IRCloggy #git 2010-03-08

Provider of IRC logs since 2005.

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

	2010-03-08
	← dj_tjerk left	00:01
	→ reprore joined	00:01
	← christophsturm left	00:01
	← thiago_home left	00:04
	→ svilayphiou joined	00:06
	← alley_cat left	00:07
	→ marcelinollano joined	00:09
	← robinr left	00:09
svilayphiou	hi, I was working on some branch "experimentel" and wanted to merge it with master. So I checked out master and run git merge experimentl. But in "experimental I renamed a folder and now git is complaining. Here is the git status log: http://dpaste.com/169233/	00:10
	How can I tel git to reflect the "experimental" branch and forget about how master was organized ?	00:10
bob2	renaming, by itself, is fine	00:11
	← topi` left	00:11
	← smuf left	00:11
wereHamster	looks like you renamed the files in both branches	00:11
	← Chaser_ left	00:12
	← bj2 left	00:13
svilayphiou	wereHamster: any solution to fix that? I'd like master to reflect "experimental"	00:15
wereHamster	reflect which changes? All of them?	00:15
	.. completely overwrite master with what is in experimental?	00:15
svilayphiou	yes	00:15
wereHamster	then take a look at the 'ours' merge strategy	00:17
	f00li5h[HireMe] → f00li5h	00:17
	← quag left	00:19
svilayphiou	wereHamster: so if I'm on master and I want to merge experimental and reflect the experimental branch should I run git merge -s ours ?	00:20
	← metalfan_ left	00:24
wereHamster	I guess you didn't read the man page.. (man git-merge -> MERGE STRATEGIES)	00:28
Gitbot	wereHamster: the git-merge manpage can be found at http://git.or.cz/man/git-merge	00:28
	← Sonderblade left	00:32
	→ johnw joined	00:32
	← daaku left	00:36
	→ jasonn1234 joined	00:36
	→ alexmchale joined	00:41
cxreg	hrm	00:44
	i wonder if there's any way to push a set of refspecs atomically	00:44
	← alexmchale left	00:44
cxreg	i'm guessing not	00:44
	→ alexmchale joined	00:45
	← pantsman left	00:45
wereHamster	nope	00:46
	← lemonchicken left	00:46
svilayphiou	wereHamster: found my way out; Thanks	00:47
	→ fool_ joined	00:48
	← fool_ left	00:48
	→ fool_ joined	00:48
	← alexmchale left	00:48
	← fool__ left	00:51
	← tildeequals left	00:55
	← Hackykid_ left	00:58
	→ flaguy48 joined	00:59
	→ richcollins joined	00:59
	← jasonn1234 left	01:00
	→ krawek joined	01:02
	claudine → claudine_away	01:03
	← ntoll left	01:04
	→ zehrique joined	01:05
	→ Mooseous joined	01:07
	← themill left	01:07
	← Mooseous left	01:10
	→ jinzhu joined	01:11
	Mike_lifeguard → Dr_Mike	01:11
	→ Mooseous joined	01:12
	← Alagar left	01:13
	← pilif\|away left	01:18
	→ jasonn1234 joined	01:18
	← halfdime left	01:19
	→ pilif joined	01:20
	pilif → pilif\|away	01:20
	← ttvd left	01:20
	← LRN left	01:23
	← svilayphiou left	01:25
	→ LRN joined	01:25
	← jjuran left	01:26
	→ jjuran joined	01:26
	→ rburton- joined	01:27
rburton-	re	01:27
	small channel :)	01:27
	← segher left	01:27
rburton-	Are there any good tutorials for working with SVN from git. I have git-svn, when issusing git-svn clone -s .... all I see created is a local .git folder	01:27
	← jjuran left	01:28
	→ jjuran joined	01:28
	→ squisher joined	01:29
offby1	dunno about tutorials ...	01:31
	that sounds odd.	01:31
	I too would have expected you'd get a working tree.	01:32
	← unreal left	01:32
offby1	run gitk --all, see if you have anything at all	01:32
rburton-	git-svn clone -s https://svn.apache.org/repos/asf/cassandra/trunk .	01:32
	waiting for it to finish again ;)	01:32
offby1	I don't think it makes sense to use "-s" if you're specifying "trunk" in the URL.	01:32
	claudine_away → claudine	01:32
offby1	I think "-s" means that it'll expect to find "trunk", "branches", and "tags" _beneath_ the URL that you give	01:33
rburton-	I've ran that now because without it I still didn't get it working it's just an ignorant test	01:33
	← zehrique left	01:33
rburton-	removed it and running it again. itll take while	01:34
	so git-svn clone https://svn.apache.org/repos/asf/cassandra/trunk should be fine right	01:34
	→ unreal joined	01:36
	← johnw left	01:36
	← simplechat left	01:36
	← disappearedng left	01:36
	← fr left	01:36
	← bigpresh left	01:36
	← dsop left	01:36
	← Wizzup left	01:36
	← badboy_ left	01:36
	← stuffcorpse left	01:36
	← Plasmastar left	01:36
	← tychoish left	01:36
	← navap left	01:36
	← yzlin left	01:36
	← blasdelf left	01:36
	← jacobat left	01:36
	← trilobyte left	01:36
	← merlin83 left	01:36
	← sw4 left	01:36
	← kozmikreis left	01:36
	← aef left	01:36
	← tk left	01:36
offby1	I would hope so.	01:37
	you'll just get one branch	01:37
rburton-	Which is fine	01:37
	Running it on trunk, branches, etc.. would take a lot of time	01:37
offby1	yep	01:37
	it's painfully slow as it is	01:37
rburton-	right :)	01:38
	→ johnw joined	01:38
	→ simplechat joined	01:38
	→ disappearedng joined	01:38
	→ fr joined	01:38
	→ bigpresh joined	01:38
	→ dsop joined	01:38
	→ Wizzup joined	01:38
	→ badboy_ joined	01:38
	→ stuffcorpse joined	01:38
	→ Plasmastar joined	01:38
	→ aef joined	01:38
	→ tychoish joined	01:38
	→ navap joined	01:38
	→ yzlin joined	01:38
	→ blasdelf joined	01:38
	→ jacobat joined	01:38
	→ trilobyte joined	01:38
	→ merlin83 joined	01:38
	→ sw4 joined	01:38
	→ kozmikreis joined	01:38
	→ tk joined	01:38
	→ LotR joined	01:38
	→ Morasique joined	01:38
unregistered	is it possible to have "git log --stat", along with some indication of deleted files? like a combination of --stat and --name-status	01:38
	← infogulch left	01:39
	→ infogulch joined	01:39
	← squisher left	01:40
offby1	try "git log --raw"	01:40
	not really sure what you want, but I like it	01:40
Dr_Mike	How can I have 'git remote' to always do 'git remote -v' ? Setting an alias seems to have no effect, presumably because it is a git builtin?	01:42
unregistered	offby1: well it combines with stat, not quite how pictured it, not as compact, but it has everything i want	01:43
	thanks	01:43
	← richcollins left	01:43
offby1	Dr_Mike: yeah, the builtin probably trumps your alias. I'd just rename the alias, and form the habit of using it.	01:44
	Dr_Mike pouts :P	01:44
SpookyET	I must be some sort of mutant. I prefer Ruby to Python, but I prefer Django to Rails, and I prefer Git's branching to Mercurial's.	01:45
	← infogulch left	01:45
unregistered	SpookyET: I don't get how the first two relate to the third?	01:47
	inappropriate question mark..	01:47
	→ infogulch joined	01:48
SpookyET	Django is written in Python. Rails in Ruby, Mercurial in Python. git it?	01:48
	← jasonn1234 left	01:48
unregistered	mercurial's language is irrelevant when you use it. and Git is in C. so the relation doesn't fit well :)	01:49
_rane	language doesn't have any relation :p	01:50
SpookyET	Python projects generally use Mercurial.	01:50
	→ beejeebus joined	01:50
	← DavidKlein left	01:51
offby1	SpookyET: OK, OK; you're a mutant.	01:53
	← marcelinollano left	01:53
	← DrNick left	01:53
	→ marcelinollano joined	01:53
	← marcelinollano left	01:53
	→ marcelinollano joined	01:53
	← marcelinollano left	01:53
	→ marcelinollano joined	01:54
	← marcelinollano left	01:54
	→ marcelinollano joined	01:54
	← marcelinollano left	01:54
	→ marcelinollano joined	01:54
	→ tildeequals joined	01:54
	← marcelinollano left	01:55
	→ DrNick joined	01:55
	→ ohadlevy joined	01:56
ohadlevy	hello, I have a feeling that I'm working wrongly with git, I'm trying to have a master branch (aka stable) and an edge branch(where new features comes in), many time, the edge branch will include stuff that I dont want in master, so I end up having to branches which are getting far away from each other, eventually, I cherry pick bug fixes and new stable features into both branches, but that doesnt seem right - am I missing something here?	01:57
unregistered	SpookyET: You are the twisted spawn of a git mascot and the django pony. And you poops out red gems.	01:58
	← Plasmastar left	01:58
	← Yuuhi left	01:58
unregistered	i was trying to make that in gimp and gave up.	01:59
SpookyET	ohadlevy: there is a graph you should check out	01:59
	http://nvie.com/git-model	01:59
	unregistered: Cheap bastard. Use photoshop.	01:59
	→ squisher joined	01:59
ohadlevy	SpookyET: thanks, looking	02:00
	→ alexmchale joined	02:00
	→ Plasmastar joined	02:00
	← aryounce left	02:00
	→ nadavoid joined	02:01
_rane	ohadlevy: maybe you should have feature-specific branches instead of a single edge branch	02:01
ohadlevy	SpookyET: but all of those are cherry-pics?	02:01
beejeebus	hi, i'm trying to pick up work on a feature branch someone else created. i'm looking for the commands to tell me if the feature branch is tracking our HEAD branch	02:02
	← stuffcorpse left	02:02
	← LRN left	02:02
	→ faulhel joined	02:03
	→ TannerF joined	02:03
ohadlevy	_rane: you might be right, but it seems like an overhead	02:03
	→ stuffcorpse joined	02:03
	← coolcat left	02:03
_rane	cherry picking lots of single commits is more overhead, imo	02:03
bob2	ohadlevy: what overhead?	02:04
_rane	with feature branches you can just merge/rebase	02:04
ohadlevy	_rane: because at the end I still want to have an edge branch, so I would need to cherry-pick again regardless	02:04
bob2	ohadlevy: no	02:04
	ohadlevy: you can merge them entirely in to edge, too	02:04
	→ KB1JWQ joined	02:04
	→ sivel joined	02:04
	← KB1JWQ left	02:04
_rane	if you need to cherry-pick from feature branches, you're probably doing something wrong	02:04
	← Morasique left	02:04
	← navap left	02:04
	→ Morasique joined	02:04
	→ navap joined	02:04
ohadlevy	_rane: i guess the question is what is the base branch for the feature branches, is it edge or master?	02:05
	→ pbevin joined	02:06
	ChanServ set mode: +o	02:06
	← CO_gairah_tinggi left	02:06
	sivel set mode: -o	02:06
	← alexmchale left	02:07
	← faulhel left	02:08
squisher	Hi, I'm on branch A, created branch B, renamed a file 'foo' to 'bar', and changed bar. Now I switched back to A, and merged branch B, thinking that it would add bar to the repository. It did, but since bar was based on foo, it disappeared.	02:08
	← nadavoid left	02:08
squisher	What's the best way to bring foo back?	02:09
bob2	was there a delete in there somewhere?	02:10
	or was bar an existing file that you obliterated?	02:10
squisher	I think I did a 'git mv foo bar"	02:10
	the strange thing is, I didn't get a 'merge' entry in the logs	02:12
	← Mooseous left	02:12
bob2	perhaps pastebin your terminal session?	02:12
	→ nadavoid joined	02:12
GyrosGeier	squisher, since B was based on A and A was unchanged, B was a descendant of A	02:13
squisher	bob2, unfortunately it's all over several days, so I don't have that. The outline should be pretty accurate though	02:13
GyrosGeier	merging a descendant yields a fast-forward	02:13
squisher	GyrosGeier, yes, exactly. I understand that, and I know git did the correct thing	02:13
	but I want something else, namely to have foo and bar both present...	02:13
bob2	oh wait	02:13
	you're complaining that foo is gone	02:13
squisher	yes :)	02:13
bob2	I thought it = bar	02:13
	no ambiguous pronouns	02:14
squisher	I know, my bad :-\	02:14
bob2	git did fine, you needed to cp foo if you wanted them to both exist	02:14
squisher	bob2, yes, I know... so the easiest way to recover would be to just check out the previous version, and then copy it into master?	02:15
bob2	git co somerevisionwherefooexisted -- foo ; git add foo	02:16
	I'd say	02:16
	← fr left	02:17
	→ fr joined	02:18
	← aef left	02:19
squisher	bob2, great that worked fine. Thanks for the help, and sorry again about the confusion...	02:19
	→ ttvd joined	02:20
	← mbroeker left	02:21
	← pbevin left	02:23
	→ alexmchale joined	02:23
	← Bombe left	02:23
	→ VVSiz_ joined	02:24
	→ Bombe joined	02:24
	→ jaysern joined	02:25
	← VVSiz left	02:27
	→ spaceonline joined	02:27
	← alexmchale left	02:29
	→ rickjarr joined	02:30
	← sivel left	02:30
	← rickjarr left	02:31
	← nadavoid left	02:37
cxreg	i just realized that the stash keeps your index separate from your work tree with a merge. that's slick.	02:39
	→ alexmchale joined	02:40
	← alexmchale left	02:41
bremner	sitaram: fyi, pretty much what I pasted before. http://www.cs.unb.ca/~bremner/blog/posts/gitolite-mirror/	02:43
	→ _Tristan joined	02:43
_Tristan	how do I rollback to recent commits?	02:43
bremner	sitaram: I'll send you the source after I find a few more errors like the one I just fixed.	02:45
	← tychoish left	02:47
	← disappearedng left	02:48
	→ mr_daniel joined	02:48
	→ tychoish joined	02:48
	→ mdel joined	02:48
sitaram	bremner: how did you fix the pushurl problem?	02:48
Dr_Mike	_Tristan: git revert, I think	02:48
	← mr_danie1 left	02:50
	→ alexmchale joined	02:52
	→ altogether joined	02:53
	→ rawrio joined	02:53
	← parasti left	02:53
	← flaguy48 left	02:57
	→ jasonn1234 joined	02:57
	← altogether left	03:01
	→ disappearedng joined	03:01
	← alexmchale left	03:03
	← neunon left	03:04
	← kingfishr left	03:05
	→ altogether joined	03:06
	→ hyperair joined	03:07
	← kukks left	03:11
	→ alexmchale joined	03:14
	← alexmchale left	03:15
	→ peanut_bu joined	03:15
_Tristan	Dr_Mike: thanks	03:18
	→ munichlinux joined	03:20
	→ cytrinox_ joined	03:20
	→ bcardarella joined	03:20
	→ aryounce joined	03:22
	→ Hannahaus joined	03:22
Hannahaus	http://imgnow.info/DSC-7-03-10.jpg/ do my tits look small?	03:22
	← Hannahaus left	03:22
	← cytrinox left	03:23
	cytrinox_ → cytrinox	03:23
	← beejeebus left	03:24
	← hgimenez left	03:27
	→ hgimenez joined	03:28
Dr_Mike	I can say "yes" to that without even looking. Now GTFO.	03:28
	← hyperair left	03:31
_Tristan	I just got " ! [rejected] master -> master (non-fast forward)" when trying to git push origin master	03:32
	what do I do?	03:32
bob2	you have to pull first	03:32
	you're missing things that the remote master has	03:32
	so pushing would obliterate them	03:32
cbreak	you can force the push and kill everything the other people have done	03:32
	they will hate you for it and call you names	03:33
unregistered	if two people push to a remote repo at about the same time will one of them get an error message?	03:33
cbreak	(usual way is to merge the changes or rebase yours on top of theirs, then push so fast-forward works)	03:33
	← aryounce left	03:34
_Tristan	cbreak: it's my repo, nobody else is helping	03:34
	← Mage left	03:35
cbreak	you only ever pushed from the repository you push now?	03:35
	you never did any history rewriting like --amend or rebase -i or reset?	03:35
_Tristan	mhm, but I just did git checkout to undo a boo boo I made	03:35
	→ Mage joined	03:35
cbreak	checkout doesn't undo anything	03:35
_Tristan	it did for me	03:35
cbreak	it doesn't	03:36
	checkout checks out something	03:36
	it doesn't change history	03:36
_Tristan	I made changes in my files, bad ones. I got old versions.	03:36
cbreak	so it can't undo things	03:36
	did you remove the bad changes from history somehow?	03:36
	that can make fast-forwarding impossible	03:37
_Tristan	I dunno. It works now.	03:37
	→ aresnick joined	03:37
cbreak	(by amending a commit)	03:37
	→ hyperair joined	03:37
_Tristan	I did git pull origin master and it played nice	03:37
unregistered	oh, i just realized i asked the completely wrong/stupid question. If someone pulls frome a repo over ssh, while I'm in the middle of a push, does the puller get an error message?	03:37
cbreak	do a git log to see how nice	03:37
bob2	_Tristan: the thing to remember is that by default you can only push to a branch when your local version is up to date	03:38
	_Tristan: you can force it, but you'll destroy stuff on the other side	03:38
unregistered	I got a call from a friend saying he couldn't push or pull. I had been pushing a lot of changes at the time, and am wondering if that was why.	03:38
	he had no problems after that	03:39
_Tristan	here's my commits: http://github.com/devTristan/Damien/commits/master	03:39
	→ jony joined	03:40
	→ aryounce joined	03:40
cbreak	seems "added unit testing" and "fixed typos" have the same parent	03:42
	so you did that at the same time, from different hosts	03:42
	or	03:42
	you did it parallel in some other way	03:42
	→ glimpse_nirvana_ joined	03:44
_Tristan	shrug I'm just getting to know git. My code started doing segfaults so I rolled back a bit. That could be it.	03:44
	→ alexmchale joined	03:45
	← joeconyers left	03:45
	← SpookyET left	03:46
bob2	rolled back = ?	03:46
	← glimpse_nirvana_ left	03:47
	→ Mooseous joined	03:47
	→ pantsman joined	03:48
	← pantsman left	03:48
	→ pantsman joined	03:48
	→ glimpse_nirvana_ joined	03:51
	← scarabx left	03:52
	→ ankit9 joined	03:54
	← reprore left	03:54
	← alexmchale left	03:56
	← aryounce left	03:56
_Tristan	bob2: uh... I ran git checkout * I think	03:56
	← rawrio left	03:56
	← TannerF left	03:56
	→ TannerF joined	03:56
	← pantsman left	03:57
_Tristan	no, git checkout -f	03:57
	← timja left	03:57
	← TannerF left	03:57
bob2	can't really see how that could cause problems	03:58
_Tristan	I'm definitely not an expert on the matter	03:58
bob2	me either	03:58
unregistered	wouldn't that only throw away local changes.. not actual commits?	03:58
bob2	fixed now, right?	03:58
_Tristan	sort of, there's something weird in my commit log, but I'll probably scrap it once the project is stable anyways	03:59
	← Twisted_Mentat left	03:59
	→ aryounce joined	04:03
	← glimpse_nirvana_ left	04:03
	cxreg ponders a script which includes a "git rebase", and how to resume execution when the user fixes conflicts	04:05
cbreak	./script --resume	04:05
	or --continue	04:06
cxreg	cbreak: um	04:06
	if only life were so simple :P	04:07
cbreak	normal rebase has that	04:07
cxreg	yes thank you, you're very helpful	04:07
	← mr_daniel left	04:08
cxreg	i find myself wishing rebase started a subshell when it errored out	04:09
	→ mr_daniel joined	04:09
	← spaceonline left	04:09
	← Mooseous left	04:09
	→ timja joined	04:09
	← aryounce left	04:09
	← kelvie left	04:10
	→ Mooseous joined	04:11
	← bronson left	04:12
	→ kelvie joined	04:16
	→ Davey joined	04:21
	← bradb left	04:22
	← Mooseous left	04:22
	→ bradb joined	04:22
	→ nevyn joined	04:23
	→ Mooseous joined	04:25
	← hyperair left	04:27
	← mdel left	04:27
	→ kingfishr joined	04:28
	← bradb left	04:31
	→ bradb joined	04:31
	← bradb left	04:31
	← jasonn1234 left	04:32
	→ cedricv joined	04:35
owen1	git clone git@localhost:test.git and i asked for git@localhost's password. it never asks me this when i do it from my laptop. what is this password? the git user?	04:36
bob2	it's ssh	04:37
	so yes	04:37
	presumably you set up keys on your laptop	04:37
_Tristan	don't you have to use keys?	04:37
owen1	bob2: what keys?	04:37
bob2	_Tristan: no	04:37
owen1	and how to set it on the desktop as well	04:37
bob2	owen1: ssh keys	04:37
_Tristan	public keys... they have a tutorial on github	04:37
owen1	it's gitosis, and i am on the same machine where my repo exist.	04:38
	→ SpookyET joined	04:38
SpookyET	What's the deal with the big files git project? I feel dirty saying SVN has an edge on Git in that regard.	04:39
bob2	you get to feel wrong, too, I guess	04:39
	since it's basically the same (except in svn you might not check out the part of the tree that has them)	04:39
owen1	bob2: both my laptop and desktop's keys are in keydir folder	04:39
bob2	ok	04:40
SpookyET	I'd like to version 50-100MB PSD files in Git	04:40
bob2	that will suck in any common vcs	04:40
owen1	bob2: so what is this password? i don't think i have password for git user	04:40
bob2	owen1: perhaps that is the problem then	04:40
owen1	git:x:114:121:git version control,,,:/home/git:/bin/sh	04:40
	not sure what it means..	04:41
bob2	passwords are in /etc/shadow-	04:41
	er no -	04:41
	→ alexmchale joined	04:42
	← Davey left	04:43
	→ taufiqkh joined	04:44
	← alexmchale left	04:44
	→ alexmchale joined	04:45
MetaCosm	SpookyET, wow, let me know how that works out	04:45
SpookyET	MetaCosm: It's sucky	04:45
	Though, In SVN, you can commit 1GB file; it won't die on you.	04:46
owen1	bob2: also, i see my desktop in gitosis.conf	04:46
	← kelvie left	04:48
MetaCosm	SpookyET, do PDFs to binary diff'ing well?	04:49
bob2	pdfs are generally deflated afaik	04:49
_Tristan	PDFs are the devil	04:50
	(pdf is satan spelled backwards)	04:50
MetaCosm	well, I meant PSD's... hehe	04:51
	so, skip that.	04:51
	also, I thought PSD's had thier undo history saved in them	04:51
_Tristan	pretty sure git will avoid non-code	04:52
bob2	git will (try to) commit whatever you ask	04:53
	psds etc are not very binary diffable, thouygh, so will suck up lots of space	04:53
MetaCosm	bob2, that is too bad	04:53
	bob2, do you know if they keep long undo paths in them?	04:53
bob2	dunno, don't think I've ever used photoshop	04:54
chuck	How do I revert back to a certain commit?	04:59
	(Without using git-reset)	04:59
	→ mr_danie1 joined	05:00
	→ aussiegeek joined	05:00
	→ psankar joined	05:01
	← squisher left	05:03
	← jmcantrell left	05:03
	← mr_daniel left	05:03
fryguy	chuck: ?	05:04
	you use git reset	05:04
	→ kelvie joined	05:04
	→ anxiolytic joined	05:05
	← bcardarella left	05:07
	← disappearedng left	05:13
	← alexmchale left	05:14
anxiolytic	My friend and I merged off of a branch awhile back. I want one of his commits to "replace" one of mine. I can merge his branch into mine, but I want to undo some changes I did from a commit awhile back before doing so. Not sure how to do it? I can manually remove/edit the files but I'd rather not	05:15
	← Mooseous left	05:15
	→ dogmatic69 joined	05:15
dogmatic69	hi all	05:15
fryguy	anxiolytic: cherry-pick	05:15
	← jfkw left	05:16
	← aresnick left	05:16
dogmatic69	what does git mean when its says "ahead of origin/dev by 50 commits"? if i push or pull it says everything up to date	05:16
bob2	you're not pushing or pulling to origin/dev	05:16
	presumably	05:16
fryguy	dogmatic69: it means that you have done 50 commits that aren't in origin/ yet	05:16
	origin/dev specifically	05:17
dogmatic69	but i just pushed	05:17
fryguy	where did you push?	05:17
dogmatic69	to dev	05:17
	im on dev	05:17
	and did git push	05:18
bob2	that doesn't indicate that	05:18
	pastebin the whole thing git said	05:18
dogmatic69	well the last push to master http://github.com/infinitas/infinitas	05:18
anxiolytic	fryguy: thanks	05:18
dogmatic69	more than a month ago, so i DID go to dev	05:18
	← kokx left	05:18
fryguy	dogmatic69: pastebin "git status ; git push origin dev"	05:19
	→ kokx joined	05:19
dogmatic69	i cant copy paste from git bash :/	05:19
	← kokx left	05:19
bob2	yeah you can	05:19
dogmatic69	screenie on the way	05:19
	no i cant	05:19
fryguy	fpaste, curlpaste, screen buffer, there's tons of options	05:19
	pick one	05:19
dogmatic69	?	05:19
bob2	why can't you?	05:20
	→ rich97 joined	05:20
	→ kokx joined	05:20
	← kokx left	05:20
	→ kokx joined	05:21
dogmatic69	fryguy: so how i do something like you said?	05:22
bob2	pastebin the output	05:22
fryguy	you install one of the first 2, and then pipe output to it	05:22
dogmatic69	meh	05:23
	like i said, i cant copy paste	05:23
bob2	why	05:23
fryguy	dogmatic69: are you retarded? install curlpaste and pipe output to it	05:23
	or fpaste, or wgetpaste, or whatever your distro provides	05:24
dogmatic69	fryguy: are you a fuck up?	05:24
	→ project2501a joined	05:24
bob2	dogmatic69: chillax and pastebin the information asked for	05:24
fryguy	or just select text and middle click into a pastebin if you are in X11	05:24
dogmatic69	bob2: ye, just dont need to be called a retard	05:24
bob2	dogmatic69: sure, that was out of line	05:25
	← kokx left	05:25
bob2	dogmatic69: but you're being pretty frustrating for someone who's asking for help	05:25
	→ kokx joined	05:25
dogmatic69	well im new to git, running on windows and there is no copy paste option i know of	05:25
	← kokx left	05:25
fryguy	dogmatic69: click the terminal icon	05:25
bob2	you can easily copy and paste from cmd.exe	05:25
fryguy	top left corner	05:25
	→ kokx joined	05:26
	← kokx left	05:26
	→ Mooseous joined	05:26
dogmatic69	http://i45.tinypic.com/30sy15s.jpg	05:26
	→ kokx joined	05:27
	← kokx left	05:27
	→ disappearedng joined	05:27
	→ rchavik joined	05:27
rburton-	I'm hitting the following "RA layer request failed: REPORT of '/repos/asf/!svn/vcc/default': SSL negotiation failed: SSL error: parse tlsext (https://svn.apache.org) at /opt/local/libexec/git-core/git-svn line 4938" with git-svn clone	05:27
	→ kokx joined	05:28
fryguy	dogmatic69: looks like you are pushing to not dev	05:28
rburton-	Not sure how to go about resolving this	05:28
	← andres_ left	05:28
	← kokx left	05:28
dogmatic69	fryguy: like i said, it is pushing to dev	05:28
	← cloudhead left	05:28
fryguy	dogmatic69: then git status doesn't say you are ahead by 57 commits	05:29
	→ kokx joined	05:29
	← kokx left	05:29
dogmatic69	so where did all my commits go?	05:30
	http://github.com/infinitas/infinitas/commits/dev	05:30
fryguy	they are in your local repository, just like git says	05:30
dogmatic69	there is my dev	05:30
	→ kokx joined	05:30
dogmatic69	now im ahead with 58 commits	05:30
	← kokx left	05:30
	← _Tristan left	05:30
bob2	wow, what a dick	05:30
	→ kokx joined	05:31
fryguy	dogmatic69: git remote show origin	05:31
dogmatic69	?	05:31
	← kokx left	05:31
fryguy	dogmatic69: it's a command, run it	05:31
	→ kokx joined	05:32
Fissure	dogmatic69: pulling with a url and branch argument won't update your remote tracking branches	05:32
dogmatic69	you want to see it?	05:32
fryguy	sure, run git fetch while you are at it	05:33
	seems like what's happening is what fissure is saying: you aren't letting git do the tracking for you, so it doesn't know when to update tracking references	05:33
	→ ericlee joined	05:34
	→ abstortedminds joined	05:34
	← abstortedminds left	05:34
dogmatic69	well how i set the repos up is git clone url, git checkout with -- track origiv/dev	05:35
fryguy	dogmatic69: right, and since you are specifying a remote manually when you push, you are bypassing all of the tracking stuff	05:35
dogmatic69	so how would i push?	05:36
fryguy	dogmatic69: git push origin	05:36
Fissure	you don't need any arguments to push	05:36
	or pull for that matter	05:36
fryguy	or, depending on how you have things in .gitconfig and .git/config, just git push	05:36
dogmatic69	git push gives errors	05:36
Fissure	oh?	05:36
fryguy	dogmatic69: git push origin, not git push	05:37
	→ SkyPocket joined	05:37
dogmatic69	nope, the addresses are wrong	05:37
fryguy	so fix them	05:37
dogmatic69	says i cant push to git://..., use git@...	05:37
	yes i will just fix them...	05:38
	← rburton- left	05:38
	→ rburton- joined	05:39
Fissure	well, of course things will break if you clone from the read-only url :P	05:39
fryguy	isn't there an introductory howto on github that explains all of this?	05:40
	→ andres_ joined	05:40
	← SkyPocket left	05:40
	→ roop joined	05:40
	→ alexmchale joined	05:42
	← alexmchale left	05:45
	→ alexmchale joined	05:46
	← ttvd left	05:49
	→ ttvd joined	05:50
Fissure	fryguy: people reading the documentation? you're asking too much	05:50
	→ khmarbaise joined	05:52
	← gdsx_ left	05:53
	→ gdsx joined	05:53
	← gdsx left	05:54
	→ gdsx joined	05:54
	← Mooseous left	05:55
	→ xsdgdsx joined	05:55
owen1	git clone git@localhost:test.git and i asked for git@localhost's password. it never asks me this when i do it from my laptop. i use gitosis and i have the public keys in kefdir folder. any tips?	05:57
gdsx	any chanops around?	05:57
	← xsdgdsx left	05:59
	← munichlinux left	05:59
fryguy	owen1: do you have a private key for the git user onn this machine?	05:59
owen1	fryguy: i don't know, but i will be happy to provide any info you need to figure it out.	06:00
	git:x:114:121:git version control,,,:/home/git:/bin/sh	06:00
	→ werdan7 joined	06:00
fryguy	ls -l /home/git/.ssh	06:00
owen1	fryguy: drwx------ 2 git git 4096 2010-02-20 20:25 .ssh how to access it? it belongs and accessable only to git user	06:02
	fryguy: can i type su?	06:02
fryguy	owen1: become root, or become the git user	06:02
owen1	fryguy: do i become root by typing su?	06:03
fryguy	owen1: sure	06:03
	specifics will depend on various implementation details of your operating system	06:03
owen1	fryguy: i can't become su since root has no password in ubuntu. let me try the other option you mentioned.	06:03
fryguy	sometimes you need to use sudo, sometimes you need to add yourself to a special file first	06:04
	owen1: just type sudo su	06:04
	← disappearedng left	06:04
owen1	fryguy: works. here it is: -rw-r--r-- 1 git git 1052 2010-02-20 20:25 authorized_keys	06:04
fryguy	and have you added your public key to this file?	06:05
owen1	fryguy: not sure, maybe i only added the key of the laptop. 1 sec	06:05
	→ priidu_ joined	06:05
fryguy	wc -l /home/git/.ssh/authorized_keys	06:06
	← alexmchale left	06:06
	← sergio_ left	06:06
owen1	fryguy: i see 3. let me cat the public key of this machine and compare. 1 sec.	06:06
	→ munichlinux joined	06:07
owen1	fryguy: yes. i see it there.	06:07
	→ webus joined	06:08
fryguy	dunno then	06:08
owen1	fryguy: when i am running git clone git@localhost:test.git, i try to clone it as git user, right?	06:09
fryguy	yes	06:10
	← project2501a left	06:10
owen1	fryguy: is there a way to find the password of this user? i don't mind typing it.	06:10
fryguy	no	06:10
	you can just change it to something	06:11
	← dogmatic69 left	06:11
	← webchick left	06:12
	→ askedrelic joined	06:13
owen1	fryguy: works now, i updated the password of git user. now, do i need to enter the public key again to make it skip the password part?	06:14
fryguy	i guess	06:14
owen1	(so i will not have to enter it each time)	06:14
	ok	06:14
	← askedrelic left	06:14
anxiolytic	can I remove everything that was done in an earlier commit, in my current branch?	06:14
	→ askedrelic joined	06:15
fryguy	anxiolytic: git revert	06:15
	or use git rebase -i and make it disappear, depending on if the commit is publicly visible or not	06:15
anxiolytic	fryguy: thx	06:15
	← werdan7 left	06:16
	→ werdan7 joined	06:16
	← Rolenun left	06:16
	← engla left	06:16
	→ Axius joined	06:19
	→ hyperair joined	06:19
	← Plasmastar left	06:19
	→ robinr joined	06:20
	← krawek left	06:21
owen1	fryguy: when i open authorize_keys i see: #autogenerated by gitosis, DO NOT EDIT	06:21
	fryguy: so i am scared (-:	06:21
	→ chid joined	06:22
	→ disappearedng joined	06:22
	→ Plasmastar joined	06:22
owen1	fryguy: all i want to do is replace a key and add it again. i know it sounds stupid, since it looks like the same key.	06:22
chid	damn peanut_bu spammer	06:23
	← chid left	06:23
	Dr_Mike → Mike\|\|gone	06:24
	→ [greg] joined	06:26
[greg]	hi	06:27
	→ chid joined	06:27
[greg]	How can use git+https with custom signed ssl cert ?	06:28
	← PointMan left	06:28
	← khmarbaise left	06:30
	→ lemonchicken joined	06:30
	→ fmardini joined	06:32
	← munichlinux left	06:36
	→ ArdRigh joined	06:37
	← webus left	06:38
	→ bronson joined	06:41
	← claudine left	06:44
	→ archis joined	06:45
	← chid left	06:46
	→ bradb joined	06:47
	← wolog left	06:48
andres_	anyone used git submodules with gitolite?	06:50
	or how if you are the admin for gitolite, how do you refer to an arbitrary repo s.t. it is not gitolite:repo.git	06:51
	← kipras left	06:53
	→ r11t joined	06:54
	← robinr left	06:54
johnw	i'm using submodules with gitolite	06:54
	← priidu_ left	06:55
	← madewokherd left	06:55
	→ rseifert joined	06:55
	← Octalot left	06:56
andres_	johnw: what's the correct way to add a submodule then, would it be git submodule add gitolite:repo.git repo if I other people in working with the parent repo also need the submodule?	06:56
johnw	i used: git submodule add ssh://gitolitehost/project.git project	06:56
	← rburton- left	06:57
andres_	johnw: the other concern is that when I added the submodules like that, they would not appear in clones of the repo, although maybe it would be fixed by what you just shared, let me try..	06:57
	← [greg] left	06:57
johnw	they won't appear in clones until the user runs git submodule --init update	06:57
	→ robinr joined	06:58
andres_	was following this tutorial http://git.wiki.kernel.org/index.php/GitSubmoduleTutorial which made it seem like once you pull, you get the submodule directory, but not the submodule. I wasn't even getting the .gitmodule	06:59
erikvold	so I made a dump dev branch in a github repo, and the branch was pushed to github, then I went git branch -D dev_branch locally, then I did a git push, and got 'Everything is up-to-date'. How can I delete the branch on the remote repo?	07:00
	← mark[oz] left	07:01
erikvold	should I git push -f ?	07:01
	→ rad182 joined	07:02
andres_	johnw: the path to the submodule should it be fully qualified ssh://host.com/home/git/repositories/module.git or is ssh://hos.com/module.git enough?	07:02
rad182	what whould be the difference if i initialize my git repo with	07:02
	init and --bare init ?	07:02
Fissure	erikvold: use the full refspec syntax and push the empty string to the branch	07:03
	← robinr left	07:03
Fissure	e.g. git push origin :deleted-branch	07:03
andres_	rad182: bare will not contain the actual files in a way you can browse them	07:03
	→ aspotashev joined	07:03
johnw	on my system i do not need the full path	07:04
	← bambule left	07:04
andres_	johnw: is that a public repo?	07:04
johnw	no	07:04
	← aspotashev\|eeepc left	07:04
Fissure	andres: gitosis effectively chroot's you to the repositories dir	07:05
	and it's treated as your home directory	07:05
rad182	andres_: will there be any differnece with storage usage?	07:05
andres_	rad182: not sure about that....	07:05
	you could test it out :)	07:05
erikvold	Fissure: hmm I'm confused, so you think 'git push origin :dev_branch_name' where dev_branch_name is the name of the branch should work?	07:05
Fissure	yes	07:06
erikvold	what's the ':' about?	07:06
Fissure	assuming origin is the proper remote	07:06
	← jinzhu left	07:06
Fissure	read the manpage	07:06
erikvold	yep it is	07:06
rad182	cuz without the --bare, when i push i can see a compressing objects but with --bare i don't	07:06
andres_	Fissure: but I'm using gitolite, is there a difference?	07:06
Fissure	andres: probably not, but i've never used gitolite myself	07:06
	→ dpino joined	07:07
	→ WebDragon joined	07:07
andres_	Fissure: but I would still have to give it the user so something like git submodule add [email@hidden.address] ?	07:08
johnw	you can easily use a cron job to gc all the bare repos	07:08
sitaram	andres_: gitolite is no different from any other ssh-based server you may be using; it and the mechanism of submodules are quite unrelated to each other. Please do not confuse yourself by asking about how gitolite works with submodules :)	07:08
erikvold	Fissure: ic now, thx	07:08
	→ maxthoursie joined	07:08
	← archis left	07:09
sitaram	andres_: just ask the following question to people who know submodules: how can different people refer to the same repo by different URLs when the repo is actually a submodule?	07:09
andres_	sitaram: I agree, but being the admin, I usually use gitolite:to access repositories, or what should I be using as the url, [email@hidden.address]	07:09
	→ jinzhu joined	07:11
	→ hipe joined	07:11
	← dreiss left	07:12
	← dropdrive left	07:14
	→ dropdrive joined	07:14
	← mahound left	07:14
	→ ntoll joined	07:15
	← rad182 left	07:16
johnw	hey there sitaram	07:16
	i e-mailed you about the .spec file	07:16
	→ mastro joined	07:16
	→ wasmahen joined	07:16
wasmahen	hi, i have created a url shortener, anyone would want it?	07:16
	→ mtkd joined	07:17
andres_	sitaram: how should I be cloning repos?	07:17
	← wasmahen left	07:18
	← bauruine left	07:18
	← Pewpewarrows left	07:19
	← VVSiz_ left	07:20
	→ VVSiz joined	07:20
	→ r11t_ joined	07:21
	→ aziz joined	07:21
	← r11t left	07:22
VVSiz	hi guys, peanut_bu spams me when I join #git. is that OK? :)	07:22
	→ thiago_home joined	07:22
	← ohadlevy left	07:23
	← Bombe left	07:23
	→ Bombe joined	07:23
thiago_home	peanut_bu has an on-join	07:25
	← Pupeno left	07:26
	→ bentob0x joined	07:27
	← SpookyET left	07:29
	← greg8si left	07:30
	→ ph^ joined	07:30
	← altogether left	07:31
	→ mklappstuhl joined	07:32
	→ gebi joined	07:33
	→ napster_123 joined	07:34
napster_123	hey i'm trying to set a git repo on windows and have used msysgit + followed it's instructions to make a central repo on windows	07:35
	everything looks set up now... but whenever i try to clone the git repo i get "fatal: The remote end hung up unexpectedly"	07:35
	→ priidu_ joined	07:36
	← Garen left	07:36
napster_123	for the sake of trial i am using "git clone user@localhost..." (meaning cloning the git repo being on server machine)	07:36
VVSiz	naderman: git clone C:/path/to/repo should work though, right/	07:37
	oops, I meant napster_123	07:37
napster_123	VVSiz, yeh sure... but i was testing for the case of remote client	07:37
	VVSiz, i think it should work since you can always ssh to your same machine	07:38
VVSiz	and you have ssh-server running and can conect to it?	07:39
	→ voxpelli joined	07:39
napster_123	VVSiz, yes i've tried connecting through ssh using putty successfully	07:40
VVSiz	weird	07:40
napster_123	VVSiz, but i've noticed that when i connect through putty ... on login the shell says "-bash: $'\r': command not found"	07:41
	→ psoo joined	07:41
napster_123	VVSiz, does it have something possibly to do when i try to clone my repo?	07:42
VVSiz	but do you really need to setup a remote repo on Windows? :) I have my central repos on Linux (in virtual machine on Windows host) or on github, if they are public.	07:42
	← jaysern left	07:42
VVSiz	napster_123: quite possibly	07:42
napster_123	VVSiz, in my office, unfortunately, i have only windows machines all around... otherwise linux would've been my obvious choice for private repo hosting	07:43
andres_	johnw: aside from doing a git submodule add , and then commiting, is there anything else you need to do?	07:43
johnw	what about linux in a VM?	07:43
	andres_: not really, no	07:44
	→ Garen joined	07:44
andres_	johnw: so I should be able to pull from another clone of the repo, and it WOULD create the submodule directory, or do I have to do anything else in other clones?	07:44
napster_123	sure i can try that... but i was looking for a no-weight way introducing git ... (we are on svn ... and i'm trying to convince then to use git)	07:44
johnw	after cloning, git submodule --init update	07:45
	→ giallu joined	07:45
johnw	napster_123: ah, I just did that convincing myself	07:45
	napster_123: i finally switched our house from svn->git last weekend	07:45
napster_123	johnw, :)	07:45
johnw	i had the IT setup a dedicated Linux box for our version control	07:45
	→ dreiss joined	07:46
johnw	but all the users are Windows, except me	07:46
napster_123	johnw, i've got one box approved as CI server but currently it doens't exist you see	07:46
VVSiz	I wonder if hosting a git repo on samba share would be a simple alternative...	07:47
napster_123	johnw, when that box comes into existance only then i'll have an opportunity to host my git repo cleanly on a linux box	07:47
	VVSiz, yeh sure but git push wouldn't be available on that right ?	07:48
johnw	a VM is pretty clean too	07:48
	and it gives you a chance to play/prove	07:48
VVSiz	why not, if the share allows write access. the only question is that safe? since multiple users might try to commit concurrently, but I think git is smart enough to handle that.	07:48
napster_123	johnw, i guess you'r correct i'll have to fallback to a vm after more couple of hours wasted... so why not do it now	07:48
	← r11t_ left	07:48
napster_123	johnw, vbox is more lightweight than vmware right ?	07:49
VVSiz	vbox works just great here	07:49
johnw	i don't know, I use vmware only	07:49
VVSiz	along with Ubuntu server VM	07:49
johnw	since I use VMware Server on some of our servers, it lets me just rsync over VMs that I want to deploy live	07:49
	← WebDragon left	07:49
napster_123	VVSiz, ok i'm going for vbox	07:49
VVSiz	vbox is definitely more lightweight and faster for me (and free) :)	07:50
napster_123	ok now on windows i just need the latest msysgit and that's it right?	07:50
johnw	yeah	07:50
	i like TortoiseGit too	07:50
	it makes convincing much easier	07:50
sitaram	andres_: whatever you use as the URL, it should not affect anyone else's use of the same repo/subrepos, because they get to change their URLs at some point if they need to	07:50
napster_123	no putty no nothing right?	07:50
sitaram	johnw: ??	07:50
johnw	no, it comes with OpenSSH	07:50
	← ph^ left	07:50
johnw	sitaram: the RPM spec file for gitolite	07:50
napster_123	johnw, alrighto	07:50
andres_	sitaram: thanks	07:50
	→ roop_ joined	07:51
	← roop left	07:51
andres_	sitaram: on another note, trying to clone a repo, if I use the gitolite:repo.git shortcut, it works fine, but if I actually use the full form, I have to use git clone [email@hidden.address]	07:51
	sitaram: bug or feature?	07:51
thiago_home	andres_: what does "gitolite" expand to?	07:53
	→ WebDragon joined	07:54
andres_	thiago_home: http://dpaste.com/169346/	07:54
thiago_home	I doubt it expands to that URL...	07:55
	→ christophsturm joined	07:55
thiago_home	andres_: are you sure you're not using url.<base>.insteadOf config?	07:56
sitaram	johnw: aah yes, I did reply to that didn't I?	07:56
	andres_: doc/6	07:56
johnw	indeed you did	07:56
	i was just saying hi :)	07:56
andres_	thiago_home: not sure what you mean	07:57
sitaram	johnw: oops! sorry :) I'm a little disoriented right now... hi!	07:57
johnw	i'm loving gitolite, thanks	07:57
	i get to confine our newbie interns to read-only branches	07:57
	and read-write branches they can deposit code to	07:58
	→ kkoehne joined	07:58
	→ jaysern joined	07:58
andres_	sitaram: great, my ssh problems did not end yesterday....	07:58
johnw	that feature is actually what let me convince the CEO that we had to use Git	07:58
	→ patgodo joined	07:58
sitaram	johnw: :) you're welcome!	07:59
	MacGyverNL_ → MacGyverNL	07:59
	→ wolog joined	07:59
sitaram	johnw: wait wait... you're saying gitolite helped sell git? wow man, you made my day (if I understood correctly!)	08:00
	← christophsturm left	08:00
sitaram	andres_: going back to your original question, using the full form bypasses gitolite completely, and you will find pushes blocked :)	08:01
	← thiago_home left	08:01
andres_	yup, read the doc :)	08:01
	ok, I'm scratching this...	08:01
	starting from the beginning...	08:01
	→ dj_tjerk joined	08:01
andres_	this is absurd	08:01
sitaram	andres_: if this is getting too confusing, why don't you just use some other [workstation] account for the gitolite admin and make yourself just a normal user? Then you'd use the git@server:repo syntax	08:01
	← magn3ts left	08:02
sitaram	andres_: and the special gitolite:repo syntax would be used only by that new user you created, to work on gitolite-admin repo exclusively	08:02
andres_	sitaram: I'm looking at the sever now. I had started installing gitosis, on the same git account I use with gitolite. Could things be getting messed up?	08:02
sitaram	andres_: http://sitaramc.github.com/0-installing/9-gitolite-basics.html also has some (hopefully well written) explanation of all the ssh stuff	08:02
	← mklappstuhl left	08:03
sitaram	andres_: it's been ages since I used gitosis but I guess if you don't get rid of it properly (I have a migration guide in my doc/ you know) then anyone pushing the gitosis-admin repo can possibly screw up the authkeys file	08:03
	→ Pupeno joined	08:04
andres_	sitaram: yes, I saw that guide. I'm going to try to uninstall it and then start from scratch...	08:04
	sitaram: thanks for the other pointer	08:04
sitaram	you're welcome!	08:06
	← mastro left	08:07
	← jaysern left	08:07
	← anxiolytic left	08:08
	→ steffkes joined	08:09
rich97	If git says "On branch 0.2-page-management. Your branch and '0.2' have diverged." what is the bet way I can get the changes from 0.2 into 0.2-page-management?	08:09
	← Bass10 left	08:09
	→ trebor_dki joined	08:09
	→ lucsky joined	08:10
	← taufiqkh left	08:10
	→ magn3ts joined	08:11
	→ nmemon_ joined	08:12
	→ r11t joined	08:13
	→ anxiolytic joined	08:15
	→ j416 joined	08:16
	← nmemon_ left	08:16
	← napster_123 left	08:16
	← Axius left	08:17
	← stringo0 left	08:19
	→ palisv joined	08:19
	← WebDragon left	08:20
	← gebi left	08:20
	← owen1 left	08:21
	← hyperair left	08:22
	→ roll joined	08:22
	← patgodo left	08:22
	← rchavik left	08:24
	→ ph^ joined	08:25
	→ thiago_home joined	08:25
	→ mattayers joined	08:25
	→ christophsturm joined	08:26
	→ JonathanT joined	08:28
maxthoursie	rich97: if it's not published, a git rebase 0.2 should do	08:31
rich97	maxthoursie: Thanks	08:32
maxthoursie	rich97: if it is published, git merge 0.2 would be better	08:32
rich97	I'll rebase :)	08:32
	I just wanted to know the proper way, thanks	08:32
	← magn3ts left	08:34
	→ darwin joined	08:34
	→ aatanasov joined	08:40
	→ munichlinux joined	08:41
	→ napster_123 joined	08:41
	→ hyperair joined	08:42
	← palisv left	08:44
	→ nicoulaj joined	08:44
	← thiago_home left	08:45
	→ doobry joined	08:45
	→ Hackykid joined	08:45
	← erikvold left	08:45
	→ project2501a joined	08:45
	→ erikvold joined	08:47
	→ Enoptromancy joined	08:48
	← isf_ left	08:51
	← hachi left	08:52
	→ hachi joined	08:52
	→ tobiassjosten joined	08:52
	→ isf_ joined	08:52
	ChanServ set mode: +o	08:53
	← peanut_bu left	08:53
	werdan7 set mode: -o	08:53
munichlinux	git fetch gitosis@xxx:xxx its ask for a password rather than keyphrase	08:53
	← psoo left	08:53
	→ psoo joined	08:54
	← r11t left	08:54
christophsturm	munichlinux: try ssh gitosis@xxx	08:58
	→ tbf joined	08:59
	← project2501a left	09:01
	← Enoptromancy left	09:01
	← johnw left	09:02
	→ poseid joined	09:02
	→ JustDark joined	09:03
	← darwin left	09:04
	← munichlinux left	09:05
	← domcat left	09:07
	→ Paraselene_ joined	09:08
	→ BearPerson joined	09:08
	→ PointMan joined	09:09
	→ shruggar joined	09:10
	→ munichlinux joined	09:10
	← isf_ left	09:11
	← GarethAdams\|Home left	09:11
BearPerson	moo.	09:11
	→ isf_ joined	09:11
	← JustDark left	09:11
	→ great_cogan joined	09:12
	→ elyob joined	09:12
	→ radarek joined	09:13
	→ roop joined	09:13
	→ segher joined	09:14
	← roop_ left	09:14
	← radarek left	09:15
	→ radarek joined	09:15
	→ hakunin joined	09:15
	sitaram clutches his heart... munichlinux attended a talk by him on git and gitolite and is using gitosis :)	09:17
	→ project2501a joined	09:18
	← cedricv left	09:20
	f00li5h → moar-f00li5h	09:22
	moar-f00li5h → f00li5h	09:22
	← dj_tjerk left	09:25
	← sylr left	09:29
	→ tvw joined	09:31
	→ themill joined	09:32
	← dreiss left	09:33
	→ jelmer_ joined	09:35
	← jelmer left	09:36
	→ g0bl1n joined	09:36
	← project2501a left	09:39
	← PointMan left	09:42
	← hipe left	09:42
Pieter	oaeu	09:42
	→ hipe joined	09:42
	→ gitte joined	09:43
	← tango_ left	09:43
	← Plasmastar left	09:44
	→ PointMan joined	09:44
	← ArdRigh left	09:46
	→ bauruine joined	09:46
	← erikvold left	09:48
	jelmer_ → jelmer	09:50
	→ pharm joined	09:51
	→ fool__ joined	09:51
	← fool__ left	09:51
	→ fool__ joined	09:51
pharm	Help!	09:51
	I have a branch I've been working on in two locations (different git repositories)	09:51
	pushing back & forth to keep them in sync.	09:52
	← fool_ left	09:52
pharm	Now I seem to have ended up in a state where I have a file which is different in the two repos, even though the log is the same & the checksum is the same.	09:52
	This isn't supposed to be possible is it?	09:52
	I've done a complete git log on the two ends & all the checksums match	09:55
	and git diff shows no differences at either end	09:55
	yet I have differing files!	09:55
	→ Pewpewarrows joined	09:59
	→ engla joined	10:00
locklace	different how	10:00
	← engla left	10:00
	→ engla joined	10:00
pharm	OK. I'm going to go tear my hair out somewhere else.	10:01
	← daed left	10:02
	← godsflaw left	10:02
	← hipe left	10:03
	→ hipe joined	10:03
	← SRabbelier left	10:04
	→ intripoon joined	10:05
unregistered	is "git checkout root_directory" equivalent to doing a "git reset --hard"?	10:08
	the former seems more intuitive	10:09
Saur	unregistered: No, they are not the same	10:10
	→ mbroeker joined	10:10
	← tvw left	10:11
unregistered	what's the difference? i've been using the former on those rare occasions I wanted to throw away the changes in my working directory	10:11
	← arekm left	10:12
	← uau left	10:12
unregistered	Oh is it that. "reset --hard" will throw away the staged files too?	10:12
	or is there more to it than that?	10:13
	← pharm left	10:16
	→ DavidKlein joined	10:17
	← engla left	10:18
wereHamster	reset --hard has an implicet HEAD, checkout uses the index	10:20
	so if you have staged changes, the two commands would copy different things into your working tree	10:21
	← orafu left	10:27
	→ orafu joined	10:27
	← anxiolytic left	10:32
	→ ixti joined	10:32
	← munichlinux left	10:33
	→ PointMan_ joined	10:35
	→ mklappstuhl joined	10:35
	← andres_ left	10:36
	← PointMan left	10:37
	→ corni joined	10:37
	PointMan_ → PointMan	10:38
	→ guyzmo joined	10:40
guyzmo	hi	10:40
	how can I make git ignore changes on one line in a file ? (I got versionned generated files that have a "Generated on XX/XX/XX" in them)	10:41
	← unijunction left	10:41
	→ phoenix24 joined	10:41
phoenix24	I've forked a project on github., and now i need to update my fork, with the changes made at the original master	10:43
	how can i do that ?	10:43
ben_h	there are two main ways, phoenix24	10:43
wereHamster	guyzmo: is that file generated from other files which are in the repository?	10:43
ben_h	you can merge the original, or rebase on it	10:43
	if you plan to contribute back, and you haven't already pushed your code, i'd suggest rebasing.	10:44
	merge is a traditional merge	10:44
	rebase means, take the commits you've made, and re-apply them onto the updated original	10:44
	→ Mandar joined	10:45
phoenix24	ok	10:45
ben_h	i'd do a	10:45
	'git fetch remote_name'	10:45
guyzmo	wereHamster - well it's complicated, those files shouldn't be versionned, but for "comfort" the others don't want to generate them and want them versionned	10:45
ben_h	'git rebase remote_name/master'	10:45
guyzmo	and the project manager is always right, even when he's wrong...	10:45
ben_h	lol	10:45
Aides	guyzmo: you can, for example, remove and readd that line with smudge/clean	10:46
wereHamster	guyzmo: well, to answer your question: no, not possible	10:46
phoenix24	ok, thanks a lot ben_h	10:46
wereHamster	or not put that line into the generated fil	10:47
guyzmo	wereHamster - oh, ok, I'm pretty sure it's possible in svn with annotations :)	10:47
	→ BryanWB joined	10:48
BryanWB	i did git pull on my current branch and not happy w/ the results of the auto-merge	10:49
	how can I revert the merge commit and replay the merge w/ a different strategy?	10:49
wereHamster	git reset --hard ORIG_HEAD	10:49
BryanWB	wereHamster, is "ORIG_HEAD" what I should literally type ?	10:50
wereHamster	yes	10:50
BryanWB	wereHamster, tks	10:50
	what is the difference between ORIG_HEAD and HEAD?	10:53
	← napster_123 left	10:53
locklace	HEAD is HEAD, ORIG_HEAD is what HEAD was before you did something dangerous and possibly stupid	10:53
	→ abstortedminds joined	10:54
	← abstortedminds left	10:54
bremner	sitaram: the "pushurl problem" was me. I had two host aliases pointing to the same host	10:54
	→ napster_123 joined	10:55
BryanWB	locklace, tks	10:55
	← j416 left	10:55
	→ unijunction joined	10:56
	← Mandar left	10:56
	→ andres_ joined	10:57
	→ d0k joined	11:00
	← Wicked left	11:00
	← jspiros left	11:01
	→ tomekw joined	11:02
	→ nithin joined	11:02
	← mklappstuhl left	11:04
	→ mklappstuhl joined	11:04
	→ timj_ joined	11:05
sitaram	bremner: in the ssh config file?	11:05
bremner	yeah, Ilari finally guessed that was the problem. Needless to say I feel pretty silly.	11:06
	especially since my instructions say to switch the gitolite alias back after the second install	11:08
	← BryanWB left	11:08
sitaram	bremner: yeah this is why I'm beginning to hate ssh. Deceptively simple, but stray off the beaten path and you're screwed	11:08
	← timja left	11:09
Rhonda	sitaram: I stumbled into a different issue this weekend: Had a Host * entry at the end, and gitolite puts itself at the end too, but because first match wins it didn't ever get there. :)	11:10
bremner	well, ssh -v gitolite could have found this pretty quick	11:10
sitaram	Rhonda: yeah basically ssh will trip if you're not careful. And often, even if you are	11:10
	Rhonda: that's why doc/6 is there... and it'll usually catch everything if you go through the whole doc :)	11:11
	← _Vi left	11:11
bremner	is the gitolite ssh alias needed for anything during the install other than cloning gitolite-admin?	11:11
sitaram	no	11:12
	Rhonda: someone asked if I could put the gitolite entry at the top instead. I guess there's no risk, so I added it to todo	11:12
Rhonda	I would think that the alias isn't needed during the install because the install should set commandline options appropriately. :)	11:12
	← rseifert left	11:12
sitaram	Rhonda: it's needed during the last step of the install, when easy-install does a "courtesy clone" of your admin repo. Not really needed, but still...	11:13
	→ squentin joined	11:13
	← dpino left	11:15
	→ rgr joined	11:17
	→ s_edrik joined	11:20
	← mklappstuhl left	11:21
s_edrik	how can I recall an old version of a git repository, I want to see if a feature worked in that revision or if I screwed up.	11:21
	→ mosno joined	11:22
wereHamster	s_edrik: git checkout <commit>	11:22
	→ kumbayo joined	11:23
s_edrik	hmm	11:23
	thanks	11:24
	=)	11:24
	← alezandro left	11:24
Rhonda	sitaram: Hmm, have to check if that happened over here. :)	11:25
	→ j416 joined	11:26
	→ tatsuya joined	11:31
	← Theravadan left	11:32
	→ dpino joined	11:32
	← frogonwheels left	11:32
	← PointMan left	11:33
shruggar	s_edrik, also look into "bisect"	11:35
s_edrik	will do'	11:35
	← bronson left	11:37
	→ project2501a joined	11:37
	→ carolyn joined	11:37
	→ PointMan joined	11:38
	→ frogonwheels joined	11:42
	→ Textmode joined	11:42
	← project2501a left	11:44
	→ jln_1 joined	11:45
disappearedng	help!	11:46
	Unifier/data.short/bakup/db.pickle \|1751349 ++++++++++++++++++++++++++	11:46
	→ Yuuhi joined	11:46
disappearedng	I realize that I added a huge file into my repos	11:46
	and commited	11:46
	how do I remove from the repos?	11:47
	is git rm --cached what I want?	11:47
	→ poona joined	11:47
shruggar	disappearedng, have you pushed?	11:48
disappearedng	no	11:48
	not yet	11:48
	→ rolfb joined	11:48
shruggar	disappearedng, is it the last commit you made?	11:48
disappearedng	it is	11:48
	just commited	11:48
shruggar	git reset HEAD^	11:48
wereHamster	git rm --cached <file>; git commit --amend	11:48
disappearedng	ok there was 2 branch	11:49
	ZOL and master	11:49
	it was ZOL's last commit	11:49
	← bauruine left	11:49
disappearedng	and I merged ZOL into master	11:49
	so both branch's last commit	11:49
poona	Hi. I am merging 2 branches which results in a conflict. The git tutorials says during a merge conflict we have three versions of the file. If I do a	11:49
	git show :1:my_file.c	11:49
	I get these errors - fatal: ambiguous argument ':2:util-error.h': unknown revision or path not in th	11:49
	e working tree.	11:49
	Use '--' to separate paths from revisions	11:49
	What am I doing wrong?	11:49
wereHamster	poona: does git status list that file as unmerged?	11:50
poona	Yes	11:50
	wereHamster: yes	11:50
	← PointMan left	11:50
	→ parasti joined	11:50
wereHamster	disappearedng: then you will have to undo the merge, amend the commit and do the merge again	11:50
	poona: is 'util-error.h' the full path to the file?	11:51
disappearedng	how do I undo a merge	11:51
	git reset --hard HEAD^?	11:51
wereHamster	yes	11:51
disappearedng	git checkout master	11:51
	error: Untracked working tree file 'Unifier/data.short/bakup/db.pickle' would be overwritten by merge.	11:51
wereHamster	git reset Unifier/data.short/bakup/db.pickle; git checkout mastre	11:52
poona	wereHamster: Yes. I am inside the source directory, which holds the file util-error.h	11:52
wereHamster	poona: you need to use the full path in :2:<path>	11:52
	→ andre____ joined	11:52
andre____	hello everyone	11:52
disappearedng	still wereHamster: error: Untracked working tree file 'Unifier/data.short/bakup/db.pickle' would be overwritten by merge.	11:53
wereHamster	disappearedng: sorry, git reset HEAD -- Unifier/data.short/bakup/db.pickle	11:53
andre____	I'm having a problem with preserving a merge commit when rebasing	11:53
	→ Ryback_ joined	11:53
poona	wereHamster: Tried the full path. the same error	11:53
disappearedng	error: Untracked working tree file 'Unifier/data.short/bakup/db.pickle' would be overwritten by merge.	11:54
	still	11:54
wereHamster	andre____: I assume you are using rebase -p, right?	11:54
mr_danie1	I use git for SCM, which is a great tool. But despite a system to track source code history and make code syncronisation between developers easy, I think also need some sort of todo-list management. Every day a developer has a new idea for a feature or proposal to change the architecture in a way, and do not those ideas to be lost. Every other develoepr should be able so see those ideaas and todos.	11:54
wereHamster	disappearedng: argh, just remove the file then	11:54
andre____	wereHamster: correct. A minimal example can be found on http://github.com/awendt/git-task	11:54
Textmode	morning all	11:54
mr_danie1	git is a greate tool for SCM, but is there also a great tool for 'idea and todo-list management'?	11:54
andre____	I've tried `git rebase -i -p --onto master master^` which still shows the merge in the interactive listing	11:55
disappearedng	wereHamster: so now I am on master	11:55
	do I have to ammend again?	11:55
andre____	but after the rebase, the merge commit is gone. it's a fast-forward	11:55
	this happens on 1.7.0.2 and 1.6.5	11:56
wereHamster	disappearedng: well, amend that commit in which you added that big file	11:56
poona	wereHamster: Tried the full path in git show :1:<full_path_to_file>. Still get the same error	11:57
Textmode	mr_danie1: I usually use a wiki for that sort of thing.	11:57
wereHamster	andre____: it could be that rebase -p doesn't preserve fast-forward merges	11:57
disappearedng	is there a command in git that shows all the files being tracked ?	11:57
wereHamster	disappearedng: git ls-files	11:57
andre____	wereHamster: bummer	11:58
Textmode	mr_danie1: though I suppose if you were really familiar with git's commands, and all the documents were related (conceptually), git would do fine.	11:58
disappearedng	thx	11:58
Textmode	mr_danie1: the only real issue is that git manages entire sets of files, rather than individual documents, depending on what you need that might actually be what you want.	11:59
mr_danie1	you mean I should create a plain text file 'todo' it let it track by git Textmode ?	12:00
andre____	wereHamster: is there any way I can preserve fast-forward merges when rebasing?	12:00
locklace	i use it for documents, and deal with that by putting unrelated docs in different branches	12:00
	→ Chillance joined	12:00
wereHamster	andre____: can you describe how you would like the history to look after the rebase? (maybe in a pastie using an ascii graph)?	12:00
disappearedng	ok cool thx man	12:01
	← roll left	12:01
	← jinzhu left	12:02
Textmode	mr_danie1: if you are just talking about adding some todo/changelogs (or similar-such) to the project file set, then have at it, thats common, and there won't be any problems, since the state of those documents is logically related to the state of the rest of the project, which is git's job to a t	12:02
wereHamster	poona: can you paste the output of git ls-files --stage?	12:02
mr_danie1	that sounds reasonable Textmode	12:03
Textmode	mr_danie1: but if you are actually looking for versioned management of large collections of documentation and texts, you might be happier with one of the many wikis or document management systems out there.	12:03
poona	wereHamster: Yeah. I can see 3 versions for util-error.h	12:04
	100644 a9c048eaeb3640070f8990246957fdf79b96efd9 1 util-error.h	12:04
	100644 7184189fca5df1aec1b02ba693f20e2384d9d424 2 util-error.h	12:04
	100644 eab621eb44c2cddf0cd128af19d87f8d5399b833 3 util-error.h	12:04
Textmode	particularly if the documents are disparate	12:04
	→ froschi joined	12:05
wereHamster	poona: does git show :0:util-error.h work?	12:05
	← jln_1 left	12:05
	→ kukks joined	12:05
poona	wereHamster: nope. the same error message - fatal: ambiguous argument ':0:util-error.h': unknown revision or path not in th	12:06
	e working tree.	12:06
andre____	wereHamster: sure -- http://gist.github.com/325096	12:06
	→ DamienCassou joined	12:07
DamienCassou	hi	12:07
wereHamster	andre____: I'd say that's impossible	12:07
	→ _Vi joined	12:07
	→ ringlej joined	12:07
andre____	wereHamster: But why oh why?	12:08
wereHamster	andre____: how shoud git know that the new parent of F should be M? You are rebasing T, not F...	12:08
ringlej	Is there a way to list local branch list with last access/modify time for each branch?	12:08
	→ solow joined	12:10
solow	Is it possible that my browser is no longer displaying changes on my localhost because of git?	12:10
wereHamster	ringlej: you could list the commit dates/times for each branch	12:10
andre____	wereHamster: Well, git-rebase seems to know about the merges when listing the commits for -i	12:11
	→ alex____ joined	12:13
alex____	hi! anybody is using nbgit? I only commit locally, my changes aren't pushing to the server.	12:14
andre____	wereHamster: Sorry, of course I messed up the graph, I've updated it, please have another look at http://gist.github.com/325096	12:14
wereHamster	andre____: if you really want to find out if that is possible, you'll have to ask on the mailing list	12:15
poona	wereHamster: any tips that can help me with the git show?	12:16
wereHamster	poona: git --version?	12:17
napster_123	is there any way i could clone a mercurial repository? i just want to init a git repo on my old mercurial repo but don't want to lose the history	12:17
poona	wereHamster: 1.5.5	12:17
locklace	figures	12:17
wereHamster	napster_123: <contrib/hg-to-git>	12:18
	er, <HEAD:contrib/hg-to-git>	12:18
Gitbot	[git hg-to-git]: http://tinyurl.com/yflbdnf [tree]	12:18
wereHamster	poona: can you update? 1.5 is _really_ old	12:18
poona	wereHamster: too old a release to have this feature?	12:18
	wereHamster: okay. will update and try it	12:18
	thanks	12:18
	→ LRN joined	12:18
	← dcampano left	12:18
	→ dcampano joined	12:19
	→ parigaudi joined	12:19
andre____	wereHamster: Thanks, I will try the ML	12:19
solow	Is it possible that my browser is no longer displaying changes on my localhost because of git?	12:21
napster_123	i tried fast-export and got this .... File "/usr/lib/pymodules/python2.6/mercurial/localrepo.py", line 2188, in instance	12:23
	return localrepository(ui, util.drop_scheme('file', path), create)	12:23
	alongwith four other errors	12:23
DamienCassou	hi	12:23
	I have a very simple (bare) repository. I also have a working copy. I did a git push from the working copy to the bare repository. When I do clone in a new directory, I can't see the change. Is that normal?	12:24
	→ alexmchale joined	12:24
	→ uau joined	12:24
poona	wereHamster: Tried the latest version 1.7.0.2. Still get the same error	12:25
	wereHamster: oh, sorry. It is working :) Thanks	12:26
ringlej	wereHamster: How do I show the commit date/time for a branch?	12:28
	← warthog9 left	12:30
	→ zoniq joined	12:30
solow	anyone?	12:30
wereHamster	DamienCassou: maybe you pushed to a different branch than master	12:31
	solow: what does a browser have to do with git?	12:31
solow	Well that's what I'm asking you	12:31
wereHamster	ringlej: git log --pretty=tformat:"%cd" -1 master	12:31
	solow: nothign	12:32
solow	Changes are made, but the browser doesn't display them. So that's why I', asking IF git coul have anything to do with it. like, maybe it wants me to commit or something	12:32
	Okay thanks.	12:32
	→ tvw joined	12:34
DamienCassou	wereHamster: no, I checked	12:34
	wereHamster: I'm only using the master branch	12:35
wereHamster	DamienCassou: then use gitk --all to see where your branches are at (in both the original repo and the clone)	12:35
	→ rseifert joined	12:36
ringlej	wereHamster: thanks	12:37
DamienCassou	wereHamster: they are both at the top	12:38
	but one has 2 more lines on the top	12:38
wereHamster	'at the top' ?	12:38
DamienCassou	(2 commits)	12:38
	wereHamster: they are both at the first line	12:38
wereHamster	they = who/what?	12:39
	← priidu_ left	12:39
	→ Raska joined	12:39
	← andre____ left	12:39
	→ marienz joined	12:39
DamienCassou	wereHamster: both gitk windows highlight the first commit line	12:39
	← poona left	12:39
	→ DiNja^ joined	12:40
	→ frend joined	12:40
	→ MasterMav joined	12:40
	→ GLUS4E joined	12:40
	→ Camille11 joined	12:40
	→ mehere953 joined	12:40
wereHamster	and does the first line in both repos refer to the same commit?	12:40
	→ Anohail joined	12:40
	→ Black^Jack joined	12:40
	→ Lonely`Kid joined	12:41
	← frend left	12:41
	← MasterMav left	12:41
	← GLUS4E left	12:41
	← Camille11 left	12:41
	← mehere953 left	12:41
	← Black^Jack left	12:41
	← Anohail left	12:41
	→ girl17 joined	12:41
	→ Intuneric joined	12:41
	← girl17 left	12:41
	→ Danger^-^Girl joined	12:41
	→ mei^happy^ joined	12:41
	→ idoru joined	12:41
DamienCassou	no. In one, the first line refers to my last commit. The other window's first line is equal to the first window's third line	12:41
	← Danger^-^Girl left	12:41
	→ majorcool joined	12:41
	← majorcool left	12:41
	→ FARM4N joined	12:41
	← FARM4N left	12:41
	← Intuneric left	12:42
	→ sternocera joined	12:42
wereHamster	was the push successful?	12:42
	← jelmer left	12:43
DamienCassou	wereHamster: yes it was	12:44
	← Lonely`Kid left	12:44
	← ringlej left	12:44
wereHamster	does the 'origin/master' branch in the first repo (the one you pushed from) point to the same commit as the 'master' branch?	12:45
	← _Vi left	12:45
DamienCassou	wereHamster: how do I know that?	12:45
	→ _Vi joined	12:46
wereHamster	gitk --all	12:46
	→ nanoo joined	12:46
	← aatanasov left	12:47
DamienCassou	wereHamster: I don't understand the question. Can you guide me a bit more please?	12:48
	→ Scotto_ joined	12:48
steveire	How do I find the common ancestor of branch1 and branch2	12:48
	← simplechat left	12:48
	← DiNja^ left	12:48
Scotto_	how do i drop changed in a file?	12:48
wereHamster	DamienCassou: git rev-parse master; git rev-parse origin/master (run that in the repo you pushed from and paste the output here)	12:49
j416	steveire: mergebase	12:49
wereHamster	steveire: git merge-base	12:49
	ArjenL_ → ArjenL	12:49
	→ sergio_ joined	12:49
wereHamster	Scotto_: what do you mean by drop?	12:49
steveire	Thanks	12:50
Scotto_	i added an empty file, and now its picking it up as an indexed change	12:50
wereHamster	Scotto_: git rm <file>	12:50
DamienCassou	git rev-parse origin/master	12:50
	origin/master	12:50
	fatal: ambiguous argument 'origin/master': unknown revision or path not in the working tree.	12:50
j416	Scotto_: "git status" will give you information on how to add/remove files	12:50
Scotto_	instead of committing the empty file or deleting it manually, i thought id be able to checkout the branhc after commiting everything	12:50
j416	(from the index)	12:50
Scotto_	im using git gui	12:50
	:D	12:51
j416	after commit, do as wereHamster says	12:51
Scotto_	you guys probably despise the gui	12:51
wereHamster	DamienCassou: how did you push to the repo?	12:51
DamienCassou	wereHamster: git push idurand master	12:51
j416	Scotto_: I have hardly used the gui, but I think it is quite limited	12:51
	I sometimes use it for viewing history, but not more than that	12:51
Scotto_	j416: seems like it	12:51
DamienCassou	wereHamster: idurand being a remote	12:51
Scotto_	:(	12:51
	i guess i coul give it a try	12:52
mattayers	WOOP	12:52
	← froschi left	12:52
wereHamster	DamienCassou: ah, then 'git rev-parse master; git rev-parse idurand/master'	12:52
Scotto_	just nice to see the branches and the staged and indexed areas illustrated like that	12:52
DamienCassou	wereHamster: same hash	12:52
Scotto_	i wish the gui would actually show the commands being used	12:52
wereHamster	DamienCassou: and 'git rev-parse origin/master' in the cloned repo?	12:53
	← great_cogan left	12:53
	→ palisv joined	12:53
DamienCassou	wereHamster: different hash	12:53
wereHamster	DamienCassou: what happens when you run 'git fetch origin' in the cloned repo?	12:54
	→ munichlinux joined	12:54
	← rgr left	12:54
DamienCassou	wereHamster: nothing is printed, and rev-parse doesn't change	12:54
wereHamster	DamienCassou: 'git ls-remote origin \| grep master' in the cloned repo?	12:55
DamienCassou	wereHamster: same hash than with rev-parse and 'refs/heads/master'	12:56
wereHamster	DamienCassou: are you fetching over http?	12:56
DamienCassou	wereHamster: yes I am	12:56
wereHamster	DamienCassou: don't do that...	12:56
DamienCassou	wereHamster: pushing through ssh and fetching from http	12:57
wereHamster	use ssh:// or git://	12:57
DamienCassou	wereHamster: it's mandatory because some people don't have ssh access and nead read-only access to the repo	12:57
	wereHamster: s/nead/need/	12:57
	→ hugoxrosa joined	12:57
wereHamster	DamienCassou: you need to run git-update-server-info after each push	12:58
	→ Tureba joined	12:58
wereHamster	otherwise the new changes aren't visible to clients fetching via http	12:58
	take a look at the default post-receive hook	12:58
	sorry, post-update	12:58
DamienCassou	wereHamster: thank you very much. That was it. I will read some info about it, thank you thank you thank you	12:59
	← hipe left	12:59
Scotto_	any good gui aps that encourage command line use?	12:59
	→ vall joined	13:00
	→ hipe joined	13:00
	→ magn3ts joined	13:00
vall	ping	13:00
	→ PointMan joined	13:00
vall	hi, I would like to manage submodules in my git project	13:02
	← JackWinter left	13:02
vall	on the root : git submodule status	13:02
	2ad51de0344ba395bfa52a847a1e5589658a0742 plop (heads/master)	13:03
	→ froschi joined	13:03
	→ jelmer joined	13:03
vall	but when I add files in submodules directory, I list no changes on subdirectory with git status (on root)	13:03
	→ niko joined	13:03
	← mosno left	13:04
	→ JackWinter joined	13:05
	← jaql left	13:07
	→ jasonn1234 joined	13:08
	← alexmchale left	13:08
	→ auslander joined	13:08
	← auslander left	13:08
	→ auslander joined	13:08
	← alex____ left	13:08
	→ aspecto joined	13:09
	← tobiassjosten left	13:09
	→ cenuij joined	13:10
	← cenuij left	13:10
	→ cenuij joined	13:10
	← parasti left	13:10
	← unijunction left	13:10
	← frogonwheels left	13:11
	← DamienCassou left	13:12
	→ seathief joined	13:13
rolfb	vall, you have to treat the submodule as its own git repository	13:14
Scotto_	what are changed but not updated files?	13:14
	→ jaql joined	13:14
Scotto_	they show up when i do git status, but they dont show up in the gui at all	13:15
j416	Scotto_: files tracked but not yet added to your index after you changed them	13:15
Scotto_	how do i check the diff on an indvidual file, i have a feelign they werent changed by me, but by gitgui to make them all unix line enedings	13:16
	→ tobiassjosten joined	13:16
wereHamster	git diff -- <path>	13:17
Scotto_	yep, nasty looking diff	13:17
	minus all lines, plus all lines	13:17
	:/	13:17
	are these files upposed to get committed?	13:17
	because the gui hadnt done so	13:18
	how do i commit line ending changes like this?	13:19
vall	rolfb: thank you	13:19
Scotto_	should i just ignore these unstaged uncommited files?	13:20
	← [a] left	13:20
	← aziz left	13:21
	→ tstclair joined	13:23
	← magn3ts left	13:23
	← mr_danie1 left	13:23
Scotto_	i just wanna correct this so i can move on, someone please help	13:23
	→ frogonwheels joined	13:23
j416	Scotto_: I think you would benefit from reading a bit about git	13:24
	there is lots of really good documentation	13:24
	Scotto_: http://git-scm.com/documentation	13:25
Scotto_	j416: that doesnt help me	13:25
j416	did you try?	13:25
	← hgimenez left	13:25
	→ unijunction joined	13:25
Scotto_	seriously im not going to go digging through the docs on a single issue	13:25
j416	ok	13:25
Scotto_	ive used SVN before	13:25
j416	so don't :)	13:25
Scotto_	everythgins been going well so far	13:25
	and ive read the docs, but that doent mean im goign to retain them all in memory	13:26
	and it takes less time to go on without git than it does to dig through the docs, then so be it	13:27
	i just thought it would be good practice	13:27
j416	Scotto_: you might want to check config option core.autocrlf / core.safecrlf	13:27
	→ Bass10 joined	13:27
j416	(man git-config for info)	13:27
Gitbot	j416: the git-config manpage can be found at http://git.or.cz/man/git-config	13:27
Scotto_	j416: i want it to change the line endings for me	13:27
	everythign should be in unix line endings	13:27
j416	then that is your solution ^	13:28
Scotto_	but... that dosnt change the fact that git status is showing me files in the unstaged/changed area that the gui does not	13:28
	i need to know why, and how to correct it	13:28
	so i can move on	13:28
j416	not sure about that	13:28
	sorry	13:28
ambuli	what is the command to see the tags in local ?	13:29
j416	ambuli: git tag	13:29
Scotto_	im trying to use the gitgui most of the time and then use command line only when necessary	13:29
	j416: seems like git is only developed to be used by git experts	13:29
	not very intuitive at all	13:30
j416	Scotto_: I think it's mainly developed to be used though command line	13:30
	through*	13:30
	I thought it was very intuitive. :)	13:30
Scotto_	j416: j416 thats because youve been staring at the manual for weeks on end	13:30
j416	I read the git book for a few hours	13:31
	then I experimented	13:31
	→ flazz joined	13:31
locklace	Scotto_: you can read http://book.git-scm.com/ in about half an hour	13:31
Scotto_	each time i experiment, i run into new problems and issues	13:31
flazz	how can i backtick in git aliases?	13:31
Scotto_	that no one can explain	13:31
j416	Scotto_: that's kind of the purpose	13:31
	of experimenting	13:31
Scotto_	no, i shouldnt run into unexplained problems	13:31
	things not covered by the docs	13:31
j416	Scotto_: I don't handle files with CRLF, that's why I don't know	13:31
	→ dc joined	13:31
	← frogonwheels left	13:32
Scotto_	j416: a lot of people use windows to edit their files	13:32
j416	sure	13:32
Scotto_	j416: its not uncommon	13:32
	→ frogonwheels joined	13:32
j416	that doesn't change the fact that I don't know anything about it, though :)	13:32
	I'm sure there are people more knowledgeable on this than me.	13:33
	wait and see :)	13:33
Scotto_	j416: ive changed my IDE to default to unix line endings, just switched to a new IDE and it defaulted to CRLFs	13:33
	← dcampano left	13:33
Scotto_	its just frustrating...	13:33
	how do i stop it from showing the results of diff or status in vi?	13:34
	id rather it cat the results	13:34
j416	Scotto_: you can set the default pager	13:34
	man git-config	13:34
Gitbot	j416: the git-config manpage can be found at http://git.or.cz/man/git-config	13:34
Scotto_	j416: does this default change randomly by itself through versions?	13:35
	because i didnt change it, but it started doing this by itself recently	13:35
j416	randomly? no	13:35
	but things change, of course.	13:35
Scotto_	thats annoying	13:35
j416	I don't know if that particular thing has changed	13:35
	you can just set it in config	13:35
	← frogonwheels left	13:36
Scotto_	you make the config sound like a gui	13:36
j416	Scotto_: git config core.pager cat	13:36
	run that and you're done	13:36
Scotto_	j416: so, how do you remember all the settings?	13:36
j416	Scotto_: I don't	13:36
	Scotto_: man git-config	13:36
Gitbot	Scotto_: the git-config manpage can be found at http://git.or.cz/man/git-config	13:36
	→ tango_ joined	13:36
Scotto_	j416: then how do you know what to enter in the config	13:37
j416	read what I wrote	13:37
	^	13:37
Scotto_	j416: so each time you setup a new repo, you go searching through the manpages?	13:37
j416	no	13:37
	← hipe left	13:37
j416	I do it when I need to change some setting	13:37
	→ hipe joined	13:38
	← hipe left	13:38
	→ frogonwheels joined	13:38
	→ hipe joined	13:38
	← palisv left	13:38
	← hipe left	13:39
	→ flaguy48 joined	13:40
	← aspecto left	13:40
	→ priidu_ joined	13:40
	→ hipe joined	13:40
	← seathief left	13:40
	→ cedricv joined	13:41
	← Hackykid left	13:41
Scotto_	j416: now using cat i dont get color coded output	13:41
	j416: before it started loading in the gui i would get color coded cat :(	13:42
	→ Hackykid joined	13:42
	→ palisv joined	13:43
sternocera	Hi. How can I create a git branch from within Qt creator, and which that to my working copy? Thanks.	13:44
j416	Scotto_: check the man page, everything is in there	13:44
	Scotto_: try: git config color.ui auto	13:45
	→ csiegert joined	13:45
	→ aziz joined	13:45
	← nicoulaj left	13:47
	→ hipe_ joined	13:48
	→ joelmoss joined	13:49
joelmoss	hi all	13:49
	si just upgraded to 1.7.0.2	13:49
	trying to bare clone a repo over http and getting this error:	13:49
	fatal: Unable to find remote helper for 'http'	13:49
	anyone have any ideas?	13:50
	i know the smart-http protocal is all new	13:50
	but backwards compat is kept ain't it?	13:50
	anyone?	13:50
Rhonda	Did you run git update-server-info on the server part?	13:50
	← hipe left	13:51
joelmoss	??	13:51
	i dont have access to repo on the seerver	13:51
douilspee	how can i make git diff not to list file permission change and binary change ?	13:52
Rhonda	joelmoss: Did git clone work from there prior to your git upgrade?	13:52
jMyles	So I've been trying to get into Git for the past week or so, and I'm starting to feel pretty good about it. A nagging question I have, though, is - how do I test out a development branch on a server? Can I have two branches correspond to two directories of files and look at each? What's the typical practice here?	13:52
joelmoss	Rhonda: no, thats why I upgraded	13:52
	seems curl wasn't compiled in	13:52
	so thought I'd upgrade while doing the recompile	13:52
Rhonda	joelmoss: I rather expect the http exposure there never did run git update-server-info	13:53
	And so it's no regression if it didn't work before, so your statement about "backwards compat is kept" isn't really called for. It is backwards compat in the way that it still doesn't work. ;)	13:53
locklace	jMyles: what are you doing, web stuff?	13:54
jMyles	locklace: Yeah, for now.	13:54
	← psankar left	13:54
joelmoss	Rhonda: so why doesn't iit wortk?	13:54
Rhonda	jMyles: You can just git clone again, or work in a git branch until you are happy to git merge back into master.	13:54
locklace	jMyles: good idea to have both a production and a dev/staging server. then you have two clones, each server pointing to one	13:54
Rhonda	joelmoss: … I thought I mentioned that twice already? :)	13:55
joelmoss	?	13:55
	Rhonda: sorry, ya lost me	13:55
Rhonda	joelmoss: "git update-server-info"	13:55
joelmoss	yeah but that has to run on the repo that I'm cloning from right?	13:56
	which I don't have access to	13:56
Rhonda	Exactly.	13:56
	Then you can't clone from there.	13:56
joelmoss	i only have read only access to it	13:56
	not over http?	13:56
Rhonda	No.	13:56
joelmoss	hmm	13:57
	← mydog2 left	13:57
joelmoss	so git update-server-info is needed to be run on all repos when upgrading to 1.7?	13:57
	oh no, I see	13:57
	just for http	13:58
Rhonda	No, it's needed to be run on all repos, no matter what version.	13:58
jMyles	Rhonda, locklace: Thanks - I'm not quite getting it, though. So, I have LAMP server. I have /var/www/project with the master branch. Then I'll clone it into /var/www/devel or something? Then have the team push to devel?	13:58
Rhonda	Yes, for dumb http	13:58
	jMyles: http://nvie.com/git-model - that gives a nice workflow model for git	13:59
locklace	jMyles: have one "central" repo that acts as master, then create two clones of it, one for each web server, and point each server at its clone	13:59
Rhonda	jMyles: And … erm, you don't push to non-bares.	13:59
	→ nicoulaj joined	13:59
jMyles	Rhonda: I don't know what "non-bares" means. I'm checking out your link - thank you very much. :-)	14:00
Rhonda	There are hacks to be able to do so but that requires discipline and no changes in those non-bares at all to which you want to push.	14:00
	← csiegert left	14:00
locklace	jMyles: central is the master repo for all your stuff, and has branches for dev, staging, production, whatever you need. the clones pull (or fetch+reset) from central as needed, and have the appropriate branch checked out	14:00
	→ mydog2 joined	14:00
Rhonda	jMyles: In principle: If you push to the git repository in which someone else has a working copy and is working in there are many bad things happening, including that god eats a kitten.	14:01
	← Hackykid left	14:01
joelmoss	Rhonda: cheers	14:01
Rhonda	jMyles: A bare clone is one without a working copy and is the only place one pushes to.	14:01
flazz	say i have two projects, A & B. i want to move files from B into A and deprecate B, but i want the history of those. how can i do it?	14:01
	→ malumalu joined	14:02
	→ project2501a joined	14:02
jY	i have 1 file that didn't make it into a tag.. is there anyway to put that file into the tag?	14:02
Rhonda	flazz: You git merge B into A? :)	14:02
douilspee	is there any git diff option like : show only file with text change ?	14:02
locklace	jY: a tag is just a pointer to a commit	14:02
flazz	Rhonda: i can merge across repositories?	14:02
joelmoss	Rhonda: so hold on tho... I can git clone that very same repo on another machine fine	14:02
	→ munichpython joined	14:03
	→ bushwakko joined	14:03
joelmoss	so that doesnt make sense	14:03
bushwakko	hey, how can I do git checkout --ours on all unmerged files?	14:03
	I have a ton of files that are just modified and/or merged, and then I have another ton that has conflicts. In these conflicts I wish to keep mine only	14:04
	→ rgr joined	14:05
	← dpino left	14:05
	→ LaFi joined	14:06
	← munichlinux left	14:06
	← _Vi left	14:06
	← bushwakko left	14:06
	→ bushwakko joined	14:06
bushwakko	whops...	14:07
	← flaguy48 left	14:07
	← bushwakko left	14:07
	→ bushwakko joined	14:07
bushwakko	Is --ours not my fixes, because I did a --theirs, and got my changes instead of the incoming ones...	14:08
	← rseifert left	14:09
	← ben_alman left	14:09
	← joelmoss left	14:09
	→ rseifert joined	14:10
	→ drubies joined	14:12
	← roop left	14:12
	→ dforsten joined	14:14
Scotto_	why is the git gui showing different staged files than what git status shows?	14:16
	this is absolutely confusing	14:16
	ok so one thing i need to know before i go further, should i just stop using the gui all together?	14:20
	→ mklappstuhl joined	14:20
	→ ben_alman joined	14:20
sternocera	Hi. When I create a new branch with git, must I create a new physical, working copy on the filesystem? Thanks	14:23
_rane	no	14:24
	git branch <branchname>	14:24
fr0sty-away	Scotto_: can you elaborate on your troubles?	14:25
dforsten	Is it possible to use git-svn to merge two svn branches bi-directionally?	14:26
	→ patthoyts joined	14:26
Scotto_	fr0sty-away: not really, i resolved them so i cant go back and give you the play by play	14:26
	→ ToreadorVampire joined	14:26
Scotto_	fr0sty-away: had line ending problems, but gitgui was showing different files in the unstaged changes area than what git status was showing	14:27
	fr0sty-away: they showed up in git status but not in the git gui	14:27
	← mei^happy^ left	14:27
	← Alex left	14:28
	→ Alex joined	14:28
	← cedricv left	14:28
	→ t0rc joined	14:28
	← sternocera left	14:28
	← Alex left	14:29
	← JackWinter left	14:29
	← jony left	14:29
	→ Axius joined	14:29
	→ JackWinter joined	14:30
	→ Alex joined	14:31
	→ _Vi joined	14:31
	→ anxiolytic joined	14:31
ToreadorVampire	Bugger :( I think I may have shot myself in the foot with git ... it uses some case-sensitive filenames ... I archived all of my git repos' up to a FAT32 volume for rebuilding my server ... now I have rebuilt it I am getting "does not appear to be a git repository" messages all over the place as I try to clone those backed-up repositories :(	14:31
	ToreadorVampire cries	14:31
	← frogonwheels left	14:31
	← LaFi left	14:32
ToreadorVampire	Ohwait, maybe the only case-sensitive filename in each repos is HEAD ??!!	14:32
	(I mean, uppercase filename)	14:32
	→ jfkw joined	14:33
ToreadorVampire	In which case it's fairly trivial to batch-rename those with `find` after I move the repositories back off of the FAT32 volume onto my ext3 volume	14:33
	Phew, that's good news at least	14:33
	← Axius left	14:34
	ToreadorVampire tries it on one repos, is able to clone it again, cheers and thanks his lucky stars :D	14:34
	→ maxthoursie_ joined	14:35
	← maxthoursie left	14:35
	maxthoursie_ → maxthoursie	14:35
ToreadorVampire	Yay!	14:36
	← ToreadorVampire left	14:36
	← napster_123 left	14:37
	→ LiamH1 joined	14:37
	← AkumaStreak left	14:38
	← poseid left	14:38
	fr0sty is happy, but feels compelled to point out that it was FAT32, not git that did the foot shooting.	14:39
	→ cedricv joined	14:39
	→ maxthoursie_ joined	14:40
	← maxthoursie left	14:40
	maxthoursie_ → maxthoursie	14:40
	→ mm_work joined	14:42
	← mm_work left	14:42
	→ mm_work joined	14:42
bushwakko	hey, I have a gazillion unmerged that I want to do git checkout --theirs on	14:43
	how can I do that	14:43
	is it possible to get an interactive merge like in gentoo's dispatch-conf=?	14:43
	← zoniq left	14:44
	→ spearce joined	14:44
	→ zoniq joined	14:45
wereHamster	git ls-files -u \| xargs git checkout --theirs ?	14:45
	→ alezandro joined	14:46
lack	wereHamster: Doesn't 'git ls-files -u' have a bunch of cruft that will screw that up? I think you need 'git ls-files -u \| cut -f 1 \| sort -u' or somesuch...	14:46
bushwakko	-f 2 seems to work	14:46
	← anxiolytic left	14:47
lack	bushwakko: That's the one.	14:47
	← mcdebugger left	14:47
	← malumalu left	14:47
	→ mrchrisadams_ joined	14:47
	→ mattayers_ joined	14:50
	→ cannonball joined	14:51
	← mattayers left	14:53
	mattayers_ → mattayers	14:53
	← maxthoursie left	14:54
	← alezandro left	14:54
	← tomekw left	14:55
	→ alezandro joined	14:55
	→ sagsousuke joined	14:56
	→ maxthoursie joined	14:57
	→ Mezner joined	14:57
	→ poseid joined	14:59
jMyles	OK, so for a web development purpose, let's say I want two main branches - master (production) and devel (development). I want both of them to be live on the web server. How can I accomplish this?	15:00
	← mithro left	15:02
bushwakko	I'm really getting myself deeper and deeper into crap here now	15:03
	at one time I took an git svn rebase, and I merged it wrong	15:04
	but I can't seem to get it back	15:04
	I need to get back in the state before that rebase, so I can redo it	15:05
	← hipe_ left	15:05
	→ hipe joined	15:06
GyrosGeier	is the rebase still active?	15:06
	oh wait	15:06
	git svn rebase	15:06
	no idea then	15:06
locklace	jMyles: i told you above	15:07
jMyles	locklace: I took a look at the article, and it made good sense, but in practical terms, I'm still not sure what to do.	15:07
	← Mezner left	15:08
locklace	jMyles: i'm not the one who gave you that link. it's good, but only marginally relevant to you	15:09
bushwakko	I cannot get to the point where the index is dirty and I haven't commited	15:09
	→ engla joined	15:09
	← engla left	15:09
	→ engla joined	15:09
bushwakko	if I do git checkout <id> I get before	15:09
	and the other id is after	15:09
	I need to get to unmerged index part	15:09
	→ SpookyET joined	15:09
EvanR-work	bushwakko: git reset --hard <ref>	15:09
	→ Axius joined	15:09
bushwakko	ok	15:09
	crosses fingers	15:09
	← SpookyET left	15:10
	← alezandro left	15:10
EvanR-work	your working tree and index will be equal to <ref>	15:10
jMyles	locklace: Ahh, yes, my bad. OK, well I understand the notion of "central" as a master repo with the respective purposes for each of the other repos, but I still don't understand how. I can make a production and dev branch easily enough, but I don't understand how to make them both live on my server so we can look at dev as we progress.	15:10
	→ alester joined	15:10
	← hipe left	15:10
cbreak	jMyles: you want two working directories?	15:10
locklace	jMyles: you're going to create two clones of the central repo with git clone, one for each web server	15:11
cbreak	then just create two repositories, each checking out one branch	15:11
lack	How do I (commandline) get a list of commits whose subjects match a specific string or regex?	15:11
bushwakko	EvanR-work: the index is always clean...	15:11
locklace	jMyles: (or each project, if you don't actually have separate servers)	15:11
EvanR-work	bushwakko: yes	15:11
cbreak	lack: man git-log	15:11
Gitbot	lack: the git-log manpage can be found at http://git.or.cz/man/git-log	15:11
cbreak	look for --grep	15:11
EvanR-work	bushwakko: what do you want it to have	15:11
	→ alezandro joined	15:11
bushwakko	I want to get to an uncommited part of it	15:11
EvanR-work	which part	15:12
bushwakko	whatever I do I get the a commited state, I want to get the uncommited states	15:12
	so I can start from there	15:12
trebor_dki	hello, are there emacs-git users here, who can give me some advices on vc-git.el and git.el?	15:12
jMyles	locklace: OK, I get that, but... and I know this is stupid noob stuff, but when I clone the directory with git clone and then run git branch, I only see one branch. For example, if I clone my production directory into project-devel, I want that branch to be called devel. Then when I run git branch, I want to see master and devel, no?	15:12
EvanR-work	bushwakko: git reset --hard to what you want the working tree to be. then git reset --mixed to where you want HEAD to be	15:13
	← themill left	15:13
bushwakko	EvanR-work: sorry, I don't really know what that means	15:13
	← froschi left	15:13
	→ themill joined	15:13
	→ jrmuizel joined	15:13
bushwakko	can I undo git reset --hard btw?	15:14
locklace	jMyles: the best working model here is to actually have 3+n repos, where n is the number of boxes you want to actually do development on	15:14
cbreak	sometimes	15:14
EvanR-work	bushwakko: yes	15:14
cbreak	git reset can kill changes that are not commited	15:14
EvanR-work	as long as where ever you want to reset too is in some object, either stashed	15:14
locklace	jMyles: the 3 repos are central, prod, and devel. central is bare, and is the hub repo that all the other repos are spokes of	15:14
cbreak	those are gone for ever, mostly	15:14
bushwakko	I have commited lots	15:14
EvanR-work	or in the reflog	15:15
bushwakko	I just did reset --hard back to an earlier commit	15:15
EvanR-work	to go forward, look in git reflog	15:15
locklace	jMyles: prod, devel, and the working repos are just non-bare clones of central	15:15
	→ dannas joined	15:15
	→ j416_ joined	15:16
bushwakko	this started when I did git commit instead of git rebase --continue in the middle of a rebase	15:16
locklace	jMyles: to make changes, you develop on one of the n working boxes, then push to central. then you get the latest versions on the prod and devel clones with fetch+reset	15:16
	→ gergap joined	15:16
	← gergap left	15:16
EvanR-work	bushwakko: tricky, anyway all your commits should still but in git reflog, so git reset --hard to that one	15:16
	be in *	15:16
cbreak	have you considered doing git rebase --abort?	15:16
locklace	jMyles: since all repos are clones of central, all branches are available in each. the "prod" clone has the prod branch checked out, and the "devel" clone has the devel branch checked out	15:17
bushwakko	so basically what I ideally would like to do is get back to the dirty index of the rebase before i did git commit	15:17
EvanR-work	you need to restart the rebase	15:17
	← rich97 left	15:17
	→ lhz joined	15:17
	→ xperience joined	15:17
EvanR-work	to get individual changes to files before committing, do git checkout <ref> -- file1 file2 etc	15:18
	→ mithro joined	15:18
EvanR-work	where ref is some nuts commit you made during rebase, i suppose	15:18
	← j416 left	15:18
	← dc left	15:19
	← Axius left	15:20
bushwakko	I always seem to get to a commited state though... I want to checkout right before a commit	15:20
	my index is always clean whatever I do right now	15:21
EvanR-work	that does not make sense	15:21
mrchrisadams_	is ssh forwarding a specific option you need to swtich on with gitosis?	15:21
	← jMyles left	15:21
wereHamster	gitosis doesn't require ssh forwarding	15:22
EvanR-work	bushwakko: if you have commit A, and commit B. and you want to be 'on' commit A, but have changes in your working tree from commit B, do git reset --hard A, then git reset --mixed B	15:22
	(where A is B's parent)	15:22
	→ mssm joined	15:22
	← phoenix24 left	15:23
	← mattayers left	15:23
EvanR-work	bushwakko: --hard, --soft, and --mixed determine whether your HEAD, index, and working tree are modified. man git-reset	15:24
	← zoniq left	15:24
Gitbot	bushwakko: the git-reset manpage can be found at http://git.or.cz/man/git-reset	15:24
cbreak	if I was you, I'd go back to where I was before the whole problem started	15:25
	→ zoniq joined	15:25
cbreak	nothing beats being at a known working state	15:25
	→ camonz joined	15:25
	← tbf left	15:25
	← rgr left	15:25
EvanR-work	things are peachy if no one has pulled your broken changes ;)	15:25
bushwakko	cbreak, that is what I'	15:26
	m trying to accomplish	15:26
	but whatever I do it seems like the rebase is mixed in there somehow	15:26
EvanR-work	did you do git rebase --abort	15:26
cbreak	did you abort the rebase as I told you?	15:26
	← marienz left	15:26
	→ bronson joined	15:26
bushwakko	I've been jumping around so much that this rebase is long fogotten	15:27
	→ aschueler joined	15:27
	→ bauruine joined	15:27
EvanR-work	do you have .git/rebase-apply	15:27
	if so, its not forgotten	15:27
bushwakko	nope	15:27
	→ TeckniX joined	15:28
EvanR-work	its possible your rebase succeeded	15:28
bushwakko	most likely	15:28
	→ parasti joined	15:28
EvanR-work	then you need to reset to the original branch somewhere	15:28
bushwakko	I think it succeeded somehow when I did that commit	15:28
EvanR-work	unless youre ok with it	15:28
cbreak	do a git reflog > backupreflog	15:29
	and maybe look at the output too :)	15:29
	← RobertMeta left	15:29
	← disappearedng left	15:29
cbreak	with a bit of guessing you should find a time where you were before the rebase started	15:29
	you can reset to that if you want	15:29
EvanR-work	fun with git-rebase	15:30
bushwakko	I've been trying to do that	15:30
	I know what i commited before the rebase, but doing reset --hard to that still contains stuff that I know came with the rebase	15:30
EvanR-work	that doesnt seem possible, they would have to have different hashes	15:31
	← palisv left	15:31
bushwakko	http://pastebin.ca/1828579	15:32
	→ joeconyers joined	15:32
EvanR-work	i see a lot of checkouts	15:32
bushwakko	3031cc0 HEAD@{28}: commit: cleaned up a gazillion conflicts <-- this is the fatal commit i did _during_ the rebase	15:33
	← tasslehoff left	15:33
	→ reprore joined	15:33
EvanR-work	so dont use that one	15:33
	you seem to have been rebasing non stop	15:34
bushwakko	I would like to get back to that runner.args state	15:34
EvanR-work	that is after some rebase, are you sure	15:35
bushwakko	I think so	15:35
	→ Mezner joined	15:35
	← rseifert left	15:36
bushwakko	I'm using svn, so git svn rebase is the way to get updates from svn	15:36
EvanR-work	if you reset to 27fa3f3 and the unwanted changes are there, then i think you shouldnt be sure	15:36
	→ dcampano joined	15:36
bushwakko	which reset should I do there then?	15:36
EvanR-work	hard	15:36
	well, i just noticed there are two runner.args commits	15:37
	which one is which	15:37
bushwakko	that seems better	15:37
	27fa3f3 <-- this ones	15:37
	the one I just check out	15:37
	I'm currently in (no branch) atm though	15:38
	but this seems like a place I'd like to start	15:38
	← drubies left	15:39
EvanR-work	no branch probably because of the checkouts you did	15:39
bushwakko	no i did git checkout -b smsfix	15:39
	if I can fix this one, I can delete the rest	15:40
EvanR-work	try checking out master or smsfix and then resetting	15:40
	← _Vi left	15:40
bushwakko	the big question now is if I can do a rebase on this one, to try to merge in the other changes (it's during this merge i did a commit earlier)	15:41
	j416_ → j416	15:41
	→ yatiohi joined	15:42
bushwakko	ahh	15:42
	I'm back	15:42
	git ls-files -u\|cut -f 2\|uniq\| git checkout --theirs	15:43
	→ Carneus joined	15:43
	← alezandro left	15:43
bushwakko	this is what I did earlier	15:43
	← trochala left	15:43
bushwakko	but now I got: --ours/--theirs is incompatible with switching branches.	15:43
Carneus	how do i fast forward to most recent commit?	15:43
	→ avar joined	15:44
	→ alezandro joined	15:44
locklace	Carneus: git checkout <branch>	15:44
EvanR-work	fast forward merge from upstream?	15:44
	← bushwakko left	15:45
avar	When I run `git foo` e.g. `git add` can I somehow see what git-add binary it uses? I'm bisecting git itself and I think my compile dir is being poisoned by my system dirs	15:45
	→ bushwakko joined	15:45
Carneus	locklace, thank you	15:45
bushwakko	whops, messing up my shortcut keys here	15:45
	→ disappearedng joined	15:45
	→ lavamind joined	15:45
lavamind	hello	15:46
bushwakko	why can't I do --theirs/--ours now?	15:46
	I would think branch stuff would be irrelevant there?	15:47
	← lamont left	15:47
bushwakko	I just want to either choose the file from the branch i'm in, or from the rebase	15:47
lavamind	I'm trying to work out a pretty big merge, big but essentielly only a fast-forward merge: my tree has a few extra files. but for some reason lots of files are not able to auto-merge properly, even though I've never touched them in my tree, so I'm trying to understand why it's happening	15:48
	→ jmcantrell joined	15:48
EvanR-work	bushwakko: git checkout <ref> -- filename	15:49
drizzd	lavamind: does "git diff --name-status ...upstream" show any changes to these files?	15:49
	← m104 left	15:49
	→ lamont joined	15:50
bushwakko	EvanR-work: which ref is mine and which is theirs?	15:50
drizzd	lavamind: oops, that should have been upstream...	15:50
	"git diff --name-status upstream..."	15:50
EvanR-work	bushwakko: i dont know? pay attention!	15:50
lavamind	drizzd: by "upstream..." you mean something like "origin/MOODLE_19_STABLE...", right	15:51
EvanR-work	this rebase thing sounds awful	15:51
drizzd	lavamind: Yes, I mean the branch you're merging	15:51
lavamind	it doesn't show anything	15:51
drizzd	ok, what about renames? have the files you were editing been renamed upstream?	15:52
lavamind	drizzd: wait I was on the wrong checkout, sorry	15:52
drizzd	the merge should mention that before spilling out auto-merge errors	15:52
	→ tedoc2000 joined	15:52
	→ aryounce joined	15:52
drizzd	s/auto-merge errors/merge failures/	15:52
bushwakko	I'm trying something different now. I'm trying to merge master into this branch. master seems to be rebased correctly	15:52
EvanR-work	eww	15:52
	i would only merge master if its a fast forward	15:53
	← alezandro left	15:53
lavamind	drizzd: ok it shows lots of files either M or A, and my example, "version.php" is "M"	15:53
drizzd	lavamind: ok, so the syntax upstream... basically tells you what changed since your branch diverged from upstream	15:54
	→ lucid_ joined	15:54
	← rolfb left	15:54
drizzd	so IOW, version.php _was_ modified in your branch wrt. upstream	15:54
	you should see the same changes in "git log --name-status upstream..."	15:54
	oops, "git log --name-status upstream.." (only two dots this time, this is a little strange)	15:55
lavamind	ok I think I know why this is happening	15:55
	I did a squash commit last time	15:55
	← lucid left	15:55
	→ owen1 joined	15:56
EvanR-work	bushwakko: you sure you dont want to merge the other way	15:56
lavamind	so the detailed commit history is lost, and I suppse that's what would otherwise resolve the conflict	15:56
drizzd	lavamind: exactly	15:56
bushwakko	seems to be working... only problem I'm having now is that : "git ls-files -u\|cut -f 2\|uniq\| xargs git checkout --ours" isn't working	15:57
drizzd	you can replace the squashed commit by a real merge and then rebase onto that	15:57
bushwakko	EvanR-work: what's the drawbacks of merging master into my branch?	15:57
EvanR-work	it makes the history nuts	15:57
lavamind	drizzd: I've never done a rebase so I'll have to read up on that	15:58
EvanR-work	but i guess that might be preferrable	15:58
drizzd	replace as in, create a new branch from squashed-merge-commit^, then merge the same version you merged in the squashed commit (not sure how you can find out which one that was), then maybe check if the result is the same as squashed-merge-commit.	15:58
EvanR-work	especially if youre just going to merge that branch into master afterward, just do it that way first and fast forward	15:58
Carneus	is there a way to throw away last commit/merge	15:58
EvanR-work	git reset --hard	15:59
	← ph^ left	15:59
drizzd	lavamind: does that make sense so far?	15:59
lavamind	drizzd: I think that might be too complicated from my needs, the squashed commit is the last commit in my log for the troubled branch	16:00
drizzd	lavamind: oh	16:00
lavamind	so I suppose I can just reset --hard and re-merge the new upstream stuff	16:00
drizzd	lavamind: yep	16:00
	but it's still strange that the merge would fail in this case, since the merged files should be the same, even though you squashed the merge	16:01
	← radarek left	16:01
lavamind	what I don't like though about not using squashed commit is that upstream is very busy and my local-only commits get quickly buried	16:01
drizzd	(unless you merged an old state of the other branch)	16:01
	lavamind: git log --first-parent only shows your commits	16:02
lavamind	drizzd: well there's about 500 commits there, and most changes are auto-merged OK, only 20-or so files have problems	16:02
	← simNIX left	16:02
locklace	lavamind: you probably want to be pull --rebase-ing instead of merging	16:02
bushwakko	when I've solved conflicts during a git merge, what is the proper thing to do, it's a git commit -a -m "something" then?	16:03
drizzd	bushwakko: you probably don't want to overwrite the commit message with -m	16:03
EvanR-work	yes you must git commit your fixes to the merge conflicts	16:04
	git commit -a	16:04
lavamind	git-log --first-parent shows the same commits as simple git log, is that what pull --rebase would affect ?	16:04
	← disappearedng left	16:05
bushwakko	PHEW! seems to be fixed now! We're using git on top of svn, just because some of us like the git toolkit better. but god is it a hassle sometimes :)	16:05
	→ chaitanya joined	16:05
drizzd	lavamind: no, nothing to do with rebase. but if you did not do any real merges, then --first-parent has no effect	16:05
locklace	lavamind: pull --rebase will push your local changes onto the stack, update the branch with all the new upstream commits, then pop the stack and reapply your changes on top	16:05
drizzd	lavamind: it hides the commits from the merged branch	16:06
locklace	lavamind: iow, after the pull you will again see all of your changes first in the output of git log	16:06
EvanR-work	bushwakko: yeah, that sounds terrible	16:07
drizzd	lavamind: yeah, rebase seems to be what you want	16:07
EvanR-work	why not just use git normally	16:07
	→ martoche joined	16:07
	→ jMyles joined	16:07
	→ _Vi joined	16:07
lavamind	drizzd, locklace: okay, thanks for the info, I will try and undo my squasehd commits taht prevent a clean merge, and try using pull --rebase to see how that works	16:08
	← munichpython left	16:08
	← psoo left	16:08
bushwakko	My co-workers doesn't see the point in using git over svn, and think git isn't right for the task anyway because it isn't centralized	16:09
	but I mean, we could all work towards a central git couldn't we?	16:09
EvanR-work	you can pretend it is	16:09
Textmode	bushwakko: aye, thats a common layout.	16:09
locklace	git can do centralized workflows just fine	16:09
EvanR-work	designate one of the repos as 'the central repo'	16:09
	← camonz left	16:10
EvanR-work	just a matter of marketting ;)	16:10
	→ wolog_ joined	16:10
martoche	I've removed a bunch of files from a repo with filter-branch, then forced pushed to origin master. But when another clone of the repo makes a push, it pushes back all the files I wanted to delete. How do I tell all the clones of the repo that they should take into account the filter-branch?	16:10
Textmode	and the princible advantage of git over svn is extremely cheap branching, and local branching. which, imho, trumps svn any day of the week.	16:10
	→ sea-gull joined	16:11
	← wolog left	16:11
drizzd	martoche: you mean a forced push?	16:11
martoche	yes	16:11
locklace	martoche: fetch and reset --hard from the clones would be good	16:11
	← EricInBNE left	16:11
EvanR-work	git is faster, i heard	16:11
	→ alezandro joined	16:12
	→ rduplain joined	16:12
drizzd	martoche: basically, you need to rebase the local changes in all your clones, or discard them	16:12
	← _Vi left	16:12
Textmode	EvanR-work: having recently been forced to checkout some large svn repos, I second that. that took forever...	16:13
	→ munichpython joined	16:13
martoche	ok, I'll try to rebase	16:13
	← chaitanya left	16:13
martoche	and is there a way to prevent anybody from pushing back all those files I've removed? can I blacklist some filenames?	16:13
EvanR-work	yes, you disable ssh for any user who is an idiot	16:14
	aryounce → aryounce\|afk	16:14
EvanR-work	only allow trusted people to push to the main repo	16:14
martoche	EvanR-work: :)	16:14
	← aryounce\|afk left	16:14
	← zoniq left	16:14
EvanR-work	sounds like you are about to recover from upstream rebase, thats a bigger issue	16:14
	← sea-gull left	16:15
drizzd	martoche: educate people not to force pushes	16:15
	← trebor_dki left	16:15
	→ zoniq joined	16:15
drizzd	(mindlessly)	16:15
martoche	drizzd: they're not doing it, if you pull from origin master, then push a change, the deleted files are pushed back	16:16
	without --force	16:16
drizzd	martoche: I can't see how that could happen	16:16
	ah, wait, ok...	16:17
	they're merging the rebased branch	16:17
	→ palisv joined	16:17
drizzd	that's awful	16:17
martoche	oh, ok, it's when they merge that everything breaks	16:17
EvanR-work	yes.	16:17
locklace	that's what you get for using filter-branch on a shared repo	16:17
	→ patrik_ joined	16:17
EvanR-work	'recovering from upstream rebase'	16:17
	← patrik_ left	16:18
martoche	so the clean way to do it is a rebase?	16:18
EvanR-work	man git-rebase	16:18
Gitbot	EvanR-work: the git-rebase manpage can be found at http://git.or.cz/man/git-rebase	16:18
EvanR-work	martoche: there is no clean way to rebase something everyone is sharing	16:18
drizzd	martoche: yeah, and a rebase --onto at that	16:18
	→ webchick joined	16:19
EvanR-work	i like the solution 'revert commit what you did and admit you screwed up'	16:19
	← avar left	16:20
	← giallu left	16:21
martoche	drizzd: EvanR-work: locklace: thanks for your help	16:22
drizzd	I have to say, git doesn't handle this case very graciously. But I'm not sure how to improve it. Would require some kind of "rebase tracking."	16:23
	→ anxiolytic joined	16:23
	→ disappearedng joined	16:23
	← patthoyts left	16:23
	→ froschi joined	16:24
EvanR-work	git does not make history rewriting easy	16:24
drizzd	or maybe at least a warning that you're merging a branch without a common ancestor, which is an unusual thing to do and in this case certainly the wrong thing to do.	16:24
	← dannas left	16:24
EvanR-work	ah you mean you doesnt give enough warning signs	16:24
drizzd	there is no warning, and there is no automatic way to recover	16:24
	sure I don't want someone else to silently rewrite my history	16:25
EvanR-work	clearly rebasing on a regular basis is the wrong thing to do	16:25
drizzd	but in this case I'd probably want to trust the maintainer and let git "do the right thing"	16:25
	(i.e., rebasing all branches onto the rewritten history)	16:25
	→ dannas joined	16:26
	→ ph^ joined	16:26
EvanR-work	so a giant warning next to a config option 'allow automatic upstream rebasing'	16:26
	← priidu_ left	16:26
drizzd	or a very long command line option like, "git pull --rebase-with-magic"	16:26
	→ yevlempy joined	16:27
	← yatiohi left	16:28
	→ ringlej joined	16:29
	← voxpelli left	16:29
ringlej	how do I abort a conflicted merge after 'git merge master'?	16:29
	→ mw joined	16:30
EvanR-work	git reset --hard	16:30
ringlej	thanks	16:30
drizzd	newer git has git reset --merge for that	16:31
	although I'm not sure what it does, it's probably safer than --hard	16:31
	→ tbf joined	16:31
	→ voxpelli joined	16:31
EvanR-work	HEAD is changed during a failed merge?	16:31
	:S	16:31
drizzd	EvanR-work: no, why?	16:32
EvanR-work	so --hard should work?	16:32
drizzd	yes, it definitely works	16:32
	← LiamH1 left	16:32
drizzd	actually, I'm beginning to think that --merge is _not_ what you want here	16:32
	→ gebi joined	16:32
EvanR-work	im still trying to understand what --merge does...	16:32
drizzd	but the documentation is very cryptic	16:32
	seems to be the same as just "git reset" without any other options	16:33
EvanR-work	it changes working tree	16:33
	--mixed does not	16:33
	Mike\|\|gone → Mike_lifeguard	16:34
EvanR-work	im confusetated	16:34
	← bushwakko left	16:34
drizzd	"git reset --merge HEAD" does not change the working tree, because the "named commit and the current commit" are the same	16:34
anxiolytic	What are the relative ways to reference a commit called, i.e., head^2, head~2, etc..?	16:34
	→ amerine joined	16:34
	→ nsussman joined	16:35
	← nsussman left	16:35
EvanR-work	man git-rev-parse	16:35
Gitbot	EvanR-work: the git-rev-parse manpage can be found at http://git.or.cz/man/git-rev-parse	16:35
MetaCosm	if working with git on a share, with multiple people possible accessing it, is the right thing to do to setup the share as a bare and use a local disk, or is it fairly safe to use the share directly?	16:36
	→ nsussman joined	16:36
EvanR-work	depending on your network, it might be simpler if you used the 'blessed repo' and pulled from contributors, and they pulled from you	16:36
	pushing is more complicated	16:37
	← voxpelli left	16:37
	nsussman → textarcana	16:37
MetaCosm	EvanR-work, well, this is for a friend who is sorta starting to use git secretly on configuration / build / image scripts at his work, they keep them on a SMB share	16:37
drizzd	MetaCosm: It is _not_ safe for multiple people to use the same repository.	16:37
MetaCosm	and he did the git init right on the share	16:38
jMyles	locklace: Thanks again for your help - sorry if I'm being a pest. I'm still just not sure how to do what you suggest. I don't know what commands to run. I need an "examples" list to learn from I think.	16:38
	← j416 left	16:38
drizzd	MetaCosm: and to set up a shared repository, you should use init --shared	16:38
	(shared as in, for pushing/fetching only, not for sharing working copy and committing directly)	16:39
	← sh1mmer left	16:39
MetaCosm	drizzd, mostly a non-issue for now, as he is the only user, but I wanted to guide him in the right direction	16:40
	he is actually using the --author to check in the work other people are doing right now	16:40
drizzd	MetaCosm: he probably doesn't want a working copy where veryone else can access it	16:40
	→ psoo joined	16:40
drizzd	that's a recipe for disaster, that's like clearcase	16:40
	→ jackdempsey joined	16:41
	← bauruine left	16:41
	drizzd ducks	16:41
MetaCosm	drizzd, so the recommended setup would be to setup a --shared on the share, and work locally on his box	16:41
	and then pull/push	16:42
	→ hobodave joined	16:42
drizzd	that would work, yes. If he's the only one to push to it, no need to make it --shared (you can always make it world readable)	16:42
MetaCosm	well, access to this directory is highly restricted anyway	16:42
	← ToxicFrog left	16:42
	← disappearedng left	16:43
EvanR-work	git via smb :S ?	16:43
	→ ToxicFrog joined	16:43
	← munichpython left	16:44
	→ m104 joined	16:44
drizzd	EvanR-work: should work. If you only use it for push/pull, it's not a performance issue either	16:45
	actually, at work we don't even have local disk space	16:45
	everything is on shares	16:45
	← aschueler left	16:46
drizzd	but that's terrible, yes	16:46
	← yevlempy left	16:46
	← badboy_ left	16:48
	→ dj_tjerk joined	16:48
	→ camonz joined	16:49
	→ LiamH joined	16:50
	← pipegeek left	16:51
	→ ft joined	16:51
	→ hipe joined	16:52
	→ amchale joined	16:52
	← amchale left	16:53
	→ rgr joined	16:53
	→ badboy_ joined	16:53
	→ jepster joined	16:53
	← sagsousuke left	16:53
jepster	which command fetches a svn repo? git-svn is already installed.	16:53
	→ sh1mmer joined	16:54
	→ lresende joined	16:54
wereHamster	jepster: git svn clone ...	16:54
jepster	thx	16:54
	→ aresnick joined	16:55
	→ sherr joined	16:56
	← hipe left	16:57
	→ drubies joined	16:57
	→ kar8nga joined	16:58
	← project2501a left	16:59
	← ph^ left	16:59
	← flazz left	16:59
	→ ph^ joined	17:00
	→ juan_arandaalvar joined	17:01
	← nanoo left	17:01
	dwmw2_gone → dwmw2_HIO	17:02
	← intripoon left	17:02
	→ yevlempy joined	17:04
sherr	Question: I am a new user of Git and am trying to use it to version control some docs I work on. Unfotunately, the docs are mainly binary and so do not automatically merge (I need to manually copy/paste between any branches).	17:04
	→ ph^_ joined	17:05
EvanR-work	git is really for source code	17:05
locklace	git is not a great choice for tracking chunks of binary data	17:05
	← joeconyers left	17:05
EvanR-work	generally, text files	17:05
locklace	EvanR-work: that's too restrictive, it's fine for lots of things besides code	17:05
	→ ssm2017 joined	17:06
ssm2017	hello	17:06
sherr	I have 2 branches in the repo - "2009" and "master" (current docs). I stillo need to occasionally switch to "2009" and then manually merge to master. I am thinking that it would haqve been better to have 2 separate repos - one for 2009 and one for "current". Can I split the repo on branch - how? filter-branch somehow?	17:06
MetaCosm	sherr, change to latex?	17:06
	hehe :)	17:06
EvanR-work	yeah, that would work	17:06
	← fmardini left	17:07
sherr	I know it is best for text .. and I am looking at switching to docbook. But there are a lot of docs now and only one of me.	17:07
	← ph^ left	17:07
	→ riz2 joined	17:07
sherr	I will be dealing with binary for a long time.	17:07
EvanR-work	what are they, microsoft word?	17:08
	← textarcana left	17:08
sherr	It seems a bit painful switching between branches for manual merges (or saved copies). Maybe 2 repos - 2 "clones" would be easier.	17:08
ssm2017	i hav copied a folder containing a git repo, after copying, git status is telling me that some files have been modified but they are not modified, how to clean this ?	17:08
	← mrchrisadams_ left	17:08
sherr	Not MS Word - something called "Help and Manual" http://www.ec-software.com/	17:08
EvanR-work	ouch	17:09
	← corni left	17:09
	← zodiak left	17:09
	→ kipras joined	17:09
sherr	I inherited this - wanted to reformat but ... pressing need to get docs out, add content etc. Too much current content and would take too long to reformat it all ...	17:09
	→ smuf joined	17:10
EvanR-work	another thing is that if your documents are compressed too, git wont be able to be as disk efficient	17:10
	→ loincloth joined	17:10
locklace	the only thing worse than an opaque binary document format is a little-used opaque binary document format	17:10
sherr	I tested how efficient it was and it seems OK	17:10
	← Pupeno left	17:10
sherr	Can I switch 2 branches out (master / 2009) to separate repos easily?	17:11
GyrosGeier	hmm	17:11
	GyrosGeier needs a three-way bisect	17:11
EvanR-work	sherr: git clone ?	17:11
	← mtkd left	17:11
EvanR-work	GyrosGeier: trisect!	17:11
	→ avar joined	17:11
GyrosGeier	i.e. I have two different behaviours that are wrong, and need to find the commits where each of the wrong behaviours were introduced	17:11
	however these are related	17:12
	→ ph^ joined	17:12
sherr	EvanR-work: git clone just leaves me with 2 repos, each with the same 2 branches. Then I would need to trim each to the single branch.	17:12
EvanR-work	you can delete the branches if you want	17:12
	or ignore them	17:12
	but this use of git is terrible	17:13
MetaCosm	sherr, maybe at some point in the future, you can script the convertion :)	17:13
sherr	EvanR-work: So delete 2009 from the "master" repo and "master" form the 2009 repo (maybe rename 2009 branch to "master")	17:13
EvanR-work	is it really two branches	17:13
	or two projects	17:14
sherr	Put to a new "2009" bare repo on the server	17:14
	*push	17:14
	→ ph^__ joined	17:14
sherr	It's the same document but branched as a "release" for 2009 - like you'd do in CVS ... :-)	17:14
	← yevlempy left	17:15
EvanR-work	cant deal with this so early in the morning, good luck ;)	17:15
sherr	Well, my simple-minded attempt at trying Git anyway. I know it is not ideal for this.	17:15
	← ph^ left	17:15
	← ph^__ left	17:15
	→ ph^ joined	17:16
	← jepster left	17:16
	→ ph^__ joined	17:16
sherr	EvanR-work: no prob. Might try the mailing list - easier to document and wait :-) thanks.	17:16
	Mike_lifeguard → Mike\|\|gone	17:17
	← aziz left	17:18
	← zoniq left	17:19
	→ enherit joined	17:19
	← ph^_ left	17:19
	→ zoniq joined	17:19
	← TeckniX left	17:20
	fsk → frsk	17:20
	→ TeckniX joined	17:20
	→ krawek joined	17:21
	← tobiassjosten left	17:21
	← loincloth left	17:22
	← ehabkost left	17:23
	← krunk-\|work left	17:23
	→ ehabkost joined	17:24
	→ kmartiin joined	17:26
	← bronson left	17:26
	← kmartiin left	17:26
	← jackdempsey left	17:27
	← palisv left	17:27
	→ corni joined	17:28
jMyles	I'm a little lost. I have a repo whose file structure is on a live server. Yet, when I commit a change, the change shows up in git log, but the file that I modified is unchanged on the server.	17:29
	← tvw left	17:30
locklace	pushing to a non-bare repo?	17:30
Espen-_-	hehe been there done that :P	17:30
	jMyles: http://toolmantim.com/thoughts/setting_up_a_new_remote_git_repository	17:30
	check that out	17:30
locklace	http://git.wiki.kernel.org/index.php/GitFaq#Why_won.27t_I_see_changes_in_the_remote_repo_after_.22git_push.22.3F	17:30
	← steffkes left	17:31
	→ defectiv joined	17:31
Espen-_-	and what locklace said	17:31
	→ yevlempy joined	17:31
jMyles	locklace: I have looked for a simple explanation of what a "non-bare" repo is, but I'm afraid this is still above me.	17:31
	← ringlej left	17:32
	→ Fullmoon joined	17:32
	← Fullmoon left	17:32
	→ brodyberg joined	17:32
	← tetha left	17:33
	← corni left	17:33
	← bcoca left	17:33
Espen-_-	jMyles: non-bare repository: a normal clone, has a working directory with checked out files	17:33
	jMyles: bare repository: a repository cloned using the --bare option, only includes the files/folders inside of the .git directory	17:33
	source: http://www.gitready.com/advanced/2009/02/01/push-to-only-bare-repositories.html	17:33
	→ bcoca joined	17:33
jMyles	locklace: OK, I understand the (good) reason why changes are not pushed to the file tree. How then can I create a workflow wherein everybody's changes are viewable on the live development server?	17:34
locklace	jMyles: a repo is "non-bare" if it has a worktree, "bare" if it doesn't	17:35
jMyles	Espen-_- : so a "bare" repo doesn't have the actual file structure?	17:35
	→ tetha joined	17:35
	← aresnick left	17:35
EvanR-work	which means you cant use it as the live server	17:35
jMyles	worktree - OK :-)	17:35
Espen-_-	jMyles: you need two repos then on your development server	17:35
	both bare and non-bare	17:35
	← martoche left	17:35
jMyles	Espen-_- : Yes I think locklace was trying to explain this to me, but I didn't get it. So I'll have a non-bare that will actually serve the files? And a bare that is for developer commits?	17:36
bremner	or, better, don't deploy with push	17:36
locklace	jMyles: i explained the best workflow for your situation in detail	17:36
	← JonathanT left	17:36
jMyles	bremmer: I'm interested - tell or show me more?	17:36
bremner	use git archive	17:36
jMyles	locklace: I know :-\ I'm sorry I'm just not turning on to this quickly. It's just barely out-of-reach for me still.	17:36
Espen-_-	jMyles: I had the same "issue" as you, but I ended up paying for github and get a private repo there	17:37
	→ khmarbaise joined	17:37
	→ parappa joined	17:37
	→ Theravadan joined	17:37
jMyles	Espen-_-: Really? What did github do for you that a server didn't?	17:37
Espen-_-	Gitweb never gets as good as github	17:38
	→ priidu_ joined	17:38
	→ justin-george joined	17:38
jMyles	locklace: this comment is still mysterious to me: jMyles: you're going to create two clones of the central repo with git clone, one for each web server	17:39
Espen-_-	And I was wasting to much time on trying get stuff to work and github was cheap.	17:39
	← shruggar left	17:39
	jotik^^ → jotik	17:39
locklace	jMyles: picture a circle called "central". now picture a couple of circles off to the southwest called "prod" and "dev" connected to central by lines with arrows on the prod and dev ends.	17:40
	← ehabkost left	17:40
	→ aresnick joined	17:40
	← parasti left	17:41
	→ nanoo joined	17:41
	← wolog_ left	17:42
	→ corni joined	17:42
	← froschi left	17:42
	→ bushwakko joined	17:43
	→ ringlej joined	17:43
	→ ehabkost joined	17:43
	→ rduplain_ joined	17:43
	← rduplain_ left	17:44
	← rduplain left	17:44
	→ rduplain joined	17:44
	← locklace left	17:47
	← aresnick left	17:48
	→ aresnick joined	17:48
jMyles	locklace, et. al.: Does it make more sense for us to just edit the files on the server? (like in /var/www/project-devel) ?	17:50
	→ locklace joined	17:51
	→ Tank` joined	17:51
	← sh1mmer left	17:51
	← Vortex35 left	17:52
	→ ddp_ joined	17:53
	→ Pupeno joined	17:53
	→ froio joined	17:54
	→ fool_ joined	17:54
	← fool_ left	17:54
	→ fool_ joined	17:54
	← t0rc left	17:54
froio	any free git repository ?	17:55
wereHamster	froio: github, gitorious	17:55
	→ zodiak joined	17:56
	→ doener joined	17:56
eli	When I create a new branch, and then `git push origin branch-name' it creates the banch on the upstream, but then `git pull' won't work without an explicit `git pull origin branch-name' -- is there some obvious way to do that when the branch is created? (Obvious = not some explicit use of `git remote' to set it up later, something that makes the connection from the start.)	17:57
	← fool__ left	17:57
	← ddp_ left	17:58
	← parappa left	17:58
	→ WebDragon joined	17:58
	← ehabkost left	17:59
WebDragon	so I'm still having this lovely rebase issue	17:59
	core.autocrlf=true and core.whitespace=nowarn are set	17:59
	→ flazz joined	17:59
Aides	eli: i don't know any, but you just have to use git branch --set-upstream once	17:59
	eli: (after git-push that is)	17:59
WebDragon	www.webdragon.net/sgrab.jpg is the current tree in gitk --all	17:59
	→ sh1mmer joined	18:00
	→ ehabkost joined	18:00
	→ bambule joined	18:00
WebDragon	the feature-footer changes res/css/stylehome.css and res/includes/inc_footerhome.php. whereas the file changed in the develop merge is res/includes/inc_javascript.php so there shouldn't even BE any conflict	18:00
	and yet...	18:00
	if I git checkout feature-footer; git rebase develop;	18:01
	I don't get a rebase. I get error messages.	18:01
	← christophsturm left	18:01
WebDragon	granted this is 1.6.2.5 (I've got F12 installed on the new 320g drive but haven't migrated myself over fully yet cuz I still need to do my job while everything's in transition)	18:02
	but still. this should "just work"	18:02
	← vall left	18:02
	← flazz left	18:03
WebDragon	http://fpaste.org/oucE/ this is the crazy nonsense I get	18:03
	→ Vortex35 joined	18:03
eli	Aides: How should I use it? (I'm trying a few things they don't do what I want...)	18:03
	→ icwiener joined	18:04
WebDragon	I'm _definitely_ not doing anything complex here, and I'm reasonably sure I'm not doing anything specifically _wrong_	18:04
	so can anyone help me figure out just what's happening here?	18:04
przemoc	eli: pushing/pulling has nothing to do with tracking. so you cannot expect this commands to do things beyond their scope.	18:05
	eli: --set-upstream requires git 1.7	18:06
WebDragon	there doesn't seem to be any particular reason for this to happen like it is	18:06
	← Pupeno left	18:07
	→ malumalu joined	18:07
eli	przemoc: I know; but the thing is that pulling on a different clone and switching to the new branch (with just `git pull; git checkout new-branch') does set up push/pull from the remote -- it's only on the original clone where the branch was made that it needs to be set explicitly.	18:07
	(And yes, I'm on 1.7.0.)	18:07
WebDragon	about the only thing I can do is create a patch from my feature branch, checkout develop and delete the branch, create a new one, and reapply the patch	18:07
	← gebi left	18:07
pasky	WebDragon: I don't think "nowarn" is valid value for core.whitespace	18:09
	WebDragon checks	18:09
pasky	WebDragon: you might want something like -space-before-tab,-blank-at-eol	18:10
	→ brizly joined	18:10
WebDragon	wow I wonder how the hell I screwed that up. the manpage is pretty clear on that in git.config	18:10
pasky	WebDragon: git is perhaps overly agressive at trying to help you avoid bad whitespaces	18:11
froio	any free git repository ?with private project storeing?	18:11
pasky	WebDragon: maybe newer versions work more reasonably, i'm not sure	18:11
	WebDragon goes all bug-eyed which looks quite humorous on a dragon	18:11
przemoc	eli: you can have many repositories storing your objects, therefore automatically setting all push/pull target/source _always_ wouldn't be a good thing	18:11
pasky	hehe	18:11
przemoc	froio: your computer	18:11
WebDragon	pasky: the odd thing is, since I'm working with html/css/php/javascript, whitespace is far less of an issue	18:11
froio	how ?	18:11
	→ Pupeno joined	18:11
pasky	WebDragon: compared to...? :)	18:11
froio	home git server? how i can do it?	18:12
WebDragon	but since the files that were changed don't even _conflict_ I shouldn't be seeing any crap whatsoever	18:12
pasky	WebDragon: well, git rebase works by wrapping git apply which enforces these rules	18:12
WebDragon	pasky: heh. compared to whatever reasons git would have to bitch about whitespace? :D shrug for dramatic effect	18:12
	→ godsflaw joined	18:13
pasky	WebDragon: they cause spurious diffs and conflicts with "invisible" differences	18:13
	it's just a matter of sanity i guess	18:13
WebDragon	pasky: I am still confused as to why this would conflict.. if you look at the gitk --all screengrab , there's almost NO difference in the branches	18:13
	← bushwakko left	18:13
przemoc	eli: if you hate using git remote than push once your branch, delete it, and checkout with tracking	18:13
pasky	WebDragon: sorry i don't have a working web browser handy	18:13
przemoc	eli: delete it just locally of course ;]	18:14
WebDragon	develop is one commit further than feature-footer branch and the three files changed in the two separate commits are different files	18:14
pasky	well there is no _conflict_	18:14
	is there?	18:14
	Mike\|\|gone → Mike_lifeguard	18:14
WebDragon	one file in the develop branch and two in the feature branch. none of them are the same file. there should be no conflict	18:14
pasky	just the whitespace complaint?	18:14
WebDragon	looks that way yes	18:14
	so I'm confused about the whitespace complaint	18:14
	the whitespace should be the same, in the files it's looking at	18:14
	← bluenovember left	18:15
	→ bronson joined	18:15
	→ AkumaStreak joined	18:15
WebDragon	I would just expect it to apply the patch and move on	18:15
pasky	it's not about them being different, just bad	18:15
WebDragon	hmm	18:15
pasky	and i'd say this is a bug	18:15
	maybe newer git has this fixed	18:15
WebDragon	if that were the case then I could wish it would 'warn' me about them sooner :)	18:15
pasky	yes that's the real wtf here	18:15
WebDragon	during normal commits, for example	18:15
	pasky: thedailywtf.com ? :D	18:15
pasky	;)	18:16
	→ bdiego joined	18:16
	WebDragon chuckles deeply	18:16
pasky	sorry i don't understand the whitespace magicks in git too well	18:16
WebDragon	apparently neither do I xD	18:16
froio	home git server? how i can do it?	18:16
przemoc	froio: do you really need publicly available private repository?	18:16
	← PointMan left	18:16
froio	yes	18:16
pasky	why?	18:16
DrNick	WebDragon: you confused core.whitespace with apply.whitespace	18:16
froio	secret project	18:17
WebDragon	DrNick: ahh okay	18:17
	WebDragon goes to look at that too	18:17
przemoc	but publicly available for sharing purposes with secret coworkers, right?	18:17
froio	yes	18:17
	→ tvw joined	18:17
przemoc	froio: so you don't need publicly available server	18:17
pasky	you could stenographically encode the repository content into photos on flickr	18:18
	*stegano	18:18
	even ;)	18:18
WebDragon	DrNick: so it looks like I could set apply.ignorewhitespace=change	18:18
froio	lol	18:18
WebDragon	does that make any sense?	18:18
	← corni left	18:18
DrNick	WebDragon: yes	18:19
WebDragon	and then remove my spurious core. bit	18:19
	I'll try that and reattempt the rebase	18:19
przemoc	froio: git is designed to work flawlessly "off-line" (in terms of centralised vcs) and decentralised.	18:19
	← nanoo left	18:20
przemoc	froio: so you can just share your work with team mates while you're online via ssh	18:20
	← nicoulaj left	18:21
	← AkumaStreak left	18:21
WebDragon	wow it still bitches about whitespace	18:22
	← camonz left	18:23
WebDragon	man I thought we left segregation behind us decades ago	18:23
	← Textmode left	18:23
	← rgr left	18:23
	→ Textmode joined	18:23
	→ wolog joined	18:23
	WebDragon tunes in Galaxy News Radio for a bit of "into each life some rain must fall..."	18:24
	→ dreiss joined	18:25
	← froio left	18:27
	→ PointMan joined	18:28
	→ AkumaStreak joined	18:30
WebDragon	ok. setting apply.whitespace=nowarn and apply.ignorewhitespace=change now it proceeds without all the spuriousness. however it STILL refused to rebase	18:30
	now I get this:	18:30
	http://fpaste.org/zLiT/	18:31
	feature-footer was checked out and clean according to git status	18:31
	→ corni joined	18:31
WebDragon	www.webdragon.net/sgrab.jpg is the current state of the tree in gitk --all	18:31
	feature-footer differs from develop by a single commit (merge in another feature branch)	18:32
	← bdiego left	18:32
WebDragon	(and said other feature branch also branched off the develop branch at the same point as feature-footer)	18:32
	← hyperair left	18:32
pasky	that looks very weird :(	18:33
	can't you try newer glibc version?	18:33
	is this on linux or some other os?	18:33
	well sorry i gotta go anyway	18:33
WebDragon	fedora 11	18:33
pasky	eh	18:33
	s/glibc/git/ :)	18:33
	newer git version	18:33
	hmm	18:33
	that _is_ strange	18:33
WebDragon	yeah I'm upgrading to F12 over the next day or two	18:33
	so I'll have a newer git version	18:33
	→ pantsman joined	18:34
WebDragon	but still IMHO this shouldn't be happening	18:34
	the files being merged/rebased don't conflict at all	18:34
pasky	well	18:34
biz	WebDragon: try to apply the 0001 patch manually with git apply now, it will most probably give you more insight on what's wrong with it	18:34
	→ sh1m joined	18:34
pasky	this message appears to be about uncommitted changes	18:34
	but yuo say git status produces no output	18:34
WebDragon	stylehome SHOULD be different since it's a feature branch	18:34
	and it complains that they differ.. but of course they differ, and they should STILL differ after this branch is moved forward	18:35
pasky	no	18:35
	error: res/css/stylehome.css: does not match index	18:35
	this is the key complaint	18:35
WebDragon	yeah. what precisely does that mean, to me	18:35
	it's saying there's uncommitted changes to that file in the feature branch?	18:36
	← bradb left	18:36
pasky	no, in your worknig tree	18:36
	hmm	18:36
	maybe you have some untracked version of the file before rebase start there?	18:36
	are you sure git status gives absolutely nothing?	18:36
	afk	18:37
WebDragon	only three branches exist. master, develop (which was just merged into master) and feature-footer (which was just branched off of develop)	18:37
	develop proceeded to spawn another feature-branch which had one commit on it and it was then merged into develop and deleted, but that branch diverged from develop at the same commit that feature-footer did	18:37
	← sh1mmer left	18:38
	sh1m → sh1mmer	18:38
	← pantsman left	18:39
	← Tank` left	18:40
eli	przemoc: It's not that I hate using `git remote' -- I'm just wondering why getting a branch sets up both push/pull directions, but pushing a new branch only sets up pushing -- I'd expect there to be some command line option to do the same setup conveniently.	18:43
	It just seems asymmetric enough to have some reason, and understanding why it's done so will make it easier to explain why things are as they are.	18:44
	← elyob left	18:44
przemoc	eli: getting a branch and push/pull are two seperate operations, push and pull doesn't do anything with config	18:45
	← ssm2017 left	18:45
eli	Yes, but there is some configuration happenning when you checkout a branch that also exists on the remote...	18:46
	← tildeequals left	18:46
	→ maxter joined	18:46
	→ pantsman joined	18:46
	→ Zephyrus joined	18:46
teukka	eli: you can give --set-upstream to push to set up tracking	18:47
eli	Given that `--set-upstream' is new, perhaps all of these automatic configuration things are still new though.	18:47
	← ringlej left	18:48
eli	teukka: Aha! That makes sense now -- thanks!	18:48
	← unregistered left	18:48
	← gitte left	18:50
	→ dpino joined	18:50
	→ hipe joined	18:51
	→ chrisadams joined	18:51
	← themill left	18:52
	→ flazz joined	18:52
	→ bcardarella joined	18:53
	→ ebneter joined	18:55
	→ loincloth joined	18:57
	→ camonz joined	18:57
przemoc	eli: indeed, I wasn't aware of -u for push.	18:58
	→ joeconyers joined	18:59
WebDragon	if I format the feature branch as a patch. delete the branch, while checking out develop again. create a new feature-branch at the tip of develop, and apply it. it goes cleanly	18:59
	WebDragon facepalms. it shouldn't be this difficult :D	18:59
	→ bauruine joined	18:59
eli	przemoc: Yeah, it's nice to know about it. I'm in the process of migrating a central repository to git, and will need some description of how common things are done -- which is why I wasn't satisfied with setting it manually or removing the local branch and re-adding it from the remote: I can deal with it myself, but I need to minimize chance of confusion for people who will use these recipes (and we're pretty much	19:01
	all new to git).	19:01
lavamind	when adding a submodule to a repository, can it be tied to a particular branch or do all branches have to take in the submodule ?	19:01
	← fullerton left	19:02
wereHamster	lavamind: it's just like any other file, it can be part of one branch but not part of another branch	19:02
	→ parasti joined	19:02
	→ JonnyRo joined	19:03
	→ jorgeortiz85 joined	19:03
	→ project2501a joined	19:03
JonnyRo	I'm trying to use msysgit on windows. Does anyone have any tips for increasing the amount of memory available to it? I'm getting out of memory errors when checking out large trees	19:03
	→ coolcat joined	19:03
JonnyRo	it appears that git only has about ~250MB available memory allocated to it	19:04
	→ _iron joined	19:04
lavamind	wereHamster: that's what i thought, too, but immediately after adding a submodule, I cannot switch branches : error: Untracked working tree file 'moules/foo/bar.php' would be overwritten by merge.	19:04
wereHamster	that has nothing to do with the submodule	19:05
	→ rgr joined	19:05
lavamind	is it because I did submodule init ?	19:05
wereHamster	well, what exactly did you do?	19:05
lavamind	on my superbranch : submodule add ssh://.... modules/foo followed by submodule init followed by git add modules/foo .gitconfig followed by a regular commit	19:07
	→ friskd joined	19:08
friskd	if my remote repository has a branch that i want to checkout, how do i get to it?	19:08
	i can't see it on my local maching	19:08
	*machine	19:08
jorgeortiz85	I've been using git-svn successfully for a while, but today I got this error message while trying to do a 'git svn fetch': fatal: ambiguous argument 'aa27a1704498ca023516dea7526158f2d772bbe3^..06482d90eb5eb77254ceb6262759cf6b0e0a9d4b': unknown revision or path not in the working tree. Anyone know what the problem might be?	19:08
	→ r11t joined	19:09
	← imajes left	19:09
friskd	I figured it out.	19:10
	→ imajes joined	19:11
	→ dekroning joined	19:11
	← Cesario_ left	19:11
	→ bushwakko joined	19:12
	→ tildeequals joined	19:12
dekroning	i have a git project, which i would like to copy to a friend, however i don't want to copy all the git repository information along, how can i extract the code without the git info?	19:12
	← anxiolytic left	19:13
	← r11t left	19:13
	→ anxiolytic joined	19:13
MetaCosm	ummmm... don't copy the .git directory... done!	19:13
	← poseid left	19:14
dekroning	ow oke :-)	19:14
	→ r11t joined	19:14
	← zoniq left	19:14
	← pantsman left	19:15
	← r11t left	19:18
	→ r11t joined	19:18
	→ ddp_ joined	19:19
jorgeortiz85	anyone know why git svn fetch might be blowing up with "unknown revision or path not in the working tree"?	19:19
fryguy	jorgeortiz85: somebody screwed with the old revision history or something	19:19
jorgeortiz85	fryguy: in git or in svn?	19:20
	→ Alagar joined	19:22
	→ kumbayo_ joined	19:23
	← r11t left	19:25
	← David-T left	19:25
	→ poseid joined	19:26
	← kumbayo left	19:26
	→ git___ joined	19:27
git___	how does one track .git directory?	19:27
fryguy	git___: you don't, you name it something else	19:27
	if you have actual content in .git that you want to track that isn't metadata for the repository itself	19:28
git___	i have a bunch of pre-commit and post-commit stuffs I want to track	19:28
fryguy	.git is reserved for used by git itself, you can write wrapper scripts that change this if you really really want	19:28
git___	okie	19:28
fryguy	git___: put them in a separate directory and symlink them into .git	19:28
	→ Cesario_ joined	19:28
git___	ah	19:28
	← carolyn left	19:29
fryguy	or, as I did in my case, make a separate repo for them and create and install script to install them into a given git repository	19:29
	→ themill joined	19:29
	→ palisv joined	19:29
	→ Tank` joined	19:31
	→ scrybe joined	19:31
	→ schacon joined	19:31
	← hipe left	19:33
	→ yann joined	19:33
	→ hipe joined	19:34
schacon	anyone using notes at all?	19:34
	← scrybe left	19:35
Mike_lifeguard	where's the skeleton used by git init kept?	19:36
	Mike_lifeguard wants to add a standard pre-commit script	19:36
	← flazz left	19:38
	← poseid left	19:39
Aides	/usr/share/git-core/templates I presume	19:39
	→ scrybe joined	19:39
	← jMyles left	19:41
	→ rook2pawn joined	19:42
	← Pewpewarrows left	19:44
Mike_lifeguard	great, thanks	19:44
rook2pawn	So git clone seems to run a bit slower, so i downloaded the zip of my group's repo from github. After I unzip the repo, is there a way to link my local unzip to our groups git project? I.e. simulate git clone	19:44
	→ aziz joined	19:44
	→ pantsman joined	19:45
Mike_lifeguard	And if you do "git init" in a repo, it will just update from /usr/share/git-core/templates right?	19:46
	← scrybe left	19:46
	→ saste joined	19:47
	← drubies left	19:47
pantsman	phew, I just moved absolutely everything in a repo into one directory deeper	19:47
	let's see how git handles this :>	19:48
saste	hi all	19:48
	I just did git co OLD_REV	19:48
	→ darwin joined	19:48
saste	now I want to go back to the previous HEAD	19:48
git___	how do i use the variables in config?	19:48
saste	how can I do that?	19:49
git___	for example, [core] worktree = /home/git/	19:49
	→ iamcam joined	19:49
saste	I tried to have a look with gitk bute there is no trace of the old HEAD...	19:49
git___	echo $core.worktree?	19:49
	→ magn3ts joined	19:50
wereHamster	git___: git config core.worktree	19:50
locklace	saste: git reflog	19:51
rook2pawn	How do you link a directory with a git repository, without downloading/cloning the repo?	19:51
	← lresende left	19:51
git___	thanks wereHamster	19:51
Rhonda	rook2pawn: git init and git remote add I would think.	19:51
wereHamster	rook2pawn: can you explain a bit more?	19:51
	→ __iron joined	19:52
	← friskd left	19:52
saste	locklace: thanks fixed	19:52
	← magn3ts left	19:53
	→ flazz joined	19:53
wereHamster	rook2pawn: ah, I've read the backlog now. The zip you downloaded, was that a normal tarball?	19:54
	← alezandro left	19:54
rook2pawn	well, I downloaded the zip archive of our group's project, and after i unzip it, i need to link it so my git push and git pull are linked to the remote url	19:54
wereHamster	in that case do what Rhonda said	19:54
rook2pawn	Yes, its basically a normal tarball	19:54
wereHamster	however I don't see a difference between first downloading the tarball and then doing a git init/git fetch and a git clone	19:55
	← _iron left	19:55
	→ Twisted_Mentat joined	19:55
rook2pawn	Thank you Rhonda, wereHamster. What had happened is that our project contains thousands of files and our conection is bad, so git clone fails with tons of net retry's	19:55
wereHamster	both will use up almost exactly the same amount of bandwidth (and if git clone was slow then git fetch will be slow as well)	19:56
charon	rook2pawn: a tarball that doesn't contain the .git is useless wrt. saving clone/fetch bandwidth. see man git-bundle for the real solution	19:56
Gitbot	rook2pawn: the git-bundle manpage can be found at http://git.or.cz/man/git-bundle	19:56
	← thiago left	19:57
wereHamster	if that is a public repo I'll be happy to create a bundle for you and host it in a place where you can download it via http	19:58
	→ David-T joined	19:58
	→ novas0x2a joined	19:58
	→ mjf joined	20:01
	← novas0x2a1 left	20:01
	← parigaudi left	20:02
	← JackWinter left	20:04
	← bushwakko left	20:05
	→ JackWinter joined	20:06
	→ jceb joined	20:06
	← pantsman left	20:07
	← Vortex35 left	20:07
	→ Vortex34 joined	20:07
	← lucsky left	20:07
	→ bushwakko joined	20:08
	← g0bl1n left	20:08
	← bushwakko left	20:08
	→ maattd joined	20:08
	→ mrchrisadams_ joined	20:09
	← priidu_ left	20:10
	← Raska left	20:10
	→ magn3ts joined	20:11
	← icwiener left	20:12
	→ rudi_s joined	20:13
	← Alagar left	20:13
	← smuf left	20:13
	→ priidu_ joined	20:13
anxiolytic	what does it mean when git says I'm ahead/behind? [ahead 3, behind 1]	20:13
	→ shennyg joined	20:14
	lucid_ → lucid	20:14
	→ nanoo joined	20:14
	→ christophsturm joined	20:14
mrchrisadams_	hi guys - I'm trying to get gitweb/gitosis to work on a server with apache - anyone running this setup here?	20:16
locklace	anxiolytic: your local version of the branch relative to the remote's	20:16
	← rudi_s_ left	20:16
	← Zephyrus left	20:17
locklace	mrchrisadams_: word is you should use gitolite instead	20:17
	→ j416 joined	20:17
	→ shiggins joined	20:19
wereHamster	anxiolytic: the remote has one commit which you don't have, and you have three commits which the remote doesn't have	20:20
mrchrisadams_	locklace: I don't think that's an option right now, sadly	20:21
	→ Zephyrus joined	20:21
mrchrisadams_	would domain/cgi-bin/gitweb.cgi fail silently if it wasn't chmodded?	20:21
WebDragon	mrchrisadams_: it'd show up in error_log, most likely	20:22
	as to how silently, it depends on the error reporting	20:22
	→ pantsman joined	20:22
WebDragon	(insofar as the user is concerned)	20:22
	usually one tries not to expose one's internals to the world wherever possible	20:23
	unless you're tiger woods	20:23
	WebDragon runs	20:23
	← magn3ts left	20:24
mrchrisadams_	WebDragon: can i trace this in the logs at all? I'm seeing nothing in the apache logs	20:25
	nor in /var/log/messages	20:25
	(I'm on centos)	20:25
	← tvw left	20:25
	← lhz left	20:27
	← christophsturm left	20:27
	→ thiago_home joined	20:27
	→ kdasdasf joined	20:28
kdasdasf	hi how can i set the editor i want to use to edit commit messages?	20:29
	← pantsman left	20:29
maattd	kdasdasf: env variable EDITOR	20:30
Mike_lifeguard	kdasdasf: it uses $EDITOR, but you can use something else for composing messages by setting core.editor	20:30
kdasdasf	Mike_liefguard like git config --global core.editor "vim"	20:31
	would this work?	20:31
Mike_lifeguard	yes	20:31
shiggins	I have gitosis running on ubuntu just fine. I have created a repository that I am sharing with multiple parties just fine. Question: How do I configure gitosis to support an additional root repository in addition to ~/repository? I see there is an option in gitosis.conf	20:31
	repositories = repositories but not sure how to use it?	20:32
	→ bushwakko joined	20:32
WebDragon	mrchrisadams_: that's what I meant. apache error_log not /var/log/messags	20:32
	I'd not expect to see a failed web cgi script show up there	20:32
mrchrisadams_	WebDragon: where woudl you normally look when you're trying to troubleshoot something like gitweb?	20:34
WebDragon	what you'd normally see in there is something along the lines of 'premature end of script headers' if the script has a problem.	20:34
	if it's permissions then you'd see a normal error for apache, like the 500/404/403 etc	20:34
	the access_log will also indicate whether it was 200 (OK) or whatever, as well, so look for that file mentioned there, and see what the result-code is	20:35
	← anxiolytic left	20:35
WebDragon	grep cgi-bin/gitweb.cgi access_log	20:35
	see what turns up for the result code	20:35
	→ anxiolytic joined	20:35
	→ Adaptee joined	20:35
Mike_lifeguard	When working with windows and linux users, autoclrf should be "input" for the linux folks? What for windows folk?	20:36
	→ migge joined	20:36
	→ pantsman joined	20:36
	← __iron left	20:37
WebDragon	Mike_lifeguard: the difference between true and input can be found here: http://www.kernel.org/pub/software/scm/git/docs/git-config.html	20:37
	I'm not sure which would be most preferable from the windows-programmer's side	20:38
	I try to avoid programming while on windows, wherever possible	20:38
	← anxiolytic left	20:38
WebDragon	it's just marginally useful enough to play a few games on :P	20:38
mrchrisadams_	oh ffs	20:39
	this server had stale dns	20:39
WebDragon	switch to google's public dns servers	20:39
mrchrisadams_	when i cranked up the verbosity on curl, i saw"Premature end of script headers: gitweb.cgi"	20:40
	→ WALoeIII joined	20:40
	→ __iron joined	20:40
	→ WebDragon\|away joined	20:40
	→ Cr`eme\|Ubuntu joined	20:40
Cr`eme\|Ubuntu	Hello.	20:41
WebDragon\|away	was wondering when that was going to happen	20:41
	← WebDragon left	20:41
Cr`eme\|Ubuntu	Does anyone know if there is a way to integrate git with WEBSITE?	20:41
	WebDragon\|away → WebDragon	20:41
WebDragon	switch to google's public dns servers : 8.8.8.8 and 8.8.4.4	20:41
	← kdasdasf left	20:42
	→ froschi joined	20:43
lack	WebDragon: Good plan! That way they know where I'm going even when I didn't search first...	20:43
WebDragon	lack: that'd be a trick. half the time I don't even know where I'm going	20:43
	xD	20:44
	trilobyte → vinse	20:44
WebDragon	vinse: man evolutions' really accelerating these days	20:44
	WebDragon hides a chuckle in his coffeemug	20:44
	← krawek left	20:44
vinse	yeah man, i was pre-cambrian just a few minutes ago!~	20:44
	→ dougmencken joined	20:45
	→ krawek joined	20:45
dougmencken	hello again; can you congratulate me? I build git! well yes, only using make \|\| make -k ; mkdir DEST && make -ki install DESTDIR=`pwd`/DEST (coz building man requires python)	20:46
	← d0k left	20:46
dougmencken	http://manulix.wikidot.com/build-howto	20:46
vinse	dougmencken: grats! now build it for me on sparc thx	20:46
WebDragon	dougmencken: "building man requires python" -- a soliloquy on the garden of eden theory	20:47
	WebDragon hides yet another chuckle	20:47
vinse	that was pretty good, you can chuckle openly on that imo	20:47
dougmencken	WebDragon, do you need error messages if I will not use -k ?	20:47
	→ gebi joined	20:48
WebDragon	vinse: bows	20:48
dougmencken	vinse, I will produce shell scripts for building	20:48
	← brizly left	20:48
	← eletuchy left	20:48
dougmencken	like http://manulix.wikidot.com/local--files/files:pbs/autoconf.pkgbs	20:48
	→ eletuchy joined	20:49
	← rgr left	20:52
	← __iron left	20:52
	yevlempy → yev\|away	20:52
	→ EvanCarroll joined	20:53
EvanCarroll	http://stackoverflow.com/questions/2404628/how-do-i-join-two-git-repos-without-a-common-root-where-all-modified-files-are-t	20:53
	← mrchrisadams_ left	20:53
	← chrisadams left	20:53
	← brodyberg left	20:54
	→ chrisadams joined	20:54
	→ mrchrisadams_ joined	20:54
	→ __iron joined	20:54
	← jceb left	20:55
	→ g0bl1n joined	20:56
dougmencken	so can you fix this please? for system which don't have any python...	20:56
	→ jspiros joined	20:57
	→ Wicked joined	20:57
	← tstclair left	20:58
	→ tstclair joined	20:59
	→ Pewpewarrows joined	20:59
	← ntoll left	21:00
EvanCarroll	gah	21:01
	→ NET\|\|abuse joined	21:01
	← camonz left	21:02
	iamcam → iamcam\|afk	21:03
	→ johnw joined	21:03
	← iamcam\|afk left	21:03
	→ wwwald joined	21:03
	→ christophsturm joined	21:04
	← nanoo left	21:05
	← mithro left	21:06
	← Tureba left	21:06
	→ Markus23 joined	21:08
git___	what is the variable for commit_msg?	21:08
	i'm trying to add the commit message to my pre-commit script	21:08
przemoc	git___: -F	21:09
	← christophsturm left	21:09
	→ girasquid joined	21:09
girasquid	I'm having some trouble doing a pull - here's the error, something about not being able to merge because one file isn't uptodate: http://dpaste.com/169615/	21:10
	← tatsuya left	21:11
dougmencken	http://pastebin.org/106021	21:11
	← mssm left	21:11
Markus23	Hi, How can git-svn clone get all commits and not only of one specific trunk/branch?	21:12
	when I do git-svn clone url/to/unstable it only gets the commits from unstable branch	21:12
	when I do git-svn clone url/to/trunk only from trunk ;(	21:12
lack	girasquid: You can't merge because you have a local edit in that file that would be touched if git were to go ahead with the merge. What does 'git status' say?	21:13
	→ spaceonline joined	21:13
girasquid	looks like there is a local edit	21:14
	would stashing work here?	21:14
	→ tiglionabbit joined	21:14
wereHamster	Markus23: --stdlayout or use -b/-t/-T	21:14
lack	girasquid: Yes. You have 3 choices: Commit it, stash it, or blow it away.	21:14
girasquid	Alright - thanks	21:14
Markus23	wereHamster: were are the -b/-t/-T documented? man git-svn does not tell much about it?	21:14
Gitbot	wereHamster: the git-svn manpage can be found at http://git.or.cz/man/git-svn	21:14
git___	is there way to see commit history of a file?	21:15
lack	git___: git log -- path/to/file	21:16
git___	thanks lack	21:16
	← dougmencken left	21:16
lack	git___: Or, if you like pretty graphics: gitk -- path/to/file	21:16
	git___: Technically speaking that's not really the "commit history of the file", but it limits the history shown to only those commits who touched that file.	21:16
wereHamster	Markus23: it does, maybe your man page is outdated	21:16
	← bushwakko left	21:17
git___	ah	21:17
Markus23	wereHamster: I would like to get rid of svn on https://svn.libelektra.org/svn/elektra	21:17
	← flazz left	21:18
	← tiglionabbit left	21:18
Markus23	so I need something like git-svn clone -T trunk -t tags -b unstable https://svn.libelektra.org/svn/elektra/	21:19
	wereHamster: the problem is that the unstable branch is deleted already, but git-svn clone https://svn.libelektra.org/svn/elektra/unstable works	21:19
	← malumalu left	21:20
wereHamster	why is that a problem?	21:20
Markus23	wereHamster: I don't know the path for branches? is it . or unstable?	21:20
	← girasquid left	21:21
wereHamster	Markus23: you might want to use git init and then init the .git/config file manually. I have no experience with git-svn and no idea how well it handles non-standard layouts	21:22
	s/then init/then edit/	21:22
	→ fool__ joined	21:22
	← fool__ left	21:22
	→ fool__ joined	21:22
Markus23	wereHamster: -T -b and -t are specifically for non-standard layouts, or?	21:22
wereHamster	yes	21:23
	→ mssm joined	21:24
	→ killerchicken joined	21:24
killerchicken	where do I learn how git notes are merged/pushed etc?	21:24
	← fool_ left	21:25
	→ magn3ts joined	21:26
Markus23	wereHamster: however, I have here a git-svn checkout from trunk and unstable, isn't it possible to get all commits together with git-pull?	21:26
	→ SRabbelier joined	21:26
	← hugoxrosa left	21:27
schacon	killerchicken: i'm writing a blog post on that now, but it seems to be somewhat ill-defined	21:28
wereHamster	Markus23: sure you can do that, but since the two repositories were created separately they don't share the same root commit. So it likely won't represent the history in the svn repo	21:28
killerchicken	schacon: please ping me when you're done, that'd be great	21:28
	→ rgr joined	21:28
schacon	basically it seems that you have to setup your refspecs manually	21:28
	i'll try to remember - you can subscribe to the progit.org feed if you want	21:29
killerchicken	ok	21:29
schacon	should be tomorrow or wed if i can figure it out properly	21:29
Markus23	wereHamster: you are right, but it should be fixable with git-rebase, or?	21:29
	→ hornairs joined	21:30
	dwmw2_HIO → dwmw2_gone	21:31
	→ _nickel joined	21:31
_nickel	when I do a rebase and I have to fix a merge conflict and then add the resolution to the index sometimes I'll end up with no changes to be committed at which point I can't continue my rebase, what am I doing wrong?	21:32
	_nickel typed too much.	21:32
AkumaStreak	how do i check in a staged file	21:33
	but not the version i'm currently changing	21:33
	commit file puts the one with the most recent changes up for commit	21:33
Markus23	How do I get a clean git repos without the git-svn config stuff? git-clone?	21:34
hornairs	Would anyone mind giving me some assistance deleting some large binary files throughout my repo's history? I've run these commands so far: http://gist.github.com/325776, but when I push to origin (github) there are still these huge objects that take quite a while to compress and send that I never should have added.	21:34
wereHamster	Markus23: why?	21:34
	hornairs: man git-filter-branch (it has an example how to do that)	21:35
Gitbot	hornairs: the git-filter-branch manpage can be found at http://git.or.cz/man/git-filter-branch	21:35
Markus23	wereHamster: the idea is that if someone clones my repo he should get the same, and not be bothered that it has been in a svn repo before	21:36
	wereHamster: there needs to be some meta-info otherwise git-svn could not work?	21:36
	← rook2pawn left	21:36
Markus23	git-svn is really great btw...	21:38
	← corni left	21:38
EvanCarroll	does rebasing work if there in no common root?	21:39
	← tstclair left	21:39
EvanCarroll	I'm having a git quagmire here and i don't think the advice i got to rebase works if there is no common root	21:39
hornairs	wereHamster: thanks, the object count is still the same as before, thats not right is it? it doesn't seem to have worked	21:39
EvanCarroll	http://stackoverflow.com/questions/2404628/how-do-i-join-two-git-repos-without-a-common-root-where-all-modified-files-are-t/	21:40
	→ tstclair joined	21:40
	→ sagsousuke joined	21:41
	→ rook2pawn joined	21:41
	← cenuij left	21:43
	→ bradb joined	21:43
	← Wicked left	21:45
	→ Ademan joined	21:45
	← ehabkost left	21:48
	← migge left	21:49
	→ tvw joined	21:49
	← rduplain left	21:50
	marcus__ → imarcusthis	21:50
	→ Wicked joined	21:50
	← parasti left	21:50
	→ parasti joined	21:50
	→ friskd joined	21:51
	→ kassah joined	21:52
	→ Beaver joined	21:52
Beaver	www.search2.net (new search engine)	21:54
	← Beaver left	21:54
	→ mmattice joined	21:54
kassah	is there any way with git svn fetch to store a project default author map file? so I don't have to do -A every time I want to fetch it?	21:57
EvanCarroll	if i search for google, and it shows analytics before the homepage, I know there is a problem	21:57
	not just analytics, but a javascript	21:57
	← johnw left	21:57
kassah	at least google.co.jp is the second result =P	21:58
	→ Fullmoon joined	21:58
	← Ademan left	21:59
	← j416 left	22:00
	→ alezandro joined	22:01
	→ ambientsound joined	22:01
	← shiggins left	22:01
	→ migge joined	22:01
	← eletuchy left	22:02
	← juan_arandaalvar left	22:02
	→ eletuchy joined	22:02
ambientsound	hi, guys, I've searched all over for the solution but I can't find any relevant solution: "gic gc" dies with "error: failed to run repack". Repository is about 318M. When trying to pull the repo from remote, it gives me "error: git upload-pack: git-pack-objects died with error."	22:03
	any suggestions?	22:03
	→ j416 joined	22:03
	← Mezner left	22:03
ambientsound	I've tried pack.threads = 1, pack.deltaCacheSize = 1, pack.windowMemory = different sizes on both sides	22:03
	→ eFfeM1 joined	22:04
ambientsound	that's "git gc", sorry. :-)	22:04
	kassah wishes he could help, but he only started trying to figure out git a few days ago...	22:04
ambientsound	"git repack" runs fine.	22:05
	← palisv left	22:05
	← magn3ts left	22:05
ambientsound	However, when running "git repack" it counts 134 objects. On "git gc" it counds 4211 objects.	22:06
	← Zephyrus left	22:06
	← eletuchy left	22:07
	← priidu_ left	22:08
	← Fullmoon left	22:09
	→ priidu_ joined	22:09
	← xperience left	22:10
	→ erikvold joined	22:12
	← j416 left	22:15
	← priidu_ left	22:15
	← pantsman left	22:15
	← lavamind left	22:16
	← WebDragon left	22:16
	→ priidu_ joined	22:16
	→ ntoll joined	22:17
	← JonnyRo left	22:17
sherr	ambientsound: Have you looked at the "fixing a broken repos" at http://git.wiki.kernel.org/index.php/GitFaq#How_to_fix_a_broken_repository.3F	22:18
	← kpreid_ left	22:19
	→ bushwakko joined	22:20
	← bushwakko left	22:20
	← jmcantrell left	22:22
ambientsound	$ git fsck --full	22:22
	Killed	22:22
	I've tried gdb'ing it, but backtrace gives nothing	22:22
	← maxter left	22:22
spearce	rtyler: remember that array index out of bounds exception in gerrit? we think it might have to do with syntax highlighting, someone just found a crash that appears to be caused by it	22:23
	← ben_alman left	22:23
	← drizzd left	22:23
	← hakunin left	22:24
	← amerine left	22:24
	← eFfeM1 left	22:24
rtyler	spearce: woohoo!	22:25
	spearce: how did a JS error become a crash? :/	22:25
spearce	well, js error in the client is a crash of the client :-)	22:26
rtyler	spearce: not technically, unless your browser is piss-poor :)	22:27
	← dcampano left	22:27
	→ rolfb joined	22:27
sherr	ambientsound: tried latest Git?	22:28
ambientsound	sherr, no, still running 1.6.2.3.	22:28
	→ kpreid_ joined	22:29
	← mm_work left	22:29
sherr	ambientsound: could be a bug - could be fixed (hopeful). maybe ask on the mailing list? Bad news, whatever it is.	22:29
	→ magn3ts joined	22:29
	← git___ left	22:29
	← jasonn1234 left	22:29
ambientsound	well, working tree is in good condition on at least 2 sides, so I have all the data backed up.	22:29
	→ git___ joined	22:30
ambientsound	however I would like to keep the dev history :-)	22:30
sherr	ambientsound: sorry I can't help - I am new to it. But good luck.	22:30
ambientsound	thanks.	22:30
	I'll ask them to install 1.7.0.2 and see if that fixes anything	22:31
	→ jfarcand joined	22:31
	← khmarbaise left	22:31
	← aziz left	22:31
jfarcand	Anobody have seen: error: git-remote-curl died of signal 13	22:32
rtyler	spearce: I'm glad we got things with Hudson all set up, we're looking good: http://hudson.brokenco.de/job/Gerrit/15/	22:33
	← kpreid_ left	22:34
	→ en1gm4 joined	22:36
	← Ryback_ left	22:36
	← en1gm4 left	22:37
	→ frogonwheels joined	22:37
	webchick → webchick\|supper	22:37
	→ jasonn1234 joined	22:37
	← imajes left	22:38
	← dannas left	22:38
	← git___ left	22:39
	→ Rolenun joined	22:39
	→ supaman joined	22:40
	→ mark[oz] joined	22:40
	→ imajes joined	22:40
supaman	hey, I am trying out git and it seems to be working well (not that I know much about it, just fiddling) but when I issue git send-email I get "git: 'send-email' is not a git-command. See 'git --help'." but this is supposed to be a supported feature, any idea of why this does not work for me?	22:41
	← kar8nga left	22:41
DrNick	supaman: you probably haven't installed that package	22:42
supaman	DrNick: ok, thought it was coupled with the main program ... Ill check it out	22:42
	→ scarabx joined	22:42
	← ntoll left	22:43
	← m104 left	22:43
supaman	DrNick: your right, it is in a sepperate package	22:43
DrNick	yes, that's why I told you that	22:43
supaman	even without knowing my distribution ;-) (so it must be like that on all distros)	22:44
	→ ereslibre joined	22:45
	← ereslibre left	22:45
	→ ereslibre joined	22:45
	← g0bl1n left	22:45
	→ dcampano joined	22:46
	→ disturb joined	22:48
disturb	hi folks	22:48
	→ rsyring joined	22:49
	← ph^__ left	22:50
	← LiamH left	22:50
rsyring	I have a local repository and a remote repository on projectlocker. I should have cloned the project locker one down first and used that instead of creating my own, but didn't. How do I get a similiar state out of my current situation?	22:50
	→ EricInBNE joined	22:50
rsyring	I have pushed my master branch to the project locker repository, but that hasn't given me the remote refs like I thought it would.	22:51
bob2	are you asking "how do I make two independent repositories that have similar files belive they are related to one another" or "how do I configure git such that 'git push' and 'git pull' don't need extra args"	22:52
	← jfarcand left	22:52
frogonwheels	hmm.. do you think an hour is excessive for a git repack on win32 approx 1gb freshly exported repo ?	22:53
rsyring	bob2: I think I am asking: "how do I take the current local repository and push it to the remote repository in such a way that the result would be the same as if I had cloned the empty remote repository and then made my changes to that." Does that help?	22:53
	← laj left	22:53
	← jailbox left	22:54
rsyring	I probably have to do more than a push...but not sure what.	22:54
bob2	ah, the remote is empty?	22:54
rsyring	bob2: it was, until I just pushed my master branch to it.	22:55
	bob2: but I don't have the remote refs like I thought I would and am not sure how to create them.	22:55
	→ eletuchy joined	22:59
	→ killerchicken_ joined	23:00
	← ixti left	23:01
	← jelmer left	23:03
	← killerchicken left	23:03
davr	maybe you want the --mirror flag, check the man git-push page	23:05
Gitbot	davr: the git-push manpage can be found at http://git.or.cz/man/git-push	23:05
	dwmw2_gone → dwmw2_HIO	23:06
	→ flaguy48 joined	23:07
	← PointMan left	23:07
	← kumbayo_ left	23:07
	→ ntoll joined	23:08
	← Yuuhi left	23:09
	← eletuchy left	23:10
	← bcardarella left	23:11
	← WALoeIII left	23:11
	→ WALoeIII joined	23:11
	→ eletuchy joined	23:13
	→ jailbox joined	23:13
	→ GodSpeed9 joined	23:13
	→ bitcrave joined	23:14
	← loincloth left	23:15
	← supaman left	23:15
	→ kpreid_ joined	23:16
	← andres_ left	23:17
	← wwwald left	23:17
	thiago_home → thiago	23:17
bitcrave	Has anyone seen a branch in their git repo named "-" every?	23:18
bob2	bet you typoed a co -b	23:18
	← ntoll left	23:18
Textmode	night all	23:18
	→ novas0x2a1 joined	23:18
	← Textmode left	23:19
	← Theravadan left	23:20
	← novas0x2a left	23:22
	← bentob0x left	23:22
	← kpreid_ left	23:22
	→ kpreid_ joined	23:22
bitcrave	bob2: so im looking through the log history of commands in linux and im not seeing a typo... which doesn't really mean anything cause i don't know when this problem started.... is someone did make a mistake... how would one do a git checkout - ????	23:23
	→ andres_ joined	23:24
bob2	maybe -- -	23:24
	or fix it in .git	23:24
	← scarabx left	23:25
	maattd → maattd\|away	23:26
	← kipras left	23:27
	← kpreid_ left	23:27
foo	Let's say I'm working with git to manage a website. If I am working in the repository (which is set as the webroot), then how would I use "git push" ?	23:28
bob2	too vague	23:29
	what do you want 'git push' to do	23:29
	→ jelmer joined	23:30
foo	bob2: not sure, that's a good question. Can I have it push to another repository?	23:30
	← mssm left	23:30
bob2	yes, that's what push does	23:30
	← priidu_ left	23:30
foo	oh, then nevermind. :)	23:31
bob2	however, as we said the other day, pushing to non-bare repositories is a bad idea	23:31
	→ hgimenez joined	23:31
foo	bob2: fwiw, I think you were the one that told me, what's the disadvantage of setting my webroot (vhost) to be a repository? You just said it was a bad idea, but I don't think I got the reasoning behind that	23:31
bob2	I don't think I did say that	23:32
	I think I probably said that pushing to it directly was a bad idea	23:32
foo	aha, I see	23:32
bob2	so you need /another/ repository	23:32
	→ loincloth joined	23:32
	→ priidu_ joined	23:33
foo	hm, just checked logs, my mistake	23:33
	bob2: so, the idea behind what you're saying is to create another repository, and when things get pushed into that, I manually cp files over to the actual webroot?	23:35
bob2	no	23:35
	pretty sure I suggested a hook script that 'cd /webroot ; git pull'	23:35
	← eletuchy left	23:35
	← psoo left	23:35
	→ cloudhead joined	23:36
	→ ben_alman joined	23:36
	webchick\|supper → webchick	23:36
	← rook2pawn left	23:36
	← andres_ left	23:36
	← navap left	23:36
foo	bob2: hm, and that would pull from another repository. Yeah, but for it to pull from another repository, doesn't /webroot have to be a repository? Or can I do git pull anywhere? (as long as I specify which repository to pull from)	23:36
bob2	:(	23:37
	yes	23:37
	no	23:37
	I thought we just agreed that having the webroot be a repository was fine	23:38
	← przemoc left	23:38
	← dforsten left	23:38
foo	err, darn it, why is this taking me so long to grasp. right, we did	23:38
	foo rereads your responses	23:38
	→ przemoc joined	23:39
	← GodSpeed9 left	23:40
foo	bob2: ok, to avoid pushing to a non-bare repository, I should be fine if I: create directory, git init; copy over the files I want in there (eg. the webroot), then git add them, then I can rename the directory to be my webroot and change the other one to be webroot-old or something - yeah?	23:40
	→ rook2pawn joined	23:41
bob2	that makes a single non-bare git repository, but has nothing to do with all the other things you have to do	23:41
foo	bob2: oh. if I already have an existing webroot/website and I want to add it to git, how do I start if it can't be non-bare?	23:42
bob2	I'm lost	23:43
	so	23:43
	you need a non-bare repository in your webroot	23:43
	that's fine, and the above steps do that	23:43
	→ andres_ joined	23:43
	← rolfb left	23:46
	→ level09 joined	23:46
	← loincloth left	23:46
	foo rereads	23:46
	← TeckniX left	23:47
	→ anxiolytic joined	23:47
level09	I'm total new to GIT, what is the best workflow to use GIT for a team of developers	23:47
	← darwin left	23:47
	→ rctay joined	23:47
	foo says to self, "pushing to webroot directly is a bad idea".	23:48
foo	bob2: so, if I understand, pushing to a webroot directory is bad, it's better to pull, correct? Or no?	23:49
bob2	since it would need to be non-bare, yes	23:50
	← Mage_Dude left	23:51
	← tvw left	23:52
	← thiago left	23:53
frogonwheels	level09: I believe one of the strengths of git is that it supports a number of development styles and scenarios	23:53
	.. or better, 'any number of'	23:54
	← gebi left	23:55
level09	frogonwheels: great, I'm looking for a way to manage a website	23:55
	frogonwheels: what would be the best scenario for that ?	23:55
	I was thinking of having one server for repository, and let developers clone it and commit to it	23:56
	→ Wandrewvious joined	23:56
level09	does that make sense ?	23:56
	← Paraselene_ left	23:56
	→ Mage_Dude joined	23:56
frogonwheels	certainly, that's a fairly standard way, and the way I'm setting up ours.	23:57
	err.. not 'commit' to it.. push to it	23:57
	.. so you clone it, then pull updates from it..	23:57
	commit to your local clone, then push changes up to the server	23:57
	← Markus23 left	23:58
	← tbf left	23:58
	→ madewokherd joined	23:58
level09	frogonwheels: thanks, how is the changes pushed to the server ?	23:58
frogonwheels	git push	23:59
level09	so that is equivalent to commiting a change on the server?	23:59
	← Wandrewvious left	23:59
frogonwheels	level09: kinda.	23:59
	level09: better in most ways than commiting to a server	23:59
	← cannonball left	23:59
	→ Wandrewvious joined	23:59

Logs Search ←Prev date Next date→

Channels Documentation