IRCloggy #git 2010-06-30

Provider of IRC logs since 2005.

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

	2010-06-30
	→ ereslibre joined	00:00
	← ereslibre left	00:00
	→ ereslibre joined	00:00
	← d0k left	00:01
	← adamv left	00:01
	→ iruediger joined	00:02
	→ JEEB joined	00:02
	← Paraselene_ left	00:02
dgs	nm, think I just found out to do it on stackoverflow	00:03
	← nadavoid left	00:04
	→ adam12 joined	00:05
	← kenichi left	00:08
	← carrus85 left	00:10
	← dreiss left	00:11
	← sftp left	00:11
	→ gsan1 joined	00:12
	← fualo left	00:13
	← tatsuya left	00:13
	← dj_tjerk left	00:15
	→ brosner joined	00:17
	← brosner left	00:18
	→ brosner joined	00:18
	← brosner left	00:19
	→ warthog9 joined	00:19
	← ajpiano left	00:19
	→ fualo joined	00:20
	→ jrmuizel joined	00:21
Tricia	I have a config file with a password on it. I don't want to commit the file with the password in it.	00:22
	I want to commit it with the password place blank	00:22
	is there a way to commit it with it blank and then tell it to ignore future changes without removing it from the repo?	00:22
coppro	no	00:23
	you can always not add it to the index	00:24
	but you can't tell git to ignore a tracked file	00:24
Tricia	http://www.gitready.com/intermediate/2009/02/18/temporarily-ignoring-files.html seems to disagree	00:24
	→ brosner joined	00:25
	← brosner left	00:25
	→ warlock_mza joined	00:25
	→ brosner joined	00:25
	← hyperair left	00:25
	← patmaddox left	00:28
	← warlock_mza left	00:29
	→ parasti joined	00:30
wereHamster	Tricia: yes, it's kindof possible.	00:30
	← iruediger left	00:30
wereHamster	but instaed of using assume-unchanged you should put a template config file into the repo and then copy it and put the password file into the copy	00:30
	← mattwynne left	00:32
	← OFICINA7 left	00:34
jmspeex	I'm trying to figure out how git-submodule works. What I'd like to understand is how submodules themselves can be distributed	00:34
	My understanding is that the superproject tracks "I want to use version XYZ" of the submodule in this repo.	00:35
	So what happens if XYZ is developed in a distributed way with no central repository?	00:36
	← RobertLaptop left	00:37
	← JEEB left	00:38
Tricia	wereHamster: is there something wrong with the assume-unchanged thing?	00:38
	→ Remixman joined	00:40
	→ warlock_mza joined	00:42
	→ johnw joined	00:42
	← Chillance left	00:43
wereHamster	yes. Everybody using your repo needs to know aoubt how assume-unchanged works (and that it exists). If you supply a template then everybody can just copy it and add their own stuff without having to do anything special with their local repo	00:47
	→ OFICINA7 joined	00:48
	→ briahn10 joined	00:49
	← briahn10 left	00:49
	← hal left	00:49
	→ JEEB joined	00:51
	→ stringo0 joined	00:55
	← hobodave left	00:55
	→ patmaddox joined	00:58
	← _Vi left	00:59
	→ _Vi joined	00:59
	← engla\|esp left	01:00
	→ anathematic joined	01:01
	→ sh1mmer joined	01:03
	→ alezandro joined	01:04
	← ereslibre left	01:09
	→ scarabx joined	01:09
	→ dreiss joined	01:09
	→ Smirnov_ joined	01:11
Smirnov_	how do i make a .gitignore entry that is the regex equivalent of /^examples/[^.]+$/ aka ignore all files without a dot	01:11
	← hakunin left	01:12
	← gbacon left	01:13
	→ magicbronson joined	01:15
Smirnov_	ah got it, examples/* followed by !examples/.	01:15
	if only powerusers use git though, why not just use regex for .gitignore syntax? so weird.	01:15
	→ ereslibre joined	01:17
	← ereslibre left	01:17
	→ ereslibre joined	01:17
	← Vampire0_ left	01:18
	→ yashi joined	01:19
	← spearce left	01:22
	→ RobertLaptop joined	01:24
	← sh1mmer left	01:26
	→ treak007 joined	01:26
	← spaceonline left	01:27
	→ VVSiz_ joined	01:27
	→ hagabaka joined	01:29
	← hagabaka left	01:29
	→ hagabaka joined	01:29
	← enherit left	01:30
	← gsan1 left	01:31
	← _VVSiz_ left	01:31
	← ereslibre left	01:32
	← brett_h left	01:32
	← hasrb left	01:35
	← patmaddox left	01:38
	→ hobodave joined	01:39
	← hobodave left	01:39
	→ hobodave joined	01:39
	→ agnathan joined	01:39
	← AAA_awright left	01:42
	← zenazn left	01:42
	→ nadavoid joined	01:43
	← adam12 left	01:44
	← jensn left	01:45
	← jrmuizel left	01:54
	← dreiss left	01:54
	← hagabaka left	01:54
	← parasti left	01:58
	← pfc left	01:58
	→ homiziado joined	01:58
	← stringoO left	02:01
	→ hyperair joined	02:02
	→ consolers joined	02:02
	← tildeequals left	02:03
	← yashi left	02:03
	→ adymitruk joined	02:04
	← bcardarella left	02:08
	→ [1]intripoon joined	02:08
	→ bcardarella joined	02:08
	← Jochen_Hayek left	02:10
	← intripoon left	02:11
	[1]intripoon → intripoon	02:11
	→ zenazn joined	02:11
	thunk → thunk\|smoking	02:11
	← tsundsted left	02:11
	← hyperair left	02:13
	← rettub left	02:13
	← coppro left	02:14
	→ coppro joined	02:15
	→ rettub joined	02:15
	← treak007 left	02:16
	→ hyperair joined	02:17
	← flaguy48 left	02:18
	← davr left	02:18
	→ AAA_awright joined	02:18
	← warlock_mza left	02:19
	→ davr joined	02:19
	→ treak007 joined	02:19
	→ AngryParsley joined	02:20
	→ cytrinox_ joined	02:20
OFICINA7	where's the gitignore file again plz?	02:20
AngryParsley	I have a branch that I want to merge with master, but I only want to merge changes to one directory in that branch. is there an easy way to do that?	02:21
	← zenazn left	02:21
	→ nsussman joined	02:21
AngryParsley	that is, I made a branch of master, made changes in /foo, then made changes in /bar. now I want to merge changes to /foo with master	02:22
	but not changes to /bar	02:22
Hunner	OFICINA7: `man gitignore`	02:22
Gitbot	OFICINA7: the gitignore manpage can be found at http://git.or.cz/man/gitignore	02:22
Hunner	AngryParsley: you can cherry pick the commits that change files in /foo, but there is not an explicit way to specify that. Git works with changesets, not with files, so it doesn't make that kind of distinction	02:23
	← cytrinox left	02:23
	cytrinox_ → cytrinox	02:23
AngryParsley	but some of my commits include changes to both /foo and /bar	02:23
OFICINA7	weird, I think it didn't create the gitignore file	02:24
Hunner	This would be a case that having 2 branches (or a branch of a branch) would have been more helpful. (one for /foo, one for /bar). You could play the rebase -i game and split those commits into two separate commits	02:25
	← warthog9 left	02:25
AngryParsley	yeah I didn't realize I wanted only changes to /foo until it was too late	02:25
Hunner	OFICINA7: It doesn't create .gitignore files by default. You can just make a file called that and it's smart enough to obey it	02:25
	AngryParsley: yeah, common problem :/	02:25
AngryParsley	so, clone again then copy stuff over manually	02:26
OFICINA7	:) thx	02:26
	← jmcantrell left	02:27
	→ jmcantrell joined	02:27
Hunner	You could branch, cherry pick your /foo and /foo+/bar commits, revert the /bar changes left over by hand, and merge that	02:27
	I'm not a git expert though... there may be a better way	02:27
AngryParsley	I just cloned master again and am copying the files over manually. hopefully nobody else changes them until I'm done with this	02:28
Hunner	okay. Sometimes that's just the easiest way :)	02:28
	← homiziado left	02:30
	→ ige joined	02:31
adymitruk	You may want investigate filter-branch	02:31
	→ Adaptee joined	02:36
	← AngryParsley left	02:36
	← hyperair left	02:40
	→ JohnFlux joined	02:41
	← iizukaway left	02:42
	← alezandro left	02:42
	→ tstclair joined	02:42
	→ iizukaway joined	02:42
	→ tildeequals joined	02:42
	← fod left	02:46
	← d1zzy left	02:47
	→ d1zzy joined	02:47
	→ mklappstuhl joined	02:50
mklappstuhl	hey	02:50
	← valros left	02:50
mklappstuhl	i want to get the last commit for a submodule i use	02:50
	do i really need to cd into the submodule's dir and do "git pull" ?	02:51
	→ timj_ joined	02:52
	thunk\|smoking → thunk	02:52
	← pmcnabb left	02:53
parndt	git submodules:update or something isn't it	02:53
	← aresnick left	02:54
	← timj__ left	02:54
mklappstuhl	?	02:54
parndt	git submodule --help	02:54
mklappstuhl	i read the manpages	02:54
	i understand the default behaviour "git submodule update" …	02:54
	← Elfe left	02:54
parndt	git submodule update	02:54
	oh okay	02:54
	doesn't that do what you want?	02:54
mklappstuhl	parndt: not really, in my case it should just pull ...	02:55
	thunk → thunk\|smoking-ag	02:55
	← mw left	02:55
	thunk\|smoking-ag → thunk\|smkn-agn	02:55
	← intripoon left	02:58
	← treak007 left	02:58
	→ alezandro joined	02:59
frogonwheels	mklappstuhl: submodules are always tied to a commit. changing which commit it points involves (yes) going into the directory and (pulling) / checking out the required commit	03:02
mklappstuhl	frogonwheels: okay thanks, didnt know that before	03:02
frogonwheels	mklappstuhl: then from the main dir, doing the git add mysubmodule/ (don't forget the /)	03:03
	.. and checking it in.	03:03
	← Remixman left	03:03
mklappstuhl	i'm mostly doing git submodule add git://… vendor/…	03:03
	what i find a bit confusing is that i need to enter the actual dir the submodule should be cloned into	03:04
frogonwheels	mklappstuhl: yep - and that will add the latest commit of the default branch at that point in time to your repo	03:04
mklappstuhl	from my perspective this should be optional, just like git clone	03:04
frogonwheels	mklappstuhl: you can use --git-dir="" or whatever :) but if you want to modify the submodule somehow, you treat it as a normal git project	03:06
mklappstuhl	frogonwheels: whats --git-dir	03:07
	thunk\|smkn-agn → thunk	03:07
	→ Elfe joined	03:08
frogonwheels	mklappstuhl: you probably don't need to know about it - it's more used for plugins &C.	03:08
mklappstuhl	frogonwheels: where can i read more about it	03:09
frogonwheels	google?	03:09
consolers	8 Peters in /who #git. is that exhaustive I wonder	03:09
mklappstuhl	http://lmgtfy.com/?q=--git-dir	03:09
	:P	03:10
frogonwheels	yeah :)	03:10
mklappstuhl	tried google before	03:10
	← kukks left	03:10
frogonwheels	mklappstuhl: why are you so against cd ?	03:10
mklappstuhl	frogonwheels: I'm not against it, it's totally okay – I just didnt know --git-dir :P	03:11
	and in some way the default behaviour couldnt be better	03:12
	→ lpolzer__ joined	03:12
mklappstuhl	frogonwheels: i just thought that there might be a switch like for using rebase always when pulling	03:13
frogonwheels	mklappstuhl: you always want to be at the 'head' of a submodule?	03:13
	← justin_george left	03:13
mklappstuhl	head of the remote - yup	03:14
	→ sh1mmer joined	03:14
consolers	how do git-added-but-not-committed objects escape gc --prune , even when they are not referenced in index or logs ??	03:14
frogonwheels	mklappstuhl: you could make an alias to git submodule foreach git pull	03:14
	→ Vampire0 joined	03:15
	← Dawgmatix left	03:15
mklappstuhl	yeah I could, but i think i will just use it as you pointed it out	03:16
	← lpolzer_ left	03:16
mklappstuhl	time for bed now (5am), thanks for the support frogonwheels	03:17
consolers	afaict nothing references the object....	03:18
	← broquaint left	03:20
	→ broquaint joined	03:21
	← anjen left	03:23
	← berserker left	03:28
	→ hyperair joined	03:28
	→ Aaaarg_ joined	03:28
	← tstclair left	03:29
	→ berserker joined	03:30
	← Aaaarg left	03:32
	Aaaarg_ → Aaaarg	03:32
	→ mklappst1hl joined	03:32
	← jmcantrell left	03:34
	← mklappstuhl left	03:34
	→ aresnick joined	03:36
	→ hakunin joined	03:36
	← mklappst1hl left	03:37
	← hakunin left	03:40
	→ hakunin joined	03:41
	← madewokherd left	03:41
	← stringo0 left	03:43
	← hyperair left	03:46
	→ MattDiPasquale joined	03:46
	→ hyperair joined	03:47
	← magicbronson left	03:49
	→ madewokherd joined	03:55
	← RobertLaptop left	04:01
	→ justin-george joined	04:03
	→ patmaddox joined	04:03
	→ dreiss joined	04:08
	jjcm → jokeruiner	04:09
	← madewokherd left	04:11
	← patmaddox left	04:14
	→ patmaddox joined	04:14
	← ede left	04:15
	← dreiss left	04:16
	→ patmaddo_ joined	04:17
	← masa- left	04:17
	← aresnick left	04:18
	→ masa- joined	04:18
	← patmaddox left	04:18
	→ dreiss joined	04:19
	→ stringo0 joined	04:20
	→ Textmode joined	04:20
	→ mtaylor joined	04:20
mtaylor	hey guys - I'm getting a crash while running git fast-import	04:20
	← justin-george left	04:20
	→ rchavik joined	04:21
mtaylor	it seems to be related to a file rename in the original bzr branch I'm importing from ... is there a sensible place to report it?	04:21
	← rchavik left	04:21
	→ rchavik joined	04:21
	→ justin-george joined	04:23
	← patmaddo_ left	04:24
	← MattDiPasquale left	04:24
	→ patmaddox joined	04:24
	→ simplechat joined	04:26
	← patmaddox left	04:28
	→ rayners joined	04:28
Ilari	mtaylor: Mailing list?	04:29
	← marc_kdab left	04:29
mtaylor	Ilari: ok. thanks	04:31
	→ SomeoneE1se joined	04:31
	jokeruiner → jjcm	04:33
SomeoneE1se	I've setup a remote git server (access via ssh) and I want to get an email everytime someone pushes to it, what hook do I want to use?	04:33
wereHamster	consolers: make sure the objects are not referenced by anything (remove them from the index etc), then run git gc	04:33
	SomeoneE1se: post-receive or post-update	04:34
	← Eaven left	04:34
SomeoneE1se	I read about post-receive-email, but I can't find a copy. it didn't come with my bare init	04:35
	← mtaylor left	04:35
	thunk → thunk\|smoking	04:35
wereHamster	SomeoneE1se: <HEAD:contrib/hooks/post-receive-email>	04:36
Gitbot	[git post-receive-email]: http://tinyurl.com/yjnbnvx [blob]	04:36
	← anathematic left	04:36
	← hyperair left	04:36
SomeoneE1se	wereHamster: thanks	04:36
	→ hyperair joined	04:37
	← niki left	04:38
	→ mercy_____ joined	04:38
mercy_____	hey guys, sorry if im overlooking something obvious in the man page. can i do a git reset --hard on just a single file, to revert it back to what it was and trash all recent changes	04:39
	← coppro left	04:39
	→ coppro joined	04:40
	← comawhite left	04:40
	← coppro left	04:40
	← Joeconyers left	04:41
rchavik	mercy_____: git checkout <file>	04:41
mercy_____	>.<	04:41
	thanks	04:41
	→ coppro joined	04:42
	→ comawhite joined	04:43
	→ pallaert joined	04:45
	← AAA_awright left	04:48
	→ AAA_awright joined	04:48
	thunk\|smoking → thunk	04:48
Ilari	SomeoneE1se: If you want to log all changes, one way would be to use post-receive and put everything received from stdin into E-Mail message and send it (together with identification of user and timestamp).	04:49
consolers	wereHamster: after rm -rf index logs gitk.cash and other HEAD s	04:49
	→ hasrb joined	04:49
consolers	gitk.cash is interesting typo	04:49
wereHamster	consolers: and which command are you running that doesn't get rid of those files? plain 'git gc'?	04:50
	← berserker left	04:50
consolers	gc --prune	04:50
	→ radioman-lt joined	04:50
consolers	the objects remain dangling ...	04:50
wereHamster	how about gc --prune=now ?	04:50
consolers	havent looked at code .	04:50
	ah thx	04:50
wereHamster	consolers: as always, man git-gc and read the description of --prune	04:51
Gitbot	consolers: the git-gc manpage can be found at http://git.or.cz/man/git-gc	04:51
consolers	i've gone through this rigmarole before i think, sorry	04:51
	→ maleknet joined	04:51
	→ berserker joined	04:51
	← maleknet left	04:52
consolers	the file timestamp is used?	04:52
	→ maleknet joined	04:52
wereHamster	yes, there' nothing else than that	04:52
consolers	if so, how incongruous!	04:52
wereHamster	eh, why?	04:53
	← mercy_____ left	04:53
consolers	considering the hatred for git towards using filetimestamps inside git	04:53
	← pallaert left	04:53
consolers	hatred of git ...	04:54
wereHamster	is there? I haven't noticed any	04:54
consolers	werehamster, git destroys all filestamp data	04:54
	the information is destroyed. there is no way i can get "when was this content last modified" information from git	04:54
DrNick	sure you can	04:55
wereHamster	git log -1 -- <file>	04:55
consolers	that gives incorrect dates	04:55
	if i merge it'll give the date of the commit	04:55
	← nadavoid left	04:55
wereHamster	--no-merges ?	04:56
	← ignatp left	04:56
consolers	relative timestamp ordering of files is lost in every commit	04:57
	a file may not have changed since 1993	04:57
	but if i get a tarball from a project using git, it'll show a timestamp of yesterday	04:57
wereHamster	why are timestamps important for the actual checkout?	04:57
consolers	right that is the hatred i'm talking about	04:58
	what you cannot understand, destroy!	04:58
	→ ArdRigh joined	04:58
wereHamster	no, it's rather: what doesn't make sense, don't keep track of it	04:59
consolers	it wont make sense if you wont let anyone keep track of it	04:59
wereHamster	and you still haven't answered my question, why the timestamps are important	05:00
	← hyperair left	05:00
consolers	i want to know when the file content was last modified	05:00
	not when some idiot decided to use git to track it	05:00
wereHamster	if you have the git history available, you can look it up there	05:01
consolers	no, its lost because the idiot chose git, and that data gets rewritten	05:01
	← frakturfreak left	05:02
wereHamster	ah, so the files were in another SCM before and you want to get that date.. ?	05:02
consolers	and i get a meaningless date, because it is meaningless	05:02
	→ frakturfreak joined	05:02
DrNick	the commit date isn't meaningless	05:02
consolers	like i said i'm dealing with hatred here, that can only be compared to genociding dictators	05:02
	→ girishr joined	05:02
	← dgs left	05:03
	→ RobertLaptop joined	05:03
	→ freimatz joined	05:03
	→ nadavoid joined	05:03
consolers	obviously if you destroy all evidence by design, there will be no evidence	05:03
DrNick	man, you have issues	05:04
	→ pedrobel_ joined	05:04
SomeoneE1se	when post-receive is run, what box runs it, the box making the push, or the server that's getting the push?	05:04
	← pedrobel_ left	05:04
PerlJam	SomeoneE1se: the machine doing the receiving :)	05:04
SomeoneE1se	thanks PerlJam	05:04
PerlJam	SomeoneE1se: (i.e. the server)	05:04
consolers	yes indeed, information is lost by git!	05:04
	so thats why i find it incogrous that git uses the mtime in this case	05:05
wereHamster	sure, some informatino is destroyed if you import source from a SCM that doesn't have atomic commits. But other than that, the information is not lost, merely not presented to you in a way you want	05:07
	btw, there is a script which you can run on your working tree and it sets the mtimes according to the commits which touched the files last	05:07
	← tedoc2000 left	05:07
consolers	yeah i use wong's script	05:08
	but commit times change even when content doesnt change	05:08
	for example during rebase	05:09
wereHamster	the commit date changes, but not the author date	05:09
consolers	and you do a :wq of a commit youre editing, nothing changes except the commit time	05:09
	author date is meaningless when content changes	05:09
wereHamster	eh?	05:09
consolers	the author date does not change during rebase for example	05:10
	but the rebase might involve an edit	05:10
	so the file contents have changed, retaining an author date from 2006	05:10
	it is meaningless to set the file time to 2006	05:10
wereHamster	uhm. true. But how should rebase figure out whether you actually changed the contents or not?	05:11
consolers	at the same time, commit dates can change without changing content	05:11
	again, see my original requirement	05:11
	i want to know when the content changed last	05:11
	ACTUALLY!	05:11
wereHamster	I know that. But you'd ahve to supply that information to rebase, as it can't figure that out on its own	05:12
consolers	what?	05:12
	it cant figure out content is not changing?	05:12
	all design decisions support the HATRED	05:13
wereHamster	if you rebase a commit, and the commit applies with a fuzz, is that considered a file change or not?	05:13
consolers	why dont you come out and admit it	05:13
PerlJam	consolers: okay. I admit it ... you're crazy :)	05:13
consolers	dude, does the sha1 change or not?	05:13
	if the sha1 does not change, give me the date when this object came first into this repo	05:14
wereHamster	git doesn't store per-file dates, as commits are atomic	05:14
consolers	the only reason this flies is because of the idiot audience that git bets on to adopt and use it	05:14
SomeoneE1se	thanks for the help guys	05:14
	← SomeoneE1se left	05:15
consolers	they dont care because they are not discerning/clueless	05:15
PerlJam	consolers: so ... go find a scm that does care and use that.	05:15
consolers	engineering them as a userbase in an outwardly democratic setting promotes the DICTATOR GENOCIDE	05:15
	i cite perljam as an example	05:15
	← hasrb left	05:16
	→ harinath joined	05:16
	→ oficina7_ joined	05:16
	→ hyperair joined	05:17
	← OFICINA7 left	05:17
	→ LongBeach joined	05:18
consolers	as with rebase, so with merge...	05:18
	ok	05:19
	← friskd left	05:19
	→ fmardini joined	05:19
	oficina7_ → OFICINA7	05:21
	← bcardarella left	05:22
consolers	wereHamster wong's script has to be patched do git-log -- file and check shas all the way back	05:22
	to do*	05:22
	← hyperair left	05:22
	→ marc_kdab joined	05:23
	← doener left	05:23
	← Yuffster left	05:23
	← bartek left	05:23
consolers	but even then the spurious rewriting of commit dates, which bad feature the adopting adopters can not help just because they use it, makes it meaningless	05:24
	→ bartek joined	05:24
	→ doener joined	05:25
	→ bambule joined	05:25
	blueshift wakes up to a rant	05:26
	← dr_win left	05:28
	← girishr left	05:29
	→ girishr joined	05:30
	→ vu3rdd joined	05:31
	← nadavoid left	05:32
	← rchavik left	05:33
	→ rchavik joined	05:33
	← rchavik left	05:35
	→ patmaddox joined	05:35
	→ jony joined	05:37
	← scarabx left	05:38
	→ thiago_home joined	05:39
	← girishr left	05:42
	→ Belna joined	05:42
	← jailbox left	05:46
	← agnathan left	05:47
	→ florianb joined	05:47
	← frakturfreak left	05:48
	→ Bennid joined	05:50
	→ frakturfreak joined	05:50
	← johnw left	05:54
	← jfkw left	05:58
	← codejunky left	05:59
	→ codejunky joined	06:00
	→ roop joined	06:01
	← simplechat left	06:02
	→ daaku joined	06:02
	→ rseifert joined	06:03
	→ ajpiano joined	06:06
	← thiago_home left	06:09
	← ajpiano left	06:11
	→ kar8nga joined	06:13
	→ fivetwentysix joined	06:13
	→ priidu joined	06:13
	→ ajpiano joined	06:15
	→ cbreak_work joined	06:16
	→ ustunozgur joined	06:17
	← bambule left	06:18
	← patmaddox left	06:18
	→ eighty4 joined	06:19
	← frogonwheels left	06:22
	→ frogonwheels joined	06:23
	← nsussman left	06:25
	← Mage left	06:25
	→ whalesalad joined	06:26
	→ gmarselis__ joined	06:27
whalesalad	hey guys is there a way to look at a specific file through time? I want to get some old code out of a file, but I have no idea when It was in the file	06:27
ustunozgur	whalesalad: git log -p filename ?	06:28
	→ psankar joined	06:28
whalesalad	ustunozgur: thanks a lot	06:29
	→ codeshepherd joined	06:29
whalesalad	crap... what if the file was renamed/moved at one point in time? trying that command with the old path doesn't work =/	06:30
	← project2501a left	06:30
	← harinath left	06:31
consolers	log --follow	06:31
nevyn	whalesalad: do you know what the old code looked like?	06:31
whalesalad	nevyn: not really =/	06:31
	it's been a few months, I need to roll back some real old code, like 20 or so lines. Not a lot, I could re-write it, but it was working great before and I figure, what's the point of version control if I can't do somethign like this	06:32
nevyn	cause you could always grep -r the .git directory to find the commitish	06:32
	so you changed some lines?	06:32
	→ ExtraSpice joined	06:32
nevyn	you know where in the file it used to be?	06:32
whalesalad	yep	06:32
nevyn	git blame file will show the commit where you changed the lines	06:32
	→ thiago_home joined	06:33
nevyn	if you then git checkout that commitish and do git blame again it should show you back from that.. etc	06:33
whalesalad	ahhh right on	06:33
	luckily, github saved the day	06:33
	just as you said that, I managed to find what I was looking for through the web interface	06:33
	the commit was on april 20th, hah. quite a while ago	06:34
	← waveclaw left	06:35
	thiago_home → thiago_office	06:35
	← thiago_office left	06:35
whalesalad	btw... just throwing this out there, how do you guys do mail testing typically? without actually sending the mail somewhere	06:36
	→ thiago_office joined	06:36
	→ rox` joined	06:36
	→ codeshepherd_ joined	06:37
	← thiago_office left	06:37
	→ thiago_office joined	06:37
	← daaku left	06:38
	← codeshepherd left	06:39
	codeshepherd_ → codeshepherd	06:39
	→ hegge joined	06:39
	parndt → parndt_away	06:40
	← disappearedng left	06:44
	→ girishr joined	06:44
	← ArdRigh left	06:45
	← codeshepherd left	06:46
	→ harinath joined	06:46
	→ nataraj joined	06:46
	→ villemv joined	06:47
	→ p3rg2z joined	06:50
	→ codeshepherd joined	06:51
	← Aaaarg left	06:52
	← troyt left	06:53
	→ troyt joined	06:53
	dwmw2_gone → dwmw2	06:54
	→ disappearedng joined	06:56
	→ apsysoev joined	06:57
	← berserker left	06:57
	→ skoop joined	06:58
	→ Aaaarg joined	06:58
	→ tatsuya joined	06:59
	→ berserker joined	06:59
	→ hyperair joined	06:59
	← codeshepherd left	06:59
	← ajpiano left	07:00
	← p3rg2z left	07:00
	← petercoulton left	07:00
	← MetaCosm left	07:00
	→ MetaCosm joined	07:01
	→ pantsman joined	07:03
	← tatsuya left	07:03
	← hyperair left	07:04
	→ codeshepherd joined	07:04
	→ ChUrP joined	07:05
	→ hasrb joined	07:05
	← disappearedng left	07:05
	→ disappearedng joined	07:06
	→ drizzd joined	07:07
	← stringo0 left	07:08
	← whalesalad left	07:09
	← saiyr9000 left	07:10
	← apsysoev left	07:11
	→ cilly joined	07:14
	← girishr left	07:15
	→ daaku joined	07:18
	← neunon left	07:19
	→ neunon joined	07:19
	← troyt left	07:19
	→ steffkes joined	07:19
	→ troyt joined	07:19
	← daaku left	07:20
	→ royalty joined	07:20
	→ lhz joined	07:21
	← dreiss left	07:21
	← harinath left	07:21
	← DrNick left	07:23
	→ masterkorp joined	07:26
	← AAA_awright left	07:26
	→ AAA_awright joined	07:26
	← ChUrP left	07:27
	→ nukem joined	07:27
	← florianb left	07:27
	→ girishr joined	07:28
	→ nevyn^ joined	07:28
nukem	im using git and a script I wrote accidentally deleted a file via rm file	07:28
	the file was checked in and commited	07:28
	how can i restore it?	07:28
	i've tried rebase	07:28
thiago_office	git checkout -- filename	07:28
	git status will tell you what to do, btw	07:29
	← girishr left	07:29
	→ stuartrexking joined	07:29
nukem	thiago_office: thanks, i tried git checkout but didn't add the file	07:29
	also git status doesn't tell me how to restore the file, just how to commit the change	07:30
thiago_office	when you say "the file was checked in and committed"	07:30
	do you mean that the deletion was checked in and committed?	07:30
nukem	no	07:30
thiago_office	it should then say how to "unstage" the change	07:31
nukem	git checkout file worked though	07:31
	→ giskard joined	07:31
	→ daaku joined	07:31
nukem	it just says git reset HEAD <file>	07:31
	which didn't work	07:32
thiago_office	here it tells me:	07:32
	# (use "git checkout -- <file>..." to discard changes in working directory)	07:32
nukem	if i do rm file	07:33
	then git status it says	07:33
	use git add/rm <file>..." to update what will be committed	07:33
	→ airborn joined	07:34
nukem	im using git 1.5.5.6 from CentOS 5.5	07:34
	← hakunin left	07:34
	→ DrNick joined	07:35
	← troyt left	07:35
	→ girishr joined	07:35
	→ troyt joined	07:35
	→ Jochen_Hayek joined	07:36
	← berserker left	07:36
	→ mastro joined	07:37
	→ berserker joined	07:39
	→ sgh joined	07:39
	→ bentob0x joined	07:41
	→ simplechat joined	07:41
	→ barismetin joined	07:41
	→ kent\n joined	07:42
	← krawek left	07:43
	→ root joined	07:43
	root → j0rd	07:44
	← kar8nga left	07:46
	→ trivol joined	07:48
	→ shruggar joined	07:48
	→ saiyr9000 joined	07:48
	← j0rd left	07:49
	→ tvw joined	07:50
eMBee	good afternoon	07:50
	← OFICINA7 left	07:50
	← sgh left	07:51
	eMBee is puzzled by git branch -r listing a branch that he can not find referenced anywhere in .git/refs/	07:51
wanders	git pack-refs	07:51
	← rayners left	07:51
thiago_office	nukem: you should upgrade	07:51
eMBee	duh, just thoyught of that. for some reason i expected packed-refs to be inside refs/	07:52
nukem	thiago_office: my main system has 1.7.0.1 but for this project I have to use CentOS 5.5	07:52
	→ ph^ joined	07:52
thiago_office	nukem: and? :-)	07:52
	nukem: why can't you upgrade?	07:52
nukem	im working with other people and will probably have to support them	07:53
	and a couple of them have never run linux any version control or command line	07:53
	→ psoo joined	07:53
nukem	tell them to dl and compile git might make their head explode	07:53
thiago_office	give them a pre-compiled package then	07:53
	→ sgh joined	07:53
	← mord left	07:53
nukem	after i get some work done i might create a package though	07:54
	← justin-george left	07:55
Rhonda	Eeks. eMBee! Ages! Not seen!	07:55
eMBee	hi Rhonda!	07:55
	→ mabrand joined	07:56
	→ dpino joined	07:56
ribasushi	when rebasing, is there a way to display the rebase-plan?	07:58
eMBee	nukem: i found 1.7.0.5 rpms that work on centos 5.4	07:58
nukem	eMBee: where?	07:59
ribasushi	(the thing where I changed 'pick' to other things)	07:59
nukem	I'm using the epel repro to get it	07:59
	← sondreb left	07:59
ribasushi	s/rebasing/interactive rebasing/	08:00
eMBee	some private repo, hold on	08:00
	← girishr left	08:00
	kent\n assumes the average #git person probably knows more tricks about github than the average #github person, so I wonder if somebody in here knows how to manually associate 2 github projects so they show as forks	08:01
	← limohua left	08:01
eMBee	nukem: http://repo.webtatic.com/yum/centos/	08:01
nukem	eMBee: cool thanks!	08:02
eMBee	works with yum	08:02
	→ radarek joined	08:02
	→ tatsuya joined	08:02
	→ ntoll joined	08:03
eMBee	now i need to find 1.7 rpms for fedora 9...	08:04
muep	better to find yourself a newer Fedora... 9 is ancient and dead	08:05
nukem	i dont see why redhat doesn't have the latest git in	08:05
	they dont even have it in their main repos	08:05
muep	RHEL 5 does not include git at all	08:05
nukem	which doesn't make sense	08:06
muep	git wasn't that popular in 2007	08:06
eMBee	muep; not a trivial task i am on an olpc XO, they are not trivial to update	08:06
nukem	RHEL 5.5 just came out	08:06
	you'd think they could add it	08:06
	eMBee nods	08:06
muep	maybe git still moves so fast that they don't want to add it there	08:07
	← daaku left	08:07
muep	they'd have to support it for many years	08:07
	without breaking backwards compatibility	08:07
	preferably	08:08
	→ MrSpooge joined	08:08
	← masterkorp left	08:08
	← MrSpooge left	08:08
	← iizukaway left	08:08
muep	it's still quite easy to just compile and install or even package git for RHEL5 & clones	08:09
	→ iizukaway joined	08:09
	← thiago_office left	08:11
	← eli left	08:11
	→ mattwynne joined	08:11
	→ sondreb joined	08:12
	→ jamesmacleod joined	08:14
	← tatsuya left	08:14
	← consolers left	08:15
	→ girishr joined	08:15
	← hasrb left	08:16
nukem	so why is it when i check something in and push it the server realizes it in the log but doesn't show any changed file	08:17
ciaranm	nukem: are you trying to faq non-bare ?	08:17
Gitbot	nukem: Pushing to non-bare repositories is discouraged. See https://git.wiki.kernel.org/index.php/GitFaq#non-bare	08:17
	→ poseid joined	08:18
nukem	so whats the suggested way of doing it then?	08:19
	← yann left	08:21
nukem	ahh i see nm	08:21
shruggar	nukem: there are two solutions: 1) make a bare repository 2) don't push to a checked-out ref. in general: realize that "push" does not effect work-trees in any way	08:21
	→ Titosemi joined	08:22
nukem	so if i want to clone from upstream and commit to that just do git clone --bare git://...	08:22
	← disappearedng left	08:25
	→ limohua joined	08:25
	← freshtonic left	08:25
	← pantsman left	08:26
	← ige left	08:26
Ilari	nukem: You "can't" commit to bare repo. Bare repos are usually updated by pushing.	08:27
	← Heimidal left	08:27
nukem	well the bare repro would be on the server i'd check out from that on my client and push back to the server	08:27
	← bauruine left	08:28
Ilari	nukem: Also possible would be just creating empty bare repo and populating it by pushing.	08:29
	→ harinath joined	08:30
	← berserker left	08:31
	→ eli joined	08:31
Ilari	nukem: If client has slow upstream, then doing initial clone is better (as client doesn't have to upload a lot of data).	08:31
	→ bousquet joined	08:32
nukem	but does it really matter?	08:32
	→ ralf_e joined	08:32
Ilari	nukem: No...	08:32
	→ berserker joined	08:33
	eMBee can't find git 1.7 for f9 but found 1.6.5 directly from kernel.org, that will have to do. better than 1.5.x anyways	08:34
frogonwheels	eMBee: can't you just compile it yourself?	08:34
	← jamesmacleod left	08:35
	← poseid left	08:36
	→ jetienne_ joined	08:37
	→ ralf_e_ joined	08:37
muep	eMBee: using Fedora 9 a year after end of support is not recommended. Installing random maybe-maintained third-party packages may also lead into problems.	08:37
	→ disappearedng joined	08:38
	← niko left	08:38
ribasushi	so anyone? how to see the current interactive rebase-plan when half way through the rebase?	08:40
	← ralf_e left	08:41
	ralf_e_ → ralf_e	08:41
	← Mage_Dude_ left	08:41
	→ robin joined	08:41
	robin → Guest68680	08:42
	→ xiong joined	08:42
	→ niko joined	08:45
	← neunon left	08:46
	← royalty left	08:47
	→ neunon joined	08:47
	→ Mage_Dude_ joined	08:47
	→ tcurdt joined	08:48
	← Guest68680 left	08:52
	← coppro left	08:54
Ilari	ribasushi: Maybe its in some file in .git/rebase-apply (or something)?	08:54
	→ Guest68680 joined	08:55
	→ engla\|esp joined	08:57
eMBee	hmmm, what is "fatal: Needed a single revision" trying to tell me when i am doing 'git rebase --onto somebranch'?	08:59
	it also says invalid upstream	08:59
	→ masterkorp joined	09:00
	← airborn left	09:01
	→ EugeneKay joined	09:02
thiago	eMBee: you need two options to --onto	09:03
	rather, you need one option to --onto, plus the normal rebase argument	09:03
EugeneKay	Hi. Im searching for a GitHub or Gitorious style system that I can install locally, but for free. Anybody use a system like this / care to point me in the right direction?	09:04
	← lpolzer__ left	09:04
eMBee	oh, duh, sorry	09:04
	→ Paraselene_ joined	09:04
EugeneKay	(GitHub:FI, but without the price tag)	09:04
eMBee	rebase confuses me sometimes	09:05
Ilari	EugeneKay: Gitorious (its software package too).	09:05
EugeneKay	Oh, there is a package version available?	09:05
	rapha_ → rapha	09:07
	← rapha left	09:08
	→ rapha joined	09:08
EugeneKay	Ilari - Looks like nobody provides an (easy) Fedora package :-/	09:10
	→ bluenovember joined	09:11
Ilari	EugeneKay: Those packages are rather heavy... Are you sure you need something that heavy caliber?	09:11
EugeneKay	All I'm really after is a GUI-way to git that I can lock down to run on my VPN	09:12
	I have a MediaWiki setup for documentation already	09:12
	← harv left	09:12
shruggar	meh, trying to fork git.git on repo.or.cz, and apparently I fail basic astronomy. "please enter the name of our nearest star": I've tried Sol, the sun, Proxima Centauri, does it want me to write it in Czech?	09:12
Ilari	shruggar: 'Sun'?	09:13
EugeneKay	Even a decent X app would fit the bill	09:13
shruggar	Ilari: agh, "Sun" was right	09:13
kent\n	shruggar: thats not an astronomy fail, thats a turing test fail =P	09:13
EugeneKay	In-browser would keep total cross-compatibility, which is somewhat important to me. We run Windows on our desktops, for the most part.	09:14
shruggar	apparently I have proven that the computer is not a human	09:14
	← ustunozgur left	09:14
Ilari	EugeneKay: If you don't need something that heavy, I would use gitolite and gitweb...	09:14
	→ snearch joined	09:15
	→ akiva joined	09:15
akiva	Is there a way I can copy .gitmodules and copy and paste from another to save time rather than adding like 100 entries manually?	09:16
	I tried, but it didn';t work. It only took the manually entered ones. The pasted entries were also not present in .git/config	09:16
	← ralf_e left	09:16
akiva	So I assumed I could not... But there are just too many to add my hand	09:17
eMBee	can you not copy .git/config too?	09:17
	eMBee is doing a blind guess here	09:18
akiva	I tried, it still would not git submodule init	09:18
Ilari	EugeneKay: You need users to be able to edit access authorizations from the GUI?	09:18
	← snearch left	09:18
EugeneKay	Ilari - Nope. Just be able to do commits, forks, etc.	09:18
	→ gebi joined	09:18
EugeneKay	git-php looks like it's about spot-on.... Im gonna play with the demo a bit and see	09:19
akiva	Any Git smarties who can help?	09:19
	→ plq joined	09:19
plq	hi, is there a way to mark a file as executable so that it automatically has +x once pulled ?	09:20
Ilari	akiva: There's also submodule index entries...	09:20
	→ curtana joined	09:20
Ilari	plq: Yes. Make it +x and add & commit the change.	09:20
shruggar	I just read http://blog.red-bean.com/sussman/?p=96 , which had me thinking "but that's not true at all! If anything, a distributed model /encourages/ contributors to share their code, because you don't need to get your branch approved for inclusion into a central repository before sharing it!" then I realized that I don't actually have a publicly-accessible repository for my work-in-progress branches, so I registered with repo.or.cz and forked	09:20
plq	Ilari, I should've tried that :) thanks	09:21
	→ dj_tjerk joined	09:21
akiva	Ilari, what do ou mean?	09:21
EugeneKay	Ilari <3 for pointing me in the right direction. So happy to get actual, useful information from a OSS #channel for once. :-p	09:22
Ilari	akiva: The references to submodules: 1) .git/config, 2) .gitmodules, 3) entry for the submodule itself (in index, so they are committable).	09:22
EugeneKay	(most of it is RTFM or LMGTFY, which is no help to somebody who has already done that, and is lost)	09:23
	thunk → thunk\|away	09:23
shruggar	#git is a pretty helpful place. Once we take over the world, though, we'll act like you should already know everything we do :)	09:24
Ilari	EugeneKay: Well, forks need heavier system (well, one could use gitolite and force everyone to have access to everything).	09:24
	EugeneKay: (maybe excluding admin repos).	09:24
	EugeneKay: Or maybe even only readonly accesss.	09:25
EugeneKay	Having ACLs really isn't an issue for me right now, small dev team who all know better.	09:25
	But, future proofing ftw	09:25
	Right now we're just using a master repo stored on a server under /data/git/$project-name/	09:25
Ilari	EugeneKay: Well, gitolite could do that 'foo' can create 'foo/baz' just by pushing and 'bar' can read that repo. In turn 'bar' can create 'bar/baz' and 'foo' can read that repo. Etc...	09:26
	← plq left	09:27
Ilari	EugeneKay: And then gitweb shows all the repos.	09:27
EugeneKay	Gotcha.	09:28
	Im using ext3's setfacl right now, which is a lot easier than user/group perms.	09:29
Ilari	EugeneKay: In local gitolite setups I have, I can create repositories under 'Ilari/' just by pushing (no need to precreate them). But I don't have automatic read permissions set up.	09:29
	EugeneKay: Gitolite can also do ACLs.	09:29
EugeneKay	Reading up on it right now.	09:30
Ilari	Last time I needed new repo the only problem was me mistyping the URL few times (I use some pretty exotic git-related stuff, so "URLs" are pretty bizarre). Usually gitolite URLs are easy.	09:31
	→ tatsuya joined	09:33
Ilari	EugeneKay: They are especially easy if you hostalias the thing in .ssh/config	09:34
EugeneKay	How does git-web(or git-php) work with gitolite for users that don't exist in *nix? Ie, I create a git user named Bob and he logs in to git-web. Does it "just work", and show all the projects/branches he has access to?	09:36
Ilari	EugeneKay: I don't know about gitweb auth support...	09:43
EugeneKay	Hrm. Well, Im looking at a coupla other copycats/forks.... gitosis is one I found via yum search.	09:44
Ilari	EugeneKay: Gitosis is obsolete.	09:47
	→ mklappstuhl joined	09:47
EugeneKay	Ah.	09:47
	→ happypolla joined	09:47
EugeneKay	Yeah, the release date is 20080825 o.O	09:47
gebi	EugeneKay: no bugs, though no new features ;)	09:48
	→ lucsky joined	09:48
EugeneKay	I call those "Finished" programs.	09:48
	Like PuTTY.	09:48
	Hasnt been updated since 2007..... works great.	09:49
	← bousquet left	09:49
	← berserker left	09:49
	→ berserker joined	09:51
shruggar	I'd call any project which ends development at version "beta 0.60" to be "abandoned" rather than "finished" :)	09:52
EugeneKay	There's that.	09:52
	→ bauruine joined	09:53
	→ Crozin joined	09:57
sitaram	EugeneKay: I've been trying to call gitolite "finished" (in the good sense) for more than 6 months now. Doesn't happen... people want stuff and it keeps gaining more features.	09:58
	→ kar8nga joined	09:58
EugeneKay	Heh.	09:59
	← mastro left	09:59
sitaram	EugeneKay: almost none of the features added in 2010 are in use at my sites... they're all for folks who asked. So for me it is "finished" but...!	09:59
mabrand	PuTTy sitll has one serious problem that seems hard to fix.	09:59
shruggar	mabrand, what's that?	10:00
mabrand	reverse tunnels hang just after rdesktop logs in. since 2004.	10:00
	ssh and tunnelier don't have this behavior.	10:00
	← d1zzy left	10:00
EugeneKay	I use NX for my rdesktop tunneling.	10:01
	← froschi1 left	10:01
EugeneKay	I've never really mucked with PuTTY's ssh tunneling, actually.	10:01
mabrand	I can reproduce that problem ever time I try.	10:01
	It's maddening.	10:01
EugeneKay	Tried one of the forks?	10:01
	→ d0k joined	10:01
EugeneKay	They'd likely accept a patch for it.	10:02
mabrand	Yes. Tried a few and they have the same problem.	10:02
Crozin	Hi there. Does anybody know some FREE (or at least quite cheap) web-based project managment soft for private (non-commerical) projects with: 1 private GIT repo, some kind of board/bug tracking, at least 3 members	10:02
EugeneKay	Crozin - Im looking for a similar thing, for fully-private use, and I keep coming back to Gitorious.	10:02
mabrand	Who can write the patch? :)	10:03
Crozin	EugeneKay - Gitorious is fine but you cannot create private repo, can you?	10:04
EugeneKay	Crozin - Install it to private server.	10:04
	EugeneKay is still in the hunting phase	10:05
	shruggar will wait until someone invents a bug-tracker that doesn't suck :)	10:05
	← xiong left	10:06
Crozin	Great idea! Thanks!	10:06
	→ d1zzy joined	10:07
	← whaley left	10:09
	→ martinciu joined	10:10
	→ mastro joined	10:10
	← mattwynne left	10:10
rvsjoen	crozin: redmine ?	10:11
	thunk\|away → tunk`	10:11
	tunk` → thunk	10:11
EugeneKay	Any thoughts on SmartGit? Closed-source, but eh.	10:11
shruggar	using a closed-source hosting solution is like using a centralized repos	10:12
	→ froschi joined	10:12
	→ ereslibre joined	10:12
	← ereslibre left	10:12
	→ ereslibre joined	10:12
shruggar	ie: may be fine for now, but eventually, somewhere down the line, if your project gets old enough, you're going to wish you'd used something that let you move around more easily	10:13
EugeneKay	Its just a client, afaik.	10:13
shruggar	must have gotten then conversations mixed up	10:13
	← harinath left	10:15
	→ ziro` joined	10:18
	← ptor left	10:19
	→ ralf_e joined	10:20
	← mastro left	10:21
	← avar left	10:21
	→ Theravadan joined	10:22
	→ corni joined	10:23
	→ Yuuhi joined	10:24
	← cenuij left	10:24
	← gcl left	10:25
	jettero_ → jettero	10:26
	← jettero left	10:26
	→ jettero joined	10:26
	→ ptor joined	10:27
	→ Titosemi_ joined	10:27
	→ elTigre joined	10:28
	← fivetwentysix left	10:28
	← Titosemi left	10:28
	Titosemi_ → Titosemi	10:28
elTigre	I just did a squash, but one commit too much, so I'd like to split this commit into separate files... how would I do that?	10:29
Pieter	edit it, commit part of it, commit other part of it	10:30
	← Crozin left	10:33
shruggar	or undo the squash, then redo the squash	10:34
	→ harinath joined	10:34
	← curtana left	10:34
elTigre	how can I undo the squash?	10:35
shruggar	if you haven't done anything since, git reset --hard HEAD@{1} may be correct (you can check with "git log HEAD@{1}"). Otherwise, use "git reflog" to find the appropriate "before-the-squash" point	10:36
elTigre	thank you	10:37
	→ icwiener joined	10:38
	→ curtana joined	10:39
	← harinath left	10:41
	← Jochen_Hayek left	10:42
	← d1zzy left	10:42
impulze	hm is it somewhat possible to create a "reverse" patch? i.e. i'm on HEAD which contains 2 big changes, i want to separate those into 2 smaller changes and rebase on HEAD^, so i want to split out all the relevant changes now, create a "reverse" patch, commit --amend and apply the patch	10:42
	→ d1zzy joined	10:43
	← froschi left	10:43
_rane	git-revert	10:43
impulze	could you eleborate? :p	10:43
	← ntoll left	10:43
shruggar	impulze, man git-revert	10:44
Gitbot	impulze: the git-revert manpage can be found at http://git.or.cz/man/git-revert	10:44
shruggar	creates a commit which is the reverse of another commit	10:44
	not much to explain	10:44
impulze	ah	10:44
	← maleknet left	10:45
	→ maleknet joined	10:46
	→ ntoll joined	10:47
	→ avar joined	10:48
	→ aziz joined	10:48
	← avar left	10:50
	← kar8nga left	10:51
	→ avar joined	10:52
	→ harinath joined	10:57
	→ dknight joined	10:57
dknight	I need help with branching stuff.	10:58
curtana	go on :)	10:59
	→ mastro joined	11:00
	→ flaguy48 joined	11:01
	→ pellera joined	11:01
	→ gcl joined	11:01
	→ malumalu joined	11:03
dknight	curtana: well I am working on amarok. earlier amarok was hosted on gitoriois.org but now it is hosted at git.kde.org. I had created a cloned repo at my gitorious a/c	11:05
	curtana: now when I try git pull --rebase origin master, I do not get anything as there are no more commits to the gitorious amarok	11:05
	← berserker left	11:06
shruggar	dknight, sounds like you need to change where your remote points	11:06
	dknight, try "git config --edit", and it may be obvious what you need to do	11:06
	← ziro` left	11:07
dknight	curtana: so I changed my config remote URL to g.k.o which is showing the following msg http://paste.lisp.org/display/112026	11:07
	shruggar: i have done it. check the paste	11:07
	→ berserker joined	11:08
dknight	shruggar, curtana : how do I resolve these issues?	11:09
shruggar	dknight, what issues?	11:09
dknight	shruggar: I want to update my amarok to latest amarok repo.	11:09
shruggar	dknight, is it possible that there's just really no new updates?	11:09
dknight	shruggar: yes its about a months time	11:10
	→ ziro` joined	11:10
dknight	shruggar: as it says I have made 3 commits and origin/master has 120 commits	11:10
	shruggar: the git status says that	11:10
	→ radarek_ joined	11:10
shruggar	dknight, does it still say that?	11:11
dknight	shruggar: yes	11:11
	shruggar: sorry. not any more	11:12
	shruggar: i don't understand. has it updated the code.	11:12
	← radarek left	11:13
curtana	forced update?	11:13
	radarek_ → radarek	11:13
shruggar	dknight, I am guessing that the one on g.k.o is actually an older version than what you had	11:13
curtana	did they rewrite history?	11:13
	→ smuf joined	11:13
dknight	shruggar: I have no idea	11:14
	curtana: can't say	11:14
curtana	gitk master...origin/master	11:14
	← radarek left	11:14
curtana	you _should_ see your commits, and the commits on origin/master, to the point where they diverged	11:14
	← flaguy48 left	11:15
dknight	curtana: well since I committed new files, they diverged from the beginning I guess	11:15
	→ icwiener_ joined	11:15
	→ ralf_e_ joined	11:17
curtana	yeah. so you should see a common point (the point that you branched off) and that should have to descendents: your line of commits and origin/master's	11:17
	← icwiener left	11:18
	→ froschi joined	11:19
dknight	curtana: yes I do see. as soon as I started off coding, I diverged.	11:19
	← ralf_e left	11:20
	ralf_e_ → ralf_e	11:20
dknight	curtana: but I do not know what to do	11:21
wereHamster	/t	11:23
	ChanServ set mode: +o	11:24
	wereHamster changed the topic to: 1.7.1.1 \| Homepage: http://git-scm.com \| Everyone asleep or clueless? Try [email@hidden.address] \| Channel log http://tinyurl.com/gitlog \| Mailing list archives: http://tinyurl.com/gitml \| Gits on git: http://tinyurl.com/gittalks \| Pastebin: http://gist.github.com/	11:24
	ChanServ set mode: -o	11:24
curtana	dknight, git rebase origin/master _should_ be waht you want	11:24
	or git merge origin/master	11:24
	depends how amarok runs its project	11:24
dknight	ok	11:25
curtana	having said that; your original pull --rebase should have done that	11:26
dknight	curtana: hmmm but it did not	11:26
curtana	and given that it says 'origin/master (forced update)' i don't understand why you now see sensible divergence rather than confusion	11:26
	also: 120 commits on your branch, 3 on origin/master? according to that message	11:26
	← tvw left	11:27
dknight	curtana: well	11:29
	← bauruine left	11:30
dknight	git does not have revision numbers like svn. so I can't check that too.	11:31
	← dj_tjerk left	11:31
dknight	curtana: if you know of anything similar then I can try that	11:33
	→ harv joined	11:34
bremner	dknight: see the examples in the gitk man page	11:34
	← psil left	11:34
curtana	dknight, in the gitk view, find the two commits that were in your pull message	11:34
	origin/master used to be at 36aa01a, now it is at 7fbdae5	11:35
	so did your commits branch off 36aa01a? and how many are there?	11:35
	→ psil joined	11:36
curtana	and is 36aa01a an ancestor of 7fbdae5? if so, I don't see why it would say 'forced update'	11:37
	← Basstard` left	11:37
	→ Basstard` joined	11:38
_Vi	Is it considered good practice to add and commit files that still has conflict markers and then resolve that conflicts in further commits? What user should do if he has already resolved about 40% of a lot of conflicts and wants to store the result in repository (to prevent loss if he break the working copy)? "git stash" doesn't work in this case.	11:38
	← Theravadan left	11:38
	→ winterweaver joined	11:39
	→ unreal_ joined	11:40
ciaranm	_Vi: committing stuff that doesn't work is never a good idea	11:41
rvsjoen	you can do it in a different branch though	11:41
_Vi	ciaranm, Yes, of course in different branch.	11:41
	→ sftp joined	11:41
shruggar	ciaranm, I disagree. comitting stuff is always a good idea. But committing conflict-markers is /not/ a good idea	11:41
	→ intripoon joined	11:42
_Vi	Is there something like "git add --do-not-consider-this-in-rerere ."?	11:42
shruggar	_Vi: I don't believe there's a good solution to your situation	11:42
dknight	curtana: gitk is returning wish not found	11:42
_Vi	Or "git rerere" knows what is conflict markers and what they mean?	11:42
shruggar	_Vi: Have you considered attempting to merge an earlier commit?	11:42
muep	maybe commit, then later commit --amend?	11:43
	→ InTheWings joined	11:43
	← unreal left	11:43
	unreal_ → unreal	11:43
	← sftp left	11:43
muep	as the work on solving the conflicts progresses	11:43
_Vi	shruggar, Yes, I know that user should retry merge with earlier, but we expect that he 1. already done a lot of work, 2. "rerere" is off.	11:43
	→ ziro`` joined	11:45
	← psankar left	11:46
	← AAA_awright left	11:46
	Unode\|NA → Unode	11:46
_Vi	"If (rerere is turned on) { add only really resolved parts and retry merge with earlier commit; } else { commit to separate branch, then squash things together as one "merge commit"; }" Is this approach good?	11:46
	→ AAA_awright joined	11:46
bremner	hey git-experts. I have a workflow question. The background is a bit involved, so I pasted it: http://paste.debian.net/79170/ Any comments?	11:46
	← ziro` left	11:47
	← happypolla left	11:48
	← psil left	11:50
	← dknight left	11:51
avar	bremner: Look at how git itself does it	11:52
	→ jensn joined	11:52
	→ sftp joined	11:52
	← ralf_e left	11:53
	← bremner left	11:53
	→ bremner joined	11:53
bremner	avar: do you mean something more specific that "read the gitworkflows man page"?	11:54
avar	not really, it looks like your question is pretty much covered by that. Or am I missing something?	11:55
	→ bauruine joined	11:57
	→ hyperair joined	11:58
	← corni left	12:01
	→ snucko joined	12:03
bremner	well, I don't think anyone claims that the git.git workflow is perfect for all projects; in particular the man page disclaims that. So, rtfm is not a complete answer.	12:03
	← Adaptee left	12:04
bremner	hence my question if you meant anything specific about that workflow.	12:04
	← snucko left	12:05
	← elTigre left	12:05
	→ lazylisper joined	12:06
wereHamster	bremner: 'trying to stay as close to their original workflow' is usually not the best way	12:07
	→ radarek joined	12:07
	→ ramusara joined	12:08
bremner	wereHamster: well, given they already looked at gitworkflows and threw their hands up, and I am an outsider, I am pretty constrained.	12:08
	→ radarek_ joined	12:08
	→ cjs joined	12:08
wereHamster	bremner: http://nvie.com/git-model	12:08
bremner	wereHamster: thanks, that looks nice.	12:10
	and not too different from what I proposed :)	12:11
Rhonda	Yeah, the git-model workflow is extremely nice because it has graphics that even management can be convinced to understand. :)	12:11
	← radarek left	12:12
	radarek_ → radarek	12:12
	→ ChUrP joined	12:12
	← hyperair left	12:13
bremner	ok, I see the important difference is a develop branch that keeps going, and the tagging happening on master.	12:13
	→ [Po]lentino joined	12:15
	← nataraj left	12:15
sitaram	Rhonda++ on the "management" thing!	12:16
	→ nataraj joined	12:16
Rhonda	;)	12:16
	→ keyvan joined	12:16
	→ psil joined	12:17
	→ radarek_ joined	12:18
	← mklappstuhl left	12:18
	← lazylisper left	12:19
	→ DavidKlein joined	12:19
	← keyvan left	12:19
	→ lazylisper joined	12:20
pragma_	pragma_++ because pragma_++	12:21
	← radarek left	12:22
	radarek_ → radarek	12:22
	→ parasti joined	12:22
	→ keyvan joined	12:24
	→ codeshepherd_ joined	12:24
bremner	so do others really buy into making --no-ff default for merging feature branches (according the the git-model page)	12:25
	→ flaguy48 joined	12:25
bremner	I guess it doesn't hurt. I am just used to people obsessing with avoiding merge commits.	12:26
	← ijuma_ left	12:27
kokx	bremner: depends on which way you're avoiding merge commits	12:27
	→ warlock_mza joined	12:27
	→ IslandUsurper joined	12:27
kokx	if you're rebasing public commits for that reason, it isn't a good ides	12:27
	* idea	12:27
	← codeshepherd left	12:28
	codeshepherd_ → codeshepherd	12:28
	← d0k left	12:28
	→ dknight joined	12:28
dknight	curtana: well i have updated to current amarok; but the files I committed are gone	12:29
	← maleknet left	12:29
dknight	I want to get back my files and put them on a different branch altogether.	12:29
	curtana: how do I do that?	12:29
	curtana: I can retrieve my files from my gitorious a/c	12:30
	→ iocor joined	12:30
iocor	how do I stash everything apart from what is staged?	12:30
	← EugeneKay left	12:30
	→ Jaxan joined	12:31
wereHamster	stash --keep-index ?	12:31
	← kent\n left	12:32
	→ kent\n joined	12:34
	← kent\n left	12:34
	→ kent\n joined	12:34
	→ d0k joined	12:34
	← dknight left	12:35
	← lazylisper left	12:39
	→ Joeconyers joined	12:45
	→ Jochen_Hayek joined	12:46
	→ hugoxrosa joined	12:46
	→ anli_ joined	12:47
anli_	I tried git in eclipse, when I have shared a project using git, I am not able to see it afterwards in git repository exploring	12:48
	Maybe the new project becomes a new repo?	12:48
	→ mw joined	12:48
	← codeshepherd left	12:48
	← _Vi left	12:49
	→ jfkw joined	12:50
	← masterkorp left	12:51
	← ntoll left	12:52
	→ hebz0rl joined	12:52
	→ MattDiPasquale joined	12:53
	→ ntoll joined	12:55
impulze	could someone recommend a nice gui for git add --interactive?	12:56
	→ dj_tjerk joined	12:56
	→ sukima joined	12:56
jast	git-gui can do comparable things	12:58
	it may not be nice but it's a GUI :)	12:58
	→ tvw joined	12:59
	→ coderdad joined	12:59
impulze	heh i wouldn't mind if it would be an application merely for that purpose because i want to add ~800 hunks selectively without breaking my fingers ;)	12:59
	→ happynoff joined	13:00
happynoff	Hi there. using git svn is there a way to start from a given revesion instead of getting the full history ? thanks :)	13:00
MadCoder	impulze: tig does that	13:01
happynoff	I'm talking about the cloning*	13:01
	← intripoon left	13:01
sukima	happynoff: --depth 1 sould do that. I'll check	13:02
happynoff	sukima: I test it :)	13:02
	→ yaotti_aw joined	13:02
sukima	git svn clone --depth 1 svn://repo	13:02
	yaotti_aw → yaotti	13:02
sukima	that should get only the last revision	13:02
happynoff	Unknown option: depth :/	13:03
sukima	hmmm	13:03
happynoff	I found this, I'll test it http://stackoverflow.com/questions/747075/how-to-git-svn-clone-the-last-n-revisions-from-a-subversion-repository	13:03
	← MattDiPasquale left	13:04
	← jetienne_ left	13:05
str1ngs	happynoff: git svn clone -r<number> <url>	13:05
	happynoff: you can also do git svn init. and git svn fetch -r<number>	13:05
happynoff	the git svn clone -rXXX:HEAD repo seems to work :)	13:06
	← winterweaver left	13:06
happynoff	sukima, str1ngs: thanks :)	13:06
	→ frakturfreak_ joined	13:06
	← ChUrP left	13:09
	→ jailbox joined	13:09
	← frakturfreak left	13:11
	frakturfreak_ → frakturfreak	13:12
sondreb	do you need to do a checkout on a branch, when the branchname was a parameter when cloning?	13:13
	→ MattDiPasquale joined	13:13
	→ bitkiller joined	13:14
	← nevyn^ left	13:15
bitkiller	anybody here using gitolite (fedora rpm)?	13:15
	→ ive joined	13:15
sitaram	no, but I can probably help... what version are you using?	13:15
bitkiller	1.4.2	13:16
str1ngs	sondreb: if you use -b it implies checkout also.	13:16
sitaram	bitkiller: ok what's up?	13:16
bitkiller	It doesnt set environment variables	13:16
	GL_RC etc	13:17
sitaram	bitkiller: you're using the wrong ssh key	13:17
	→ dr_win joined	13:17
sitaram	bitkiller: http://github.com/sitaramc/gitolite/blob/pu/doc/6-ssh-troubleshooting.mkd	13:18
bitkiller	sitaram, oh...now i see...you are the gitolite developer :)	13:18
sitaram	I've been known to dabble in it, yes ;-)	13:18
	→ gbacon joined	13:18
	→ ChUrP joined	13:20
	← ph^ left	13:21
	← curtana left	13:21
	→ flazz joined	13:23
	→ giallu joined	13:23
	→ ehabkost joined	13:24
	→ tstclair joined	13:24
	→ HG` joined	13:27
	→ MadProfessor joined	13:27
	← tstclair left	13:27
bitkiller	sitaram, wonderful...thanks...	13:27
	→ tstclair joined	13:27
MadProfessor	anyone know if gitosis is supposed to work with git 1.7.1?	13:28
bitkiller	sitaram, i passed through that doc yesterday solving the clone problem by creating the second one and not returning to the doc...	13:28
sitaram	bitkiller: so did it work now?	13:29
bitkiller	yes...I forgot to copy the key to .ssh...it wasnt saved there when I created	13:30
	→ Shapeshifter joined	13:30
Shapeshifter	Hi. I'm quite new to git. I opened my own project on gitorious and committed some stuff back in april and everything worked as I followed a guide I can't remember. Now, on another machine, I did git clone, modified the source, then did git commit -a, and typed something into the changelog. Then I try doing git push or git push origin master but it says "You do not have write access to this repository" and " ! [remote rejected] master ...	13:30
	... -> master (pre-receive hook declined)" but I had added the public key of this box to my account.	13:30
sitaram	bitkiller: ok	13:30
	Alien_Fr1ak → csgeek	13:31
bitkiller	sitaram, thks very much, nice job	13:31
	→ nsussman joined	13:31
sitaram	you're welcome!	13:32
	→ TheNewAndy joined	13:34
	→ aresnick joined	13:35
	← bitkiller left	13:36
	→ bitkiller joined	13:36
	→ curtana joined	13:36
	← sukima left	13:37
	→ nevyn^ joined	13:38
	→ mrzombie joined	13:39
	← nataraj left	13:39
	← aresnick left	13:40
	→ js2 joined	13:41
	← Jaxan left	13:42
	← js2 left	13:42
	→ kpreid__ joined	13:42
	← martinciu left	13:42
	→ hyperair joined	13:43
	← girishr left	13:44
	← parasti left	13:44
	→ kukks joined	13:44
	← GodEater left	13:45
	← simplechat left	13:45
	← AAA_awright left	13:47
	→ AAA_awright joined	13:47
	← staale left	13:48
happynoff	How can I manage files with special chars in their names ? with svn I was using the unicode-path patch but how do I do it with git ? thanks	13:50
	→ GodEater joined	13:54
curtana	happynoff, are you on windows?	13:54
	→ mineiro joined	13:55
happynoff	curtana: nope, on mac, which uses decomposed utf8 if I'm right	13:55
	→ spearce joined	13:55
curtana	oh, i have no idea about mac stuff. i'd assume that it would work as well as it does on any other POSIX system though--that is, UTF-8 should work fine	13:56
	← nsussman left	13:57
	← henkboom left	13:57
IslandUsurper	happynoff, I've read that git doesn't do anything special with filenames, so as long as you always view them with the same encoding, it should work	13:58
	→ beatak joined	13:58
happynoff	IslandUsurper: the problem is that I'm working with git-svn, I think the issue is in the svn layer of git :/	13:58
	→ scarabx joined	13:59
happynoff	when I use git status it indicates 3 new files just after the checkout	13:59
	I had the same problem using a svn client	13:59
	→ _Vi joined	14:00
	→ henkboom joined	14:00
	← kpreid__ left	14:01
	→ nsussman joined	14:01
	← freimatz left	14:01
avar	Is there an easier way to get rid of redundant merge commits that have already happened than git format-patch begin-work-sha, then checkout the upstream and git am on top of that?	14:02
	← beatak left	14:02
	→ beatak joined	14:02
	← HG` left	14:03
	→ bcardarella joined	14:05
	→ Titosemi_ joined	14:06
	← vu3rdd left	14:06
bnovc	I have two branches A and B. Right now A points to commit C and B points to commit D. I want B to point to the same commit as A. How can I do this without using `git checkout` [since git doesn't let you without a checkout of it]?	14:06
	→ aless joined	14:06
wereHamster	bnovc: update-ref	14:07
shruggar	bnovc: git branch -d B && git branch B A	14:07
wereHamster	or branch -f B A	14:08
	← Titosemi left	14:08
	Titosemi_ → Titosemi	14:08
bnovc	hm, guess thast pretty easy :) thanks	14:08
	→ Ryback_ joined	14:08
bnovc	i was thinking <oldbranch> was only meant for -m	14:08
shruggar	this would be [<start-point>]	14:09
	git branch [--set-upstream \| --track \| --no-track] [-l] [-f] <branchname> [<start-point>]	14:09
s0enke	consider you are in branch "feature" and do "git rebase master" - now rebase has conflicts and there is a file "deleted by us" - but the file was deleted in master, not in "feature" branch. confused	14:09
fr0sty	avar: why are they redundant?	14:09
avar	fr0sty: Because they're just git pull automerges	14:10
bnovc	so if I don't plan to have any association with the other branch except a common starting point, I should also --no-track, right?	14:10
	← Shapeshifter left	14:10
shruggar	s0enke, I'm pretty sure the "ours" and "theirs" labels are reversed in rebase. At least, I've noticed that before, but never had time to confirm and report	14:10
	← shader left	14:11
shruggar	or confirm and think-through to see if it actually makes sense	14:11
s0enke	shruggar, ok at least i now know that i'm not completely stupid	14:11
shruggar	s0enke, rule 1: never take anything I agree with as evidence /against/ being completely stupid ;)	14:12
fr0sty	avar: but they must be doing something if they are non-ff	14:13
	→ nighthwk1 joined	14:14
fr0sty	If you don't want that to happen the branch set up to rebase on pull instead of merge.	14:14
avar	Yes, but that's something to do before you screw things up, not after.	14:14
	← Belna left	14:15
fr0sty	true... how many commits are you trying to bring forward?	14:15
avar	I don't know. I don't actually have this problem, I'm just wondering if there's a better way to solve it than the format-patch + am solution.	14:16
	→ WebDragon joined	14:16
fr0sty	for 'n' commits (where n is small), git cherry-pick will retain your history. (because each individual commit will remain separate.	14:17
	there is probably a set of arguments you could give to 'git rebase' but nothing is springing immediately to mind.	14:18
	→ cirwin joined	14:19
avar	yeah	14:19
cirwin	I'd like to use git to edit source code that is stored on a wiki - I can easily set up a repo that has a post-commit hook that updates the wiki, is there a hook I can use to fetch the latest version from the wiki when I do git pull?	14:20
	← sgh left	14:21
TBH	cirwin: http://ikiwiki.info/rcs/git/	14:21
	← hobodave left	14:21
TBH	Maybe it will solve your issue.	14:21
cirwin	wrong wiki software, unfortunately :)	14:22
Stummi	According his hostmask, I am thinkin he means a mediawiki	14:22
TBH	Nothing comes to my mind ;/	14:23
pasky	Hmm, can I gitignore a directory, but exclude a glob within that directory from the ignore?	14:23
cirwin	I suppose the next question is, how complicated would it be to set up something that pretends to be a remote git repository that I can pull from	14:23
	← spearce left	14:24
TBH	but I can post you a very nice frog instead http://nik.bot.nu/img/orig/4/5/455910.jpg	14:24
	:D	14:24
rkreis	that really is a nice frog	14:24
fr0sty	pasky: !glob	14:25
pragma_	fr0sty: that's disgusting	14:26
rkreis	what's that reflection in his left eye? looks like the lady in red from the matrix	14:26
fr0sty	man gitignore for more details.	14:26
Gitbot	fr0sty: the gitignore manpage can be found at http://git.or.cz/man/gitignore	14:26
	→ spearce joined	14:27
pragma_	man test	14:27
	← aluink_ left	14:28
pasky	fr0sty: can you elaborate?	14:28
	→ kpreid_ joined	14:28
pasky	fr0sty: if I ignore a directory, simple !glob afterwards has no effect	14:29
	so I was wondering if there is some common trickery to get around that	14:29
pragma_	man gittest	14:29
	man git-stash	14:29
Gitbot	pragma_: the git-stash manpage can be found at http://git.or.cz/man/git-stash	14:29
	→ tedoc2000 joined	14:29
rkreis	man ./git-stash	14:30
	← TheNewAndy left	14:31
fr0sty	in your .gitignore file:	14:31
	directory/* #ignore all files in directory	14:31
	← alezandro left	14:32
fr0sty	!directory/foobar #except foobar	14:32
	→ aluink joined	14:33
	← malumalu left	14:33
luis	Hi. I'm having trouble with a shared git repository. I've set core.sharedRepository=0777 but the subdirectories inside .objects/ still have 755 permissions.	14:33
	← beatak left	14:34
luis	Any hints?	14:34
rkreis	i guess you have to manually chmod them	14:34
	you could have used git init --shared instead	14:34
	← MattDiPasquale left	14:35
luis	ah, but the problem is that whenever a new commit generates a new subdirectory, that subdirectory has 755 permissions, instead of 777.	14:35
	→ leo2007 joined	14:35
rkreis	i think you also have to set the sticky bit, i'm not sure	14:36
leo2007	how to show a list of files changed in a commit?	14:36
	→ ijuma_ joined	14:36
luis	rkreis: 1777?	14:37
rkreis	i'd say so	14:37
	i'd just re-init it and pull the changes over or so :D	14:37
pasky	fr0sty: but then git status will stil show the directory/ as untracked	14:37
	← harinath left	14:37
	→ MattDiPasquale joined	14:37
	← ijuma_ left	14:37
luis	rkreis: I'll try that	14:37
pasky	fr0sty: instead of just directory/foobar	14:38
rkreis	directory/?* maybe :D	14:38
pasky	would that help?	14:38
rkreis	i have no idea	14:38
pasky	i can't see how	14:38
	hmm, i think there's no way to do this	14:38
rkreis	oops	14:39
	← warlock_mza left	14:39
rkreis	i thought directory/* is the except rule and directory/ would match it	14:39
fr0sty	pasky: git does not track directories, only files.	14:39
	→ ijuma_ joined	14:39
fr0sty	lets back up: "what are you trying to accomplish?"	14:39
	→ oriba joined	14:39
	→ Jaxan joined	14:39
	← mineiro left	14:40
	→ mineiro joined	14:40
luis	rkreis: "I did a git init --bare --shared=0777 foo" and .e.g., .objects/info has permissions drwxr-xr-x+.	14:40
fr0sty	leo2007: 'git log -1 --stat <commit>' might work	14:40
pasky	fr0sty: there're some junk directories being created that are uninteresting, except when junk/output file is there, in that case I might want to commit it	14:40
	← scarabx left	14:40
	← ijuma_ left	14:40
pasky	fr0sty: however, git won't descend to a gitignored subdirectory,never matching the exclude pattern	14:41
	← ntoll left	14:41
	← psoo left	14:41
fr0sty	are the junk directories 'durable' (do their names change)?	14:41
leo2007	fr0sty: that worked.	14:41
pasky	_and_ git status will always chop off the filename of the path if it's in a subdirectory	14:41
	fr0sty: they are globbale	14:42
	← bauruine left	14:42
luis	rkreis: hmm, it works if I use a local directory, but not on a remote directory. Must be some samba or cygwin weirdness. :-/	14:42
fr0sty	leo2007: there are other options to use --name-status, --diffstat, --stat . Look in man git-log for more	14:42
Gitbot	leo2007: the git-log manpage can be found at http://git.or.cz/man/git-log	14:42
pasky	I guess I will have to live with the shorter listing in git status output	14:42
fr0sty	but not always the same, right?	14:42
pasky	not always	14:42
rkreis	luis, samba and cygwin? i'm sorry	14:43
	did you try it locally as another user?	14:43
	→ ijuma_ joined	14:43
	← happynoff left	14:43
	→ madewokherd joined	14:44
luis	rkreis: should I be using msysgit instead?	14:44
fr0sty	pasky: if the directories are globbable then run : 'git add <glob>/output' before you commit.	14:44
rkreis	luis, i don't use windows	14:44
	luis, i just find samba confusing	14:44
luis	rkreis: lucky you :)	14:44
rkreis	i'd rather have git over ssh	14:44
offby1	what rkreis said	14:44
	→ Yuffster joined	14:44
leo2007	fr0sty: thanks. I looked at ls-files and show.	14:45
mrzombie	Hi guys, quick question regarding Grit: Anyone knows how I can get a commit log tree out of it?	14:45
	→ masterkorp joined	14:45
pasky	fr0sty: huh? that's orthogonal to my question	14:45
luis	rkreis: oh, that's a good idea. I might find a way to get ssh around here. Thanks!	14:45
pasky	fr0sty: first I need to _see_ them first in git status output	14:45
fr0sty	do you want to only include them conditionally?	14:46
	→ ajpiano joined	14:46
pasky	yes	14:46
	e.g. in a separate commit or such	14:46
offby1	mrzombie: what is a "commit log tree"?	14:46
	mrzombie: I wonder if you're looking for "git log --graph --pretty=oneline --abbrev-commit --decorate"	14:47
luis	rkreis: ah, but I'd need git on the server, meh.	14:48
mrzombie	offby1: This makes me want to execute the command right now, but Grit is a Ruby library for Git. I'm pretty much certain what you suggest is what I want, though. Only, not in command-line-world :P	14:48
offby1	aaah ... I thought "grit" was a typo, or a joke	14:49
	never heard of it before	14:49
blueshift	mrzombie: gitk?	14:49
	→ ntoll joined	14:49
blueshift	oh, "grit".. blank look	14:49
sondreb	why am i getting this error? fatal: Could not switch to 'win32-builder@git.oslo.osa/var/git/desktop': No such file or directory	14:50
GarethAdams\|Work	sondreb: are you trying to checkout when you meant to clone?	14:50
sondreb	no im trying to clone!	14:51
fr0sty	pasky: if I have (glob/* followed by !glob/output) git status will show the directory name iff 'glob/output' exists.	14:51
	← radarek left	14:51
fr0sty	that might be the best you can do.	14:51
	→ jmworx joined	14:51
sondreb	git clone --branch taskbranch-DSK-287666-3 --depth=1 lalalal@lalalala/var/git/desktop/work	14:51
offby1	sondreb: I wonder if you're missing a colon there somewhere	14:51
	lalalal@lalalala:/var/git/desktop/work	14:52
	→ meder joined	14:52
offby1	----------------^	14:52
	← steffkes left	14:52
jmworx	can someone confirm whether the following is correct:	14:52
	→ harinath joined	14:52
offby1	it's wrong.	14:52
	offby1 whistles innocently	14:52
	← leo2007 left	14:52
sondreb	where oofby1?	14:52
jmworx	If I want to use a superproject that has submodules, I can just checkout the super project and call git submodule init	14:52
sondreb	offby1	14:52
offby1	sondreb: after the host name, before the file name	14:52
	jmworx: I think you also have to do "git submodule update" as well	14:53
jmworx	then I can change the URL for the submodules in .git/config	14:53
offby1	jmworx: but yes, more or less	14:53
jmworx	only then I do git submodule update	14:53
	Is that correct?	14:53
offby1	jmworx: sounds reasonable. Why not Try It And See™? Should be risk-free	14:53
jmworx	But I should not modify the .gitmodules ?	14:53
sondreb	thanks you offby1.. it seems like I was off by 1 :D	14:54
offby1	jmworx: I think .gitmodules is where the URLs live, actually.	14:54
jmworx	offby1: I hate "try it and see" because you end up doing a lot of stupid things that sort of work and you get screwed in the end	14:54
offby1	jmworx: in my experience, most things with git either work perfectly, or fail obviously, but ...	14:54
jmworx	i.e. just because it happens to work today doesn't mean it's the right thing	14:55
offby1	jmworx: anyway, then, the real answer is "I am not sure, since I haven't futzed with submodules in a while"	14:55
pasky	fr0sty: yes, that's what I said :)	14:55
offby1	_I_ would do the Try It And See™ thing myself.	14:55
	jmworx: in the one project I have that uses submodules, there's no reference at all to the submodules in .git/config, but plenty in .gitmodules.	14:56
	← aziz left	14:56
jmworx	Well, I guess I have no other option anyway. Maybe I just got exposed to too much C code that was written as "try it and see"	14:56
offby1	ah, well, C code doesn't generally have that nice 'works perfectly or else fails in immediately and obviously' property :)	14:56
	← tcurdt left	14:56
jmworx	offby1: the issus is that the .gitmodules lists a read-only git url and each developer will want to point to his own writable url	14:57
	→ hobodave joined	14:59
	→ rlb3 joined	15:00
	← henkboom left	15:00
offby1	sounds like the simple solution is: have each developer edit .gitmodules, as you say	15:00
	and yet ...	15:00
	there's something unusual about that approach	15:01
	→ psankar joined	15:01
offby1	.gitmodules is typically checked in to the parent project.	15:01
	So it sounds like you're suggesting that each developer change that file, and _not_ check those changes in.	15:01
	that smells funny	15:01
	→ spaceonline joined	15:02
	← priidu left	15:02
	← bcardarella left	15:02
	→ quellhorst joined	15:03
	← anli_ left	15:03
	→ henkboom joined	15:04
	← chrowe left	15:04
quellhorst	i was working on a 'unknown branch' committed some changes, didn't push them, switched to a different local branch. did i just lose the changes i made?	15:04
curtana	quellhorst, run 'git reflog'	15:04
quellhorst	my commit said: [detached HEAD 5a4423c] fix mistake of changing <title> logic	15:04
curtana	if your last commit was 5a4423c then you can 'git checkout -b newbranchname 5a4423c'	15:05
quellhorst	thanks :)	15:05
curtana	anyway, reflog records all these changes. so you can get your missing commit id from it anyway :)	15:06
	← ntoll left	15:07
	← _Vi left	15:07
	← Textmode left	15:07
	← roop left	15:08
	← eighty4 left	15:08
	→ chrowe joined	15:08
	→ bgy joined	15:08
	→ boazsender joined	15:08
bgy	hi	15:08
	I'd like to do search for some terms I think I modified and then pushed in git, is there any way to do such search through diff history ?	15:09
	← giallu left	15:09
fr0sty	'git diff -p -S<search string>'	15:10
	bgy: man git log for more details	15:10
	crap,	15:10
	bgy: should be: 'git log -p -S<search string>' and man git-log	15:11
Gitbot	bgy: the git-log manpage can be found at http://git.or.cz/man/git-log	15:11
	← spaceonline left	15:12
bgy	fr0sty, ok, thanks I'll take a look at that	15:12
	← cbreak_work left	15:13
	→ bcardarella joined	15:14
	← rlb3 left	15:16
	→ rlb3 joined	15:16
	→ t0rc joined	15:18
	← frakturfreak left	15:20
	→ alezandro joined	15:20
	← marc_kdab left	15:22
	→ frakturfreak joined	15:22
	← nevyn^ left	15:27
	→ k3yvn joined	15:29
	yaotti → yaotti_aw	15:29
	← keyvan left	15:30
	← henkboom left	15:30
	← RobertLaptop left	15:31
	→ raichoo joined	15:31
	→ petercoulton joined	15:31
	→ henkboom joined	15:33
	meder → paparazzi	15:36
	→ Bass10 joined	15:36
	← skoop left	15:36
	← jony left	15:37
	← Bass10 left	15:37
	← EricInBNE left	15:37
	paparazzi → meder	15:37
	→ Bass10 joined	15:38
	offby1 → offby_something_	15:39
	→ eighty4 joined	15:40
	→ altrux joined	15:41
	← Guest68680 left	15:42
	← oriba left	15:43
	→ rayners joined	15:43
	→ Zephyrus joined	15:44
	← Zephyrus left	15:44
	→ Zephyrus joined	15:44
	← rlb3 left	15:44
	→ rlb3 joined	15:45
	→ devinus joined	15:46
devinus	anybody know how to set the username to use when the git repo is behind ssh during pull/etc ?	15:46
	← mastro left	15:47
cirwin	~/.ssh/config	15:47
	offby_something_ → offby1	15:47
	offby1 → park	15:48
	park → offby1	15:48
hyperair	devinus: · [user@]host.xz:/path/to/repo.git/	15:48
	offby1 → someone	15:48
hyperair	devinus: that was in git-fetch(1)	15:48
	someone → offby1	15:49
cirwin	hyperair: ah, the obvious solution, I like it	15:49
devinus	hyperair: thanks, i know that syntax, but i couldnt add that as the remote origin or my coworkers would get my username, right?	15:49
hyperair	devinus: er.	15:50
	→ corecode joined	15:50
corecode	hey	15:50
	is there a way to pull a stash from one repo to another?	15:50
hyperair	devinus: okay, see here. if you're concerned about people knowing your username, then your security is screwed to begin with.	15:50
corecode	or fetch a stash	15:50
cirwin	hyperair: I think he's more concerned that the co-workers should use their own usernames, so it shouldn't have his username in the config	15:50
devinus	hyperair: no, no, what i mean is i did a git remote add origin ssh://hosename	15:51
	now when i try to push it's like	15:51
	devin's password:	15:51
	curtana worked on a web application for an idiot client that thought that user names were equally as secret as the user's password	15:51
devinus	but my username on the box is dtorres	15:51
	:-/	15:51
hyperair	devinus: oh, there's a way to set a different push url.	15:51
	→ aziz joined	15:51
curtana	devinus, in that case, you can store the username for a particular host name in ~/.ssh/config	15:51
cirwin	devinus: you can add it to your .ssh/config	15:51
devinus	curtana: cirwin : thanks i'll check that out	15:51
	hyperair: thank you too	15:51
	→ kar8nga joined	15:51
	← kar8nga left	15:51
mrzombie	you can use ssh://dtorres@hostname too	15:51
	(just sayin')	15:52
hyperair	devinus: remote.<name>.pushurl	15:52
	mrzombie: wasn't that syntax deprecated?	15:52
	i think it was replaced by git+ssh	15:52
	and that ssh:// is just extra cruft.	15:52
	IslandUsurper → IslandUsurperAFK	15:53
mrzombie	The main point I was illustrating was the specification of the remote username within the url.	15:53
shruggar	corecode, if it was the last thing stashed, you could literally just fetch it: git fetch someremote refs/stash:refs/heads/fetched-stash	15:54
hyperair	mrzombie: which just reiterated what i first wrote =)	15:54
mrzombie	Oh, I didn't notice it, sorry :P Didn't mean to be redundant.	15:54
hyperair	np =)	15:55
	→ WebDragon\|away joined	15:56
corecode	shruggar: tried, didn't work	15:56
mrzombie	Does anyone know a nice way to setup a git service in a shared hosting environment so that every user in the hosting environment can host their repositories without interfering with the other users, and still provide the service to several different outer users?	15:56
_rane	private repos?	15:57
	← dj_tjerk left	15:57
shruggar	corecode, hm. Worked for me, but then I was just lazily fetching a local ref	15:57
mrzombie	And/or public.	15:57
corecode	shruggar: hum?	15:57
mrzombie	In a linux environment.	15:58
corecode	shruggar: the stash head is not created locally	15:58
shruggar	I did: git fetch shruggar@localhost:/path/to/current-repos refs/stash:refs/heads/stash-o because I don't have any stash-refs on really-remote repositories I can access at the moment	15:59
	← henkboom left	15:59
picasso	howdy folks	16:00
	→ sagsousuke joined	16:00
	← t0rc left	16:00
picasso	i'm having an issue trying to delete a remote branch on my gitolite server	16:00
	← WebDragon left	16:00
picasso	im using two remotes - gitolite and github	16:01
	→ ShooterMG joined	16:01
picasso	git push github :branch # works fine	16:01
	git push origin :branch # (gitolite) fails, says "hook declined"	16:01
curtana	your remote has a hook that is preventing you from deleting it then	16:01
	← adymitruk left	16:02
curtana	FYI you can do: 'git push --delete origin branch' in the latest git	16:02
picasso	ah that's better syntax	16:02
	→ henkboom joined	16:02
	→ kumbayo joined	16:02
	→ adymitruk joined	16:02
	→ Dvyjones joined	16:02
Dvyjones	If git thinks a file is a file rename, but it isn't, how do I tell it it isn't? :P	16:03
curtana	don't think you can. git doesn't really know about renames. it just says, when you ruin 'git-show' or similar, taht since the files contents are similar, it was a rename	16:03
jast	it doesn't really matter, does it?	16:04
Dvyjones	jast: Well, no it doesn't.	16:04
	← meder left	16:04
	← d1zzy left	16:04
Dvyjones	I guess most people can tell that git is wrong on thinking that README -> lib/foo.rb is a rename, but README -> README.md isn't a rename :P	16:04
	→ d1zzy joined	16:05
	→ xperience joined	16:05
shruggar	not without looking at the contents	16:05
	← barismetin left	16:05
curtana	are the contents of README and lib/foo.rb identical?	16:05
Dvyjones	They are both emptty.	16:05
	*empty	16:05
	:P	16:06
	→ didlix joined	16:06
curtana	solatis, yes. :)	16:06
	oops. i mean... so, yes. :)	16:06
	→ NET\|\|abuse joined	16:07
	→ kenichi joined	16:08
	→ psoo joined	16:08
shruggar	It all depends on why you want to know something is a rename	16:08
	← icwiener_ left	16:08
	← altrux left	16:08
shruggar	for git, the reason for reporting something as a rename is to be useful when viewing the history of the content. That is, it's saying "what name did this particular content have previously?", because git cares about content, and filenames are just a trivial piece of meta-data attached to that content	16:10
	→ marc_kdab joined	16:10
shruggar	however, I still agree that git should have a way of specifying "rename hints"	16:11
	→ aliceinwire joined	16:11
	WebDragon\|away → WebDragon	16:11
curtana	maybe the rename detection should special-case the empty blob	16:12
kokx	shruggar: you mean that git should have some internel metadata that says 'file X is renamed to file Y' ?	16:12
shruggar	kokx: I mostly mean that I should be able to specify "file X's content was generated independently of file Y"	16:12
	← trivol left	16:13
kokx	shruggar: why ?	16:13
	← rlb3 left	16:13
	→ rlb3 joined	16:14
luis	God dammit. msysgit doesn't seem to play well with samba either.	16:15
	← ChUrP left	16:15
shruggar	kokx, I find that git detects renames/copies a lot with XML data, for example.	16:15
kokx	shruggar: well, whats the problem of detecting the wrong rename ?	16:16
	(i notice this too, but i'm not really bothered by it)	16:16
shruggar	kokx: the problem is it makes logs ugly and hides actual renames	16:16
curtana	one problem is that you ahve to have this exact same conversation over and over again	16:16
kokx	shruggar: in my opinion, it would be better to remove the rename detection than to add more metadata	16:17
	← NET\|\|abuse left	16:18
shruggar	kokx: another (reversed) example would be "Normally, you wouldn't detect these two blobs as being related. However, they really are related, so if I'm using blame for example, look really hard in just this one other blob for changes to follow"	16:18
	→ radarek joined	16:18
shruggar	basically, a manual optimization for --find-copies-harder	16:19
bremner	do I need tk 8.5 for ttf fonts in gitk? I just "upgraded" my system and gitk fonts seem to have gone to heck.	16:20
shruggar	while over entire files "if git can't detect it, it's not interesting", individual lines or hunks may be useful where git can't detect them	16:20
	them == "them when mixed in with a bunch of other things"	16:21
kokx	shruggar: well, i still don't really see why adding some useless metadata would help	16:21
	→ justin-george joined	16:21
	← justin-george left	16:21
kokx	(hell, i don't even see a good possibility to fit the rename into the DAG without breaking backwards-compatability)	16:21
shruggar	kokx: because rather than --find-copies-harder, you have the option of saying "find copies the normal, fast way, but here are two blobs which I know are related"	16:21
	kokx: I'm just talking about .git/info/rename-hints, which would take the format: sha1 related-sha1 [related-sha1 [related-sha1]]	16:22
	kokx, and the reverse of that, specified with a ! prefix	16:22
	→ priidu joined	16:23
kokx	shruggar: i'm a bit puzzled; is the .git/info/rename-hints already built in or is it something you propose ?	16:23
shruggar	something I propose	16:23
kokx	well, in that way, it might be nice	16:24
shruggar	but which I have only thought out well enough for inane IRC chatter, not for writing a well-written RFC or patch	16:24
	→ Utkarsh joined	16:24
Utkarsh	what would be the equivalent of svn revert in git?	16:25
shruggar	Utkarsh, git checkout -- file	16:25
	Utkarsh, I recommend git checkout --patch	16:25
Utkarsh	shruggar: awesome, worked, thanks :)	16:25
	→ RobertLaptop joined	16:26
muep	hmm, would it be insane to use git to store a personal photo collection? It's only a few gigabytes since I do not photograph that much	16:28
shruggar	kokx: though what I just said wouldn't work for the case described of two empty files. Looking through my notes it seems as though I accounted for this originally by writing it as "treeish:path treeish:path", rather than the blob-sha1	16:28
	→ bauruine joined	16:28
shruggar	muep: usually, yes. Git stores the entire history of the entire project, for example, while usually for photos the individual history of each photo is more relevant	16:29
muep	yeah, maybe I should just backup my stuff even better	16:30
kokx	shruggar: somehow, still using the blob sha1's makes more sense to me	16:30
	→ ChUrP joined	16:31
shruggar	kokx: to me as well. I suppose for the empty-file case you could say "!sha1 same-sha1" to say "do not follow history unless the filename is the same for this file".	16:31
kokx	yep	16:32
shruggar	kokx: sounds like a "why is this not working?" waiting to happen, though :)	16:32
	→ juanjoc joined	16:32
shruggar	for my own purposes, blob-sha1 blob-sha1 would work fine	16:32
kokx	true	16:33
	→ kipras joined	16:33
	← adymitruk left	16:33
	← Utkarsh left	16:33
	→ yann_ joined	16:34
shruggar	I'd say just "any blob-ish" would be specified, but I'm not sure how one would handle spaces in filenames within .git/info/ files	16:35
	← rlb3 left	16:35
shruggar	"murder those who mention the oversight" sounds like a good plan of action	16:36
	← kpreid_ left	16:36
	→ kpreid__ joined	16:36
davr	Is there a way to stop a specific branch from being pushed when I type just "git push", without completely removing its remote tracking? I'd still like it to track the remote branch, I just don't want it getting auto-pushed like the rest of my branches.	16:37
	← petercoulton left	16:40
shruggar	davr: I thought there was something like specifying push = -refs/something:whatever after your main push line in config, but apparently I was wrong :/	16:40
	→ adripop joined	16:40
	← shruggar left	16:40
adripop	hi	16:40
	→ magicbronson joined	16:41
kokx	ShooterMG: is your idea to fill the file manually, or generatable from the logs, or ??	16:41
	→ agile joined	16:41
	→ spaceonline joined	16:43
adripop	can't understand how to create a git global repository	16:43
	i've create a git install --bare	16:43
	on a directory accessing through dav	16:43
kokx	*sorry, i meant shruggar	16:44
davr	adripop: If you don't understand git and just need general explanation of how it works, I suggest reading through one of the documentation / tutorials. I started with the one on progit.org, it worked pretty well for me	16:45
	→ tar- joined	16:46
davr	if you have a specific question, feel free to ask it, but "Can't understand how to create a git repo" is pretty generic	16:46
adripop	yep i know, but i stop writting because i delete all the stuff on the dav to recreate everything safely	16:47
	:)	16:47
	anyway thx	16:47
	← froschi left	16:47
adripop	http://www-cs-students.stanford.edu/~blynn/gitmagic this one is a good tutorial	16:49
	i think	16:49
	← sh1mmer left	16:50
	→ Zouppen joined	16:50
	→ froschi joined	16:51
	→ _Vi joined	16:51
	→ krawek joined	16:52
	→ coderdad_ joined	16:53
	← coderdad left	16:53
bremner	so, "git merge -s recursive -X theirs" is almost I want, except that I would like it to resolve delete-create and delete-modified constraints in terms of the other branch as well. I have vague memories of the "theirs" merge strategy, but I don't see it documented.	16:54
Zouppen	Is there a typo in the man page of git-diff-tree? In chapter RAW OUTPUT FORMAT it says: 8. sha1 for "dst"; 0{40} if creation, unmerged or "look at work tree". I suppose it should be "deletion", not "creation".	16:54
	→ jumpa joined	16:55
	← aliceinwire left	16:55
jumpa	Hello, I'm new to git. I've setup a repository on my development machine with two branches: master and local. I have deleted some files in the local branch, commited them, changed to master, changed some other stuff, and now switched back to local, but the files i have deleted are again existent.	16:56
str1ngs	jumpa: git clean -nd	16:56
	jumpa: will show you what needs cleaning. git clean -fd will clean.	16:57
	→ patmaddox joined	16:57
jumpa	thanks	16:57
	→ aspotashev joined	16:57
	← radarek left	16:58
	← radioman-lt left	16:58
str1ngs	jumpa: git status will show files not in the index or repo also	16:58
	→ radioman-lt joined	16:58
	← aspotashev\|eeepc left	16:58
jumpa	the command actually did not showed anything. I think the problem has been the commit	16:59
	← Jaxan left	16:59
str1ngs	jumpa: git log to check your last commits	16:59
	→ keyvan joined	17:00
	→ whaley joined	17:00
	← pellera left	17:01
str1ngs	jumpa: how did you delete the files?	17:01
	← bcardarella left	17:01
jumpa	rm -R folder	17:01
str1ngs	ah did you use git commit -a ?	17:02
jumpa	i dont think so	17:02
	yann_ → yann	17:02
	← phantomcircuit left	17:02
kokx	shruggar: are you planning to work out your idea a little better ?	17:02
str1ngs	ok use git rm -r <folder>	17:02
	jumpa: or if you remove with rm. then you need to commit with -a	17:02
	→ dreiss joined	17:03
jumpa	works, thank you	17:03
	→ Jaxan joined	17:03
jumpa	the next commit will contain the deletion of the folder?	17:03
str1ngs	git rm is much better since it puts it right into staging	17:03
	← k3yvn left	17:04
str1ngs	yes read the whole commit message it will make sense what being committed and whats not.	17:04
	IslandUsurperAFK → IslandUsurper	17:04
	→ ntoll joined	17:06
	→ programble joined	17:09
	← psoo left	17:10
	→ lispy joined	17:11
	→ yunosh joined	17:11
yunosh	hi, i accidentally created a branch "-t" because i mixed up the order of "-t -b". now i can't remove it anymore, because "git branch -d -t" complains about -t being an invalid flag name. halp? :)	17:12
	→ carrus85 joined	17:12
	← tedoc2000 left	17:12
curtana	yunosh, try: git branch -d -- -t	17:13
	→ tedoc2000 joined	17:13
yunosh	curtana: thanks, that did it	17:13
	← mineiro left	17:13
	→ cbreak_work joined	17:13
curtana	yunosh, the use of -- to mark the end of option arguments is pretty common, in case you ever create a directory called -rf or something :)	17:14
	← dpino left	17:14
ptor	yunosh: Tools using getopt(3) will stop parsing options when encountering '--', so everything after is passed on as non-options	17:14
yunosh	cool, so this is a general solution.	17:14
ptor	yep	17:15
adripop	ok i got it working !	17:15
	← Jaxan left	17:15
	← _Vi left	17:15
ptor	yunosh: And in Git it's also used to differ between branch/tag/whatever and filenames, thus 'git checkout HEAD -- filename', for example	17:16
	→ airborn joined	17:16
ptor	(whenever there's a possibility of confusion)	17:16
	← marc_kdab left	17:17
	→ marc_kdab joined	17:17
	→ _Vi joined	17:17
	← nuoHep left	17:23
	→ aresnick joined	17:24
	← adripop left	17:25
	→ Alagar joined	17:27
	← Alagar left	17:27
	→ nuoHep joined	17:27
	← Joeconyers left	17:28
	→ coppro joined	17:28
	← _Vi left	17:29
	→ k3yvn joined	17:29
	← giskard left	17:30
	→ _Vi joined	17:30
	← keyvan left	17:31
jmworx	offby1: Actually, just editing .git/config is sufficient AFAICT	17:31
	I don't think I need to modify .gitmodules at all	17:32
	← ramusara left	17:32
	→ phantomcircuit joined	17:32
	← bluenovember left	17:35
	→ tizbac joined	17:35
	← froschi left	17:35
	← tizbac left	17:37
yunosh	how can i find out (with git 1.5) which remote branches are tracked by my local branches?	17:37
	→ tizbac joined	17:37
	← Gitzilla left	17:38
	← fmardini left	17:38
str1ngs	yunosh: git branch -a . not sure if it works for 1.5	17:39
yunosh	no, it only shows the local and remote branches in separate lists	17:39
PerlJam	yunosh: look in .git/config	17:40
yunosh	PerlJam: where exactly?	17:40
PerlJam	yunosh: there should be some [branch "branchname"] sections that tell you which remote	17:41
	→ phlowx joined	17:41
PerlJam	I don't remember if 1.5 did that automatically or not though	17:41
yunosh	ah, the refs specified by the merge property	17:41
	→ kar8nga joined	17:42
	→ jrmuizel joined	17:43
	→ thedeeno joined	17:44
	← MattDiPasquale left	17:44
thedeeno	anyone know why git is showing 'deleted file mode' rather than just 'deleted' when rm a file?	17:44
	→ warthog9 joined	17:44
	→ Titosemi_ joined	17:45
	→ Gitzilla joined	17:45
	→ parasti joined	17:46
	← Titosemi left	17:46
	Titosemi_ → Titosemi	17:46
	→ khmarbaise joined	17:47
	← tatsuya left	17:47
	← tildeequals left	17:47
	→ zehrique joined	17:48
	→ giskard joined	17:50
	← devinus left	17:50
	→ noah joined	17:50
noah	Good morning. Anybody home?	17:52
	→ zenazn joined	17:53
	→ SlickT10 joined	17:53
	← WebDragon left	17:53
	→ radarek joined	17:54
	← drizzd left	17:54
fr0sty	noah: just ask.	17:55
	← nsussman left	17:55
	← JEEB left	17:55
kokx	noah: ask, but don't think there's anybody home	17:56
	→ lresende joined	17:56
	← sagsousuke left	17:56
	← Jochen_Hayek left	17:57
bremner	Can I commit in a script getting the default commit message listing conflicts? I realize it is not ideal, but I have a lot of these to do.	17:57
cbreak	commits don't have conflicts	17:58
bremner	yeah, so. I do a bunch of git adds and git rm's in a script, now I hope everything is cool, and try to commit.	17:59
	If that commit fails, I resolve things by hand	18:00
fr0sty	commits don't fail.	18:00
bremner	the command "git commit" does	18:00
fr0sty	with what error?	18:00
	← smuf left	18:00
fr0sty	(are you in the middle of a merge?)	18:01
bremner	yes	18:01
	→ opalka joined	18:01
	→ ChUrP1 joined	18:01
bremner	see http://paste.debian.net/79209/	18:01
noah	Have git working for a project. Mostly just on one machine. Wen't to "pull" from that machine to another in the office and ran into several errors about corruption and loose object. I tried "git gc" on the main repository, but it complained about corrupted objects. I'm new to git and don't know how to resolve this. It worries me.	18:02
	← Chaser left	18:02
	→ WebDragon joined	18:02
	← ChUrP left	18:03
	← spearce left	18:03
fr0sty	as long as you 'git add <formerly conflicted file>' in your subshell things should be fine.	18:03
	or are you trying to eliminate that step?	18:04
	btw, in what cases are you getting conflicts when using 'theirs'?	18:05
bremner	when one of the branches deletes the file	18:05
	← lispy left	18:05
	→ rseifert_ joined	18:05
	→ lispy joined	18:05
	← rseifert_ left	18:05
bremner	I'm trying to skip the step of opening up an editor on line 21 and having nothing very interesting to add to the default message.	18:06
	← Mpenz left	18:06
	→ rseifert_ joined	18:06
	← ChUrP1 left	18:06
	← rseifert left	18:07
	→ Mpenz joined	18:07
	← tvw left	18:08
	→ skoop joined	18:08
	ChanServ set mode: +o	18:09
	ChanServ set mode: -o	18:09
	← rseifert_ left	18:09
	→ enherit joined	18:09
	→ dj_tjerk joined	18:10
	→ rseifert joined	18:10
	→ ChUrP joined	18:10
fr0sty	bremner: you can provide -m"msg" to 'git commit' to prevent the editor from opening.	18:13
	→ nadavoid joined	18:14
	→ markabur joined	18:15
	← disappearedng left	18:15
noah	Anybody?	18:16
	→ sh1mmer joined	18:16
bremner	fr0sty: right. I just wanted the right value for msg so that is lists the conflict files, like git commit adds by default.	18:16
fr0sty	noah: can you paste the errors?	18:17
	← Gitzilla left	18:17
noah	From the "remote" side. (where I issue the "git pull")	18:17
	remote: error: inflate: data stream error (incorrect data check)	18:17
	remote: error: corrupt loose object '1c968b9dbead20f4f4f91c7d9a3aeb820e10ef5b'	18:17
	remote: fatal: object 1c968b9dbead20f4f4f91c7d9a3aeb820e10ef5b is corrupted	18:17
	error: git upload-pack: git-pack-objects died with error.	18:17
	fatal: git upload-pack: aborting due to possible repository corruption on the remote side.	18:17
	remote: aborting due to possible repository corruption on the remote side.	18:17
	fatal: protocol error: bad pack header	18:17
	→ Chaser joined	18:17
markabur	somehow my remote repo got set to track the wrong branch as head. i've tried git remote set-head origin master, which fixes the problem locally, but how do i make that change happen on the remote? git push says everything is up to date, and if i pull a new clone it still defaults to the wrong branch.	18:17
fr0sty	not here...	18:18
	noah: use gist.github.com	18:18
noah	Oh. Thanks. I'm new to this....	18:18
	← SlickT10 left	18:19
	← Chaser left	18:19
noah	fr0sty. Please try this: git://gist.github.com/459026.git	18:20
	Or this: http://gist.github.com/459026	18:20
	← whaley left	18:20
	→ TAsn joined	18:20
TAsn	Hey, I have active projects, and I just ran: git config --global --add core.excludesfile "*~" Though I don't see any changes with git status, I can still see the ~ terminated files. Did I miss anything?	18:22
	→ p3rg2z joined	18:23
	← sftp left	18:23
TAsn	ok nvm	18:24
fr0sty	noah, try 'git fsck --full' on the master repository side.	18:24
TAsn	I'm an idiot probably.	18:24
	sec.	18:24
	← coppro left	18:25
TAsn	ok fixed nvm :P	18:25
	skipped a word in the manual	18:25
	← TAsn left	18:25
	← priidu left	18:26
	jumpa → JUMPA	18:26
	→ sagsousuke joined	18:27
	→ coppro joined	18:27
	← _Vi left	18:28
	→ MattDiPasquale joined	18:28
	→ _Vi joined	18:29
	← psankar left	18:29
	← _Vi left	18:31
	← nuoHep left	18:31
	← pedrobelo left	18:32
	→ pedrobelo joined	18:33
	← pedrobelo left	18:33
	→ nuoHep joined	18:33
	→ mjf joined	18:33
	→ villemv_ joined	18:34
	→ _Vi joined	18:34
	← didlix left	18:34
noah	TAsn. Running now. Thanks!	18:36
	→ bamccaig joined	18:36
	→ keyvan joined	18:37
	VVSiz_ → VVSiz	18:37
bamccaig	For the first time ever I tried pushing to a completely new remote (a working repository on a remote server).	18:37
fr0sty	bremner: not sure how to get the real message, but you could probably fake a similar one by cutting up 'git status' output.	18:37
bamccaig	I got that to work, but now when I look at the status from that remote it shows files staged that undo the commit that I pushed into it.	18:37
	What is going on?	18:38
fr0sty	bamccaig: git faq non-bare	18:38
Gitbot	bamccaig: Pushing to non-bare repositories is discouraged. See https://git.wiki.kernel.org/index.php/GitFaq#non-bare	18:38
bamccaig	:-[	18:38
	← k3yvn left	18:38
	→ pantsman joined	18:39
	← dj_tjerk left	18:39
	→ tildeequals joined	18:39
	← radarek left	18:40
bamccaig	So because I had no changes on the remote end I can "fix" it by just doing a git reset --hard?	18:40
	→ altrux joined	18:40
	← altrux left	18:40
bamccaig	Yep, that worked. \o/ So I guess the /correct/ way to do that would be to use a buffer repository in between?	18:41
noah	git tsck --full failes with errors.	18:41
ciaranm	git tsck. is that what git says when it sees you've been using subversion on the side?	18:42
fr0sty	bamccaig: the other option is to only pull from both places.	18:45
	(or fetch+rebase, your option)	18:45
	← coppro left	18:45
	→ coppro joined	18:48
noah	type. "GIT FSCK --FULL"	18:48
	Unode → Unode\|NA	18:49
	→ disappearedng joined	18:49
	← bamccaig left	18:50
	→ Fullmoon joined	18:53
fr0sty	noah: what errors did 'fsck --full' return? (add them to your gist)	18:53
	← andrewl left	18:53
	→ hal joined	18:54
noah	OK. Errors added to gist.github	18:54
	→ wad joined	18:54
	← masterkorp left	18:54
wad	I'm wanting to look at some code from github. I'm on an Ubuntu 10.04 box. Which package should I install to use git? There are many different ones, and none of them are obviously the right one.	18:55
	← xenoterracide left	18:55
phlowx	git-core	18:56
wad	thnaks!	18:56
hal	I have changes in my working tree that I want to compare (using diff) with a serveral commits ago (before a merge with master). Does anyone know how I should do this, please?	18:56
fr0sty	git diff <commit> ?	18:57
	← harinath left	18:59
	← disappearedng left	18:59
fr0sty	noah: try the steps listed here: https://git.wiki.kernel.org/index.php/GitFaq#How_to_fix_a_broken_repository.3F	19:00
	be sure to make a copy of your repository first (that is listed as step one).	19:01
noah	fr0sty: Thanks!	19:01
	In general, this has happened before. Our repo consists of many text files (code) AND many large binary objects (10M or more). We always see files getting corrupted. Is there some kind of regular maintenance we should be performing, or some other technique to prevent this type of issue??	19:02
	→ harinath joined	19:02
David-T	yes. fix your broken hardware.	19:03
	→ tvw joined	19:05
noah	What?	19:05
pantsman	shield your machine from the cosmic rays	19:06
IslandUsurper	net some butterflies	19:07
ciaranm	stop licking your harddrives	19:07
David-T	if files written to your filesystem are different when they are read back (and your computer is always shutdown cleanly), you most probably have a hardware problem. e.g. bad memory. or an OS/driver bug. or perhaps a git bug.	19:07
	actually scratch that. a git bug wont break your filesystem. it could still break git, though.	19:07
noah	Interesting. Since we don't like our harddrive that often, I'll assume it is something else.	19:08
	Any ideas on how to diagnose this (hardware vs. software)??	19:09
	→ xiong joined	19:09
	→ Jaxan joined	19:10
	→ shader joined	19:10
	→ disappearedng joined	19:11
ft	um, is it just me, or does test 17 of `t9119-git-svn-info.sh' fail currently?	19:11
shader	I recently started cloning a svn repo using git-svn clone, but I started it at a late revision to get to work quickly. Unfortunately, it doesn't seem willing to fetch revisions older than the one I started it with. How do I fetch older revisions?	19:12
bremner	fr0sty: yeah, it's down to fighting with shell quoting now :)	19:12
	← wad left	19:12
fr0sty	yay!	19:12
	→ Mukophage joined	19:12
	← _Vi left	19:13
fr0sty	you can save the hassle and invoke another script to create the commit message and then use -F<filename> instead of -m<message>	19:13
	though that may be more, instead of less, trouble...	19:13
bremner	this (bashism) seems to work: conflicts=$'resolved conflicts:\n\n'`git status -s \| egrep '^(UD\|UA\|AU\|DD\|DU)'`	19:13
	← enherit left	19:14
	← berserker left	19:14
fr0sty	whatever works for you,	19:14
	← Jaxan left	19:15
	→ bc0906524 joined	19:15
	→ _Vi joined	19:16
	→ berserker joined	19:16
	← magicbronson left	19:16
	→ dp joined	19:17
	← coderdad_ left	19:18
noah	OK. That worked :) :) :)	19:18
	Thanks everyone!!	19:18
dp	I'm going to be forcing my coworkers to use git, but only after I've started using it for a while. if they make changes directly within the repo, and then I git commit, I'll be forced to "fix" my erroneous code, right?	19:18
fr0sty	shader: look at this: http://utsl.gen.nz/talks/git-svn/intro.html#grafting it might help...	19:18
	→ coderdad joined	19:18
	→ rubyruy joined	19:19
PerlJam	dp: "the repo"?	19:19
	dp: they should each have their own repo	19:19
	→ magicbronson joined	19:20
dp	PerlJam: they should, but assume we're not to that level yet	19:20
bremner	that is the zeroth level	19:21
	otherwise it gets much harder	19:21
dp	IE, I have a repo, but they're editing directly within where the data lives	19:21
	→ project2501a joined	19:21
fr0sty	dp: 'git clone', post haste.	19:21
noah	OK. Question about day-to-day workflow	19:21
PerlJam	heh	19:21
project2501a	they guys, puzzling error: error: pack-objects died of signal 13 error: pack-objects died with strange error	19:22
noah	I work on 3 machines. Server, Desktop, Laptop. Using git to keep everything in sync.	19:22
project2501a	google doesn't come up with something. can i have a clue what it may be?	19:22
fr0sty	huh, that's strange...	19:22
	;-)	19:22
project2501a	yes, it's a strange error	19:22
	strangest than most	19:22
	you can call it dr strange	19:22
noah	Normally, I use "git pull" on the laptop to get the latest version. Then do my work. At the end of the day, I use "git commit' followed by "git push". Does that sound correct?	19:22
	← magicbronson left	19:22
fr0sty	out-of-memory, perhaps.	19:23
	← skoop left	19:23
project2501a	fr0sty: on the client or the git server?	19:23
fr0sty	failure of a network share?	19:23
PerlJam	dp: if you and they are all working in/commiting to the same repo, you're going to step on each others toes unless you can guarantee that multiple people aren't editing the same file at the same time.	19:23
wereHamster	noah: yes	19:23
fr0sty	or the network, in that case (signal 13 is SIGPIPE)	19:23
_rane	noah: sounds correct, if you use git commit more than once	19:23
project2501a	fr0sty: ... i'm on ssh, i don't have "network shares" :P	19:23
	→ apsysoev joined	19:23
noah	But, I read someting in the FAQ bout pushe changes not affecting the work tree in the main repo??	19:24
	→ JEEB joined	19:24
noah	https://git.wiki.kernel.org/index.php/GitFaq#Why_won.27t_I_see_changes_in_the_remote_repo_after_.22git_push.22.3F	19:24
fr0sty	noah: also, if the place you are pushing to is a bare repository.	19:24
	→ petercoulton joined	19:24
fr0sty	then run-of-the-mill networking errors may be to blame.	19:24
noah	Not a bare repo. The server version is the actual "live" version we use for running and testing.	19:24
	So I need the server to always be the "master" with the most current version of everything	19:24
wereHamster	noah: then create another, bare repo which you psh into	19:25
noah	Pleaes forgive the newbie questions. I'm coming from rsync, so this is all a lot more	19:25
	→ aliceinwire joined	19:25
fr0sty	project2501a: what command were you using.	19:25
noah	wereHamster: Not sure I understand. How do I know if the server version is "bare" or "regular"	19:25
wereHamster	noah: because you know whethre you created it with git init or git init --bare	19:26
	← harinath left	19:26
project2501a	fr0sty: git push	19:26
noah	Honestly don't remember. Anyway to look in the .git directory for a clue?	19:26
wereHamster	git config core.bare	19:27
IslandUsurper	noah, if you have a separate .git directory plus all of the files, it's not bare	19:27
project2501a	does git add a .git suffix to a repo when it cannot find repo.git ?	19:27
noah	Ahhhh. OK	19:27
fr0sty	project2501a: did you get the same error when attempting to push again? (also, how much memory is available on the remote host, and how big is your repository?)	19:28
project2501a	just a minute, i'll answer that	19:28
	remote repo is 38Mb and the host has 24Gb mem	19:29
	→ javaJake joined	19:29
javaJake	How do I update my local listing of git-svn branches?	19:29
wereHamster	javaJake: git svn fetch ?	19:29
	← petercoulton left	19:29
javaJake	Oh my	19:29
noah	Will having a bare repo on the server still track changes, diff, etc.??	19:29
javaJake	I thought git svn rebase handled the fetch bit	19:29
wereHamster	noah: 'track changes'?	19:30
Ilari	noah: Yes. Bare repos don't have working copy nor index but all the other stuff is there.	19:30
noah	"track changes" - All the "good stuff" that git does.	19:30
	llari: Will the actual files be in the bare repo, so we can compile and run them from there?	19:31
wereHamster	no	19:31
_rane	no, because you shouldn't do that	19:31
IslandUsurper	noah, the place you want to compile should pull from the bare repo	19:31
	but the development work should be pushed to the bare repo instead of directly to production	19:32
noah	Ahhh. So to be sure I understand. I really need TWO directories on the server. 1) a bar directory for all the remote commits. 2) A working directory that is checked out from the bare ???	19:32
fr0sty	project2501a: ok, worth asking at least... did the subsequent push fare any better? (btw, try -v next time	19:32
wereHamster	noah: yes	19:32
	← Basstard` left	19:32
noah	Understood. Thanks :)	19:32
	← dp left	19:33
IslandUsurper	noah, the benefit is that you can still commit hotfixes on the server if you really need to, and push those back to the bare repo	19:33
project2501a	fr0sty: apparently, no.	19:33
	→ spearce joined	19:33
noah	IslandUsurper: That makes sense.	19:33
project2501a	does git add a .git suffix to a repo when it cannot find repo.git ?	19:33
noah	I've been using one directory for both. Probably made a mess.	19:33
project2501a	i mean, does it add .git to /foo/repo/bar when it can't find /foo/repo/bar ?	19:34
	i mean, does it add .git to /foo/repo/bar when it can't find /foo/repo/bar.git ?	19:34
	→ j0shua joined	19:34
fr0sty	project2501a: I don't think so.	19:34
	are you pushing to an existing, extablished repository?	19:35
wereHamster	project2501a: try it and see	19:35
project2501a	i just experimented pulling and apparently it does	19:35
	fr0sty: yeah i am	19:35
fr0sty	any hints from running 'git push -v?	19:36
project2501a	maybe tomorrow morning	19:36
	i can'd deal with it right now	19:36
	i'm falling asleep :D	19:36
	wait a sec	19:36
	← p3rg2z left	19:36
project2501a	fr0sty: can i privmsg you?	19:36
	→ stringo0 joined	19:37
Ilari	noah: Note that changing repositories that are updated by hooks is a bad idea (because merging sanely is impossible).	19:37
picasso	I'm trying to delete a remote branch managed by gitolite, and somehow I'm being denied permission	19:37
project2501a	fr0sty: actual error http://pastebin.com/zgTeu70e	19:37
Ilari	picasso: Hook rejected? If so, any error messages above?	19:38
	← rubyruy left	19:38
picasso	yea: + refs/heads/canvas jquery-wheelgame mike_2a-mac-pro DENIED by fallthru	19:38
	but i can write to that branch	19:38
	im thinking gitolite maybe has different permissions for deleting branches versus just writing?	19:39
	but i dont know how to debug, i dont see anything in .gitolite/log/	19:39
wereHamster	project2501a: line 5 ... ?	19:40
Ilari	picasso: The operation would need '+' permission. And you don't have it.	19:40
	→ OFICINA7 joined	19:40
project2501a	wereHamster: yeah, i figured it out	19:41
	← cilly left	19:41
	→ harinath joined	19:41
picasso	Ilari: ahh. got it :) thank you	19:42
	missed that character	19:42
	← xperience left	19:43
Ilari	picasso: Newer gitolite versions have 'D' permissions (wheither '+' or 'D' permission is needed is determined from presence of any 'D' permissions).	19:43
	→ jobor joined	19:43
	→ p3rg2z joined	19:43
	← DavidKlein left	19:45
	→ drizzd joined	19:45
	→ jobor_ joined	19:45
	→ RORgasm_ joined	19:45
RORgasm_	hey guys	19:46
	so i did git svn clone to clone a svn repo to a git repo	19:46
	when i look at .git/objecrts , that folder is like a 120 something MBs	19:46
	that seems way to big to me for the actual code	19:47
	→ State joined	19:47
State	Hi: Just ran git add . when I really should't have (added a bunch of things not needed) - how can I roll back to the last state?	19:48
drizzd	RORgasm_: it's not just the code, it's the entire history	19:48
shader	RORgasm_: try running git gc	19:48
RORgasm_	right	19:48
shader	or git repack	19:48
RORgasm_	but teh things is	19:48
	← jobor left	19:48
	→ dj_tjerk joined	19:48
RORgasm_	the orignal svn repo had like 57 revisions	19:48
	and it wasn't much	19:48
	git gc	19:48
	ok	19:48
drizzd	RORgasm_: you can use faq find-large to display large files	19:49
Gitbot	RORgasm_: Howto find large files. See https://git.wiki.kernel.org/index.php/GitFaq#find-large	19:49
RORgasm_	let me try that	19:49
	← jensn left	19:49
RORgasm_	can't find git gc, is that an alias for another command	19:49
	?	19:49
drizzd	State: you can unstage everything by using git reset	19:50
	you can unstage selectively by passing files as arguments to git reset, or using git gui	19:50
	← HarryHood left	19:50
	← Tac left	19:50
drizzd	RORgasm_: where are you looking? git help?	19:50
	try git help -a	19:50
fr0sty	project2501a: it's a free country...	19:51
shader	git gc is a normal command, as is git repack	19:51
	← gebi left	19:51
	→ ilogger2 joined	19:58
Ilari	RORgasm_: Gitosis is obsolete (obsoleted by Gitolite).	19:58
wereHamster	shawn-p: by whom?	19:58
project2501a	thank you, bill gates	19:58
	* adminspotting *	19:58
shawn-p	Oh, I think I understand	19:58
	Do you add your public key when you create your github account?	19:58
	I'm using an old github username and I've since switched computers	19:58
	→ didlix joined	19:59
fr0sty	project2501a: is the repo setup as 'shared' core.sharedRepository=group	19:59
project2501a	nope	19:59
	to be frank, if i had my way, the repo would have been hg	19:59
	→ awilkins joined	20:00
project2501a	but it's not up to me.	20:00
	→ nairb774_ joined	20:00
project2501a	nonetheless, fr0sty, that's in .git/config in the remote git repo, right?	20:00
shawn-p	Yeah, that was it	20:00
	Thank you wereHamster	20:00
fr0sty	project2501a: yes, that makes sure new objects get the right umask when they are created.	20:02
	→ sh1mmer joined	20:02
project2501a	choose life. choose a big computer. choose disk drives the size of washing machines	20:02
	→ sondreb joined	20:03
project2501a	choose usenet and alt.sysadmin.recovery. choose cheap office carpet and combat boots.	20:03
	kek, i'll stop now. thank you guys.	20:03
	→ phayte joined	20:04
	← phayte left	20:05
	→ JEEB joined	20:05
	→ ggh joined	20:06
	→ Jaxan joined	20:06
	→ opalka joined	20:06
State	Is there a way to mark files / folders as always out of the tree? Such that I can run git add . and certain things will never be added?	20:06
mmattice	.gitignore	20:07
State	mmattice: Thanks/	20:07
	→ noah joined	20:08
noah	Hi. So the plot thickens.	20:09
	→ Gitzilla joined	20:09
noah	On the SAME machine. I created a new directory and issued a "git pull files/repo"	20:09
	← Jaxan left	20:09
	→ berserker joined	20:09
fr0sty	'git init'	20:09
	if you haven't done so.	20:09
	→ solaris_bell joined	20:10
noah	Git took a while to compress and receive the object, then spit out an error: "Your local changes to code/file.r would be overwritten by merge. Aborting" However there are not files in the new diretory.	20:10
	I just did a "git pull"	20:10
solaris_bell	!ops A message` to you from `BIKCMP: Uhm, excusez-moi, but as a regular on the richest channel on EFnet, I think I have the authority to deem you a pusillanimous troll. Go back to dipping your stale cheetos in your ramen, you white trash too-poor-to-afford-neutrogena too-stupid-to-know-Biore-is-better pizza-faced nerd. I'm going to go ejaculate into a fresh silk scarf, hand made by a Thai company in which my father has a controlli	20:10
	→ hasrb joined	20:10
fr0sty	noah: "ls" shows what?	20:11
noah	empty directory	20:11
fr0sty	ls -la?	20:11
noah	I literally just did "mkdir test; cd test; git pull files/repo"	20:11
	→ marienz joined	20:12
wereHamster	noah: git init before the pull	20:12
	→ KB1JWQ joined	20:12
noah	OK. Trying that.	20:12
fr0sty	or git clone	20:12
noah	ahhh.	20:12
	← solaris_bell left	20:12
fr0sty	also, files/repo is probably wrong	20:12
	you need a full path to the repository	20:13
noah	Based on previous advice here, I want to convert my existing repo to a bare one. So first I want to pul la copy as my working.	20:13
	fr0sty - I have the path, "file/repo" was just a placeholder in my IM	20:13
wereHamster	noah: create the new, bare repo (git init --bare foo.git), then cd foo; git push --all --tags ../foo.git	20:14
	→ tomsdale joined	20:14
	← tomsdale left	20:14
	→ Joeconyers joined	20:14
	→ johnw joined	20:14
	→ tomsdale joined	20:14
noah	wereHamseter: saw in the FAQ that I could just rm all the files and then copy .git/* to . change the config file and it should be bare??	20:15
	← tomsdale left	20:15
	→ sixteneighty joined	20:16
wereHamster	noah: yes, that works, too.	20:16
noah	But first need to checkout a working copy.	20:16
	Still seeing some errors. Did a git pull on a laptop over ssh. Got all the way through, but then error on uncompressing "fatal: pack has bad object at offset 273517203: inflate returned -3	20:17
	"	20:17
	→ tomsdale joined	20:17
	→ drizzd joined	20:18
tomsdale	oh dear - I just typed 'git push origin HEAD' and seemed to have overridden my working branch a not up to date remote repository - how can I revert?	20:19
Ilari	tomsdale: That's not a forced push.	20:20
	→ WebDragon\|laptop joined	20:20
tomsdale	so 'git revert ###' should do it - just afraid to do anything else until I understand what I actually did.	20:20
	→ Jaxan joined	20:21
Ilari	tomsdale: git revert is for reverting individual commits by making a new commit.	20:22
	← Jaxan left	20:22
tomsdale	llari - and it doesn't work either - tried to revert the last commit - no change.	20:24
	← JEEB left	20:26
tomsdale	I think I kind of understand what I did - I was working on mybranch und pushed the not up to date master (instead of mybranch ) to the repository. Then I executed 'git push origin HEAD' and I seemed to have overwritten mybranch with the not up to date master.	20:26
	→ thiago_home joined	20:26
	WebDragon\|laptop → WebDragon	20:27
	→ justin-george joined	20:27
	← KB1JWQ left	20:28
tomsdale	am I f*****?	20:29
noah	Question: When issuing a "git pull" either on the same machine or remotely, it appears as if git is compressing everything, transferring everything and then uncompressing. Since we only update a small number of our files, isn't this wasteful? Is there a better way?	20:29
cirwin	if you are only doing a small number of files, it won't take long anyway	20:31
	if you are doing a lot, then it's worth it	20:31
noah	It always seems to take long. It appears as if it is compressing and moving everything every time.	20:31
tomsdale	I think I am - I executed: take the not up to date master from the remote machine and overwrite my up to date local branch (Head) with it. I don't think there is any way out - is there?	20:35
	→ hukl joined	20:36
	→ priidu joined	20:37
	→ AzureWolf joined	20:38
	→ alphabeat1 joined	20:38
	→ Phyx joined	20:38
	→ SamuelofR247 joined	20:38
	← SamuelofR247 left	20:39
	← Phyx left	20:39
	← AzureWolf left	20:39
	→ State_ joined	20:39
	→ cyngs joined	20:40
	→ alphabeat joined	20:41
	→ alice\|wl joined	20:41
	State_ → State	20:41
	→ stringoO joined	20:42
	→ dgs joined	20:42
alice\|wl	hello, somehow I managed to not currently be on a branch	20:42
	but I want the state as it is to be the master	20:42
	how can I acomplish that?	20:42
	← alphabeat1 left	20:43
	→ VxJasonxV joined	20:44
	→ Zephyrus joined	20:44
	← Zephyrus left	20:44
	→ iLurk joined	20:45
	← iLurk left	20:45
	← drizzd left	20:46
	← sh1mmer left	20:46
	→ jenk joined	20:46
jenk	how do i pull an updated version of a specific file? in svn i could just type "svn up foo/bar.jpg"	20:47
	is there a git equivilant?	20:47
	← cyngs left	20:47
Rhonda	git checkout HEAD -- foo/bar.jpg	20:48
	← VxJasonxV left	20:48
Rhonda	git fetch before that	20:48
	→ VxJasonxV joined	20:49
	→ bambule joined	20:50
jenk	thanks	20:50
	← opalka left	20:50
jenk	what does the -- do?	20:51
jast	in this case, nothing	20:51
	→ justin_george joined	20:52
jast	in general, it's to make sure there are no confusions if you use a filename that also exists as a branch	20:52
	because you can use git checkout with just a filename argument, too	20:52
	then it'll give you the last staged version of the file	20:52
	← VxJasonxV left	20:52
jenk	ah i see, it means "i don't mean branch named foo/bar.jpg" ?	20:52
jast	yeah	20:52
jenk	gotcha	20:52
jast	git checkout only tries to interpret the first argument as a branch name, though	20:53
	so if you specify both branch and filename you don't really ever need it	20:53
	but if you specify only a filename, it helps to build the habit to use --	20:53
jenk	any idea when debian stable is going to include a reasonable version of git?	20:53
Nebukadneza	-wc	20:53
jast	probably with the next debian stable	20:53
	→ appleperl joined	20:53
jast	there's a recent version in lenny-backports, by the way	20:54
	→ stamina joined	20:54
	→ phantomcircuit joined	20:55
	→ whaley joined	20:55
	→ ShadeHawk joined	20:55
	→ Kloerihub joined	20:55
Arrowmaster	debian stable never gets new version, only security fixes for existing versions of software	20:55
	you should know that if you are using debian stable	20:55
Kloerihub	!ops a `special message` to you `from `BIKCMP: Still dialing 911 for emergency police services? If you're like many people in the New York area, these simple numbers can be a painful reminder of loved ones lost in the terrible tragedy of September 11, 2001.	20:55
	← justin-george left	20:55
	← Kloerihub left	20:55
	→ enherit joined	20:56
jast	hmm, spam bots do curious things sometimes	20:56
	→ funcx86 joined	20:57
	→ jensn joined	20:57
funcx86	!ops a `special message` to you `from `BIKCMP: Keep at it! I'm sure someday you'll be mentioned in a blog.	20:57
jast	sigh	20:57
	← funcx86 left	20:57
	→ anjen joined	20:57
	← appleperl left	20:58
ShadeHawk	An idea: what do you think about having "Git User's Survey 2010"? What should be in it, if it should be done?	20:58
jast	hum	20:58
	→ eletuchy joined	20:58
jast	were there any actual consequences to the previous surveys? :)	20:58
	→ unreal_ joined	20:59
	→ beatak_tksh joined	21:00
	unreal_ → unreal	21:00
	← beatak_tksh left	21:00
	→ HubComputersdron joined	21:00
	→ beatak_tksh joined	21:01
HubComputersdron	!ops a `special message` to you `from `BIKCMP: The Flu Fuck: blow that snot into your hand. wipe it on your tongue oh man. crunchy crunch, the boogers go. Gulp! you swallow. Down they go. Uh-uh-uh here comes sneeze. Achoo! You're bent, hands on your knees. A prodding's felt, betwixt your cheeks. What the heck? It's a linux geek With cock unfurl'd, he rips your pants. Nothing can stop this fag's advance. In it goes, his cock dot	21:01
	← HubComputersdron left	21:01
	→ ralphschindler joined	21:01
marienz	/12/20	21:01
	gah, sorry	21:01
ralphschindler	hey all, if i want to back out several commits at a specific path, what workflow should I be looking at?	21:02
	→ computersBsdAtar joined	21:02
computersBsdAtar	!ops a `special message` to you `from `BIKCMP: Frankly, I am appalled at the way that members here treat the underground furry community. It seems to me like nobody here really understands anything about the furry way of life. I mean, seriously, do you know how freeing it is be honest about being encased in a human body whilst having an animal soul?	21:03
	→ sunos_linux joined	21:03
sunos_linux	!ops a `special message` to you `from `BIKCMP: "If a man finds a girl who is a virgin, who is not engaged, and seizes her and lies with her and they are discovered, then the man who lay with her shall give to the girl's father fifty {shekels} of silver, and she shall become his wife because he has violated her; he cannot divorce her all his days." (Deuteronomy 22:28-29, NASB) Can you imagine the mental torment and terror such a woma	21:03
	← computersBsdAtar left	21:03
	← sunos_linux left	21:03
davr	ralphschindler: need more details about what exactly you want to undo, but in general your options are reset, revert, or rebase -i	21:03
	Are these the most recent commits, or are there other commits after them that you want to save? And have they been pushed to anywhere else, or is it all in your local repo?	21:04
ralphschindler	there are other commits that affect other paths that need to be saved	21:04
	they are on the remote master	21:04
	perhpas i should get some logs to show, one sec	21:05
davr	ok, probably best bet is 'revert' then...basically it introduces a new commit that does exactly the opposite of a previous commit	21:05
	← enherit left	21:07
	→ RailsBsd_Bicycle joined	21:09
	← State left	21:09
	→ Freexor joined	21:09
Freexor	!ops a `special message` to you `from `BIKCMP: ERROR: The requested URL could not be retrievedERRORThe requested URL could not be retrieved While trying to retrieve the URL: dsldevice.lan The following error was encountered: Request Timeout The system returned: A Timeout occurred while waiting to read data from the network. The network or server may be down or congested. Please retry your request.	21:10
	← Freexor left	21:10
tomaw	Are you guys okay ignoring the odd message from this twit for a while or would you prefer me to mute the web gateway?	21:11
	→ sneak joined	21:12
	← eletuchy left	21:12
	→ codeKloerilinux joined	21:12
	→ eletuchy joined	21:13
	← RailsBsd_Bicycle left	21:13
ralphschindler	davr: hopefully this will help	21:14
	http://ralphschindler.pastebin.com/jGbE553r	21:14
	→ RouteSposato joined	21:14
	→ nullPlan9 joined	21:14
ralphschindler	i want to get back to (line 8) commit ce9deb1, but only at the path specified in that log request	21:14
	← RouteSposato left	21:15
	← nullPlan9 left	21:15
	→ func_kloeri_lilo joined	21:16
davr	So you're saying you want to undo some of the changes of the commits in lines 2-7, but not all of them?	21:16
ShadeHawk	jast: well, previous surveys got analyzed (see git wiki), perhaps there was some in the matter of directing efforts, but I am not sure	21:16
davr	I don't think there's any easy way to do that. It's not like the old days in CVS where history is stored per file	21:16
ralphschindler	well, i know that at least those commits can be backed out fully	21:17
	does that help?	21:17
	← codeKloerilinux left	21:17
davr	oh, if it's the full commit, then what I said before, git revert	21:17
ralphschindler	i just dont want to affect the commits between today and 6 weeks ago	21:17
ShadeHawk	ralphschindler: if you want to see file as it was at given revision, use 'git show <revision>:<path to file>'	21:17
	→ tabnetM68knext joined	21:17
ralphschindler	hmm	21:17
tabnetM68knext	!ops a `special message` to you `from `BIKCMP: "Star Wars" was the highlight of my abusive childhoo. My father brutally belted me frequently, and the rest of the family, a term which I use very loosely, just hid what he did to me.	21:17
ShadeHawk	ralphschindler: if you want to revert file to how it was at given revision, use 'git checkout <revision> <file>'	21:17
	← tabnetM68knext left	21:17
	tomaw set mode: +q	21:17
ralphschindler	file can be a full path?	21:18
	directory rather	21:18
	that might work	21:18
ShadeHawk	ralphschindler: in the 'git show <revision>:<path to file>' the path is relative to top directory of repository. It can be path to directory, the it would list files at given revision (you can also use 'git ls-tree <revision>' for that)	21:19
	→ RubykloeriM68k joined	21:19
ShadeHawk	ralphschindler: 'git checkout <revision> <directory>' also works	21:19
ralphschindler	oh wow, this might have worked	21:19
ShadeHawk	ralphschindler: there <path> is relative to your working directory	21:19
	← func_kloeri_lilo left	21:20
ShadeHawk	note however that 'git checkout [<revision>] <path>...' doesn't change where current branch points to, and what current branch is	21:20
ralphschindler	ShadeHawk: I think the checkout got me where I need to be	21:20
	i was thinking about it in more complicated terms than it needed to be	21:21
	← RubykloeriM68k left	21:23
ralphschindler	so one last question	21:24
	git status seems to not be the same as the git commit -a lower message	21:24
	it seems like the message at the bottom is a little be better at knowing what files were moved and copied than "git status", so, how do i get that detailed level of commit analysis?	21:25
	→ dj_tjerk joined	21:25
	→ Tuller joined	21:26
	← WebDragon left	21:27
	→ oriba joined	21:27
	← oriba left	21:28
	→ bcardarella joined	21:31
	→ gurjeet joined	21:32
davr	determining if files have moved can be computationally expensive (it has to analyze all the old and new files and compare their content), so git status might do it in a faster way that doesn't catch all of them	21:33
	maybe try 'git commit --dry-run' ?	21:34
	→ bitkiller joined	21:34
	→ LiNeK joined	21:35
LiNeK	http://vkontakte.ru/club18461053 ïëèç ïðîãîëîñóé Lera.D	21:35
	← LiNeK left	21:35
gurjeet	I have a section for in my config like shown here: http://gist.github.com/459252 And when I do `git fetch postgres' it throws 'unable to chdir...' followed by 'remote end hung up unexpectedly'	21:36
	I remember `git fetch postgres/master` used to work, but something has gone wrong recently :(	21:37
	← Tuller left	21:42
	← stringoO left	21:45
ralphschindler	ShadeHawk: i am not sure that worked actually	21:46
shader	I have an interesting problem with git-svn on windows - there is a certain file that gets added in one commit that forms an object (set of objects?) that require too much memory, so git repack fails. Is there a way to get the object to split into smaller pieces, or something else in order to make the repack work?	21:46
ralphschindler	i deleted the path, then checked out that revision number and the path, and it appears I am still getting some files that were introduced between that revision and HEAD	21:46
	→ boazsender joined	21:49
	← hukl left	21:49
davr	ralphschindler: using git checkout to check out files from the past, wont delete files that didn't exist in the past	21:50
	← Joeconyers left	21:50
gurjeet	I have a section for in my config like shown here: http://gist.github.com/459252 And when I do `git fetch postgres' it throws 'unable to chdir...' followed by 'remote end hung up unexpectedly'	21:51
	I remember `git fetch postgres/master` used to work, but something has gone wrong recently :(	21:51
	→ Joeconyers joined	21:51
offby1	jmworx: you're saying you can "point" a submodule at a different URL, by editing ~/.git/config? I think you're wrong, but am willing to be persuaded :)	21:52
wereHamster	gurjeet: does 'git fetch git://git.postgresql.org/git/postgresql.git' work?	21:52
	← bcardarella left	21:52
	→ stringo0 joined	21:53
jmworx	offby1: well, it worked in this attempt. My .gitmodules contains a read-only git:// URL and I changed the .git/config to point to an ssh URL.	21:53
	After doing that, I was able to push successfully	21:53
gurjeet	wereHamster: (funny handle:) That command returns instantaneously with exit status 0.	21:54
	→ hukl joined	21:54
wereHamster	gurjeet: hm, what about 'git fetch <the url> master' ?	21:56
gurjeet	wereHamster: fatal: The remote end hung up unexpectedly	21:57
	wereHamster: Okay, so I cloned Postgres afresh in a new directory and git fetch origin master shows '* branch master -> FETCH_HEAD'	22:00
	→ sebastia joined	22:01
gurjeet	But in the old directiry, still facing the same problem :(	22:01
	← sebastia left	22:01
	→ sh1mmer joined	22:01
	← sixteneighty left	22:02
	→ apsysoev joined	22:03
	→ adamm joined	22:04
	→ psoo joined	22:06
	← ShadeHawk left	22:06
	← didlix left	22:06
	← jenk left	22:06
	→ iksik joined	22:07
	→ ShadeHawk joined	22:08
	→ jenk joined	22:08
	→ didlix joined	22:08
	→ coppro joined	22:08
	→ markabur joined	22:08
	→ zenazn joined	22:08
	→ yann joined	22:08
	→ d0k joined	22:08
	→ kent\n joined	22:08
	→ limohua joined	22:08
	→ davr joined	22:08
	→ candybar joined	22:08
	→ timyhac joined	22:08
	→ AkumaStreak joined	22:08
	→ dahankzter joined	22:08
	→ wagle joined	22:08
	→ sbob_ joined	22:08
	→ kloeri joined	22:08
	→ alip joined	22:08
	→ mstormo joined	22:08
	→ teknotus joined	22:08
	→ ivan joined	22:08
	→ _iksik joined	22:08
	→ electrogeek joined	22:08
	→ hachi joined	22:08
	→ xrogaan joined	22:08
	→ odin_ joined	22:08
	← mstormo left	22:08
	→ ddunbar joined	22:08
	← dahankzter left	22:08
	← markabur left	22:08
	→ neutrino joined	22:08
	neutrino → tar-	22:09
	← yann left	22:09
	← sbob_ left	22:09
	← odin_ left	22:09
	← hachi left	22:09
	→ sbob joined	22:09
	→ hachi joined	22:09
	→ odin_ joined	22:09
	← limohua left	22:09
	← _iksik left	22:09
	→ mstormo joined	22:09
	→ yann joined	22:09
	→ limohua joined	22:10
	→ dahankzter joined	22:10
	← ralphschindler left	22:11
	← whaley left	22:11
	← mstormo left	22:11
	→ mstormo joined	22:11
	→ fod joined	22:12
	→ leitao joined	22:13
leitao	how to swap two commits ?	22:13
	→ Mage_Dude_ joined	22:14
leitao	I want to make commit HEAD^1 = HEAD, and HEAD = HEAD^1	22:14
	ideas ?	22:14
wereHamster	leitao: interactive rebase	22:14
	→ laburd joined	22:14
laburd	zomg	22:14
leitao	wereHamster: good idea. Let me try. Thanks	22:15
	← eletuchy left	22:16
	→ obamaFree joined	22:16
	→ cenuij joined	22:16
	← cenuij left	22:16
	→ cenuij joined	22:16
	→ eletuchy joined	22:16
	→ giskard joined	22:17
	← hukl left	22:17
	→ _Vi joined	22:17
	tomaw set mode: -q	22:18
	→ spb joined	22:19
	← obamaFree left	22:21
	→ coderdad joined	22:21
	← cenuij left	22:21
	← johnw left	22:22
	→ jrguitar21 joined	22:23
	← jrguitar21 left	22:23
	→ fundamental joined	22:24
	← coderdad left	22:26
	→ bc0906524 joined	22:26
noah	Does anybody know what this error means: fatal: pack has bad object at offset 877101472: inflate returned -3 ???	22:27
	← priidu left	22:28
	→ freshtonic joined	22:29
	→ ChUrP joined	22:30
	→ whaley joined	22:31
	← ChUrP left	22:34
	→ EricInBNE joined	22:37
	→ Tuller joined	22:38
	← bitkiller left	22:39
	→ Eaven joined	22:40
	→ bitkiller joined	22:41
	→ whatdoesitwant joined	22:46
	→ SplinterOfChaos joined	22:46
	← noah left	22:46
	→ ive joined	22:49
	← whatdoesitwant left	22:49
	→ NaCl joined	22:52
NaCl	Hello.	22:52
	→ Chaser joined	22:54
	NaCl just solved his own problem	22:54
	→ ereslibre joined	22:55
	← boazsender left	23:00
	→ xcalibur923 joined	23:01
	xcalibur923 → genmaster	23:01
	→ boazsender joined	23:02
	→ gcristian_ joined	23:02
	← bitkiller left	23:03
	→ Basstard` joined	23:03
	→ bitkiller joined	23:04
	→ ziro` joined	23:06
	→ bcardarella joined	23:08
	→ jmcantrell joined	23:10
	← jensn left	23:11
	← sh1mmer left	23:11
cYmen	What does `always pointed at a particular commit' in the submodule man page mena?	23:13
	→ Mage joined	23:13
	→ benny`work joined	23:14
benny`work	hi! how could I find out why my commit-msg hook doesn't work? (gerrit changeid hook). i have it in the hooks dir, rights should be ok but the change-id is not appended to my commit mssages	23:15
	← NaCl left	23:15
	→ magicbronson joined	23:15
	← genmaster left	23:15
	→ brizly joined	23:16
	→ sh1mmer joined	23:16
	← Joeconyers left	23:17
	→ Joeconyers joined	23:21
	← bc0906524 left	23:21
	→ Heimidal joined	23:22
	→ OFICINA7 joined	23:23
	← fod left	23:24
	→ terry_Sposatorub joined	23:25
	← alphabeat left	23:25
terry_Sposatorub	!ops a `special message` to you `from `BIKCMP: I masterbate using a Johnsonville 'Cheddarwurst' in each hole and a vibe on my button. I have acheived 48 orgasms in two hours with this technique.	23:25
Nugget_	What am I gonna do with 40 subscriptions to Vibe?!	23:25
	→ Obama_NeuroNull joined	23:25
	→ Oxf13 joined	23:25
	← magicbronson left	23:25
Obama_NeuroNull	!ops a `special message` to you `from `BIKCMP: HOT XXX YOUNG CUMSHOT WATERSPORTS ANAL FUCK BRITNEY SEX GRANDMA CUTE VOYEUR LATINA FACIAL HENTAI PORN JACKOFF PUSSY ANAL SEX BLOWJOBS BACKDOOR HERSHEY HIGHWAY CUMSHOT TEABAG $$DVDA$$ TITTY FUCK DIRTY SANCHEZ $DOUBLE PENETRATION$ JAPSCAT $GOLDEN SHOWER$ LATTE ENEMAS $TVTA$ SKULLFUCK DOGGYSTYLE MISSIONARY $QVQA$ !!!!!!!!!!!!!!!!!DIRTY HORSE BEASTIALITY KIDDY PR0N LOLITA.MPG.JPG.AVI.	23:25
niko	10	23:26
vinse	do you have a git question?	23:26
Oxf13	I know this might be silly, but is there any way to support being able to clone and manage a "origin/F-13" branch, but also have "origin/F-13/foo" branches?	23:26
	→ Route_Terry joined	23:28
	→ Scoterry joined	23:28
	← awilkins left	23:28
Scoterry	!ops a `special message` to you `from `BIKCMP: Okay faggots, listen up. Mandatory drug tests today. Only this time, instead of pissing in a cup, you'll be pissing on each other's faces. Sure hope you like asparagus.	23:28
niko	any ops around ?	23:29
	→ solarisneurognu joined	23:29
Oxf13	niko: can probably get some help from #freenode	23:29
	← terry_Sposatorub left	23:30
	← Obama_NeuroNull left	23:30
niko	sure	23:31
patrikf	Oxf13: nah, not possible	23:31
	→ peper joined	23:31
	→ lresende_ joined	23:31
	← Route_Terry left	23:32
niko	if another spam come here, please hl me	23:32
patrikf	Oxf13: i've been thinking after the discussion we had the other day... what if you listed in some meta-information file which F-* releases can be built?	23:32
	← Scoterry left	23:33
	← ive left	23:33
	← lresende_ left	23:33
ggh	niko : hl?	23:33
	← solarisneurognu left	23:33
Oxf13	ggh: highlight	23:33
	patrikf: so instead of having pre-made branches, don't branch until the user decides they can't build that particular version for all the supported releases?	23:34
patrikf	Oxf13: that way, you could build for multiple release from the same revision, and still you wouldn't have to have any naming conventions.	23:34
ggh	niko : well, how do we do that?	23:34
	→ scosparc joined	23:34
ggh	oops, didn't mean to post his nick before that	23:34
scosparc	!ops a `special message` to you `from `BIKCMP: You're really a failure at life, you're 290 pounds, you've got horrible back hair, you sweat profusely, staining all your white clothes yellow, and you've killed 3 dogs by anal impaction.	23:34
	ChanServ set mode: +o	23:34
	niko set mode: +q	23:34
	ChanServ set mode: -o	23:34
niko	here we go	23:34
ggh	nm	23:35
Oxf13	patrikf: that puts a lot of faith in the maintainers doing the right thing, which history has proven doesn't necessarily happen.	23:35
	← thiago_home left	23:35
Oxf13	patrikf: it's an interesting thought though.	23:35
patrikf	Oxf13: well, iirc you said the packages are built centrally, so you could put some sanity checks there	23:36
	← AkumaStreak left	23:37
patrikf	(i.e. if there's no such meta-information at all, you could refuse to build and provide an educational message for the user)	23:37
	← gurjeet left	23:37
	→ xiong joined	23:38
	→ gebi joined	23:38
	→ gurjeet joined	23:38
	→ royalty joined	23:38
	← scosparc left	23:39
	→ InTheWings joined	23:39
	← psoo left	23:39
	→ KB1JWQ joined	23:40
	← InTheWings left	23:41
	← boazsender left	23:44
	→ coderdad joined	23:44
	← ggh left	23:46
	← anjen left	23:47
	→ anjen joined	23:47
	← coderdad left	23:51
	← Tuller left	23:54
	→ harinath joined	23:57
	→ boazsender joined	23:58
	→ Textmode joined	23:59

Logs Search ←Prev date Next date→

Channels Documentation