster	bilboa: no	14:35
bilboa	does that heuristic algorithm see theres a commit record "type revert" somehow?	14:36
wereHamster	well, it's possible. BUt then you have no other choice than fixing the conflicts manually	14:36
	no	14:36
bilboa	hmm all i found to solve this was git rebase -i and delete the commit, unfortunatel its a pretty old commit and rebase eventually fails with a nonfixable error	14:36
	← paradoxx left	14:37
	← sivy left	14:37
	vals_ → tango_	14:37
bilboa	i feel like i'd better extra a patch of the current state and apply it on top of a fresh clone, so that theres no conflicts on files i don't modify in the future	14:37
	← ixti left	14:37
bilboa	however it means losing history	14:37
	→ sivy joined	14:37
wereHamster	you could merge --ours to avoid losing the history	14:37
\|REM\|	hey guys. I'm setting up Git and moving from SVN. What's the best frontend for Git?	14:38
wereHamster	\|REM\|: git	14:39
charon	bash ;)	14:39
	← uau left	14:39
bilboa	wereHamster, hmm, if i get that right, i make a new branch (or clone) and i merge --ours from my current repos/branch and that takes only the side which is modified by me and keep the modifications as commits?	14:39
\|REM\|	ha. ok sounds good thanks	14:39
charon	\|REM\|: make sure you use the bash (or zsh, if that's your thing) completion and prompt scripts	14:40
bilboa	\|REM\|, if you mean web front-end somehow, there's git-web	14:40
	→ slonopotamus joined	14:40
	← tizbac left	14:40
\|REM\|	yeh, i was refering to the web frontend. Does git work with apache like svn?	14:41
wereHamster	bilboa: sorry, it's merge -s ours. And the descriptino from the man page: This resolves any number of heads, but the resulting tree of the merge is always that of the current branch head, effectively ignoring all changes from all other branches.	14:41
slonopotamus	\|REM\|: git doesn't need apache, it has own very efficient server	14:41
\|REM\|	ok cool	14:41
	thats what i was wondering	14:41
	→ uau joined	14:41
\|REM\|	thanks again for the info	14:42
bilboa	\|REM\|, git-daemon serves git without apache but its not a web front-end	14:42
	\|REM\|, git-web requires a web server afaik	14:42
	git-daemon let you serve files faster than via http, without requiring ssh (i use it for anymous checkouts) for example, git-web is an interface like http://git.kernel.org/<some git repos>	14:43
	wereHamster, im not sure to make sense of it to be honest, i will just try lol	14:43
	wereHamster, thanks by the way	14:43
wereHamster	bilboa: you make a fresh clone, apply all the patches you want so that the tree looks just like you want it to	14:44
	bilboa: then you merge -s ours your old history	14:44
	→ kukks joined	14:44
ivan-kanis	Ilari: i'll tell you as soon as it finishe cloning...	14:46
	→ radarek joined	14:46
	→ smuf_ joined	14:46
	← MattDiPasquale left	14:47
	→ MattDiPasquale joined	14:48
	← smuf left	14:49
Ilari	ivan-kanis: If you have access to server, run it on server?	14:49
	→ tureba joined	14:51
	← becomingGuru left	14:51
	← project2501b left	14:51
	← letas left	14:53
	← tstclair left	14:54
	→ mfacenet joined	14:54
	← fod left	14:58
	→ sagsousuke joined	15:00
	→ hakunin joined	15:01
ivan-kanis	Ilari: 30334024	15:01
	← khelll left	15:01
ivan-kanis	does that mean I have a 30G file ?	15:01
wereHamster	MB	15:02
	← Ryback_ left	15:04
	→ WebDragon joined	15:04
	← MattDiPasquale left	15:04
	→ MattDiPasquale joined	15:04
	→ Ryback_ joined	15:04
	← redbaritone left	15:07
	→ khelll joined	15:07
	← second left	15:08
	→ zimbres joined	15:09
	→ yaotti_aw joined	15:09
	yaotti_aw → yaotti	15:09
	← radarek left	15:09
	← \|REM\| left	15:09
	← MattDiPasquale left	15:09
	← mar77i left	15:09
Ilari	ivan-kanis: Hmm... Is the repo extremely large in number of objects, revisions or files?	15:09
	← khelll left	15:10
	→ opalka joined	15:11
	→ beni- joined	15:11
	← bilboa left	15:12
	→ hobodave joined	15:15
	→ akahn joined	15:15
akahn	Is there a way to remove a file from a commit I just made? I didn't mean to include it.	15:15
	← cbreak_work left	15:16
	→ adamv joined	15:16
wereHamster	akahn: git checkout HEAD~1 -- <file>; git commit --amend; git checkout HEAD@{1} -- <file>	15:16
	→ paradoxx joined	15:17
	→ alester joined	15:17
	← fmardini left	15:17
akahn	wereHamster: cool, thanks	15:17
ivan-kanis	Ilari: 174798 objects	15:17
charon	or just 'git reset HEAD^ -- <file>; git commit --amend' which avoids the reflog dance	15:17
Ilari	ivan-kanis: That's not pracicularly lot.	15:18
	→ LeLutin joined	15:18
	→ bremner joined	15:18
	← urkud left	15:20
	→ urkud joined	15:20
ivan-kanis	Ilari: shall I count the files?	15:20
nchase	if my working directory is clean, why would git rebase -i HEAD~5 return "Working tree is dirty" ?	15:20
	← LeLutin_ left	15:20
	zz_rubydiamond → rubydiamond	15:21
	→ wooden joined	15:21
	← zaytsev left	15:21
slonopotamus	nchase: submodules?	15:22
	→ razvand joined	15:22
	→ davidw__ joined	15:23
	davidw → Guest15788	15:23
nchase	slonopotamus: not using any submodules	15:23
	davidw__ → davidw	15:24
	← davidw left	15:24
	→ davidw joined	15:24
ivan-kanis	Ilari: 17564 files	15:24
charon	nchase: run 'git update-index --refresh'. if after that your tree is still dirty, i'd have to agree with git-rebase	15:25
	← xuxc000 left	15:25
	→ altrux joined	15:25
	← mfacenet left	15:25
	← Guest15788 left	15:26
nchase	hmm	15:26
	← thurd left	15:27
	→ vital joined	15:27
	→ MattDiPasquale joined	15:28
	→ a_l_e_x joined	15:28
	← IslandUsurper left	15:29
ivan-kanis	Ilari: the largest file is 125M, i don't see why it's taking 4G of memory on the server...	15:29
WebDragon	nchase: you using the setgitperms.perl program to track permissions via hooks ?	15:29
nchase	WebDragon: no	15:31
	WebDragon: no idea even what that is :)	15:31
WebDragon	something I ran into a while back myself. this patch fixes the problem with setgitperms.perl leaving the working tree dirty http://fpaste.org/4HTk/	15:31
	it's a contrib/hooks/ mod that tracks ALL perms not just +/-x	15:32
	for some webdevel I need to make sure to preserve the directory perms on some things as 777	15:32
	easier to have git do it than write custom makefiles to do it after the fact	15:32
	→ marc_kdab joined	15:33
nchase	yeah	15:33
	→ zaytsev joined	15:33
	→ stringoO joined	15:34
	→ tizbac joined	15:34
	→ aep joined	15:34
	← andrewwatts left	15:34
WebDragon	but if you're not using it, then something else caused it.	15:34
nchase	yeah	15:34
	← skoop left	15:34
nchase	the mind wonders	15:34
	→ andrewwatts joined	15:35
	← xiong left	15:35
	→ montylounge joined	15:35
nchase	confusing	15:36
	← jason237 left	15:38
	← adxp-2`` left	15:38
	→ adxp-2`` joined	15:38
	← ivan-kanis left	15:38
WebDragon	did charon's suggestion work, though?	15:39
	← paradoxx left	15:39
	← altrux left	15:40
	→ second3 joined	15:41
	→ j0shua joined	15:42
	→ paradoxx joined	15:42
	→ tedoc2000 joined	15:43
	← marc_kdab left	15:43
	→ marc_kdab joined	15:45
nchase	WebDragon: no	15:46
	→ mumboww joined	15:47
	→ giallu joined	15:48
mumboww	greetings everyone	15:48
	offby1 wonders if mumboww is related to the tasty Chinese dumpling	15:48
mumboww	is there a command that can show when a particular line was introduced in a source file?	15:48
	→ DjiDji joined	15:48
rane_	git-blame	15:49
offby1	or git annotate.	15:49
	Similar but subtly different.	15:49
mumboww	oh cool	15:49
	thanks guys	15:49
	← engla left	15:49
	yaotti → yaotti_aw	15:50
	→ raichoo joined	15:51
	→ zz1 joined	15:51
	← zz1 left	15:52
	← paradoxx left	15:53
	→ hipe joined	15:53
	← priidu_ left	15:53
	← andrewwatts left	15:55
	→ letas joined	15:56
	→ zorzar_ joined	15:57
	→ priidu_ joined	15:58
letas	in my exclude file I have this: /app/connect/setup.xml but git still tracks the file, any idea why?	16:00
	oops without the first /	16:00
wereHamster	letas: rm --cached app/connect/setup.xml; commit	16:01
letas	werehamster - just in case using git rm or just rm ?	16:02
wereHamster	yes, and git commit instaed of commit	16:02
	git rm that is	16:02
letas	ok - thanks let me try it now	16:02
	← parasti left	16:03
	← DjiDji left	16:03
	← davidw left	16:04
	→ Heimidal joined	16:05
	← EricInBNE left	16:05
	← Jaxan left	16:06
	→ Eaven joined	16:07
	→ blaines joined	16:12
	← [Po]lentino left	16:13
	← blaines left	16:14
	← froschi1 left	16:14
	→ friskd joined	16:14
	→ mw joined	16:15
	← mw left	16:15
	→ mw joined	16:15
	← friskd left	16:15
	→ sixteneighty joined	16:15
offby1	it was the same in #svn -- people just don't "get" that excluding a file has _no effect_ if it's already checked in	16:15
	→ tjholowaychuk joined	16:16
	→ [Po]lentino joined	16:18
	← hipe left	16:18
	← lanthan_ left	16:19
	→ fxn joined	16:19
fxn	I forked a project in github and did some commits... can I format-patch against the original repo?	16:20
	← trivol left	16:20
	→ kenichi joined	16:21
	← mastroDani left	16:22
rudi_s	fxn: Sure git format-patch origin/master (or change the master if you worked on a different branch).	16:22
	→ bnovc joined	16:23
fxn	oh cool, I assumed origin/master was my own forked repo	16:24
rudi_s	fxn: It's the master branch of the remote repo.	16:24
	← intripoon left	16:24
	← second3 left	16:26
	→ frakturfreak__ joined	16:27
	frakturfreak__ → frakturfreak	16:27
	← frakturfreak_ left	16:28
	→ ProperN[out] joined	16:29
	← ProperN[out] left	16:29
	→ ProperN[out] joined	16:29
	→ dreiss joined	16:30
	← ProperNoun left	16:30
	ProperN[out] → ProperNoun	16:30
	← adxp-2`` left	16:33
	→ adxp-2`` joined	16:33
	← rgr left	16:34
	← Nemurenai left	16:35
	→ kpreid joined	16:36
	→ kenichi_ joined	16:36
	← steph021 left	16:37
	← aspotashev left	16:38
	← mtkd left	16:39
	→ mtkd joined	16:39
	← mtkd left	16:39
	← kenichi left	16:39
	kenichi_ → kenichi	16:39
	→ LinuxFetus joined	16:40
	→ IslandUsurper joined	16:41
	→ jayam joined	16:42
	← tvw left	16:46
	← loincloth left	16:48
	← poseid left	16:48
	→ loincloth joined	16:49
	→ justin-george joined	16:49
	→ bbell joined	16:50
	→ theli0nheart joined	16:51
kenichi	need help tracking down an issue, here's the deal: changes were made to file A like a week ago, co-worker made a commit in his local branch without pulling changes to A (i.e. parent sha was old), then he pulled, which resulted in a merge, couple of conflicts, but not in file A. without saying anything, git chose "his" for file A and all changes for the past week were lost. how can this be? doesn't even show in history/github...	16:51
LinuxFetus	Hey I entered in a command into my terminal and received the following error (https://gist.github.com/300a6010701272594bb1)	16:52
	← zaytsev left	16:53
	← Lgb left	16:54
	→ g3d joined	16:55
kenichi	LinuxFetus: looks to me like you have to configure your networking properly.	16:56
	→ Lgb joined	16:57
	→ krawek joined	16:58
LinuxFetus	kenichi, What does that mean, and how do I go about doing that?	16:59
	← Lgb left	16:59
	← montylounge left	17:00
kenichi	LinuxFetus: what distro?	17:00
LinuxFetus	kenichi, Oh, and I googled the error before posting here and it looks like someone else got something identical (for a different URL). http://sourceforge.net/apps/trac/sourceforge/ticket/5953	17:00
	kenichi, AV Linux 4. http://www.bandshed.net/AVLinux.html	17:00
	→ zrbecker joined	17:00
	→ Lgb joined	17:01
	← tcurdt left	17:02
	← fxn left	17:03
	→ incandenza joined	17:03
	→ fxn joined	17:03
	← fxn left	17:03
	← Schmallon left	17:03
	← whatdoesitwant left	17:04
kenichi	LinuxFetus: well, the short answer is your box can't connect to git.or.cz. the long answer probably belongs in another channel. google something like 'how to set up networking in debian' or 'avlinux' and read up.	17:06
	→ troyt_ joined	17:06
	← ereslibre left	17:07
	← troyt_ left	17:07
	← workmad3 left	17:10
a_l_e_x	hey	17:11
	is it ok to ask for usage help in this chan?	17:11
PerlJam	yes, but it's not ok to ask to ask :)	17:11
	→ xorian joined	17:11
a_l_e_x	dammit	17:11
	← a_l_e_x left	17:11
	→ second3 joined	17:11
	→ a_l_e_x joined	17:11
a_l_e_x	trying to reset the cache on a fresh repo; "git reset" gives me: fatal: Failed to resolve 'HEAD' as a valid ref.	17:12
	→ rgr joined	17:12
	← DrNick left	17:13
jast	nothing committed at all, ever? rm .git/index	17:13
a_l_e_x	brillant, thanks	17:13
	adamholt → adamholt_away	17:13
	adamholt_away → adamholt	17:13
	slyphon_away → slyphon	17:13
a_l_e_x	parfait	17:13
	is that a bad thing to do in any other situation?	17:13
jast	well, it'll make git forget what is staged	17:14
	but then you can simply call git reset to rebuild the index from the last commit	17:14
a_l_e_x	makes sense	17:14
xorian	I have a small patch to the git merge code on which I'd like feedback: http://codepad.org/1FowZtzC	17:14
a_l_e_x	i added lots of .svn files :( they're very intrusive	17:14
	← zorzar_ left	17:15
jast	xorian: ideally you look at the document about submitting patches. the mailing list is absolutely the right place to get feedback, especially if you mark a patch as RFC (though patches always get reviewed anyway as part of the whole process)	17:15
	and fwiw I can't access that pastebin site right now	17:16
	right, works now	17:16
xorian	jast: Yeah, I just figured since this would be my first patch it wouldn't hurt to ask here first.	17:16
	← galderz left	17:17
jast	all the more reason to go to the mailing list, where you'll get sensible feedback from the people who actually know how that part of the code works	17:17
	you don't even have to subscribe	17:18
	← jayam left	17:18
	→ DrNick joined	17:18
	→ jayam joined	17:18
	→ andrewwatts joined	17:18
jast	I'm not too familiar with the merge code, so I can only guess at what that thing you wrote there is actually doing :)	17:18
bbell	Anyone here ever looked at Plastic SCM, and have an opinion to share? Some people at my company view it as "commercial git", with commercial git, fancy GUI, etc. There's not much by way of comparisons that I can find.	17:19
	er, "with commercial support"	17:19
	"Plastic SCM sucks rocks compared to git because ..." would be wonderful. :-)	17:19
PerlJam	bbell: ... it requires a relational DBMS	17:20
	well ... relational-ish if it can use mysql	17:20
	← slonopotamus left	17:20
jast	bbell: the blurbs from the website sound interesting, at least. I can't really say anything about how it performs in practice, though.	17:20
bbell	I noticed that. Apparently it has support for SQLite, so at least you don't necessary have to set one up on each machine.	17:20
jast	sqlite isn't great for extreme concurrency, but you probably don't need that anyway	17:21
	→ MattDiPa_ joined	17:22
jast	hmm, .NET-based.	17:23
	so the server part only officially runs on windows	17:23
	← adxp-2`` left	17:23
	→ adxp-2`` joined	17:23
coldboot	Is there a way to do `git log --before-some-commit=afeafea7d7 -- some_file.txt` similar to --before=<date> ?	17:24
jast	the setup guide doesn't actually mention the need to use a RDBMS	17:24
	← d1b left	17:25
jast	coldboot: git log afeafea7d7^ some_file.txt	17:25
	→ hipe joined	17:25
coldboot	jast: Oh yes, of course, thanks!	17:25
	← bentob0x left	17:25
	→ bambule___ joined	17:26
	→ d1b joined	17:26
jast	one apparent downside of plastic is that you can only configure one server for the client to use	17:27
	which makes for less of a truly decentralized scenario	17:28
	→ ProperN[out] joined	17:29
	← ProperN[out] left	17:29
	→ ProperN[out] joined	17:29
	← ProperNoun left	17:29
	ProperN[out] → ProperNoun	17:30
jast	okay, several servers is possible but cumbersome	17:30
	the whole thing reeks of unnecessary complexity if you ask me	17:30
	adamholt → adamholt_away	17:31
	adamholt_away → adamholt	17:31
	← munichlinux left	17:31
	→ parasti joined	17:31
	← Heimidal left	17:32
	← beni- left	17:32
	→ programble joined	17:34
	→ raichoo1 joined	17:34
	→ bauruine joined	17:35
	← raichoo left	17:36
	raichoo1 → raichoo	17:36
	← leo2007 left	17:36
	→ jason237 joined	17:36
	← LinuxFetus left	17:39
	→ mspitsc joined	17:39
	→ bambule joined	17:39
	← bambule___ left	17:39
	→ Heimidal joined	17:39
	← resmo left	17:40
	← jason237 left	17:41
	← jayam left	17:42
	→ jayam joined	17:42
nchase	anyone still listening: the problem that i described earlier may be because my git repo is on a windows file system and i'm running commands on a mac	17:43
	→ steph021 joined	17:46
	→ kipras joined	17:46
a_l_e_x	nchase: missed that; were you getting problems with case-sensitivity?	17:47
nchase	i was trying to rebase and it kept returning that my working tree was dirty	17:48
	← j0shua left	17:50
	→ valeo_ joined	17:50
	→ rlb3 joined	17:51
	valeo_ → valeo	17:51
xorian	jast: Branch 1 does "mv a b", branch 2 adds b with identical contents. Merging the two gives "CONFLICT (rename/add) ..." and you wind up with two files named "b" and "b~branch2" in the working tree with identical contents. This seems a little pointless.	17:52
	jast: But yeah, I'll go to the mailing list. :-)	17:52
a_l_e_x	nchase: ah. i get dirty issues with line-ending differences sometimes	17:52
	← a_l_e_x left	17:52
nchase	a_l_e_x: windows makes me want to cry :(	17:52
jast	yeah, without thinking too deeply about it, that sounds like a good change	17:52
nchase	ah, gone	17:52
	→ resmo joined	17:52
	→ patthoyts joined	17:53
	← rlb3 left	17:53
esc	/../wc	17:54
	ls	17:54
	→ beni- joined	17:54
	→ amerine joined	17:54
esc	er :-/ .. sorry	17:54
	→ aspotashev joined	17:54
patthoyts	Something bad has happened to repo.or.cz ? Seems unreachable from here.	17:54
	→ zenazn joined	17:55
	→ notjohn joined	17:55
	→ telemachus joined	17:56
	← telemachus left	17:57
	→ bambule___ joined	17:57
	← CoinOpeBoy left	17:57
	← kenyon left	17:58
	← bambule left	17:58
	← hohoho left	17:58
__name__	patthoyts: same here	17:59
	→ sammcd joined	17:59
patthoyts	thanks. not just me then :)	18:00
	→ tstclair joined	18:00
__name__	"It's not just you! http://repo.or.cz looks down from here."	18:00
	→ kenyon joined	18:03
	← MattDiPa_ left	18:03
	→ slonopotamus joined	18:03
	→ telemachus joined	18:08
	← npetrov left	18:08
	→ andrewwatts_ joined	18:08
bbell	jast: The model is distributed in that you run a local server, and use your client on a local server. OR you can use it in a centralized fashion by connecting to a remote server. I agree that using a local server seems unnecessary complex.	18:09
	← Heimidal left	18:09
	← xorian left	18:10
	← telemachus left	18:10
bbell	__name__: Is that http://downforeveryoneorjustme.com/repo.or.cz ? :-)	18:11
__name__	bbell: Indeed	18:11
	← andrewwatts left	18:11
	andrewwatts_ → andrewwatts	18:11
	← roop left	18:12
patthoyts	neat. I'll have to remember that on.	18:13
	→ kumbayo_ joined	18:13
	← adxp-2`` left	18:15
	→ adxp-2`` joined	18:15
	← jayam left	18:17
	← kumbayo left	18:17
	← curtana left	18:17
	← tatsuya left	18:17
	← sagsousuke left	18:18
	← uau left	18:18
	← SOCiETii left	18:18
	→ sagsousuke joined	18:18
	→ tobiassjosten joined	18:19
	→ radarek joined	18:19
	→ tvw joined	18:21
	→ telemach1s joined	18:24
jjuran	bbell: http://downforeveryoneorjustme.com/downforeveryoneorjustme.com	18:24
	← naquad left	18:25
	→ RG\|Lappy joined	18:25
	← hebz0rl left	18:26
	→ slackstation joined	18:26
	telemach1s → telemachus	18:27
slackstation	I set vim as my editor but, when I edit my commit message and :wq (write quit) I get an error: There was a problem with the editor 'vim'. Is there a flag I should be passing to vim?	18:27
jast	usually not	18:27
	do you get anything else?	18:27
	→ ToreadorVampire joined	18:27
slackstation	jast: Please supply the message using either -m or -F option.	18:28
	is the error I get	18:28
	← tobiassjosten left	18:28
	→ naquad joined	18:28
jast	did vim output anything interesting?	18:29
	← resmo left	18:29
	← ToreadorVampire left	18:29
slackstation	jast: nope	18:30
	just quit normally	18:30
	← telemachus left	18:30
	→ leo2007 joined	18:31
	← incandenza left	18:31
	→ drizzd joined	18:33
jast	I think for some reason vim exits with a code that isn't 0	18:33
slackstation	jast: I've tried using 'vi' instead of 'vim' for editor and it's the same result	18:34
	Is there some log that I can check what code vim exits with?	18:34
jast	not really	18:34
	← drizzd_ left	18:36
slackstation	How do people ever get work done in UNIX? I feel like I spend all my time messing with config options for every tool I touch.	18:36
drizzd	slackstation: you can just edit a random file with vim, :wq and then type "echo $?"	18:38
	→ paradoxx joined	18:38
slackstation	drizzd: trying that...	18:38
	drizzd: output: 1	18:39
__name__	slackstation: I wonder the same thing about Windows.	18:39
	← bluenovember left	18:39
__name__	How can one work with that?	18:40
slackstation	drizzd: I guess that means that vim is not exiting with a code 0	18:40
	→ drizzd_ joined	18:40
drizzd_	slakmagik: it should say 0	18:40
fr0sty	slackstation: bingo. What distro are you using? what does 'which vim' tell you?	18:40
	→ lyosha joined	18:41
drizzd_	slakmagik: oops, wrong address	18:41
slackstation	"distro": OS X 10.6.4	18:41
	← stringoO left	18:41
slackstation	"which vim": /usr/bin/vim	18:41
	→ resmo joined	18:42
jast	yeah, seems to be a problem in vim for OS X	18:42
	I just found someone else having that problem	18:42
drizzd_	well, the easy fix is writing a script which simply has "vim; exit 0"	18:42
	ahem, I mean workaround. that doesn't fix the real problem vim is having	18:42
	→ Titosemi_ joined	18:43
	→ Jaxan joined	18:43
jast	rumour says you can set EDITOR=/usr/bin/vim (full path!) or git config --global core.editor /usr/bin/vim	18:43
lyosha	Hi. I have a remote that I set up by doing "git remote add -t branch origin git://...". How can I add fetching of another branch to it?	18:43
jast	that allegedly fixes it for some reason	18:43
drizzd_	how strange	18:44
	→ ijuma_ joined	18:44
	← drizzd left	18:45
	← ijuma left	18:45
	← Titosemi left	18:45
	Titosemi_ → Titosemi	18:45
drizzd_	lyosha: you need to edit .git/config, it should have something like [remote "origin"] fetch = refs/heads/branch:refs/heads/origin/branch	18:45
	← gebi left	18:45
jast	lyosha: edit .git/config, find the line in [remote "origin"] that says fetch = <something with that branch in it>, and create an additional line that goes similarly	18:45
	← razvand left	18:45
drizzd_	jast: oh? multiple lines are concatenated?	18:45
	brosner_ → brosner	18:45
jast	don't we have this big feature of multi-value settings precisely for that?	18:46
lyosha	jast: thanks. I just add another line that looks similar? There is no formal way using git commands?	18:46
	andreaa → andreaa`afk	18:46
drizzd_	jast: I didn't know, but that's useful	18:46
jast	lyosha: none that I'm aware of	18:46
	drizzd_: see the --add switch and friends on git config	18:46
	I think it's especially for settings like the refspecs	18:46
lyosha	jast: yep, that worked. Thanks!	18:46
	→ stringoO joined	18:47
	→ uau joined	18:47
	→ CoinOpeBoy joined	18:51
	→ galderz joined	18:52
	→ drizzd joined	18:52
	← RG\|Lappy left	18:54
	← smuf_ left	18:54
	→ bcardarella joined	18:55
	← kenichi left	18:56
	← drizzd_ left	18:56
	→ rpg joined	18:56
letas	is it better to git pull remote branch_name or just git fetch?	18:56
	← lyosha left	18:57
	→ chrispitzer joined	18:58
chrispitzer	I've started making changes, and now I realize i should have created a branch. Can I move my uncommitted changes to a branch?	18:58
jast	letas: pull actually combines fetch and merge, so if you want to update your local branch with changes from elsewhere, git fetch alone won't do the trick	18:59
	← bambule___ left	18:59
jast	chrispitzer: uncommitted changes are not really attached to anything. git checkout -b newbranch and that's it.	19:00
	→ bambule___ joined	19:00
	→ mattikus joined	19:00
chrispitzer	thanks	19:00
	→ kenichi joined	19:00
letas	jast thanks - I think my question is actually: Which should I use? I remember attending a conference and the guy saying that he advises to use git fetch and then git merge	19:01
rane_	bad advice	19:01
jast	letas: the advantage of that is that you can locally review the changes you fetched before deciding whether you want to merge and what	19:01
	if you know that the changes are going to be good, just do a git pull	19:01
letas	jast gotcha - makes total sense	19:02
rane_	typically you don't have to review what you're about to merge	19:02
jast	that depends completely on your scenario	19:02
rane_	yeah	19:02
	that's why I said typically	19:02
jast	decentralized projects can have totally unpredictable changes in some remote repositories	19:02
letas	so in a regular collaboration environment with 4 to 6 branches - I think u need to review them - right?	19:03
jast	it's up to you, really :)	19:03
	if I know who can push to the repository I'm fetching from, and I trust that person to push only good things, I typically don't bother reviewing things every time	19:03
	→ pantsman joined	19:04
jast	and if I'm the only person working on that repository, just in different places, it's really a foregone conclusion :}	19:04
	because, you know, I never trust myself. I'm a really untrustworthy kind of guy.	19:04
	← adxp-2`` left	19:04
	→ adxp-2`` joined	19:04
	← hilem left	19:05
	← loincloth left	19:07
	→ masterkorp joined	19:08
	→ loincloth joined	19:08
	→ gigamonkey joined	19:09
	← hipe left	19:09
	← _Vi left	19:09
	→ hipe joined	19:09
letas	makes sense	19:10
	→ _Vi joined	19:10
letas	thanks a lot all :D	19:10
	adamholt → adamholt_away	19:10
	adamholt_away → adamholt	19:10
	→ ProperN[out] joined	19:10
	← ProperN[out] left	19:10
	→ ProperN[out] joined	19:10
	← zrbecker left	19:12
	→ calin joined	19:13
	← ProperNoun left	19:13
	ProperN[out] → ProperNoun	19:13
	← redondos left	19:13
	← hipe left	19:14
	→ redondos joined	19:14
slackstation	I have a directory (sto-rails) that is in my repo but, when I cloned it to another machine, it didn't update the contents, it was just empty when I pulled it. I put files in there manually and they don't show up when 'git status'. What could be going on?	19:14
	← Garoth left	19:14
jast	slackstation: were those files ever actually committed to the source repository?	19:15
	→ Garoth joined	19:15
	→ brntbeer joined	19:15
slackstation	I think so but, when I put files in that directory on the remote machine directly and do 'git status' they don't show up at all. I've cleared out the .gitignore but, it won't show up. Is there a file permissions issue?	19:16
jast	there shouldn't really	19:16
	← valeo left	19:16
jast	after all, you cloned the repository yourself and the rest of it got created without problems, right?	19:16
	does git ls-files -o show them?	19:17
	← sixteneighty left	19:17
slackstation	'git ls-files -o' shows nothing at all	19:18
	→ hipe joined	19:18
	→ poseid joined	19:19
	← [Po]lentino left	19:20
jast	it's supposed to show untracked files	19:21
	git ls-files -c shows tracked+unchanged files, ignoring ignore patterns	19:21
	→ mbroeker joined	19:21
	← stefl left	19:22
	Caelum → MDMA	19:24
	← Jaxan left	19:24
	MDMA → Caelum	19:25
	→ sixteneighty joined	19:26
	← mulp left	19:26
	← _Vi left	19:26
	→ _Vi joined	19:28
	← stringoO left	19:29
	→ khelll joined	19:29
	← khelll left	19:30
	← opalka left	19:34
	← unreal left	19:34
slackstation	it just lists sto-rails as one line while everything else it lists... Maybe I set sto-rails as a symlink on the original machine?	19:35
jast	that might be possible	19:35
	ls -l should show you whether it's a symlink	19:36
	→ fod joined	19:36
	→ blaines joined	19:37
mumboww	hi all	19:37
	what's the right way to "fix" whitespace diffs	19:38
slackstation	jast: no symlink on the original machine	19:38
mumboww	i have two versions of a file that the only differences are extra newlines	19:38
	but i dont want to ignore the whitespace diffs... just want to get rid of teh extra newlines in the new file	19:38
	→ bentob0x joined	19:38
	→ gebi joined	19:38
	← blaines left	19:39
	→ unreal joined	19:39
	→ blaines joined	19:40
	→ mar77i joined	19:40
	← galderz left	19:40
mumboww	actually sorry	19:41
	hmm	19:42
	← blaines left	19:42
mumboww	ok i was confused	19:42
	rubydiamond → zz_rubydiamond	19:43
	→ jeffszusz joined	19:43
	→ whatdoesitwant joined	19:44
	← mattikus left	19:45
jast	slackstation: are the files listed in: git show HEAD:sto-rails	19:45
	(assuming that sto-rails is at the top level of the repository)	19:45
	→ blaines joined	19:45
	← sixteneighty left	19:46
	← gebi left	19:46
	→ gebi joined	19:46
	→ cubuanic joined	19:47
slackstation	I did 'git show HEAD:' and sto-rails was listed without a slash after it. Does this mean it wasn't a directory?	19:47
	← cilly left	19:47
slackstation	I'm about ready to just blow this repo out and start over.	19:47
	← blaines left	19:47
	→ khmarbaise joined	19:48
	→ drizzd__ joined	19:50
	→ sam_ joined	19:50
	second3 → second	19:51
	← second left	19:51
	→ second joined	19:51
	← second left	19:51
	→ second joined	19:51
jast	slackstation: in the original repository, what does git cat-file tree HEAD: \| grep sto-rails say?	19:51
	oh wait	19:51
	I forgot something there :)	19:51
slackstation	jast: Thanks for all the help but, I'm gonna shelve this for now and get to work.	19:52
jast	it's actually: git cat-file -p HEAD: \| grep sto-rails	19:52
	oh, right	19:52
	see you around, perhaps	19:52
sam_	Git ignores the tmp directory under all .svn directories (.svn/tmp is igored) . Any reason for this? I don't have any exclusions in the repository.	19:52
	→ felipec joined	19:53
	← drizzd left	19:53
chrispitzer	i'm ending up with branches that I've deleted locally hanging around on github... how do I make it so github kills a branch when i delete it locally and push?	19:55
	→ reuteras joined	19:55
parasti	sam_: Git does not track directories	19:57
	← paradoxx left	19:58
sam_	parasti: the temp directories contain other files and directories and they are all ignored.	19:58
	← fr0sty left	19:58
	← psankar left	19:58
sam_	parasti: the ".svn/tmp"	19:58
parasti	sam_: .svn/tmp has only directories in it	19:59
	sam_: at least in the test I just did	19:59
	← tech13 left	19:59
	← radarek left	19:59
	→ anjen joined	19:59
	→ ProLoser\|Work joined	20:00
ProLoser\|Work	hello	20:00
	$ git remote rm origin	20:00
	error: Could not remove config section 'remote.origin'	20:00
	what is causing this?	20:00
	→ jeffszusz_ joined	20:00
sam_	parasti: you're right. I should have looked closer. Now I need to figure out how to get svn to deal with this.	20:00
	parasti: Thank you!	20:01
	→ sylr joined	20:01
sylr	Hi	20:01
	non git related question: When using --prefix with configure, how do I retrieve its value IN the makefile ?	20:01
jast	chrispitzer: that sort of thing is only implemented for completely mirroring a repository to somewhere else, which probably won't do you too much good. generally, the idea is that you usually push branches that are meant to keep existing on the other side, and use git push --delete to get rid of them once you decide you don't need them anymore	20:01
	→ sgh joined	20:02
	← ProLoser\|Work left	20:02
jast	sylr: @prefix@ in Makefile.in	20:03
	(do read the docs of autotools :))	20:03
	→ Heimidal joined	20:03
sylr	jast: thanks, I was looking at --prefix	20:03
	← jeffszusz left	20:04
	← felipec left	20:04
	← redondos left	20:04
	→ redondos joined	20:05
	← tvw left	20:05
	→ reflectivedev joined	20:05
	← Titosemi left	20:05
	← ft left	20:06
	→ ryoshu_ joined	20:06
	→ ft joined	20:08
	→ Jaxan joined	20:08
	zorzar → zorzar_	20:08
	→ blaines joined	20:10
	→ paradoxx joined	20:11
	→ mtkd joined	20:11
	← blaines left	20:12
	→ zorzar joined	20:14
	→ webus joined	20:15
	← ph^ left	20:15
	→ RORgasm joined	20:16
	← Jaxan left	20:17
	← patthoyts left	20:17
	← sam_ left	20:19
	← rpg left	20:19
	← rgr left	20:19
	→ rpg joined	20:19
	→ rainvt joined	20:20
	← rpg left	20:21
	adamholt → adamholt_away	20:21
	adamholt_away → adamholt	20:21
	← orafu left	20:27
__name__	is there an easy way to determine if a change has been done by pull when scripting operations?	20:27
	→ orafu joined	20:28
	→ stringoO joined	20:30
	← RORgasm left	20:30
	→ RORgasm joined	20:31
	← reuteras left	20:32
	← nchase left	20:32
	← bcardarella left	20:33
	yaotti_aw → yaotti	20:33
	← Theravadan left	20:33
drizzd__	__name__: check the sha1 before and after the pull?	20:33
__name__	drizzd__: i wondered if there was an easy way. thanks :)	20:33
	*easier	20:33
drizzd__	not sure how much easier it can get though	20:34
__name__	i thought i might have missed an option of pull	20:34
	that gives me that info in the return code	20:34
	← vital left	20:34
drizzd__	but pulling in a script comes with other problems anyways	20:34
__name__	drizzd__: merges?	20:35
drizzd__	or dirty files	20:35
	→ vital joined	20:35
	→ phlowx joined	20:35
__name__	no the repo is clean.	20:35
	I'm writing a small buildbot-like thing.	20:35
	← redondos left	20:36
	→ redondos joined	20:36
	yaotti → yaotti_aw	20:39
	← TheUni left	20:39
	yaotti_aw → yaotti	20:39
	→ apsysoev joined	20:41
	→ ChrisBuchholz joined	20:42
	← sgh left	20:43
	→ rgr joined	20:43
	rgr → Guest4850	20:44
	→ johnw joined	20:44
	→ brntbeer_ joined	20:45
	→ rgr___ joined	20:45
	← lhz left	20:45
	→ reuteras joined	20:45
	← rgr___ left	20:46
	← Guest4850 left	20:46
	← rgr__ left	20:46
	← brntbeer left	20:46
	→ ShooterMG joined	20:47
	→ eletuchy joined	20:47
	→ Lars_G joined	20:47
Lars_G	Hi all.	20:47
	Is there any way I can change config so that I push ALL my branches (new and all) to my origin?	20:48
	→ sgh joined	20:49
	→ Blackguard joined	20:49
	→ TheUni joined	20:49
	→ aresnick joined	20:51
	← tg left	20:51
	→ tg joined	20:51
	← adamv left	20:52
	← jaql left	20:53
	← sgh left	20:53
	← gigamonkey left	20:54
	← marc_kdab left	20:54
	→ marc_kdab joined	20:55
	→ rgr__ joined	20:55
	← andrewwatts left	20:57
	→ sgh joined	20:57
	→ andrewwatts joined	20:57
	← raichoo left	20:58
wereHamster	Lars_G: man git-config -> push.default	20:58
Gitbot	Lars_G: the git-config manpage can be found at http://git.or.cz/man/git-config	20:58
	← coderdad left	20:59
	→ tvw joined	20:59
	→ coderdad joined	20:59
	← rgr__ left	20:59
	← kipras left	20:59
	→ rgr__ joined	20:59
Lars_G	The thing wereHamster is that do I have to config a [Branch] for each new branch in each client?	21:00
	← mtkd left	21:00
	→ coderdad_ joined	21:01
	← coderdad left	21:01
	← andrewwatts left	21:01
wereHamster	Lars_G: ah, push.default will not do it. But 'git config remote.origin.push refs/heads/:refs/heads/' will..	21:02
Lars_G	ok	21:02
	I'll try, thanks	21:02
	← rgr__ left	21:02
	→ rgr__ joined	21:03
Lars_G	See you and thanks	21:03
	← Lars_G left	21:03
	← rgr__ left	21:03
	adam__ → adam-_-	21:03
	→ rgr__ joined	21:03
	← aspotashev left	21:04
	→ znake joined	21:04
sylr	is repo.or.cz down for you ?	21:05
	← rainvt left	21:05
	→ EricInBNE joined	21:05
	→ andrewwatts joined	21:05
	← gebi left	21:06
	← adxp-2`` left	21:06
	→ adxp-2`` joined	21:06
	→ picardo12 joined	21:08
wereHamster	yes, seems down	21:08
	← rgr__ left	21:08
	→ rgr_ joined	21:09
	rgr_ → Guest28004	21:09
	← slonopotamus left	21:10
	→ p3rg2z joined	21:11
	← mar77i left	21:12
	→ blaines_ joined	21:13
	← beni- left	21:13
thiago	sylr: http://downforeveryoneorjustme.com/repo.or.cz	21:13
	→ pantsman- joined	21:13
	← blaines_ left	21:13
	← pantsman- left	21:14
	← ajpiano left	21:14
	← hoijui left	21:15
	→ blaines_ joined	21:17
	← znake left	21:17
	← pantsman left	21:17
	← blaines_ left	21:17
	→ workmad3 joined	21:18
	← Guest28004 left	21:20
	→ Schmallon joined	21:21
	← redondos left	21:21
	→ Theravadan joined	21:22
	← unclown left	21:22
	→ redondos joined	21:22
	← gcl left	21:24
	← masterkorp left	21:24
	→ blaines joined	21:24
	→ Mage joined	21:24
	← apsysoev left	21:26
	← blaines left	21:26
	→ zenazn_ joined	21:28
	← zenazn left	21:28
	← sgh left	21:29
	→ gigamonkey joined	21:29
	← bambule___ left	21:29
	adamholt → adamholt_away	21:30
	adamholt_away → adamholt	21:30
	→ rpg joined	21:30
	→ bambule___ joined	21:31
	→ pantsman joined	21:32
	← g3d left	21:32
	← stringoO left	21:32
	→ spaceonline joined	21:33
	← devurandom left	21:33
	← resmo left	21:34
	← andrewwatts left	21:34
	← voxpelli left	21:36
	→ resmo joined	21:38
	→ blaines joined	21:38
	← mspitsc left	21:39
	← blaines left	21:41
	→ lresende joined	21:41
	← lresende left	21:41
	← mdel left	21:42
	← IslandUsurper left	21:44
	← rpg left	21:44
	← johnw left	21:44
	← tstclair left	21:44
	→ rpg joined	21:44
	→ Jaxan joined	21:46
	→ malev joined	21:50
	→ gcl joined	21:51
malev	hi everyone! is it posible to get only one file from a remote branch with git?	21:51
	→ blaines joined	21:51
	← sylr left	21:51
	→ flaguy48 joined	21:52
	→ airborn joined	21:53
	→ galderz joined	21:54
	← jeffszusz_ left	21:54
	← letas left	21:55
	← d0k left	21:55
	← Heimidal left	21:55
	← gigamonkey left	21:56
	← Schmallon left	21:56
	← picardo12 left	21:56
	adam-_- → adam__	21:56
	→ bin1010 joined	21:56
	→ Heimidal joined	21:56
	→ picardo12 joined	21:56
malev	malev: fasda	21:56
	← Reticenti left	21:57
	← Jaxan left	21:59
airborn	hello	22:00
	→ khmarbaise_ joined	22:00
malev	hello	22:00
	drizzd__ → drizzd	22:00
	← hipe left	22:00
	← Blackguard left	22:00
	← khmarbaise left	22:00
	khmarbaise_ → khmarbaise	22:00
	→ hipe joined	22:00
	→ ping-pong joined	22:02
	← p3rg2z left	22:02
	yaotti → yaotti_aw	22:03
	← unix_remote left	22:03
	← hipe left	22:05
	← q0rban left	22:05
	← zenazn_ left	22:05
	← RORgasm left	22:05
	→ hsa2 joined	22:05
	← fod left	22:06
hsa2	hello	22:06
	i installed "Full installer for official Git 1.7.0.2" on windows, but i can't make git recognize my ssh keys, where should i put them?	22:07
	← jpieper left	22:07
	← ChrisBuchholz left	22:07
	← malev left	22:07
	→ altrux_ joined	22:08
	→ jpieper joined	22:08
	→ madewokherd joined	22:08
thiago	hsa2: Git doesn't deal with ssh keys, ssh does	22:09
	→ zenazn joined	22:10
	→ lresende joined	22:10
	← lresende left	22:10
hsa2	thiago: i know, but ssh is included in windows package of git, i din't install openssh	22:10
thiago	ah, I see. So the ssh won't find your keys	22:10
hsa2	thiago: yes	22:10
	exactly :)	22:11
thiago	I don't know what ssh considers as $HOME on Windows, sorry	22:11
hsa2	windows is great, but i hate it's being gui dependent	22:11
	→ malev joined	22:11
	← altrux_ left	22:12
	← malev left	22:13
hsa2	thiago: i think i found it C:\Users\<username> works :)	22:13
	→ Nemurenai joined	22:13
	← tvw left	22:14
	→ gorozco joined	22:15
mstormo	hsa2: You can override by actually setting HOME to something	22:17
	normally uses USERPROFILE or HOMEDRIVE+HOMEPATH or HOME	22:19
	← paradoxx left	22:19
	← poseid left	22:19
	→ [reaVer] joined	22:21
	← bin1010 left	22:21
	→ engla joined	22:22
	← leo2007 left	22:23
	→ unclown joined	22:23
	← engla left	22:23
	→ engla joined	22:23
	← reaVer left	22:23
	← akahn left	22:24
	← whatdoesitwant left	22:24
	← notjohn left	22:25
	→ MacHat joined	22:25
	← resmo left	22:27
	← chittoor left	22:27
	→ Jippi_mac joined	22:28
	→ scarabx joined	22:28
	← _Vi left	22:29
	→ mar77i joined	22:31
	← beatak left	22:31
	← redondos left	22:31
	→ redondos joined	22:31
	← phlowx left	22:32
	hsa2 → hsa2\|gone	22:32
	→ ProperN[out] joined	22:33
	← ProperN[out] left	22:33
	→ ProperN[out] joined	22:33
	← amerine left	22:33
	hsa2\|gone → hsa2	22:34
	→ xiong joined	22:34
	→ batok joined	22:35
	← loincloth left	22:35
	← ProperNoun left	22:36
	ProperN[out] → ProperNoun	22:36
	← galderz left	22:36
	slyphon → slyphon_away	22:37
	slyphon_away → slyphon	22:37
	← kumbayo_ left	22:38
	← ping-pong left	22:38
	→ ping-pong joined	22:39
	→ malev joined	22:40
	→ Aaaarg joined	22:40
	andreaa`afk → andreaa	22:42
	← Nemurenai left	22:43
	← sagsousuke left	22:44
	← Jippi_mac left	22:45
	→ hilem joined	22:45
	← blinky left	22:46
	← rudi_s left	22:46
	← metze left	22:46
	← kakos left	22:46
	→ hipe joined	22:46
	← Helios left	22:46
	← khmarbaise left	22:47
	→ mjf joined	22:48
	← elmob left	22:49
	← coderdad_ left	22:49
	← Yuuhi left	22:50
	← Ryback_ left	22:50
	→ metze joined	22:50
	→ rudi_s joined	22:50
	← candybar left	22:51
	→ Helios joined	22:51
	→ blinky joined	22:51
	← Dave^\| left	22:52
	hsa2 → hsa2\|gone	22:52
	→ candybar joined	22:52
	← MacHat left	22:53
	→ jksz joined	22:53
	← jksz left	22:53
	← terinjokes left	22:55
	→ ajpiano joined	22:56
	← mjf left	22:57
	→ toddr joined	22:57
	hsa2\|gone → hsa2	22:57
	→ vinse joined	22:58
	→ MacHat joined	22:59
	→ jfkw joined	23:00
	← Heimidal left	23:02
	← toddr left	23:02
	→ freshtonic joined	23:03
	← tizbac left	23:06
	← Theravadan left	23:08
	← Eaven left	23:09
	→ coderdad joined	23:10
	→ guide_ joined	23:10
	← malev left	23:10
	← MacHat left	23:10
	← metze left	23:10
	← Helios left	23:10
	← guide left	23:11
	hsa2 → hsa2\|gone	23:11
	← hipe left	23:12
	← guide_ left	23:14
ribasushi	is there some command or whatnot	23:14
	→ guide joined	23:14
thiago	to do what?	23:14
ribasushi	that will fire my $EDITOR for evey conflict file available?	23:14
	→ Helios joined	23:14
	← MattDiPasquale left	23:15
	← sammcd left	23:15
	← samer left	23:15
	→ Heimidal joined	23:16
	← engla left	23:16
	→ metze joined	23:16
	→ gnoel joined	23:17
	→ slonopotamus joined	23:17
	→ engla joined	23:18
	← engla left	23:18
	→ engla joined	23:18
	→ paradoxx joined	23:19
gnoel	hi folks, how does one figure out from which remote branch a local branch was created?	23:22
vmiklos	git config branch.foo.remote and git config branch.foo.merge	23:23
	← paradoxx left	23:24
	← elitexray left	23:27
	→ Theravadan joined	23:27
LeLutin	does anyone know something about remote helpers? I'm trying to pull out a bzr helper and there's some details I can't quite figure out..	23:27
thiago	ribasushi: git mergetool	23:28
	← WebDragon left	23:31
	← pantsman left	23:32
	← brntbeer_ left	23:33
	→ brntbeer joined	23:33
	← lemonchicken left	23:35
	→ paradoxx joined	23:35
	← xiong left	23:35
	← Heimidal left	23:36
	→ Nightsorrow joined	23:37
	→ Fangorn_ joined	23:38
	← dr_win left	23:40
	← tureba left	23:41
	→ dr_win joined	23:41
	→ tureba joined	23:42
	← andreaa left	23:44
	← jrmuizel left	23:48
	← tjholowaychuk left	23:50
	→ lemonchicken joined	23:52
	← mbroeker left	23:52
	← LongBeach left	23:53
	hsa2\|gone → hsa2	23:53
	← hsa2 left	23:54
	→ chapter_29 joined	23:55
	← slackstation left	23:55
	← masonkramer left	23:55
	→ masonkramer joined	23:56
	→ yngress joined	23:56
	← coderdad left	23:58
	← alice\|wl left	23:59

Logs Search ←Prev date Next date→

Channels Documentation