IRCloggy #git 2008-12-29

Provider of IRC logs since 2005.

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

	2008-12-29
	ceej-away → ceej	00:02
	→ cgardner joined	00:03
	← radarek left	00:04
	← zerok_ left	00:05
	→ sverrej joined	00:05
	ceej → ceej-away	00:16
	→ quoin joined	00:18
	← quoin left	00:18
	→ Lil-WaynE joined	00:20
Lil-WaynE	LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL	00:21
	LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL	00:21
	LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL	00:21
	LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL	00:21
	LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL	00:21
	LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL	00:21
	LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL	00:21
	LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL	00:21
	LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL	00:21
	LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL	00:21
	LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL	00:21
	LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL	00:21
	← Lil-WaynE left	00:21
	→ Lil-WaynE joined	00:22
	← Lil-WaynE left	00:22
	→ ricky joined	00:24
	→ quoin joined	00:24
	→ eruijters joined	00:25
	← ashleyw_ left	00:26
	← _grepper left	00:29
	→ qrush joined	00:32
	← DrFrasierCrane left	00:34
	← zero__ left	00:37
	→ joshdavey joined	00:40
	→ mindware joined	00:42
	→ jackdempsey joined	00:43
	← jackdempsey left	00:44
	← sdboyer left	00:49
	← Yuuhi left	00:53
	→ alezandro joined	00:54
	→ dxdt joined	01:01
dxdt	I've never done any real sorta rsynching or anything like that and while I was just reading up on Rsync for the first time, it occurred to me that for the simple synchronizing of files, Git might work or be better. Umm any thoughts for the more experienced?	01:02
Wyk3d	"fatal: Unable to write new index file" - i really am trying to write a new index file, i copied .git/index to a new file and set GIT_INDEX_FILE to it, but it works for me and not for someone else, what could cause this ?	01:03
j_engelh	dxdt: git:// is faster anyway	01:03
	← chris2 left	01:04
drewr	dxdt: It depends on the nature of the content. Both have merits.	01:04
	← dmq left	01:05
dxdt	drewr: It would be for synching stuff between my desktop and laptop when I wanted. Probably more binary files than anything, but there would be some source mixed in.	01:06
	← tpiep left	01:06
drewr	I would lean toward rsync there. I've used git for that type of thing and it's generally more hassle than it's worth.	01:06
	With git you have to add/remove files manually, commit changes, etc. It's a pain.	01:07
dxdt	that's a good point.	01:07
	They both send only the diff of files, right?	01:07
drewr	I'm not sure if rsync only sends the diff or not, but it definitely only sends files if they've changed.	01:08
Ilari	dxdt: Rsync only sends the diff... And it is likely faster in computing those diffs than Git...	01:08
	← scientes left	01:09
drewr	And rsync3's comparison algorithm is screaming fast for trees with hundreds of thousands of nodes.	01:09
dxdt	interesting. It sounds like Rsync is what I should use for now then. Thanks!	01:12
	→ sword joined	01:13
	→ tpiep joined	01:17
	← k776 left	01:19
	← tpiep left	01:21
sword	Here's an easy one: I'd like for git log to show my lightweight tags in the summary line (like git-instaweb does) but the only way I seem to be able to access them is to git log <tag> individually, which shows me where they are, but only with one tag at a time. No dice in manpages or the Google so far.	01:23
	← ebzzry_ left	01:23
	→ jmspeex joined	01:24
Ilari	sword: 'git log --decorate'?	01:25
	→ sdboyer joined	01:26
	sstangl → sstangl--	01:26
sword	Ilari: Indeed -- so, I should have searched for ref instead of tag in the manpage.	01:27
	sstangl-- → sstangl	01:27
	sstangl → DeMarko--	01:28
	DeMarko-- → sstangl	01:28
	← Tr3B left	01:28
	← _slacker_ left	01:32
	← aziz left	01:50
	← parasti left	01:52
j_engelh	rtyler: I attest you severe memory loss.	01:52
	rtyler buys j_engelh an 'r'	01:53
rtyler	I knew you looked familiar	01:53
j_engelh	.oO( where should I stuff that 'r')	01:53
rtyler	your*?	01:53
j_engelh	no.	01:54
	← lolage0 left	01:54
j_engelh	I (who?) do attest (action) you (whom?) memory loss (what?)	01:55
danderson	the you thrown in does make the sentence parse a little surprisingly	01:56
rtyler	heh	01:57
	← tanek left	01:57
danderson	well, the brain expects to parse 'your memory loss' as one unit	01:57
j_engelh	hm maybe it is due to translation from german	01:57
	like, I attest you sth = I hereby state that you happen to have xyz. whereas I attest your sth = It is correct what you presumed to have.	01:58
	(med.)	01:58
danderson	hmm.	01:58
	← bryanl left	01:58
	← metajack left	01:59
	→ bryanl joined	01:59
danderson	I read that as 'attest you xyz' = I certify to you that xyz	01:59
	← cgardner left	01:59
	→ metajack joined	01:59
rtyler	j_engelh: sag es mal auf Deutsch?	02:00
j_engelh	Ich attestiere dir Ged"achtnisschwund. (Vgl. Ich attestiere dir deinen Ged"achtnisschwund.)	02:01
	danderson: yeah	02:01
rtyler	that parses far better for me than the english version	02:01
	odd	02:01
j_engelh	rtyler: but it's how it works for me :p	02:01
	→ tanek joined	02:02
danderson	yeah, the german parses better for me too	02:02
	(which is weird, since I have only survival-level skills in german)	02:02
j_engelh	survival ha	02:02
	note to self: iron curtain gone since 18 years.	02:02
danderson	survival skills: "ein Bier bitte" & "Prost" :P	02:03
rtyler	danderson: "noch ein Doener bitte"	02:03
	that's all you really need	02:03
j_engelh	thankfully	02:03
	you can easily use an alternate language (like English in Germany). Too bad it is not true for some - 1st place actually: america. not far from it: hungary.	02:04
	← tanek left	02:05
j_engelh	and france...	02:05
	→ k776 joined	02:05
	→ oudeis_ joined	02:05
danderson	j_engelh: outside Paris, you can actually get along rather well with English	02:06
	even in Paris, except that people are rude to you if you try.	02:06
j_engelh	yeah but screw it	02:07
	danderson grew up in Alsace, where French and German are both fairly common	02:07
j_engelh	I'm more like "your English sucks so much, so it's better for the both of us if we speak French"	02:07
	what was it, my french is better than your english or so	02:07
	→ patrick_ joined	02:09
	→ tpiep joined	02:11
tpiep	Is it possible to include an SVN repository inside a Git repository? Like a submodule, but for a 3rd party repo that isn't in Git.	02:12
	← oudeis left	02:12
j_engelh	tpiep: SVN ALERT. All hands battle stations!	02:13
tpiep	j_engelh: :) I'd prefer to just use submodules, but I can't wait for the rest of the world to move on from SVN.	02:13
rtyler	tpiep: just use separate git repositories and git remote	02:14
	git-svn(1) ftw	02:14
andres	tpiep: use git-svn in the submodule?	02:14
tpiep	andres: I didn't know you could do that.	02:14
	rtyler: What do you mean separate git repositories? I'm talking about other people's code that's in SVN. Are you saying create local git-svn clones and use those local clones as submodules?	02:17
Arrowmaster	yes	02:18
	← joshdavey left	02:23
	← patrick232 left	02:26
	← kukks left	02:29
	→ carllerche_ joined	02:35
tpiep	Arrowmaster: Interesting. I hate to have the repository dependent on the setup of my local machine, though.	02:36
	← drewr left	02:43
	← sword left	02:46
	→ atourino joined	03:03
	← eruijters left	03:07
	→ xinming joined	03:10
xinming	WHen I stash a change, after a while, I do git stash pop, And found conflict, After I've done editing, How do I make git stash pop to succeed please?	03:10
	After I edit the content, Every time I do 'git stash pop' there will be conflict	03:11
	← pygi left	03:11
	← mbroeker left	03:11
j_engelh	after done editing, you commit it	03:11
davidfetter	nin hao, xinming :)	03:11
j_engelh	nin?	03:11
	twt?	03:12
xinming	davidfetter: Ni hao. :-)	03:12
	j_engelh: either nin and ni is ok	03:12
davidfetter	more formal iirc	03:12
	← bryanl left	03:12
xinming	davidfetter: ni is commonly used as you, and nin is normally used in writing for the respect of the person.	03:13
davidfetter	dui dui dui ;)	03:13
j_engelh	hai hai	03:14
xinming	...	03:14
	bremner respectfully points out japan != china	03:14
offby1	wait. They're not the same?	03:15
j_engelh	no :p	03:15
davidfetter	lol!	03:15
	they're different in a very important way (to me)	03:15
ari-_-e	what are japan and china?	03:15
davidfetter	i've been to china and not japan ;)	03:15
j_engelh	ari-_-e: we're not playing Jeopardy yet.	03:15
ari-_-e	aww, darn	03:15
davidfetter	anybody see slumdog millionaire?	03:16
offby1	yep	03:16
	++	03:16
	I recommend it	03:16
	davidfetter too	03:16
j_engelh	"<Please do not ask if anyone uses, knows, is good with, can help you with or is a guru/expert of <someprogram>. Instead, ask your real question.">	03:16
	offby1 invites davidfetter to ##cinema	03:16
j_engelh	screw it, compose broken again	03:16
offby1	:)	03:16
xinming	offby1: ni and nin are almost the same. The slight difference is ni is used for people talk to people. But nin is used in Eg: a customer service talk to a customer.	03:17
j_engelh	The Knights Who Say 'Ni'!	03:17
	#git has them.	03:17
	← k776 left	03:18
xinming	Ok, It's #git, Not #chinese. ;-)	03:18
	let's stop	03:18
j_engelh	but there is nothing else to do	03:19
xinming	When I stash a change, after a while, I do git stash pop, And found conflict, After I've done editing, How do I make git stash pop to succeed please?	03:20
bremner	j_engelh: please finish subtree checkouts	03:20
	xinming: same as a merge conflict?	03:20
	→ pygi joined	03:21
bremner	xinming: i.e., edit, add, commit. Or do I misunderstand the question?	03:21
j_engelh	bremner: why don't you give me a start?	03:21
	← mgrimes left	03:22
bremner	j_engelh: you were the one complaining of nothing to do, just trying to help	03:22
	← atourino left	03:22
	davidfetter wonders whether "dui dui dui" is common outside beijing	03:22
xinming	bremner: Yes, something like, or a bit like 'git rebase', If there is a conflict in 'git rebase', We just need to 'git add' and 'git rebase --continue'	03:23
	bremner: But for 'git stash pop', If there is conflict, the pop won't succeed. After done merge, Do I have to remove that stash myself?	03:24
bremner	xinming: oh, I get it. Yes, I expect so. drop ? is it called	03:24
trupheenix	offby1: hey ...	03:24
j_engelh	bremner: I was not complaining. It was a fact.	03:24
trupheenix	offby1: i still get that errie problem i slept off	03:24
	→ jim_c joined	03:25
	→ joshdavey joined	03:26
offby1	trupheenix: what version of git are you using? The only thing I can think of is that your version is kinda old	03:31
	if it's older than v1.5.3.6, that -might- explain it.	03:31
trupheenix	offby1: it's v1.5.4.3	03:31
jim_c	Does anyone recommend (or not recommend) the Pragmatic Programmers Git book? http://www.pragprog.com/titles/tsgit/pragmatic-version-control-using-git	03:32
offby1	haven't read it, but I'd be pathetically grateful for a good book about Git	03:32
	trupheenix: nuts.	03:32
trupheenix	offby1: i dunno i'm confused and kinda jittery to hand over the repo to the team	03:32
offby1	trupheenix: did you see the paste I made a few hours ago? I did _exactly_ what you did, and got the opposite result you got :-\|	03:33
	I _could_ try again with your version of git.	03:33
	if you think that'd be worthwhile.	03:33
trupheenix	offby1: hmmmmmm	03:34
offby1	act now, because I'm drinking MacAllan's pretty fast :-\|	03:35
	(I'm probably already past the point of usefulness, but you never know)	03:35
trupheenix	offby1: let me try out on another server	03:37
offby1	sure	03:37
trupheenix	offby1: yepp... it works correctly on the server	03:37
offby1	hmph	03:38
	what version of "git" is on that box?	03:38
trupheenix	offby1: we r using version 1.5.6.3	03:38
offby1	hmph	03:38
	maybe there was a bug fix before that version, but after v1.5.4.3, that was pertinent.	03:38
trupheenix	offby1: i think there's a prblem with v 1.5.4.3 on Ubuntu Hardy Heron	03:38
offby1	are you willing to try building a newer version? (It's pretty easy)	03:38
trupheenix	offby1: sure i'm willing to put on a newer version	03:39
offby1	I'm curious.	03:39
	→ eldok joined	03:39
trupheenix	offby1: i'm curious... how do i upgrade?	03:40
	← alikins left	03:41
offby1	oh, just clone the master "git" repository and build. If you're lacking certain libraries, you'll need to pass a special flag to "make" when you invoke it, but it's not hard.	03:42
	I'll give you the URL if you want	03:42
	anyone on this channel can too	03:42
	← LiamH left	03:42
offby1	"git clone git://git.kernel.org/pub/scm/git/git.git"	03:42
	there ya go	03:42
	if the build fails, paste the error and I'll tell ya how to get past it	03:42
	→ alikins joined	03:44
	← markelikalderon left	03:45
j_engelh	installadistropackageinstalladistropackageinstalladistropackageinstalladistropackage	03:45
	← eldok left	03:46
	← jmspeex left	03:46
ari-_-e	lies	03:46
	→ Flathead joined	03:46
	← BugeyeD_ left	03:49
	→ BugeyeD joined	03:49
	→ markelikalderon joined	03:50
	→ k776 joined	03:51
	→ gasull joined	03:53
	← gasull left	03:53
	← pygi left	03:53
	→ gasull joined	03:55
gasull	Hi. I have my git assigned to no branch. How can I assign it to master? Thanks	03:55
	I tried git-branch master and it tries to create it	03:56
Ilari	gasull: Maybe 'git checkout master', 'git merge HEAD@{1}'?	03:57
gasull	Ilari: Thanks. 'git checkout master' worked	03:58
Ilari	gasull: It might have lost some commits (if you committed them to "(no branch)")..	03:59
kevlarman	git lets you commit without a branch?	03:59
	→ elmex_ joined	04:01
jim_c	llari: Having been bitten by the no branch problem yesterday, I still consider it a bug. Software shouldn't lose data (even if losing in this case means making it inaccessbielf or 90 days before tossing it.)	04:02
	← gasull left	04:03
	← greyface left	04:07
	← markelikalderon left	04:10
	→ rellik_ joined	04:10
offby1	j_engelh: well, that too :)	04:12
	j_engelh: but I assumed his distro package was kind of old.	04:12
	kevlarman: I think it does. Try It And See™.	04:12
rellik_	I'm fairly new to git, and am having a frustrating problem.. I have a repo with (among other things) a few files around 100M... When I try to clone this repo, either 1) the compression takes forever, or 2) the server crashes with OOM errors and has to be restarted. In my research, all I can find is a client-side config (core.compression), but I can't find anything server-side to lessen/disable compression for this repo.. an	04:14
	ybody able to help or point me in the right direction?	04:14
ToxicFrog	I've got nothing, sorry	04:14
offby1	me needer :-\|	04:15
	how much RAM on the server?	04:15
rellik_	the server is a slicehost VPS, with around 100M free ram	04:15
offby1	any swap?	04:15
	(that's not much RAM)	04:15
rellik_	yeah, plenty of swap (500M+)	04:16
offby1	hmph	04:16
	nothing comes to mind	04:16
rellik_	I'd hope 100M of ram would far exceed the requirements of a SCM	04:16
	→ akitada__ joined	04:16
ToxicFrog	Yeah	04:16
	Likewise the compressor, even bzip2 doesn't use _that_ much memory	04:16
	← akitada__ left	04:17
	→ akitada__ joined	04:17
	← elmex left	04:17
	elmex_ → elmex	04:17
akitada__	what us the git-core on download page?	04:18
offby1	akitada__: say that again?	04:19
akitada__	is*	04:19
	offby1: git-core-xxxxtar.gz	04:21
	← flaguy left	04:21
akitada__	on pub/software/scm/git	04:24
	← Fullmoon left	04:25
	→ ben_h joined	04:27
offby1	I assume that's the source. I don't know for sure though.	04:30
	← akitada__ left	04:31
	→ Fullmoon joined	04:35
	← ari-_-e left	04:35
	→ ari-_-e joined	04:38
	→ bergwolf joined	04:38
bergwolf	is there any way to figure out which commit modifies a specific single line of a file?	04:39
deskin	bergwolf: yes, man git-blame	04:39
Gitbot	bergwolf: the git-blame manpage can be found at http://git.or.cz/man/git-blame	04:39
bergwolf	thanks, deskin, Gitbot	04:39
	← ben_h left	04:40
offby1	Gitbot is a friendly fella	04:42
	→ aspotashev joined	04:43
	→ rhett` joined	04:45
rhett`	I have a file that I want git to ignore: config/database.yml	04:45
	i've tried every combination of putting that into .gitignore files that i can think of	04:45
	added and committed the gitignore files	04:46
	→ seangrove joined	04:46
	← trupheenix left	04:46
rhett`	i can't get git to ignore that file	04:46
deskin	rhett`: is git already tracking this file, perhaps by accident? If so you need to git rm --cached config/database.yml	04:46
rhett`	ahh	04:46
	ok, thanks	04:46
deskin	and then commit, of course	04:47
offby1	damn. Same thing came up on #svn, over and over	04:47
	second time today it's come up here.	04:47
deskin	ugh, you monitor #svn? ;)	04:47
offby1	used to	04:49
	used to live there, practically.	04:49
deskin	I'm afraid to join #svn--- for one, I'd have little practical advice for anyone, and for two, I'd probably answer everything by saying 'switch to git'	04:51
offby1	:)	04:51
	I gotta say: that channel was, hands down, the most useful channel I'd run into.	04:52
	kind of a shame, really: the people behind Subversion are first-class all the way.	04:52
	It's just that ... what with the working copies still getting hosed, and the six-year-old merge bugs ... I no longer trust Subversion to work without trouble.	04:53
	Whereas git, even though it's been kind of hard to learn, has never once done anything I'd describe as a bug.	04:53
deskin	yeah; I gave a presentation to my coworkers on git-svn (we're a svn-based shop); when they asked why to use it if it has so many rough spots, my reply was basically that no matter what, git won't lose your data, even if everything else goes wrong	04:55
offby1	deskin: interesting question; I'm _this close_ to giving a similar presentation	04:59
	← statim left	05:01
	→ statim joined	05:01
	← seangrove left	05:03
	→ seangrove joined	05:03
	→ psankar joined	05:04
	← jim_c left	05:10
	← lene left	05:11
	sdboyer → lyricnzs_wife	05:11
	lyricnzs_wife → sdboyer	05:11
	→ girishr joined	05:16
	← andreaja_ left	05:16
	→ andreaja joined	05:18
	← ari-_-e left	05:19
	← tjafk1 left	05:27
	→ tjafk2 joined	05:27
	→ foutrelis joined	05:29
	← carllerche_ left	05:32
	ceej-away → ceej	05:47
	← bdiego left	05:50
	→ girishr_ joined	05:51
	→ zwu_ joined	05:51
	→ patelc75 joined	05:57
patelc75	git submodule add git://github.com/jasherai/acts-as-rated.git /vendor/plugins/ fatal: could not create leading directories of '/vendor/plugins': Permission denied Clone of 'git://github.com/jasherai/acts-as-rated.git' into submodule path '/vendor/plugins' failed	05:57
	any ideas?	05:57
	→ plediii_ joined	05:58
deskin	patelc75: the leading slash on /vendor/plugins looks troublesome to me	05:58
patelc75	thanks deskin! that did the trick	05:59
deskin	also, if I'm reading things correctly, you probably want vendor/plugins/acts-as-rated or something, but that's just preference	05:59
	→ DrFrasierCrane joined	06:05
	→ carrus85 joined	06:05
	← plediii left	06:05
	→ cedricv joined	06:11
	→ QaDeS joined	06:11
	→ PloenK joined	06:11
	← bergwolf left	06:13
	← girishr left	06:13
	ceej → ceej-away	06:19
	→ FH` joined	06:20
	→ akitada__ joined	06:20
	→ temoto-0x8a joined	06:21
	← akitada__ left	06:23
	→ billm joined	06:24
	→ eno__ joined	06:24
	← Jacolyte left	06:26
	→ robinr joined	06:34
	← temoto-mobi-alt left	06:35
	← FH` left	06:37
	→ Arafangion joined	06:39
	← priidu left	06:42
	→ priidu_ joined	06:42
	→ ebzzry joined	06:45
	← ebzzry left	06:46
	← dwmw2_gone left	06:49
	← girishr_ left	06:54
	zwu_ → zwu	06:56
	← Flathead left	06:58
	← Arafangion left	06:58
	← QaDeS left	07:01
	→ Flathead joined	07:01
	→ simplechat joined	07:02
	→ sammcd joined	07:12
	→ jackdempsey joined	07:12
	→ chuck joined	07:13
chuck	Is there any way to provide the ssh passphrase so that you don't have to enter it in while running git pull?	07:13
DrNick	use keys instead of passwords	07:14
	→ girishr_ joined	07:15
	← etienne9 left	07:21
sammcd	git push ?	07:24
	→ chelz joined	07:24
	← sammcd left	07:24
	← oudeis_ left	07:25
	← priidu_ left	07:27
	→ clairvy joined	07:30
chuck	DrNick, we are using keys, i mean an ssh key passphrase	07:34
DrNick	use the ssh agent	07:34
	← k776 left	07:34
	← patelc75 left	07:37
	← jackdempsey left	07:38
	→ ben_h joined	07:43
	← bobesponja left	07:45
	→ resmo joined	07:47
	← DrFrasierCrane left	07:49
	→ bentob0x joined	07:52
	→ rtomayko joined	07:53
	→ pencilo joined	07:53
	← aspotashev left	07:56
	→ reel_ joined	07:58
	→ zerok_ joined	08:01
	← Fullmoon left	08:04
	→ _slacker_ joined	08:06
	→ amerine joined	08:06
	← _slacker_ left	08:06
	→ _slacker_ joined	08:07
	→ ankit9 joined	08:07
	← _slacker_ left	08:07
	← clairvy left	08:08
	→ muthu_ joined	08:10
	← branstrom left	08:11
	→ Fullmoon joined	08:13
	← schlort left	08:15
	→ Lau_of_DK joined	08:19
Lau_of_DK	Gents - Ive just downloaded a project from a webserver (non-git) and I want to version-control everything, "git init", then what ?	08:19
amerine	quiet in here	08:24
resmo	Lau_of_DK: http://git.or.cz/gitwiki/QuickStart	08:24
	→ niarbtfel joined	08:25
	← amerine left	08:25
Lau_of_DK	resmo: Thanks big guy	08:25
resmo	Lau_of_DK: np	08:26
	→ gregor__ joined	08:35
	← CodeOfficer left	08:35
	→ temoto joined	08:36
gregor__	is it possible to see the total amout of data-size stored on a gitserver?	08:39
	→ dwmw2_gone joined	08:40
resmo	gregor__: du -h /var/cache/git ?	08:41
gregor__	resmo, no, via git, i want to know how big the repository is	08:42
	← tpiep left	08:48
	← ankit9 left	08:49
	→ ankit9 joined	08:49
	→ ph^ joined	08:50
	← rhett` left	08:51
gregor__	understand what i mean?	08:53
	→ kandjar joined	08:59
kandjar	hi there	08:59
	I m encountering some permission issue with git	09:00
	the problem I have is: I have the a bare repo on a linux server, 2 diff account accessing it, lets say" userA and userB	09:01
	when userA submit, he may create a subfolder inside objects	09:01
	from there, userB wont be able to submit anything which touch that folder	09:02
	because the right of the folder are : drwxr-xr-x 2 userA userA ...	09:02
	i have to change it to drwxrwxr-x userA git-user ...	09:03
	is there a way to fix the issue other than 'by hand' ?	09:03
andres	kandjar: suid directory	09:04
	kandjar: default ACLs	09:04
kandjar	suid directory?	09:04
andres	kandjar: different (pulling) workflow	09:05
kandjar	+s?	09:05
	← Aides left	09:05
	→ Aides joined	09:05
kandjar	ACLs??	09:05
andres	kandjar: suid on directory will overwrite the owner of the created files, sgid the group	09:06
	kandjar: well, POSIX ACLs. Whats the problem there?	09:07
gregor__	http://en.wikipedia.org/wiki/Access_control_list	09:07
kandjar	andres: the problem is for example if userA crease a directory A0 inside objects, the rights are all wrong! :)	09:08
	I have userA and userB which belong to the same group: git-user	09:08
	but userA created the folder A0 with drwxr-xr-x userA userA...	09:08
andres	kandjar: yea, I do understand that.	09:08
	kandjar: setgid on the directories will solve that.	09:09
kandjar	on objects?	09:09
	chmod g+s objects?	09:09
	→ sadmac joined	09:10
	→ Tr3B joined	09:11
	→ tpiep joined	09:11
andres	kandjar: on all directories.	09:12
kandjar	including the .git one?	09:12
gregor__	hot to get the size of gits on a server(through git)	09:12
kandjar	andres, gregor__ thanks	09:14
gregor__	i only posted a link?	09:15
	← sadmac left	09:16
	← simplechat left	09:18
	→ slainer68 joined	09:19
	← girishr_ left	09:21
	→ simplechat joined	09:21
	→ akitada__ joined	09:30
	← akitada__ left	09:31
	→ yakov joined	09:31
	→ akitada__ joined	09:31
	← tpiep left	09:33
	← gregor__ left	09:33
kandjar	andres: ok i understand now the concept of the g+s stuff; however the write access is still disabled by default	09:34
	← akitada__ left	09:36
	← chalky left	09:36
andres	kandjar: umask	09:40
kandjar	yeah but umask is global	09:40
	I wish I could do somethign local to that directory :)	09:41
andres	kandjar: core.sharedRepository	09:41
	This sets a appropriate umask afair.	09:41
kandjar	oh	09:41
	← ia left	09:42
andres	kandjar: man git-config	09:42
Gitbot	kandjar: the git-config manpage can be found at http://git.or.cz/man/git-config	09:42
kandjar	and u can set that for a bare repo too?	09:42
	thanks!	09:42
	:)	09:42
	→ ia joined	09:42
	→ chalky joined	09:43
	← zerok_ left	09:52
	← Flathead left	09:54
	→ simosx joined	09:59
	→ _slacker_ joined	10:02
	← simosx left	10:05
	← gambler__ left	10:07
	→ lholden joined	10:08
	→ patrick232 joined	10:09
	← patrick_ left	10:12
xenoterracide	does git no lonager have files like git-init on system?	10:16
	s/lonager/longer	10:16
	just checking my system and I don't have binaries like that anymore	10:17
cehteh	reading the release notes (months ago) would solve that mystery	10:18
	← lholden left	10:20
xenoterracide	cehteh: I'll take that as a yes	10:21
	← ashH left	10:22
	zwu → zwu\|gone	10:24
	→ killing-joke joined	10:24
drizzd	xenoterracide: most git-* binaries are installed in libexec/git-core, so that they are not in your PATH by default	10:24
xenoterracide	ah	10:25
	→ josephwilk joined	10:25
	← efaistos left	10:26
killing-joke	i want to try to place controls on checkouts (outside of git, with credentials). i have read http://www.kernel.org/pub/software/scm/git/docs/howto/using-merge-subtree.html	10:28
	is there a specific strategy i need to take to make each module be treated separately from the others?	10:29
	the master project has the most restricted access. some subprojects will be safe for almost any emloyee to view.	10:29
	how do i create repo(s) and segment the trees so that the less-credentialed cannot infer the existence of more restricted content?	10:30
	← Lau_of_DK left	10:34
	→ peritus_ joined	10:37
	→ oudeis joined	10:38
	← peritus_ left	10:40
	→ masak joined	10:40
masak	I recently did 'git co HEAD^' in a repo, and since then I'va made a couple of commits. now it turns out that I'm 'not on any branch' according to git. how do I re-declere the current HEAD as 'master'?	10:43
	→ d0k_ joined	10:43
	→ LotR joined	10:44
	→ ashleyw joined	10:44
	→ ashH joined	10:44
drizzd	masak: when you did "git checkout HEAD^", you "detached HEAD"	10:44
masak	drizzd: ok.	10:45
drizzd	that's why you are not on any branch	10:45
masak	I need to re-attach my HEAD, then.	10:45
drizzd	you can create a new branch at your current HEAD using "git branch <branch>"	10:45
masak	yes.	10:45
drizzd	actually, "git checkout -b <branch>" is probably what you want	10:45
	(it does "git branch <branch>; git checkout <branch>")	10:45
masak	even though 'master' already exists as a branch?	10:45
drizzd	well, if you want to use the name master again, you either have to rename or delete the old one	10:46
	you can do that using "git branch -m" or "git branch -D", respectively	10:46
masak	drizzd: thank you.	10:47
	← kandjar left	10:49
	← d0k left	10:50
	← ashH left	10:50
drizzd	is there any documentation for merge-recursive in git.git?	10:51
Pieter	yes	10:52
	<HEAD:Documentation/merge-strategies.txt>	10:52
Gitbot	[git merge-strategies.txt]: http://tinyurl.com/6sqzxz [blob]	10:52
drizzd	Pieter: thank you. I was looking for something more detailed, like this: http://article.gmane.org/gmane.comp.version-control.git/11705/match=recursive	10:53
	the above does not explain how multiple common ancestors are recursed, however	10:53
	← oudeis left	10:54
	← foutrelis left	10:54
	→ gambler joined	10:54
	← killing-joke left	10:55
	→ foutrelis joined	10:56
	→ pygi joined	10:59
	→ scientes joined	10:59
	→ oudeis joined	11:00
	→ kumbayo joined	11:06
	→ glommer joined	11:06
	← simplechat left	11:06
	→ rittyan joined	11:07
rittyan	hello. please give me that git workflows man link	11:07
	→ tom joined	11:10
	→ xjjk joined	11:11
tom	i've succeeded checking out the master branch of http://git.clutter-project.org/cgit.cgi?url=clutter/ using git clone git://git.clutter-project.org/clutter now how can i get the clutter-0-8 branch instead?	11:11
	i tried git://git.clutter-project.org/clutter-0-8 and git://git.clutter-project.org/clutter/clutter-0-8 but no luck though	11:11
drizzd	tom: git checkout clutter-0-8	11:13
tom	drizzd: in the checked out master directory?	11:14
Bombe	tom, yes.	11:14
tom	error: pathspec 'clutter-0-8' did not match any file(s) known to git.	11:15
Bombe	tom, with git you always get the complete repository.	11:15
Wyk3d	git checkout -b local_branch origin/clutter-0-8 ?	11:15
tom	Wyk3d: thanks. apparently that did the magic ;)	11:17
	← not-xjjk left	11:22
	← ashleyw left	11:23
	→ lorandi joined	11:24
	← oudeis left	11:28
	→ oudeis joined	11:29
	← muthu_ left	11:29
	← quoin left	11:33
rtyler	somebody needs to update /topic; 1.6.1 FTW!	11:36
	→ mbroeker joined	11:36
	→ Acry joined	11:36
	← josephwilk left	11:37
	→ DrFrasierCrane joined	11:39
	→ quoin joined	11:39
	← ankit9 left	11:40
	→ ankit9 joined	11:40
	← psankar left	11:43
	→ henux joined	11:43
henux	Hello, I have repo in github and I would like to give a write access to someone for it. Can this be done?	11:43
	→ simplechat joined	11:44
Wyk3d	henux: admin panel -> collaborators panel -> add another collaborator	11:46
	→ josephwilk joined	11:47
henux	thx	11:48
	← henux left	11:48
	→ florian_bravo joined	11:50
	← carrus85 left	11:54
	→ ashleyw joined	11:54
	→ peritus_ joined	11:54
	← danlarkin left	11:56
	→ chris2 joined	12:05
	← bentob0x left	12:07
	← oudeis left	12:09
	→ Yuuhi joined	12:11
	→ dmq joined	12:11
	← florian_bravo left	12:13
	→ henux joined	12:14
henux	what was the command again to edit the message of the last commit?	12:14
	git commit --amend?	12:15
	got it	12:16
	← henux left	12:16
	aptituz_ → aptituz	12:17
cbreak	it doesn't only change the message, it also changes the commit itself	12:17
	→ clairvy joined	12:22
	← peritus_ left	12:23
tom	what was the command?	12:25
	→ Sigma joined	12:25
	→ ashleyw_ joined	12:31
	← ashleyw_ left	12:34
	→ snitko joined	12:34
	← Fullmoon left	12:38
	→ \ask- joined	12:39
\ask-	hi -- any tips for rewriting log messages slightly (importing a repository from svn)	12:40
	I have log messages like	12:40
	r8370@embla: ask \| 2007-09-14 03:27:52 -0700	12:40
	CNAME support	12:40
	I'd like to put the first line last (or just remove it)	12:40
	← dmq left	12:41
	← rittyan left	12:41
	← chalky left	12:42
Ilari	\ask-: filter-branch? There's filter for changing commit messages...	12:42
	← unreal left	12:43
Ilari	\ask-: --msg-filter to be exact...	12:43
	→ unreal joined	12:43
	→ oudeis joined	12:46
	← ashleyw left	12:46
	→ parasti joined	12:47
	← Acry left	12:47
\ask-	Ilari: ah, perfect. I knew about filter-branch, but not --msg-filter - thanks.	12:49
	→ cannonball joined	12:49
	→ Lau_of_DK joined	12:50
Lau_of_DK	Gents, Ive got a Git repo here in my PC, which I want my colleague to clone, we're on the same network, how do I enable him to do so without handing out my login ?	12:51
	← oudeis left	12:51
Bombe	Lau_of_DK, make him give you his login and push it to his machine. :)	12:51
Lau_of_DK	Ok, other suggestions? :)	12:52
	Im thinking, can I just add a user to my linux box and set up a deamon for that account or how is that handled?	12:52
xl0	Lau_of_DK: ssh is the easiest way.	12:52
bremner	Lau_of_DK: IMHO, the simplest is give him an account with git-shell as shell, an set permissions to group shared on the repo	12:56
Ilari	Lau_of_DK: Create new bare repo, allowing access to both your and his account, push what you want to it. Then he can clone it via ssh...	12:56
bremner	err, right. what he said	12:57
	→ Loof joined	12:58
	← alezandro left	12:59
	→ charon joined	13:01
	→ patrikf joined	13:06
	→ ashleyw joined	13:07
	→ greyface joined	13:08
	→ CIA-9 joined	13:08
	→ kukks joined	13:09
	← masak left	13:09
	→ doener joined	13:17
	→ greyface_ joined	13:18
	→ muthu_ joined	13:24
	← greyface left	13:24
	← snitko left	13:25
	→ lottadot joined	13:27
	→ bdiego joined	13:29
	→ zerok joined	13:29
	← robinr left	13:31
	← mithraic left	13:31
	← pygi left	13:34
Lau_of_DK	Ok, question. Ive made a user 'shared' and put a repo in /home/shared/myrepo, added files, made the first commit. What do I tell my colleague to type in this commandline to clone this repo?	13:34
	→ drewr joined	13:34
	→ aspotashev joined	13:34
	← greyface_ left	13:37
Ilari	Lau_of_DK: Added files and made first commit? That doesn't sound right.	13:39
	Lau_of_DK: Shared repos are usually populated by pushing and the do not have working trees.	13:43
	← rtomayko left	13:43
	→ akitada__ joined	13:45
	← akitada__ left	13:46
	→ akitada__ joined	13:46
	← akitada__ left	13:47
	→ akitada__ joined	13:48
akitada__	Topic here is a little old	13:49
	→ priidu_ joined	13:49
	→ snitko joined	13:51
	→ jmspeex joined	13:51
akitada__	Is there any Linux distro provides gut 1.6 package?	13:52
bremner	akitada__: debian has it in experimental	13:52
	→ juanjoc joined	13:53
akitada__	git*	13:53
	Ah debian. Debian provides git as separate packages and it's nice	13:55
bremner	akitada__: but why do you care exactly?	13:56
akitada__	stash	13:56
bremner	akitada__: so you are going to switch distro instead of compiling git?	13:57
parasti	akitada__: stash was available pre-1.6, too...	13:57
akitada__	nope. I'm using a little different system and want to package git myself. so i was looking for an existing pkg	13:59
	→ mw-home joined	14:00
Tr3B	does git offer something similar to svn properties for individual files? e.g. svn:license	14:03
jast	Tr3B, see man gitattributes	14:03
Gitbot	jast: the gitattributes manpage can be found at http://git.or.cz/man/gitattributes	14:03
	← akitada__ left	14:03
jast	oh, yeah, bot highlight doesn't work with nicknames that have digits in them	14:04
mw-home	I made 3 consecutive commits on Friday on a branch AAA then quit work. Now this morning, I want to make a new branch named BBB, just like AAA. Then on AAA, I want to undo all of Friday's work. Should I first git branch BBB; git checkout AAA; git reset # last thursday's commit?	14:04
Tr3B	thx :)	14:04
jast	mw-home, depends on what you want to have as the end result. if you have BBB to have the three commits then yes.	14:04
charon	mw-home: you probably want to reset --hard however, since a default (--mixed) reset would leave the changes as unstaged in the working gree	14:05
jast	grr. if you want BBB to have...	14:05
charon	s/gree/tree/	14:05
parasti	mw-home: the short version of branch/checkout is git checkout -b $branch	14:05
jast	the checkout is not necessary here, in fact	14:05
	→ ironfroggy_ joined	14:05
jast	if you're on AAA right now	14:06
mw-home	I'm in AAA right now and BBB doesn't exist.	14:06
	so, first branch, then make sure I'm in AAA, then reset --hard?	14:06
parasti	mw-home: you can also just branch off the commit directly...	14:06
mw-home	parasti: huh?	14:06
parasti	mw-home: git checkout -b BBB $lasthursdayscommit	14:07
mw-home	parasti: oh, I want BBB to be what AAA is before I reset it.	14:07
parasti	mw-home: oh, hang on. I misunderstood what you wanted	14:07
	mw-home: yeah. ignore me then :)	14:08
ironfroggy_	I have a repo im trying to clone that always locks at 30%, but not on the same object. is there something deterministic that could cause that?	14:08
	← agib left	14:10
	→ pygi joined	14:12
Ilari	ironfroggy_: Does the client side still try to read data from network after it locks?	14:13
ironfroggy_	yes, it seems so	14:14
mw-home	OK, I reset my local repo. Now I want to push to a remote bare-repo, but (of course) my push was rejected. How do I reset on the remote repo?	14:14
ironfroggy_	actually, after i said that first message, it advanced to the same object number it was at before (still within 30%)	14:14
	but i didnt think one object could take this long	14:14
Ilari	ironfroggy_: If the client version is new enough to display transfer progress, how does the transferred count and average rate progress after it locks?	14:15
	← slainer68 left	14:15
ironfroggy_	well when i say "locks" i mean the object count stops going up. its a small repo, and clones in the past always happened quickly, which is why i found that odd. the rate still is changing, up and down, appearing normal.	14:16
Ilari	ironfroggy_: And the bytes transferred count is going up?	14:17
ironfroggy_	i dont have bytes transferred count, just objects transferred.	14:17
	→ akitada joined	14:17
ironfroggy_	ok i should have done this before, but i tried running du on it a few times and i see the repo growing, so its transfering something i guess.	14:18
Ilari	ironfroggy_: What git version (git --version)? Newer versions have added bytes counter and transfer rate display...	14:18
	ironfroggy_: From 'git fetch' of 1.6.1 (and it has been there for some version series): "Receiving objects: 100% (276/276), 166.54 KiB \| 56 KiB/s, done"...	14:19
	ironfroggy_: So it counted 166.54KiB transferred, at 56KiB per second...	14:19
ironfroggy_	oh, i dont know how i didnt see that. it didnt realize it was a bytes transfered. yeah, its going up.	14:20
	its early	14:20
	and this is my first day back on code after two weeks off, so im allowed a stupid mistake or twelve this morning..	14:20
Ilari	ironfroggy_: Then its receiving something... Likely something very large...	14:20
	← tom left	14:20
	← Lau_of_DK left	14:21
ironfroggy_	ill have to look later then because there shouldn't be anything so large in this repo..	14:21
mw-home	I think I will need to revert, not reset my changes, since I'm working with a local and a remote repo.	14:21
ironfroggy_	but i might also have a renegade coder on my hands.	14:21
	→ markelikalderon joined	14:23
Ilari	ironfroggy_: "git rev-list --objects --all \| cut -d \ -f 1 \| git cat-file --batch-check" displays object name, object type and size for all objects that are in history (except tags)...	14:24
ironfroggy_	oh, thanks.	14:25
	→ slainer68 joined	14:25
ironfroggy_	if only it will finish, i think this one object is up to 550M already.	14:25
	← quoin left	14:26
mw-home	Do a lot of people use bare repositories as a way to share code?	14:26
Ilari	ironfroggy_: "(git for-each-ref --format=%$objectname$ ; git rev-list --objects --all \| cut -d \ -f 1) \| git cat-file --batch-check" also includes tags...	14:26
ironfroggy_	i have a feeling its not a (now) 600M tag :-)	14:26
Ilari	mw-home: Yes.	14:26
ironfroggy_	mw-home: see: github	14:27
mw-home	ironfroggy_: after pushing to github, i tried to reset my local repo, then (of course) when I pushed to github again, i was rejected. so instead, do i need to revert my local stuff?	14:28
ironfroggy_	you can't change history (easily) if that history has spread to other repositories already.	14:29
	← akitada left	14:29
mw-home	ironfroggy_: so, what is recommended? reverting?	14:30
Ilari	ironfroggy_: To filter the listing for objects for those larger than 1MB, use something like 'egrep " [0-9]{7}"'.	14:30
	→ Strogg joined	14:31
	← pygi left	14:31
ironfroggy_	mw-home: i think so	14:31
	Ilari: thanks even more	14:32
	→ pygi joined	14:32
	→ AngryElf joined	14:32
	← markelikalderon left	14:34
	→ robinr joined	14:34
Ilari	ironfroggy_: After you have the object ID, grep for it in 'git rev-list --objects --all'....	14:35
	→ markelikalderon joined	14:36
	→ LiamH joined	14:37
Ilari	ironfroggy_: How large do you think its now?	14:37
ironfroggy_	about 650M	14:39
Ilari	ironfroggy_: CD ISO?	14:40
ironfroggy_	hah, have to check. and maybe bust a skull.	14:40
	waiting for the checkout now	14:40
mw-home	Somebody put a CD ISO into git?	14:40
robinr	why not? I tried to put a vmware image into CVS once...	14:41
Ilari	ironfroggy_: Or maybe first searching local working tree for large files could be faster...	14:41
	m3y3ring → meyering	14:42
	→ chalky joined	14:42
	→ kev joined	14:43
	→ mibbit_user joined	14:44
mibbit_user	hey guys, i'm trying to create a tracking branch for a secondary remote	14:45
	but i keep getting the fatal error message: fatal: Not a valid object name: 'remote_name/remote_branch'	14:45
	and ideas?	14:45
Ilari	mibbit_user: 'git fetch remote_name'.	14:45
	→ mithraic joined	14:46
	→ Voker57 joined	14:47
mibbit_user	sweetness thanks Ilari!	14:47
	→ akitada joined	14:47
Ilari	mibbit_user: git remote add also has the -f switch to run that immediately after creating the remote.	14:48
	← mibbit_user left	14:48
	← priidu_ left	14:49
kev	Hi! Is it possible to assign kdiff3 or some other GUI-based program as a handler for git-diff?	14:49
Voker57	--ext-diff	14:54
	→ Tr3B_ joined	14:55
	← Tr3B left	14:56
	→ spearce joined	14:57
kev	Thank you!	14:58
	→ sh1 joined	14:59
damog	you're welcome.	14:59
sh1	how do i see what are the last git command i have used on a repo?	15:00
ironfroggy_	turns out zipcode listings can be over a gig. thanks for the help, Ilari	15:04
	→ bobmcw joined	15:06
	→ alezandro joined	15:06
	← AngryElf left	15:07
	← akitada left	15:07
sh1	is there any verbose git pull?	15:08
	← zerok left	15:08
Rhonda	sh1: You mean different to the --verbose option?	15:10
	→ zerok joined	15:10
kev	They refer to GIT_EXTERNAL_DIFF, what does it mean? Is this an environment variable?	15:10
	→ cemerick joined	15:11
ironfroggy_	whats the syntax im missing for a diff from one commit to HEAD?	15:11
jast	kev, yes	15:11
Rhonda	kev: Yes	15:11
	→ peeja joined	15:11
jast	ironfroggy_, git diff <commit> HEAD	15:11
	→ coteyr joined	15:11
	→ zero joined	15:12
	zero → Guest75801	15:12
kev	Where can I read about its format and so on??	15:12
sh1	Rhonda: Yes	15:12
	any verbose option which tells me what it is doing	15:13
coteyr	ok i have an idiot question. I am trying to move from svn to git but I can't find out how to replicate one function. I have to repos a main code repo and a text file repo that just has some basic text files. I need the text files to be added to the main repo.	15:13
sh1	git pull is on for a long time	15:13
jast	kev, see man git	15:13
ironfroggy_	tahts what i thought, but im getting "ambiguous argument" for the commit...	15:13
jast	duh, the bot doesn't get that right	15:13
	→ geiseri joined	15:13
jast	kev, http://git.or.cz/man/git	15:13
ironfroggy_	ugh, missed a character. nevermind.	15:14
	← snitko left	15:14
Guest75801	hello. I'm having some issues trying to make git-svn work on my slackware 12.2 box. Thanks for any help in anticipation. Here's the error message: http://pastebin.com/m54ffcd4d	15:14
coteyr	what i have done so far is create the code repo. Thhen attempt to do a git pull --no-commit on the text file rpo into the code repo. This mostly works except there is a merge error. I just want to use the files from the text repo no matter what.	15:14
jast	Guest75801, what is it you want to do?	15:14
geiseri	hey is there a good git howto for subversion users? i found one at http://git.or.cz/course/svn.html but i am not sure i understand enough about git to make sense of it	15:14
jast	coteyr, do you want to use the text files repo in a style like svn:externals or just merge them in and forget about that other repo?	15:15
coteyr	merge and forget	15:15
	i sued to do an svn export to get them in	15:15
Guest75801	jast, I am trying to move a svn repository to git. I installed Alien::SVN from cpan (which I had a lot of trouble with too -been here yesterday too)	15:16
jast	Guest75801, you'll probably need to use the command "git svn clone"; there's an example and documentation in man git-svn	15:16
Gitbot	jast: the git-svn manpage can be found at http://git.or.cz/man/git-svn	15:16
jast	(there ^)	15:16
Guest75801	jast, any "git-svn" or "git svn" command I use brings up an error. The one you just mentioned brings up another error	15:17
	→ doumo joined	15:17
jast	which error?	15:17
Guest75801	jast, http://pastebin.com/m73c53e4a	15:18
kev	jast: thanks!	15:18
	→ withanx joined	15:18
Guest75801	jast, looks like a perl issue though	15:18
	← _slacker_ left	15:18
parasti	Guest75801: plain git svn doesn't do anything...	15:18
jast	Guest75801, at any rate, you need to give clone actual arguments, as explained in the manpage i gave you a link to	15:18
Guest75801	parasti, I know, I'm just trying to get rid of the errors	15:19
jast	though i guess git-svn should handle that sort of invocation more gracefully	15:19
	→ _slacker_ joined	15:19
	→ Cap_J_L_Picard joined	15:20
	← plediii_ left	15:21
EgS	hm... how can I undo a git-add?	15:21
parasti	EgS: read git status output	15:21
	← geiseri left	15:21
coteyr	jast, oh and I never want changes to the files that have been "exported" into the code repo to ever see the text repo again.	15:21
	I just want them to say with the code repo	15:22
	stay*	15:22
EgS	parasti: thanks... I felt really stupid as I didn't find it in the man pages :)	15:22
jast	the very easiest way, of course, is to manually copy the files from one working tree to the other	15:22
Guest75801	jast, here's an actual "live" example. http://pastebin.com/m29b6ce4a	15:22
parasti	Guest75801: what's your Git version? (might help)	15:22
coteyr	(i.e. the text files are templates and get modified for documentation reasons)	15:22
Guest75801	parasti, 1.5.5	15:22
coteyr	jast, except they are branched and versioned as well.	15:23
Guest75801	parasti, tried with 1.6.1 too	15:23
coteyr	jast, just once a pack of templates is assigned to a project it never changes.	15:23
	→ bryanl joined	15:23
parasti	Guest75801: hmm, Perl version?	15:23
jast	parasti, 5.8.8 :)	15:23
parasti	I'd prefer to see perl --version output or something like that	15:24
Guest75801	parasti, hmm, strange, I get this when I run perl --version: This is perl, v5.10.0 built for i486-linux-thread-multi	15:24
coteyr	jast, git pull is so close if I could just make it use 'theirs' instead of trying to merge. Or better yet merge but when a confilct comes up use stupid	15:24
	that might work	15:24
	→ xenoterracide_ joined	15:24
parasti	I wonder if the files for 5.8.8 conflict with the binary for 5.10.0	15:24
Guest75801	parasti, I think some dependencies where updated through cpan while I installed cpanplus and/or Alien:":SVN	15:25
	→ sammcd joined	15:25
Guest75801	parasti, not sure though	15:25
jast	coteyr, if using 1.6.1 you can easily fix the conflicts: git checkout --theirs <all conflicted files here>	15:25
	it's a bit less intuitive in older versions: git checkout :3:file1 :3:file2 etc.	15:26
	well, actually I think you can use git checkout MERGE_HEAD -- files	15:27
	→ agib joined	15:27
coteyr	jast, hmm. Theres just not a export command floating around somewhere is there. Lol.	15:27
	jast, Let me try the last one	15:28
jast	depends on what you want export to do	15:28
	i mean, in svn it just fetches the current revision of all files into a new directory	15:28
coteyr	just copy the files down and leave me alone	15:28
	lol	15:28
	no .git directory of course	15:28
jast	git clone and then rm -rf foo/.git should do the trick	15:28
	← bobmcw left	15:29
jast	unlike svn git only has one of those dirs per repository, after all	15:29
	→ bobmcw joined	15:30
coteyr	jast, not in an existing directory. Here is my old setup (and what I am trying to replicate). rails foo; svn import foo svn://repo/; svn co foo; cd foo; svn export --force repo_for_bar; svn add .........	15:31
	jast, using real commands of course but thats the idea	15:31
	→ xenoterracide__ joined	15:31
	→ jackdempsey joined	15:31
coteyr	jast, sad part it git pull is so close let me try the git checkout MERGE_HEAD and see what that does	15:32
	← Sigma left	15:32
	← ankit9 left	15:33
coteyr	jast, MERGE_HEAD trick worked.	15:33
jast	now just commit and be happy	15:34
	problem solved, world peace achieved	15:34
coteyr	jast, well I'll commit but I'm not promissing anything about being happy.	15:34
Guest75801	parasti, any suggestion on my issue? Should I clean up somehow and try again? though I'm not sure how to do it...	15:35
	← xenoterracide_ left	15:36
	→ aziz joined	15:37
coteyr	jast, thanks for the help	15:37
jast	oh well, i gotta go	15:37
	→ jm joined	15:37
	← coteyr left	15:38
	→ ckm joined	15:38
sh1	can git work within a firewall?	15:40
danderson	within, sure. It might have problems if you need to go through it though.	15:41
	if you just need clone/pull, you can do that over http, but it's rather slow	15:41
sh1	how do i do it via http?	15:42
parasti	Guest75801: it's some kind of Perl issue, for sure, but I don't know anything about Perl stuff, so I can't help you :(	15:42
Ilari	sh1: Also, git can use SSH (for authenticated ops only) as transport. The default git:// port is 9418 (for poking holes).	15:43
sh1	i want to pull the tree via http	15:43
	how do i do it?	15:43
	→ plediii joined	15:44
	← plediii left	15:44
vuf	sh1: git clone http://...	15:44
	→ eventualbuddha joined	15:44
	→ plediii joined	15:45
Ilari	sh1: Not all repositories that are publically accessable through git:// are accessable through http://. And even if they should be, http:// can be broken while git:// is not...	15:45
sh1	i have a "pure" git repo. i want to do a git pull through http protocol	15:45
	isnt it possible?	15:46
vuf	sh1: if it is your own repo, you have to set up an http server	15:46
	→ rtomayko joined	15:46
	→ abbe joined	15:47
Ilari	sh1: And keep caches about some stuff in that repo up-to-date (git-update-server-info).	15:47
	sh1: Oops, 'git update-server-info'.	15:47
Guest75801	parasti, well, no problem, thanks so far anyway :)	15:47
	→ Jacolyte joined	15:47
	→ camwest joined	15:48
Ilari	sh1: file://, git:// and ssh:// do not use those caches. Only stuff like http://, https:// and ftp:// does...	15:48
	sh1: Also, pushing over http(s):// is quite nasty (and requires WebDAV)... Push over ftp:// is not possible at all. File:// and ssh:// both support pushing nicely, while git:// has it disabled by default (for good reasons).	15:50
	← camwest left	15:50
	← ashleyw left	15:51
	← robinr left	15:51
	→ ashleyw joined	15:51
	← yakov left	15:54
charon	what's the point of passing ${1+"$@"} in sh? isn't that the same as "$@"?	15:54
	← mithro left	15:54
	← LotR left	15:55
charon	i'm asking because junio put such a construct in git-rebase.sh back in 9a111c91, and it survived to the current form in d70b4a8f, and i'm not entirely sure if i can just substitute $var "$@" after shifting out $1 to $var	15:55
Gitbot	[git 9a111c91]: http://tinyurl.com/8gg5ul -- rebase: allow a hook to refuse rebasing.	15:55
	← eno__ left	15:57
	← abbe left	15:57
	← doener left	15:58
charon	also, do we allow branch names that contain shell-dangerous characters?	15:59
	→ LotR joined	16:01
	→ eno joined	16:02
	→ doener joined	16:02
	→ tpiep joined	16:05
	← rtomayko left	16:08
	→ Vernius joined	16:08
	→ agile joined	16:08
	← agile left	16:09
	→ agile joined	16:09
	→ eruijters joined	16:09
withanx	if I have updated the master branch on a remote fork, how would I update one of the branches to merge with master?	16:10
	→ stephens joined	16:11
	← _slacker_ left	16:12
Ilari	withanx: 'git pull <remote-nick-or-URL> <branch>' or 'git merge master'?	16:13
	→ _slacker_ joined	16:14
	← andreaja left	16:14
Ilari	charon: The branch name rules are documented in man git-check-ref-format	16:15
Gitbot	charon: the git-check-ref-format manpage can be found at http://git.or.cz/man/git-check-ref-format	16:15
	← cedricv left	16:16
charon	Ilari: ok, thanks, i guess i should have found that myself. have to wonder about a branch "foo$(evil)bar" though.	16:17
withanx	Ilari: I have a clone of the fork on my local machine, master is up to date, but the branch is not. I tried pulling the branch into the master and pushing to the branch, but that fails. I also tried doing git checkout origin/branch_name and doing a merge, but that fails.	16:18
	← ashleyw left	16:18
	→ bts- joined	16:19
	← naoshige left	16:21
Ilari	withanx: Checkout that branch first (one of branches to merge is implicitly current branch) and only current branch is updted by merge.	16:21
withanx	what?	16:22
	← chuck left	16:22
Ilari	charon: Yes, there can be branch with that name.	16:22
withanx	I tried git clone url; git checkout branch, but I'm only getting the master branch from the clone	16:23
Ilari	withanx: 'git branch -r'.	16:23
charon	Ilari: which again defeats the point of disallowing * etc in case a script accidentally uses a branch name unquoted, doesn't it?	16:23
withanx	but can I checkout a remote branch and merge it?	16:23
parasti	withanx: create a local branch from it, then merge and push	16:24
withanx	how do I do that?	16:24
	the creating a local branch part	16:24
Ilari	withanx: Push failing indicates that some local branch is not up to date... Check it out and merge the remote branch...	16:24
parasti	withanx: git checkout -b $name origin/$name...	16:24
	→ name joined	16:24
Ilari	withanx: Presumably after running 'git fetch <remote>'.	16:25
	← bts- left	16:25
	→ thiago_home joined	16:26
	→ bts- joined	16:27
Ilari	withanx: Remote tracking branches are mirrors of branches in some other repositories. They are not meant to be committed to!	16:27
	→ radarek joined	16:28
Ilari	withanx: In fact, git checkout will not check them out. If you try, it starts new unnamed scratch branch from tip of that branch and switches to it.	16:28
withanx	okay, so when I go a git merge with the local cloned branch, I get a conflict, I manually fix the conflict and I have a ton of modified/added files in git status, do I commit?	16:29
Ilari	withanx: After each file looks like what you want, do 'git add' on files having conflicts and then 'git commit'. The commit should say it's a merge commit.	16:30
	← sh1 left	16:31
withanx	ah, cool	16:31
	that worked	16:31
	thanks	16:31
	← lorandi left	16:32
Ilari	If merge command can't automerge conflicts and immediately create merge commit, it will save the extraneous parents to file, where 'git commit' will pick them up.	16:32
withanx	neet	16:33
	→ lorandi joined	16:33
	→ lholden joined	16:34
	→ plediii_ joined	16:37
Ilari	withanx: Conflict resolution in some cases can even involve deleting a file (if you want it deleted in merge result, and branches to be merged don't agree wheither it exists or not).	16:38
	→ camwest joined	16:38
	ceej-away → ceej	16:38
Ilari	withanx: In that case, it's 'git rm' instead of 'git add'.	16:38
	← ben_h left	16:40
	← \ask- left	16:41
	← daniel-w left	16:42
	← DrFrasierCrane left	16:44
	→ daniel-w joined	16:45
	← bts- left	16:47
	→ chuck joined	16:47
chuck	are there any git commands that will start up a built in web server with gitweb?	16:47
Ilari	charon: There's no builtin webserver... Instaweb uses external programs to act as HTTP servers...	16:48
	→ tango_ joined	16:48
	→ sgrimm joined	16:49
	← plediii left	16:51
	← eventualbuddha left	16:52
	← ckm left	16:52
	← Guest75801 left	16:52
	→ carllerche joined	16:55
	← camwest left	16:58
	← jackdempsey left	16:58
	← reel_ left	17:00
	→ oudeis joined	17:00
charon	?	17:00
chuck	charon, he pinged the wrong person there	17:00
	← statim left	17:00
charon	oh.	17:01
Ilari	Darn Tab completion... :-/	17:01
	→ statim joined	17:01
	foutrelis → Death-to-IE	17:02
	→ RayMorgan joined	17:02
	← plediii_ left	17:03
	→ rlorandi joined	17:04
	→ Zimm joined	17:04
	→ CodeOfficer joined	17:06
	→ AngryElf joined	17:07
	← tpiep left	17:08
AngryElf	what's the process for merging all of the commits on a branch into master and then dcommiting as a single commit instead of individual ones?	17:09
	→ netoman joined	17:10
	→ numillumi joined	17:11
AlexB	AngryElf: merge --squash ?	17:11
AngryElf	thanx, i thought it was something to do with git patch	17:12
	was looking in the wrong place	17:12
	→ Eridius\|work joined	17:13
	Eridius\|work → Eridius	17:13
	← slainer68 left	17:13
	← lorandi left	17:14
	← tango_ left	17:14
	→ slainer68 joined	17:15
	← slainer68 left	17:15
Baughn	http://brage.info/~repo_df/ <-- I've got a git repo here, but for some reason I need to add ".git/" to the url for git-clone to work. What's up?	17:15
Voker57	non-bare repo	17:16
Baughn	..but I don't want it to be bare	17:16
Voker57	then you have to add .git	17:16
Baughn	Oh well	17:16
Voker57	because bare repo is just .git directory alone	17:17
	Baughn shrugs	17:17
	→ jackdempsey joined	17:18
	← kev left	17:19
	← temoto left	17:19
bremner	Baughn: you know about faq non-bare right?	17:19
Gitbot	Baughn: Pushing to non-bare repositories is discouraged. See http://git.or.cz/gitwiki/GitFaq#non-bare	17:19
Baughn	bremner: Now I do	17:19
	bremner: I want people to be able to browse the source without checking out the git repository, though. git-web?	17:20
	→ plediii joined	17:20
wereHamster	cgit or gitweb	17:20
Baughn	So, can I convert the current non-bare repository to a bare one by just extracting the .git subdirectory?	17:21
bremner	Baughn: yes	17:21
	Baughn: or git --bare clone foo foo.git	17:21
	→ LuboC joined	17:24
	← LuboC left	17:25
Baughn	And this git-daemon thing?	17:26
wereHamster	it's if you want to access the git repo through the git:// protocol	17:27
bremner	Baughn: fast read only cloning	17:27
	→ ankit9 joined	17:27
	Death-to-IE → foutrelis	17:28
	← bryanl left	17:28
	← bdiego left	17:29
	→ rubydiamond joined	17:30
Ilari	Actually also pushing if one doesn't care about security at all...	17:31
	The primary protocol for inter-machine pushing is ssh://	17:31
	→ cgardner joined	17:32
	→ greyface joined	17:32
	← ankit9 left	17:34
	ceej → ceej-away	17:35
	→ ankit9 joined	17:35
	→ tanek joined	17:36
	xenoterracide__ → xenoterracide	17:36
	→ hobodave joined	17:37
	→ jelmer joined	17:38
	→ \ask- joined	17:40
	→ tango_ joined	17:40
	← sgrimm left	17:41
	← ankit9 left	17:42
	→ ankit9 joined	17:42
	→ halfmouse joined	17:45
	← davidfetter left	17:45
halfmouse	I got a general question. I was using git as sc and made a back of the .git directory. My main drive crashed. Is it possible to recover from the backup of .git? If so how?	17:46
Baughn	halfmouse: The .git directory contains all your repository data, so that would be "yes"	17:46
	← radarek left	17:47
	← ankit9 left	17:47
	← jelmer left	17:47
Baughn	halfmouse: It's a bare git repository; you can clone from it to get a normal one back	17:47
halfmouse	So that would be to init a new one and clone from the backup?	17:47
Baughn	Of course, half the point of using git is to keep your data on >1 /live/ computers. ;)	17:47
	→ davidfetter joined	17:47
Baughn	halfmouse: No, just clone. You don't need to init a new one first.	17:47
Ilari	halfmouse: Make empty directory and inside it, directory named '.git' now extract the backup there (inside the .git directory). Now 'git status' should work and show everything deleted...	17:48
	→ jelmer joined	17:48
	← jelmer left	17:49
Ilari	halfmouse: Then, do something like 'git checkout .'.	17:49
	→ jelmer joined	17:49
Ilari	halfmouse: That is, after the extraction, you should have directory '.git/objects' inside the empty directory.	17:49
Baughn	..the clone sounds simpler	17:49
	← AngryElf left	17:49
	← Eridius left	17:50
	→ Eridius joined	17:50
Ilari	Baughn: Clone doesn't preserve all data there is in that backup...	17:50
Baughn	What does it miss?	17:51
doener	reflogs	17:51
Ilari	Baughn: Reflogs, remotes...	17:51
doener	and all objects only reachable through them	17:51
	stash	17:51
	(well, that's basically a special reflog)	17:51
	← LotR left	17:53
Ilari	halfmouse: Like it is possible to convert repo from non-bare to bare, the opposite operation (from bare to non-bare) is also possible. Of course, the non-bare to bare conversion destroys any modifications in working tree...	17:53
halfmouse	I made the .git as per the precrash dir structure. I copied the backup there, with all sub dirs and files. The git status tells me fatal: This operation must be run in a work tree. Does this mean I must init this dir to be a git tree?	17:54
	← Eridius left	17:54
Ilari	halfmouse: Does '.git/objects' exist (when viewed from root of new repo)?	17:54
halfmouse	Yes	17:55
	→ Eridius joined	17:55
Ilari	halfmouse: And presumably 'git checkout .' also complains about working tree?	17:55
halfmouse	Correct. I tried that before coming here	17:55
	When I tried git log, it showed all logged operation, so I have the feeling my stuff is still there.	17:57
	← bobmcw left	17:57
halfmouse	But I'm trying to understand the difference between a tree and a repository	17:57
	← doumo left	17:58
	← mw-home left	17:59
	ceej-away → ceej	17:59
	→ Mario__ joined	18:00
offby1	a tree is one of the four kinds of objects stored in the repository	18:00
	along with blobs, tags, and commits	18:00
Ilari	halfmouse: You have environment variable GIT_DIR set?	18:01
	→ ankit9 joined	18:01
	→ girishr joined	18:01
halfmouse	no	18:01
	I should set this to my git path?	18:01
	→ drobbins joined	18:02
drobbins	hiya	18:02
Ilari	halfmouse: No. Setting it would cause exactly those "not in working tree" errors.	18:02
halfmouse	offby1: Thanks.	18:02
	← simplechat left	18:02
	→ LotR joined	18:03
drobbins	hey, I have a question. I have a spanish translation of some XML, and I would like to reference a SHA1 in the translated XML in order to keep track of what version of the original article it translated.	18:03
halfmouse	Ilari: Mmm. I installed git from a macport	18:03
drobbins	I'm thinking that it would be best to reference the SHA1 of the XML file itself rather than the SHA1 of the commit.	18:03
	← pygi left	18:03
	→ pygi joined	18:03
drobbins	looking for opinions, and also info for how to display the SHA1 of a file (not the commit) using git.	18:03
offby1	yeah, a commit might have a bunch of files, and hence be confusgin	18:03
	confusing	18:03
	← oudeis left	18:04
offby1	"git ls-tree" I think will show you the sha1 of the files in the cd	18:04
drobbins	ok, I'll try that	18:04
	→ ashH joined	18:05
	← Mario__ left	18:05
offby1	say "git ls-tree HEAD"	18:05
drobbins	yep	18:05
	evbox xml # git ls-tree HEAD en/articles/funtoo/quick-install-howto/index.xml	18:05
	100755 blob 3af751ed9f9bdc5953f2c185804cd6faab62240f en/articles/funtoo/quick-install-howto/index.xml	18:05
	that seemed to do it.	18:05
	thanks offby1! :)	18:06
offby1	on the other hand ...	18:06
parasti	drobbins: have a look at man gitattributes, too ("ident", specifically)	18:06
Gitbot	drobbins: the gitattributes manpage can be found at http://git.or.cz/man/gitattributes	18:06
offby1	now that you've got the blob's SHA1 into your file, how useful is that?	18:06
drobbins	offby1: hmmmm	18:07
	ideally, I want a translator to be able to use the sha1 to see how far behind a translation is.	18:07
	maybe I should reference the commit. I don't know.	18:07
offby1	if the commit doesn't have a zillion files in it, I think it'd be more convenient.	18:08
drobbins	I want the translator to be able to do a git diff and see all the changes that have not been added yet	18:08
offby1	not vastly	18:08
	ok, then: if "git diff" takes blob IDs, you're set	18:08
	dunno if it does though.	18:08
Ilari	halfmouse: What directory are you trying to run the checkout from? It must be from dir that has the .git directory as direct subdirectory (which in turn contains 'objects', 'refs', etc...)...	18:08
	← patrick232 left	18:09
	→ patrick232 joined	18:10
	ceej → ceej-away	18:10
drobbins	I think I'll do the commit.	18:11
	"git log <filename>" shows the last commit that modified that file. That works for me.	18:11
halfmouse	Ilari, I was trying do to it from the .git directory itself. It works correctly when called from the parent directory	18:11
	→ schacon joined	18:12
Ilari	halfmouse: You might want to run 'git fsck --full' on restore to see if there is any data corruption (ignore 'dangling' messages, they are not about data corruption)...	18:13
	→ __name__ joined	18:14
parasti	drobbins: maybe you could wrap something like "git diff $(git rev-list -1 HEAD -- $translation) HEAD -- $source" in a script for translators to run	18:16
halfmouse	Ilari: fsck exited silently, which I will take as a good omen!	18:16
	→ kennethlove joined	18:16
halfmouse	Ilari: thank you, internet stranger...	18:17
drobbins	parasti: yep, making notes in the XML about how to do a diff for now	18:17
	→ bryanl joined	18:17
drobbins	I could actually have a script that parses the XML, extracts the commit name and the original file name, and then works from there.	18:18
halfmouse	Ilari: Before you leave... where can I learn more about the reflogs?	18:18
	→ dreiss joined	18:18
Ilari	halfmouse: One can also keep repository on seperate media and push to it...	18:18
	halfmouse: There's man git-reflog	18:18
Gitbot	halfmouse: the git-reflog manpage can be found at http://git.or.cz/man/git-reflog	18:18
	← charon left	18:19
	→ bobmcw joined	18:19
halfmouse	Ilari: The pushing was actually what I intented to do, before I started a trip. I'm going to set that up now, to help prevent more information loss.	18:19
	← plediii left	18:21
Ilari	halfmouse: Pushing (if done using "smart" protocol) checks the received data for consistency at remote end (yes, even file:// has a "remote end"). At least in new enough git versions...	18:21
	→ sgrimm joined	18:25
	→ oudeis joined	18:26
	→ plediii joined	18:28
	← LotR left	18:28
	→ LotR joined	18:29
	→ kennethlove_ joined	18:29
	→ uau joined	18:29
	← name left	18:29
	__name__ → name	18:30
	← ankit9 left	18:31
	← seangrove left	18:31
	→ ankit9 joined	18:31
	→ Weasel[DK] joined	18:34
	← ashH left	18:36
	← BugeyeD left	18:39
	→ bdiego joined	18:40
	→ floppyears joined	18:41
floppyears	hi	18:41
	I did git-rm on a file	18:41
	I then did git reset HEAD on that file	18:41
	how can I get it back ?	18:41
__mnemoc	checkout	18:42
	→ kennethlove__ joined	18:42
floppyears	thx	18:43
__mnemoc	yw	18:43
	→ radarek joined	18:45
	← kennethlove left	18:45
	← name left	18:47
	← mbroeker left	18:47
	→ mbroeker joined	18:48
	← rlorandi left	18:50
	→ rlorandi joined	18:51
	← Vernius left	18:52
	rlorandi → lorandi	18:52
	← cods left	18:55
	→ fhobia joined	18:57
	← kennethlove_ left	18:57
	ceej-away → ceej	18:59
	← rubydiamond left	19:01
	→ bosie joined	19:01
	← \ask- left	19:02
bosie	i have a master branch and a second branch called invoice. i worked in invoice on 6 files, commited 3 of them and merged with master. realized my merge was wrong and did a "git reset --hard HEAD" in master. now the 3 non-commited files are gone	19:03
	any chances i can get them back	19:03
	→ bosie_ joined	19:04
	→ luciddream joined	19:05
	→ fabi_an_ joined	19:05
__mnemoc	bosie_: they are still on the invoice branch	19:06
bosie_	__mnemoc: no they aren't :/	19:06
	→ k776 joined	19:08
__mnemoc	bosie_: but you said you had commited them there	19:08
parasti	bosie_: your uncommitted changes are gone (that's what --hard does)	19:08
bosie_	__mnemoc: no, i commited half of my files	19:08
__mnemoc	ah, sorry, i misread	19:08
bosie_	parasti: ok, any way to get them back ?	19:09
__mnemoc	bosie_: commit everything, always	19:09
	bosie_: retyping them	19:09
parasti	bosie_: no, they're really gone	19:09
bosie_	__mnemoc: funny for you to say ;)	19:09
	parasti: wicked	19:09
__mnemoc	bosie_: even if you think the work is temporal or incomplete or whatever, commit!	19:09
parasti	bosie_: use git stash, too	19:09
Ilari	bosie_: This is why one shouldn't merge with dirty working tree (even if git allows the operation).	19:10
	→ AngryElf joined	19:10
__mnemoc	Ilari: can this be considered a bug?	19:10
Ilari	__mnemoc: No.	19:11
	← ankit9 left	19:11
__mnemoc	Ilari: holy git protect us from messing with dirty working trees on other parts	19:11
Ilari	__mnemoc: Its by design and as such by defintion isn't a bug.	19:11
__mnemoc	:(	19:11
AngryElf	so I just compiled 1.5.4 and am getting what looks like a glibc error when I try to git svn rebase -- should I get a newer version to fix this? http://pastie.org/348468	19:11
Ilari	AngryElf: Seems like APR libs are buggy...	19:13
	← _slacker_ left	19:14
Ilari	AngryElf: Or some other native libs (Perl shouldn't allow Perl code to corrupt the interpretter)...	19:14
	← temoto-0x8a left	19:15
AngryElf	so you think this is a perl issue?	19:15
Ilari	AngryElf: I think it is an APR (a library SVN uses) issue... Or that git-svn is using SVN libraries very wrong, or that SVN libraries are using APR libraries wrong...	19:17
	AngryElf: The ultimate reason for crash is that Glibc detects that memory allocator internal structures are inconsistent. It then proceeds to dump stack backtrace and process memory map.	19:19
	→ lack joined	19:21
lack	Can I rebase a merge?	19:21
Ilari	AngryElf: Decoding from that memory map, you have APR 1.2.12 and SVN 1.5.5... No idea about those versions...	19:21
bosie_	parasti: ok thank you	19:21
	__mnemoc: sure but oviously i didn't want something in the master branch ;)	19:22
__mnemoc	bosie_: that's why can can "prepare" a branch for the merge, cherry-picking only the commits you do want to merge to the master branch	19:23
	s/can/you/1	19:23
	← bosie_ left	19:23
	→ offby1 joined	19:25
	← eruijters left	19:26
	→ bosie_ joined	19:26
Ilari	But then, there's a difference between interpretter having to deal with hostile input and interpretter that has to deal with hostile code...	19:29
	← clairvy left	19:30
	← PloenK left	19:31
	→ tenpaiyomi joined	19:34
tenpaiyomi	Quick question, if somebody could help me	19:34
	→ ashH joined	19:34
Ilari	tenpaiyomi: Well, nobody can help if you don't ask...	19:34
tenpaiyomi	I have a repo that I have been working on, that was a clone of another repo. I need to update my repo with the commits from the other repo now, so I can fix problems and create a working version	19:35
	→ name joined	19:35
tenpaiyomi	Ilari: Ya, was typing as you said that	19:35
	→ PloenK joined	19:36
henr_k	tenpaiyomi: like a git pull?	19:36
Ilari	tenpaiyomi: Pull branches of clone you made from the repository you cloned from...	19:36
	tenpaiyomi: There is no inverse of 'git pull' as single command. In practicular, 'git push' is NOT inverse of 'git pull'.	19:40
	→ \ask- joined	19:42
	← bryanl left	19:43
	→ tamsler joined	19:45
	← AngryElf left	19:46
	→ bobesponja joined	19:46
	→ cmarcelo joined	19:47
tamsler	I am trying to get the gitweb $export_ok to work	19:47
	... I set $export_ok = "true" in my /etc/gitweb.conf file	19:47
	I don't know what "file" needs to be added to a repository in order for it to show up again in gitweb	19:47
	← sammcd left	19:48
tamsler	... or do I need to set the $export_ok variable to the "magic file name" in the repository?	19:48
	I have tried pretty much all of that put gitweb doesn't list/show the repository	19:48
	← pygi left	19:49
tenpaiyomi	Ilari: Ok, I'm trying to do a git pull my_current_branch the_remote_repo, and it's giving me an error on that it's unable to chdir/not a git archive	19:49
Ilari	tenpaiyomi: The first argument to pull (if specified) specifies the repo to pull from. It can be either URL or known remote name...	19:51
parasti	tenpaiyomi: git pull is git fetch + git merge. merge always works on the current branch, so you're specifying the remote branch not your current branch	19:51
Ilari	tenpaiyomi: The usual syntax is 'git pull <remote-URL-or-nick> <branch-in-remote-repo>'.	19:52
tenpaiyomi	So, just wanting to merge with the branch I am on, I should just do a git pull url_of_repo_to_pull (assuming it's the master branch)	19:53
	→ bryanl joined	19:53
	← ashH left	19:53
parasti	tenpaiyomi: if you cloned, you already have a remote defined, so it might be just git pull	19:54
Ilari	parasti: AFAIK, This is transfer in opposite direction...	19:54
tenpaiyomi	Ok, lemme try and reexplain. A pull of repo A was made (repo B), and work was done on that copy. Lots of commits, et al. repo A has done some updates since then, so we want to merge the changes of repo A in with repo B, such that repo B is up to date with the other, but we still have our changes	19:56
	← patrikf left	19:56
Ilari	Clone creates remote name 'origin' that refers to repository cloned from...	19:56
	tenpaiyomi: Most If you are updating clone, then even 'git pull' might be enough...	19:57
tenpaiyomi	Right, but repo A wasn't cloned from repo B, it was a pull	19:57
	2 separate copies	19:57
	→ gregor joined	19:57
tenpaiyomi	Well, no	19:57
	It was cloned, but the remote is not repo A's repo	19:58
Ilari	tenpaiyomi: As said, 'git pull <remote-URL-or-nick> <branch-on-remote>'...	19:58
	← luciddream left	19:59
	→ FH` joined	20:00
Ilari	tenpaiyomi: You might also want to do 'git remote add -f <new-remotenick> <remote-URL>' so you can use shorthand to refer to that repo..	20:00
	← bobmcw left	20:01
	← offby1 left	20:01
	→ gregor_ joined	20:02
	→ xenoterracide_ joined	20:03
	→ rtomayko joined	20:03
	→ Niamor joined	20:04
	← lorandi left	20:05
\ask-	I have two code bases (each in separate branches). The last commit of A removes all files; I'd like to have the history spliced so in one branch I have A and then all the B commits. Any ideas?	20:05
	I've tried a few variations of rebasing; but I can't get it to Quite Work (rebasing A on B seems to lose the first commit from B, for example).	20:06
	it's sorta ok with just merging the two branches; but I'd prefer to avoid having the merge commit.	20:06
kevlarman	\ask-: man git-filter-branch	20:06
Gitbot	kevlarman: the git-filter-branch manpage can be found at http://git.or.cz/man/git-filter-branch	20:06
kevlarman	it's not pretty though	20:06
\ask-	kevlarman: any more hints than that? It seems when branch A ends in an empty tree, there should be an easier way...	20:07
	← radarek left	20:07
	← bdiego left	20:08
	← metajack left	20:09
	→ metajack joined	20:10
	→ patrikf joined	20:11
	→ radarek joined	20:11
lack	I have 2 branches that I have merged, but I want to rebase this merge as if it happened one commit later on one of the original branches. Is that possible with 'rebase'? OR do I have to just make a new branch, start with the merge, and then 'rebase onto' for the rest of the commits?	20:12
	← daniel-w left	20:13
	→ bdiego joined	20:16
	→ daniel-w joined	20:18
	← girishr left	20:18
	→ absentia joined	20:19
absentia	ji	20:19
	hi. I have 1 git repo with 1 other dev.. every time one of us pushes, the other will gt errors.	20:19
	← markelikalderon left	20:20
absentia	we both have umask 002, directtories are set g+w and g+s .. ew are both i the same group to share git files (but it's not our default group) ... what needs to be done so I don't have to do a chown -R gitgrp git-repo; chmod -R g+w git-repo after every singlepush?	20:20
	→ weirdo joined	20:20
weirdo	hey	20:20
	got a problem	20:20
	→ bobmcw joined	20:21
weirdo	have to run git-gc after every commit, otherwise repo is unclonable	20:21
	warning: remote HEAD refers to nonexistent ref, unable to checkout.	20:21
bremner	absentia: man git-init (--shared)	20:22
Gitbot	absentia: the git-init manpage can be found at http://git.or.cz/man/git-init	20:22
	← gregor left	20:23
	← bobesponja left	20:24
weirdo	any ideas? :\|	20:25
	← lottadot left	20:25
weirdo	what i do is rsyncing the git repo to a http server and cloning from there	20:25
wereHamster	cgit strips whitespace between my name and email address when generating patches: From: Tomas Carnecky<tom@dbservice.com>	20:26
	→ offby1 joined	20:26
wereHamster	should I send an email to git@vger or is the cgit maintainer here right now?	20:27
Ilari	weirdo: Sounds familiar... Somebody else had those kind of issues with rsync transport...	20:29
	weirdo: Also, if you use HTTP, is that cached info for dumb transports (git update-server-info) up to date?	20:32
weirdo	i guess i'll just run git-gc in post-commit hook	20:32
	but i don't use rsync://, i just rsync the repo manually	20:32
	→ gregor__ joined	20:32
	← gregor_ left	20:33
	→ Tobbe joined	20:33
Tobbe	Hi. How can I tell what commits haven't yet been pushed?	20:34
Ilari	Tobbe: 'git log --branches --not --remotes'?	20:35
	Oops, that doesn't work.	20:35
Tobbe	looks like it does :)	20:36
	→ cods joined	20:36
bremner	weirdo: why do you rsync the repo manually? do you not have ssh access, but do have rsync access?	20:38
weirdo	i do have ssh access	20:38
	that's how i rsync the tree	20:38
	bremner thinks rsyncing a repo is fraught with peril if done w/o -c	20:39
absentia	thanks	20:39
	← absentia left	20:39
	← Eridius left	20:39
bremner	weirdo: don't be such a weirdo :-) use git push ssh:// instead	20:39
	→ sergiodj joined	20:40
	→ absentia joined	20:41
absentia	the git-init --shared looks good.. but I have a report with 23000 items... do I ahve to recreate it, or can i just somehow update git to shared?	20:41
Ilari	absentia: 'git config core.sharedrepository true' can be used on existing repository...	20:42
bremner	absentia: git --bare init --shared in the repo should be harmless	20:42
	Ilari: so can git init AIUI	20:42
	absentia: assuming a bare repo	20:42
absentia	I'll try the bremner one first...	20:42
bremner	absentia: in general, Ilari's answers are better than mine :-). But here I think it should fine my way.	20:43
absentia	git --bare init --shared=spygit	20:44
	fatal: bad config value for 'arg'	20:44
	ok, gonna try ilari's	20:44
	ilaris: ok, if I use your command to set to shred, how do I specify the group?	20:44
	git config core.sharedrepository groupname?	20:44
Ilari	absentia: No.	20:44
weirdo	bremner, thank you, -c works fine now :-)	20:44
Ilari	absentia: Set setgid flag for all directories, and set the owning group correctly....	20:45
absentia	hmm.. just did it.	20:45
	no error.	20:45
	ilaris: I did.	20:45
	git config core.sharedrepository true	20:45
	fatal: bad config value for 'core.sharedrepository' in .git/config	20:45
	ug	20:45
weirdo	bremner, basically, i have a script that synchronizes my whole website, i wouldn't want to add an another one especially for repos	20:45
absentia	ah, wrong box.	20:45
	→ gasbag joined	20:45
	← weirdo left	20:45
bremner	weirdo: obviously two of you doing that will destroy the universe	20:46
	→ tpiep joined	20:46
absentia	ok, gonna try it now	20:47
bremner	absentia: by the way, the correct version of git init would use one of the options specified in the man page, i.e. the string "group"	20:47
absentia	I'm still stuck with not being able to pull ...	20:48
	er, push	20:48
	says: rejected, non-fast forward... error: failed to push some refs to master.	20:48
	can't make that go away.	20:49
Ilari	absentia: Pull first and then push...	20:49
absentia	bremner: agreed, but I ahve 23000 files in the repo.. took hours for my friend to download it.	20:49
	I did a pull, then a push.. always get htat error	20:49
Ilari	absentia: Switch to branch that gets that error, pull the corresponding branch from remote repo...	20:50
absentia	I only have master.	20:50
bremner	absentia: your last remark sense makes not me to	20:50
	← oudeis left	20:50
absentia	anyway... perms look good.	20:51
	I'll see how it goes.	20:51
	← chalky left	20:51
absentia	need to wait for my friend to show up to see if he can do a pull, commit local and then pull and then push...	20:51
	← floppyears left	20:51
	← bryanl left	20:51
	→ bryanl joined	20:52
Ilari	absentia: Does 'git pull' from the repo that can't push really say 'already up to date' or something like that?	20:52
absentia	yes	20:52
	up to date	20:52
	I'll do it again, let me modify a file.	20:52
	maybe it was the file permissions issue	20:53
	if the error comes back, I'll paste a log... didn't fail that time.	20:53
	thanks for the help!	20:54
Ilari	absentia: "Not fast forward" would mean that overwriting the remote branch with your version would lose commits...	20:54
absentia	ah.	20:55
	I usually do a pull then... to get up to day.	20:55
	→ eruijters joined	20:55
absentia	but then mypush says can't push some refs.	20:55
	← FH` left	20:56
Ilari	absentia: 'git pull' only updates current branch...	20:56
	← bobmcw left	20:57
	→ bobmcw joined	20:57
	→ kbingham_ joined	21:04
	← _dave_h_d left	21:04
	← xl0 left	21:09
	← aspotashev left	21:11
	→ ashleyw joined	21:12
	→ aspotashev joined	21:12
	← halfmouse left	21:16
	→ oudeis joined	21:16
	← PloenK left	21:17
	→ PloenK joined	21:18
	→ berndj joined	21:22
	→ git_ joined	21:26
git_	hi	21:26
	i have 3 machines behind a firewall	21:26
	i use git push to push source code to a server outside the firewall	21:27
	how do I merge the code i pushed?	21:27
j_engelh	git_: yeah, that's the easiest way to duplicate corporate code :p	21:27
git_	hehe	21:27
j_engelh	unsurprisingly, git merge	21:27
git_	git merge, that's it?!?	21:27
j_engelh	rtnm?	21:28
	← up_the_irons2 left	21:30
	→ up_the_irons joined	21:31
	→ icwiener joined	21:33
git_	git merge requires <remote>	21:33
	← name left	21:34
	→ name joined	21:35
	→ charon joined	21:36
	← schacon left	21:37
	→ markelikalderon joined	21:40
	→ lolage0 joined	21:42
	→ pygi joined	21:43
	← ashleyw left	21:47
	← kbingham_ left	21:47
	← gasbag left	21:47
	← daniel-w left	21:47
	← \ask- left	21:47
	← bosie_ left	21:47
	← fabi_an_ left	21:47
	← jackdempsey left	21:47
	← numillumi left	21:47
	← Jacolyte left	21:47
	← cemerick left	21:47
	← zerok left	21:47
	← Yuuhi left	21:47
	← josephwilk left	21:47
	← d0k_ left	21:47
	← tjafk2 left	21:47
	← synsol left	21:47
	← dmlloyd left	21:47
	← beachsurfin left	21:47
	← BCarlyon left	21:47
	← niarbtfel left	21:47
	← lack left	21:47
	← pygi left	21:47
	← name left	21:47
	← cods left	21:47
	← metajack left	21:47
	← rtomayko left	21:47
	← cmarcelo left	21:47
	← k776 left	21:47
	← kennethlove__ left	21:47
	← Weasel[DK] left	21:47
	← cgardner left	21:47
	← RayMorgan left	21:47
	← doener left	21:47
	← Voker57 left	21:47
	← LiamH left	21:47
	← cannonball left	21:47
	← ia left	21:48
	← resmo left	21:48
	← zwu\|gone left	21:48
	← elmex left	21:48
	← alikins left	21:48
	← sstangl left	21:48
	← bgerber left	21:48
	← mw\|out left	21:48
	← przemoc left	21:48
	← ceej left	21:48
	← dermoth left	21:48
	← lordpil left	21:48
	← oudeis left	21:48
	← bdiego left	21:48
	← hobodave left	21:48
	← withanx left	21:48
	← kukks left	21:48
	← pencilo left	21:48
	← billm left	21:48
	← nadim left	21:48
	← dennis- left	21:48
	← josh^ left	21:48
	← wagle left	21:48
	← patrick232 left	21:48
	← Zimm left	21:48
	← thiago_home left	21:48
	← mithraic left	21:48
	← drobbins left	21:48
	← ricky left	21:48
	← EvanCarroll left	21:48
	← aeruder left	21:48
	← j_engelh left	21:48
	← sergiodj left	21:48
	← Niamor left	21:48
	← CodeOfficer left	21:48
	← spearce left	21:48
	← drewr left	21:48
	← sdboyer left	21:48
	← Simetrical left	21:48
	← vuf left	21:48
	← nottha_k left	21:48
	← meyering left	21:48
	← piroko left	21:48
	← seezer left	21:48
	← ndim left	21:48
	← bono left	21:48
	← rhett left	21:48
	← aphexer left	21:48
	← absentia left	21:48
	← a-atwood_ left	21:48
	← git_ left	21:48
	← bobmcw left	21:48
	← lholden left	21:48
	← stephens left	21:48
	← ph^ left	21:48
	← xinming left	21:48
	← tamsler left	21:48
	← davidfetter left	21:48
	← Tr3B_ left	21:48
	← Aides left	21:48
	← dwmw2_gone left	21:48
	← ftehw left	21:48
	← worch_ left	21:48
	← gregor__ left	21:48
	← patrikf left	21:48
	← tenpaiyomi left	21:48
	← Loof left	21:48
	← rellik_ left	21:48
	← sverrej left	21:48
	← numbah left	21:48
	← cbreak left	21:48
	← prh left	21:48
	← damog left	21:48
	← broonie left	21:48
	← Ademan left	21:48
	← KiBi left	21:48
	← Weiss left	21:48
	← segher left	21:48
	← selckin left	21:48
	← GyrosGeier left	21:48
	→ Sigma joined	21:49
	→ mpr joined	21:49
	→ ft joined	21:49
	→ stouset joined	21:49
	→ pygi joined	21:49
	→ name joined	21:49
	→ git_ joined	21:49
	→ oudeis joined	21:49
	→ ashleyw joined	21:49
	→ kbingham_ joined	21:49
	→ bobmcw joined	21:49
	→ gasbag joined	21:49
	→ absentia joined	21:49
	→ sergiodj joined	21:49
	→ cods joined	21:49
	→ daniel-w joined	21:49
	→ bdiego joined	21:49
	→ metajack joined	21:49
	→ Niamor joined	21:49
	→ rtomayko joined	21:49
	→ cmarcelo joined	21:49
	→ tamsler joined	21:49
	→ \ask- joined	21:49
	→ bosie_ joined	21:49
	→ lack joined	21:49
	→ k776 joined	21:49
	→ fabi_an_ joined	21:49
	→ kennethlove__ joined	21:49
	→ Weasel[DK] joined	21:49
	→ patrick232 joined	21:49
	→ drobbins joined	21:49
	→ davidfetter joined	21:49
	→ hobodave joined	21:49
	→ cgardner joined	21:49
	→ jackdempsey joined	21:49
	→ numillumi joined	21:49
	→ CodeOfficer joined	21:49
	→ Zimm joined	21:49
	→ RayMorgan joined	21:49
	→ lholden joined	21:49
	→ thiago_home joined	21:49
	→ stephens joined	21:49
	→ doener joined	21:49
	→ Jacolyte joined	21:49
	→ withanx joined	21:49
	→ cemerick joined	21:49
	→ zerok joined	21:49
	→ spearce joined	21:49
	→ Tr3B_ joined	21:49
	→ Voker57 joined	21:49
	→ mithraic joined	21:49
	→ LiamH joined	21:49
	→ drewr joined	21:49
	→ kukks joined	21:49
	→ cannonball joined	21:49
	→ Yuuhi joined	21:49
	→ josephwilk joined	21:49
	→ d0k_ joined	21:49
	→ ia joined	21:49
	→ Aides joined	21:49
	→ ph^ joined	21:49
	→ dwmw2_gone joined	21:49
	→ niarbtfel joined	21:49
	→ pencilo joined	21:49
	→ resmo joined	21:49
	→ billm joined	21:49
	→ zwu\|gone joined	21:49
	→ tjafk2 joined	21:49
	→ elmex joined	21:49
	→ alikins joined	21:49
	→ xinming joined	21:49
	→ sdboyer joined	21:49
	→ ricky joined	21:49
	→ Simetrical joined	21:49
	→ synsol joined	21:49
	→ vuf joined	21:49
	→ sstangl joined	21:49
	→ bgerber joined	21:49
	→ nadim joined	21:49
	→ dmlloyd joined	21:49
	→ dennis- joined	21:49
	→ ftehw joined	21:49
	→ nottha_k joined	21:49
	→ meyering joined	21:49
	→ EvanCarroll joined	21:49
	→ aeruder joined	21:49
	→ josh^ joined	21:49
	→ wagle joined	21:49
	→ worch_ joined	21:49
	→ a-atwood_ joined	21:49
	→ piroko joined	21:49
	→ mw\|out joined	21:49
	→ seezer joined	21:49
	→ przemoc joined	21:49
	→ beachsurfin joined	21:49
	→ rhett joined	21:49
	→ ndim joined	21:49
	→ bono joined	21:49
	→ ceej joined	21:49
	→ j_engelh joined	21:49
	→ BCarlyon joined	21:49
	→ dermoth joined	21:49
	→ lordpil joined	21:49
	→ aphexer joined	21:49
	→ warthog9 joined	21:49
	→ AAA_awright joined	21:49
	→ nothinghappens joined	21:49
	→ Illocution joined	21:49
	→ r0bby joined	21:49
	→ Tv joined	21:49
	→ yoka187 joined	21:49
	→ dholm joined	21:49
	→ PugMajere joined	21:49
	→ tmm1 joined	21:49
	→ rboyd joined	21:49
	→ ska-fan joined	21:49
	→ bartman joined	21:49
	→ stick joined	21:49
	→ nothingmuch joined	21:49
	→ corecode joined	21:49
	→ Zathrus joined	21:49
	→ orospakr joined	21:49
	→ brodie joined	21:49
	→ jpieper joined	21:49
	→ cehteh joined	21:49
	→ mutex joined	21:49
	→ Tommy[D] joined	21:49
	→ malahal joined	21:49
	→ jones- joined	21:49
	→ jql joined	21:49
	→ mika joined	21:49
	→ mugwump joined	21:49
	→ pasky joined	21:49
	→ oxdeadbeef joined	21:49
	→ kevlarman joined	21:49
	→ pastie joined	21:49
	→ ahf joined	21:49
	→ kbingham joined	21:49
	→ flexd joined	21:49
	→ dwave joined	21:49
	→ tchan joined	21:49
	→ peper joined	21:49
	→ context joined	21:49
	→ halfline joined	21:49
	→ scrooloose joined	21:49
	→ godlygeek joined	21:49
	→ diva1234 joined	21:49
	→ dfarnsworth joined	21:49
	→ vomjom joined	21:49
	→ jaswinder joined	21:49
	→ zobie joined	21:49
	→ Toba joined	21:49
	→ zumbrunn joined	21:49
	→ metafollic joined	21:49
	→ flazz joined	21:49
	→ elliottcable joined	21:49
	→ redinger_away joined	21:49
	→ ToxicFrog joined	21:49
	→ bogolisk joined	21:49
	→ bunz joined	21:49
	→ Sonderblade joined	21:49
	→ w00t_ joined	21:49
	→ gaffo joined	21:49
	→ RichiH joined	21:49
	→ rphillips joined	21:49
	→ semifor joined	21:49
	→ Rich_Morin joined	21:49
	→ machrider joined	21:49
	→ blithe joined	21:49
	→ ash_ joined	21:49
	→ petteri joined	21:49
	→ Toad joined	21:49
	→ mikearr joined	21:49
	→ cbrake joined	21:49
	→ meermanr joined	21:49
	→ jefferai joined	21:49
	→ djinni joined	21:49
	→ Bartman007 joined	21:49
	→ ReinH joined	21:49
	→ Bucciarati joined	21:49
	→ kelvie joined	21:49
	→ spb joined	21:49
	→ djpig joined	21:49
	→ wereHamster joined	21:49
	→ lu_zero joined	21:49
	→ ThaDon joined	21:49
	→ dublpaws joined	21:49
	→ CelticSoul joined	21:49
	→ argonel joined	21:49
	→ IRSeekBot joined	21:49
	→ ddollar joined	21:49
	→ cliffstah joined	21:49
	→ rtyler joined	21:49
	→ Bombe joined	21:49
	→ nomis joined	21:49
	→ majoh joined	21:49
	→ redondos joined	21:49
	→ Gitbot joined	21:49
	→ ceecil joined	21:49
	→ azuk joined	21:49
	→ precision joined	21:49
	→ Thumper_ joined	21:49
	→ AlexanderS joined	21:49
	→ cardioid joined	21:49
	→ reuss joined	21:49
	→ Rhonda joined	21:49
	→ dkr_ joined	21:49
	→ reaVer joined	21:49
	→ jayne joined	21:49
	→ sr joined	21:49
	→ gregor__ joined	21:50
	→ patrikf joined	21:50
	→ tenpaiyomi joined	21:50
	→ Loof joined	21:50
	→ rellik_ joined	21:50
	→ sverrej joined	21:50
	→ selckin joined	21:50
	→ Ademan joined	21:50
	→ GyrosGeier joined	21:50
	→ cbreak joined	21:50
	→ KiBi joined	21:50
	→ numbah joined	21:50
	→ segher joined	21:50
	→ Weiss joined	21:50
	→ damog joined	21:50
	→ prh joined	21:50
	→ broonie joined	21:50
	← mpr left	21:50
	→ mps joined	21:50
Ilari	git_: Git merge requires committish to merge. To transfer branches from other repositories, use 'git fetch'.	21:50
	→ naeu joined	21:50
	← foutrelis left	21:50
	← josephwilk left	21:51
	← semifor left	21:51
	→ semifor joined	21:51
	← Zimm left	21:52
	← gasbag left	21:54
	← bobmcw left	21:55
	← jelmer left	21:57
	→ jelmer joined	21:57
	← jelmer left	21:57
	← PloenK left	22:00
	→ jelmer joined	22:00
	← jelmer left	22:02
	← patrikf left	22:02
	→ Whtiger joined	22:02
	← netoman left	22:04
	← Tobbe left	22:08
AAA_awright	Git doesn't make it very clear what sort of workflow you are supposed to use, or make it very easy to setup a place to push your changes too. How do I explain how to create a workflow for a project with 3-5 people?	22:09
	← aspotashev left	22:09
	← cannonball left	22:10
	← cemerick left	22:11
	→ tamsler_ joined	22:11
	→ Eridius joined	22:12
	→ patrikf joined	22:14
	← Voker57 left	22:14
	→ jelmer joined	22:14
	← jelmer left	22:14
Ilari	AAA_awright: Git supports multiple workflows... And setting up places where one user can push to is quite easy...	22:14
AAA_awright	I should have pointed out, for Windows systems mostly, explaining it to mostly (not entirely) people new to version control.	22:15
	← git_ left	22:15
	→ jelmer joined	22:16
Ilari	AAA_awright: Ouch. Creating pushable repositories on various types of Unixes and unixlike systems is not that complicated... But I don't know about setting pushable stuff on Windows box.	22:16
	stouset → stouset\|away	22:17
	← tamsler left	22:18
AAA_awright	And this being new, I can't figure out which workflow would be most straightforward. But in the meantime, I was thinking, a Lighttpd+WebDAV+Perl setup	22:18
Ilari	AAA_awright: That doesn't sound "most straightforward"...	22:19
	→ luciddream joined	22:20
	→ seb_ccc joined	22:20
Ilari	AAA_awright: (maybe Unless the server box needs to run Windows)...	22:20
	← ph^ left	22:20
luciddream	hi, i've deleted a handful of files from my working copy. git status shows that they are deleted but not staged. do I need to run "git rm" on each of them individually or is there a way I can stage all the deletions at once similar to "git add ."	22:21
AAA_awright	It probably will be. The environment I have to work with is unreliable access to anything centralized, mostly just each persons (Windows) laptop.	22:21
	stouset\|away → stouset	22:21
	→ badass joined	22:22
Ilari	luciddream: There's 'git add -u', but it would also stage all other modifications in tracked files...	22:22
	→ syamajala joined	22:23
syamajala	hi	22:23
	i'm trying to setup git-http-push	22:23
seb_ccc	Hi, I have a patch history consisting of 4 patches. The goal was to move some portions of text from a website into a file, so my first commit is the content as it appeared on the website, the second is cleaning up the content, the third is another chunk of content from the website, the fourth cleans it up. None of the patches follow the "no line should be wider than 80 chars" rule that the project likes, so I'm trying to clean up my	22:23
syamajala	and i have a problem with curl	22:23
	it needs --digest in order to work, but i don't know how to add that to the .netrc	22:23
Ilari	syamajala: You work in some environment where pushing over SSH isn't possible?	22:24
luciddream	Ilari: thanks that does it	22:24
	→ Joghurt joined	22:24
syamajala	no	22:24
Ilari	syamajala: Pushing over SSH is usually preferable over pushing over HTTP (unless there is some specific reason for using HTTP)...	22:26
	← xenoterracide left	22:26
syamajala	i didn't know pushing over ssh was possible	22:26
	← xenoterracide_ left	22:26
syamajala	i'm new to git...	22:26
AAA_awright	What would be the best way to push/pull between Windows computers then?	22:27
Ilari	syamajala: Mainly requres git to be installed in target machine and in PATH for noninteractive logins.	22:28
	AAA_awright: HTTP pulling is somewhat tolerable... HTTP pushing altough supported tends to be quite nasty to set up.	22:29
	← juanjoc left	22:29
syamajala	well i kinda have it working	22:29
	← eruijters left	22:29
	← peeja left	22:29
syamajala	i just need to know how to get curl to use --digest by default instead of --basic	22:29
AAA_awright	Ilari: I already know how to setup Lighttpd with WebDAV and passwords and such, so if so, would that be a viable option?	22:30
syamajala	AAA_awright: thats what i did	22:30
AAA_awright	Git-gui doesn't like interactive prompts for passwords though, SSH or otherwise	22:30
	kennethlove__ → kennethlove	22:31
	← Niamor left	22:31
	← nothinghappens left	22:31
Ilari	AAA_awright: HTTP pulling consists of having repository files together with some extra data (generated by 'git update-server-info').	22:31
AAA_awright	syamajala: What sort of workflow did you use? centralized, push to a public/pull from public?	22:32
	So you need commit hooks	22:32
	→ git_ joined	22:33
Ilari	AAA_awright: Also, there is experimental patches for git:// server written in Java. Unfortunately, it doesn't do ANY authentication.	22:33
	AAA_awright: Not that git:// does any authentication anyway...	22:34
	→ nothinghappens joined	22:34
AAA_awright	I don't really need authentication	22:34
syamajala	ah i figured it out	22:34
	← tamsler_ left	22:34
syamajala	.curlrc	22:34
	lets see if it works	22:34
	← selckin left	22:36
	→ BugeyeD joined	22:37
	→ schlort joined	22:38
syamajala	hmm	22:41
	curl works, but git doesn't	22:41
Ilari	AAA_awright: And git:// can't do pushing (well, it does if you configure it to be really insecure)...	22:41
syamajala	error: Cannot access URL http://user@web.genesis.blogdns.net/proj.git/, return code 22	22:42
AAA_awright	Ilari: Heh, I forgot. Yeah, that won't work.	22:42
	Well, not really.	22:42
berndj	does git detect motion of individual hunks? or shall i take the opportunity to clean up whitespace in the same commit that moves functions between files?	22:43
offby1	22 is EINVAL	22:43
syamajala	according to google git seems to ignore ~/.curlrc	22:43
AAA_awright	Use WebDAV for pushing ang git:// for pulling? But why make it more complex then you need to.	22:43
	← cmarcelo left	22:43
	← Joghurt left	22:43
	→ pointlessjon joined	22:44
Ilari	AAA_awright: If you only have small development team, maybe you don't need pushing (except to central servers on those ocassions where they are available)...	22:45
	→ Niamor joined	22:45
	← Sigma left	22:45
	← radarek left	22:46
AAA_awright	How would one set up pulling from each other?	22:46
	On Windows (that is the fine print)	22:46
	← metajack left	22:46
	→ metajack joined	22:46
Ilari	AAA_awright: And then there's the sneakernet transport (git bundle)...	22:47
	← patrikf left	22:47
Wyk3d	does git daemon work reliably under cygwin ?	22:48
	→ diegoviola joined	22:52
	← pointlessjon left	22:53
	→ bg_ joined	22:54
	← LiamH left	22:55
bg_	anyone using git on OS X and can help me understand where my problem is with git? (it hangs most of the time on my home DSL connection)	22:56
Ilari	bg_: Define "Hangs"?	22:57
bg_	llari: starts "Initializing empty Git repository..." than simply either hangs forever, or continues after several minuts to fetching the data	22:58
offby1	bg_: ktrace might be useful to see where it's hanging	22:58
	→ Theory joined	22:59
bg_	I tried on two different connections... one works just fine	22:59
	it seems there's an issue only on my home connection	22:59
	← LotR left	23:01
diegoviola	hi	23:01
	can i use git for storing documents like txt files and configuration like xml, etc?	23:02
Ilari	diegoviola: Yes... Altough merging XML could be somewhat nastier...	23:02
diegoviola	i'm not exactly doing programming now but i would like to store my documents and configurations in a repo	23:02
	Ilari: what do you mean with nastier? git wont handle it?	23:03
parasti	diegoviola: watch out for special file permissions, Git doesn't try hard to preserve them	23:03
diegoviola	parasti: ok	23:03
	→ AngryElf joined	23:04
Ilari	diegoviola: The built-in git merging algorithms are fairly good for source. They aren't designed to handle XML... Also if you put text files, try to avoid changing the flowing (as that really causes conflicts when merging).	23:04
	→ jmd joined	23:05
Theory	If you don't care about merging, obviously these things don't matter.	23:05
diegoviola	i will do push/pull/commits mostly	23:05
	← AngryElf left	23:06
Ilari	diegoviola: pull is fetch + merge...	23:06
	← ironfroggy_ left	23:07
diegoviola	hm ok	23:07
Ilari	diegoviola: Altough they aren't designed to handle XML, if XML file is "nicely" formatted, then the merge algorithm maybe could get useful merge result...	23:07
	→ simplechat joined	23:07
diegoviola	i see	23:08
Ilari	diegoviola: If only one side has modified some file, then the changes are taken into result without problems. Problems only start if both sides have modified some file...	23:08
jmd	How do I change the "origin" of a repository?	23:08
Ilari	jmd: 'git config remote.origin.url <newurl>'?	23:08
shd	jmd: you can edit .git/config, for example, or do as ilari says	23:09
	i like to edit .git/config :-)	23:09
bg_	llari: this is what I could find tracing git... http://gist.github.com/41426	23:09
syamajala	hmm	23:09
	Ilari: looks like git over http with webdav doesn't work with lighttpd	23:09
diegoviola	Ilari: how about for single text files?	23:10
	plain text	23:10
shd	diegoviola: for plain text, same rules apply as to source code (or anything else for that matter)	23:11
Ilari	diegoviola: Those work fine, as long as not much flow changing occurs (use editor with dynamical wrapping)...	23:11
shd	diegoviola: merge will always be trivial when there are no conflicting changes in branches	23:11
	diegoviola: and one only gets conflicts by editing the same file in two different branches	23:11
	diegoviola: conflicts are usually easy to handle	23:11
jmd	Ilari, Should the "remote.origin.url" be entered literally ?	23:11
	← joshdavey left	23:12
Ilari	jmd: Yes (if changing remote "origin").	23:12
	← bryanl left	23:12
diegoviola	ok i will give it a try, thanks guys	23:13
Ilari	diegoviola: And if you got tools, you can extract the files needed for three-way merge... Maybe there are readymade tools that do three-way merge on XML.	23:13
diegoviola	cool	23:13
jmd	Ilari, shd Thanks.	23:13
shd	diegoviola: btw. i've been handling most my documents for years with version control.. it has been rather easy	23:13
	diegoviola: i've been using git for 1,5 years for my docs, configs, ...	23:14
diegoviola	shd: nice, using git?	23:14
	shd: nice, whats your experience with it?	23:14
Ilari	bg_: That doesn't look very useful... What the processes involved are? Where it hangs?	23:14
shd	diegoviola: it works pretty well.. and since it's sha1 checksumed and constantly checked i can feel safe about bit corruption	23:15
	→ radarek joined	23:15
Ilari	(as long as one has backups)...	23:15
diegoviola	shd: yep, cool	23:15
bg_	llari: it is running now git-index-pack, but VERY slow	23:15
shd	what backups?-) huh.. who does backups??	23:15
bg_	llari: sometimes it won't even reach git-index-pack	23:15
	llari: it looks like it is trying to read the same chunk over and over again with errors	23:16
Ilari	bg_: index-pack is invoked to process the incoming data stream... If connection isn't practicularly fast, it is likely network-bound...	23:16
bg_	llari: i'm on 5mbit DSL connection... git over http is 100 times faster here than using git protocol	23:17
	← icwiener left	23:17
Ilari	bg_: Maybe ISP does something nasty on git://?	23:17
	← carllerche left	23:18
bg_	llari: I really hope not :) it might be that port 9418 is also used by some P2P apps... :(	23:18
Ilari	bg_: It might also be that they trottle all but some ports...	23:19
	→ zerok_ joined	23:19
Ilari	bg_: And no P2P app should be using 9418, as its reserved for "Git pack transfer service"...	23:19
bg_	llari: any way I can try git clone over a proxy server? (with git protocol)	23:19
Ilari	bg_: There is proxy support for git:// but maybe easier (if possible, and SSH is not messed with) to open SSH port forwarding somewhere...	23:20
bg_	ahh	23:21
	i'll try that	23:21
Ilari	bg_: And at 5Mbps, the transfer over git:// should be network/sender bound...	23:22
bg_	llari: just tried tunneling over SSH to another server, and everything works just fine	23:23
	llari: I guess i have some shouting to do at my ISP	23:23
kbingham_	anyone use git imap-send and gmail ?	23:23
	← pencilo left	23:24
kbingham_	i added in the imap section but i keep getting Resolving imaps://imap.gmail.com... gethostbyname: Connection timed out	23:24
	failed to open store	23:24
	→ seangrove joined	23:24
	→ travlr joined	23:24
Ilari	kbingham_: Maybe imap.gmail.com uses some non-standard port for IMAPS? I have seen setups that do IMAP over SSL via some weird port...	23:25
kbingham_	[imap]	23:25
	folder = "[Gmail]/Drafts"	23:25
	host = imaps://imap.gmail.com	23:25
	user = [email@hidden.address]	23:25
	pass = p4ssw0rd	23:25
	port = 993	23:25
	sslverify = false	23:26
	← radarek left	23:26
	→ altrux joined	23:26
kbingham_	http://repo.or.cz/w/git.git?a=blob_plain;f=Documentation/SubmittingPatches;hb=HEAD	23:26
Ilari	kbingham_: Aha. Its 'gethostbyname' error... That means it can't even map imap.gmail.com to IP...	23:26
	→ quoin joined	23:27
__mnemoc	it resolves to me	23:27
kbingham_	ping imap.gmail.com	23:27
	PING gmail-imap.l.google.com (74.125.79.111) 56(84) bytes of data.	23:27
	64 bytes from ey-in-f111.google.com (74.125.79.111): icmp_seq=1 ttl=243 time=36.8 ms	23:27
	^C	23:27
	so i don't think its that - i think the gethostbyname muist be red-herring :S	23:27
	not sure how though	23:27
Ilari	kbingham_: Aha... Remove that 'imaps://' part...	23:27
kbingham_	i was thinking the ...	23:27
	ok cheers will try that	23:27
	← zerok left	23:28
kbingham_	hehehe well ... its doing something different ....	23:28
	Resolving imap.gmail.com... ok	23:29
	Connecting to 74.125.79.109:993... ok	23:29
	but now its just sat there... :(	23:29
	i'll go check my gmail drafts	23:29
	← bg_ left	23:29
kbingham_	nope nothing there yet	23:29
git_	why the only way to "merge" a git push from a laptop is "git reset --hard"?	23:29
Ilari	git_: $faq non-bare	23:30
Gitbot	git_: Pushing to non-bare repositories is discouraged. See http://git.or.cz/gitwiki/GitFaq#non-bare	23:30
cbreak	git reset --hard should never merge anything	23:30
	use pull	23:30
	reset will reset things	23:30
	it will remove local changes	23:30
	it's usually not what you want when you merge	23:30
Ilari	git_: Also maybe search for 'firewall' from git FAQ while you are at it...	23:31
	← sdboyer left	23:32
kbingham_	does imap-send have another way to specify to use ssl ? now that i took out the imaps from the [imap] section ?	23:35
	perhaps thats why the connection doesn't continue :S	23:35
	nope and if i add the imap://imap.gmail.com it connects the same as imap.gmail.com	23:38
	→ towski joined	23:38
kbingham_	but with the s there it does the gethostbyname error :(	23:38
	← dkr_ left	23:38
git_	the part I don't understand is how can you "fetch" from a machine inside a firewall as described in http://git.or.cz/gitwiki/GitFaq#push-is-reverse-of-fetch	23:40
jast	that's not really true anyway; in particular, pushing to the branch that's currently checked out in the target repository tends to be a bad idea	23:43
shd	jast: how so? works well for me	23:43
	jast: I do remember if I've pushed to repo with working dir	23:43
jast	they do describe how to use it without running into that problem though	23:43
	shd, see faq non-bare	23:44
Gitbot	shd: Pushing to non-bare repositories is discouraged. See http://git.or.cz/gitwiki/GitFaq#non-bare	23:44
Ilari	git_: Search git FAQ for 'firewall'.	23:44
shd	it's discouraged if one doesn't know how to handle it..	23:44
jast	yes, of course	23:44
	I said "tends to be"	23:44
shd	i often push to working repos between my work and home machine..	23:45
	← k776 left	23:45
cbreak	I have a bare sync repository	23:46
git_	shd, does that mean, u create 3 repos, work is one, home is two, working is three?	23:46
	ah	23:46
cbreak	svn style	23:46
git_	on the server, i need to create bare and non-bare repo	23:46
	→ flaguy joined	23:46
shd	git_: actually, i do have 3 repos, but my employer only sees one of them :)	23:46
	→ hansengel joined	23:46
cbreak	why do you need a working dir on the server?	23:46
	do you work on the server?	23:46
git_	nope, i publish on the server	23:46
	and test	23:47
	→ juanjoc joined	23:49
Ilari	git_: I have two-repo setup, where both repos have working trees (but only unidirectional connectivity is possible...)	23:49
	git_: I wouldn't use working dirs for anything but development. I would do deploy-on-push differently...	23:50
git_	how would one push a web project out?	23:51
	ceej → ceej-away	23:51
Ilari	git_: Have bare repo on server with post-update hook that does something sensible. Git-archive is good way to get snapshots out without working tree...	23:52
	← kumbayo left	23:54
	← quoin left	23:55
Ilari	kbingham_: Strange: 'imaps://host' should be the way SSL IMAPS is done... What version of git you have?	23:56
	kbingham_: And even stranger that it does something with imap:// but not imaps://	23:57
kbingham_	indeed	23:57
	1.5.6.3 is on the bottom of my format-patches	23:57
	i think thats the git version right ?	23:57
Ilari	kbingham_: Aha. Upgrade Git.	23:58
kbingham_	stupid ubuntu :(	23:58
	← agile left	23:59

Logs Search ←Prev date Next date→

Channels Documentation