IRCloggy #git 2009-08-22

Provider of IRC logs since 2005.

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

	2009-08-22
	→ bugfux joined	00:00
	→ rektide joined	00:02
	← voker57 left	00:03
	→ voker57 joined	00:04
	→ _Vi joined	00:05
	→ eletuchy joined	00:05
	← ctusar left	00:06
	→ Theravadan joined	00:06
	← riverrat66 left	00:06
	← hobodave left	00:07
	← voker57 left	00:08
	→ voker57 joined	00:08
	→ kwatz joined	00:09
	gebi_ → gebi	00:10
	← qhoxie left	00:10
	← scroogey left	00:13
	← bugfux left	00:13
	→ tonguero_ joined	00:13
	← tonguero_ left	00:13
	← voker57 left	00:14
	→ voker57 joined	00:15
	← agile left	00:15
	← joshsdc left	00:17
	→ mgualt joined	00:18
	← kibombo left	00:18
	← mgualt left	00:18
	→ kibombo joined	00:18
jonthefly	murbank: Try git-format-patch	00:19
	← jceb left	00:19
	bleything → bleything\|away	00:19
jonthefly	err...	00:19
	sorry	00:19
	← voker57 left	00:19
jonthefly	nevermind you want the history	00:19
	→ voker57 joined	00:19
	← fualo left	00:22
	← dpino left	00:22
	← kukks left	00:23
	← voker57 left	00:23
	← tedoc2000 left	00:24
	→ voker57 joined	00:24
	← kibombo left	00:24
	→ kibombo joined	00:24
	→ mgualt joined	00:25
	← justin-george left	00:25
	→ fualo joined	00:27
	← robinr left	00:27
	← voker57 left	00:28
	→ voker57 joined	00:29
	← tongueroo left	00:29
	→ xckpd7 joined	00:32
	← voker57 left	00:33
	← eletuchy left	00:35
	← kibombo left	00:35
	← mgualt left	00:35
	→ voker57 joined	00:36
	← kwatz left	00:37
	→ kibombo joined	00:38
	← kibombo left	00:39
	→ kibombo joined	00:39
	← kibombo left	00:40
	→ faithfulgeek_ joined	00:40
	→ kibombo joined	00:40
	← bgerber left	00:41
	→ juanjoc joined	00:42
	← gonzojive__ left	00:42
	← Ryback_ left	00:43
	→ bgerber joined	00:43
	→ eletuchy joined	00:43
	← eletuchy left	00:44
	→ scroogey joined	00:47
scroogey	hello	00:47
	← voker57 left	00:47
	→ voker57 joined	00:48
	← dreiss left	00:48
	← faithfulgeek left	00:48
	faithfulgeek_ → faithfulgeek	00:49
	← voker57 left	00:52
	→ voker57 joined	00:52
	→ rapha joined	00:53
rapha	hi	00:53
	→ gonzojive_ joined	00:53
	← voker57 left	00:55
rapha	when it comes to keeping a website in a repository, would you say Git is the more sensible choice over Subversion or just to stick with the latter?	00:55
	→ voker57 joined	00:55
scroogey	no idea, only used svn a few times	00:56
	→ agile joined	00:57
	→ johnnyg1 joined	00:59
	← voker57 left	00:59
	→ voker57 joined	01:00
	→ Cromulent joined	01:00
	→ aspotashev\|eeepc joined	01:01
	← altrux left	01:02
	→ altrux joined	01:02
	← cedricv left	01:03
	← porcnick left	01:03
	← reidab left	01:03
rapha	scroogey: would you feel comfortable doing something like that with git?	01:04
	← voker57 left	01:04
	← kibombo left	01:04
rudi_s	rapha: I have used both Git and Subversion, I prefer Git but for a simple website it shouldn't matter which you use. Except if you have SSH access (not only FTP) then Git has advantags as you can directly push to the version and the live page is updated.	01:04
	→ kibombo joined	01:04
rudi_s	s/version/version online/	01:05
rapha	rudi_s: yeah i have ssh ... so that's a real advantage then.	01:05
	rudi_s: would it handle symlinks gracefully?	01:06
	← crack_head left	01:06
rapha	and ownership, rights, etc?	01:06
rudi_s	rapha: Ownerships are a bit of a problem (git doesn't track them), but you could use a hook to set them correctly.	01:07
	Symlincs are handled correctly.	01:07
rapha	okay. should be good enough. the pushing thing makes me curious at the least.	01:07
	→ voker57 joined	01:08
	→ joshsdc joined	01:09
	→ akitada joined	01:09
	→ ab3 joined	01:11
PHLAK	rapha: check this link out	01:11
	rapha: http://joemaller.com/2008/11/25/a-web-focused-git-workflow/	01:11
ab3	hoi	01:11
PHLAK	rapha: this is how I manage my personal websites and how I do it at my work	01:12
	← voker57 left	01:12
rapha	wow! that looks pretty cool PHLAK!	01:12
	thanks :)	01:12
	and how did you make that drawing? is it just skill and you scanned it in?	01:12
	→ voker57 joined	01:12
ab3	ok, I have a strange problem with git, if I add changes to my local repo with commit -a -m "blabla" and then push them to my remote repo it returns Everything up-to-date	01:12
	But I don't see any changes when I visit my github account	01:13
PHLAK	rapha: at my work we even abstract from that and have a separate development and production branch and only pull to the live directory when we're sure we're good	01:13
	rapha: but I'd just stick with that for now	01:13
rapha	well, i'm setting up our new company server, and we'll probably have staging, test and production environments	01:14
PHLAK	ab3: try "git push origin master"	01:14
rapha	i think i'm gonna print that article of yours out and put it on the bedside table	01:14
PHLAK	rapha: lol, I didn't write that, but it's proven to work fairly well	01:15
rapha	oh okay :)	01:15
ab3	PHLAK: That is what I did, it just return Everything up-to-date, and no changes in github...	01:15
PHLAK	ab3: do "git remote show origin" and make sure you see your github repo there	01:16
rapha	i did it with svn on the last server and it proved to be somewhat of a chore with things becoming out of sync and what have you	01:16
PHLAK	rapha: I've never used SVN, but my boss has, and after using Git he said he'd never go back to SVN	01:16
rapha	yeah thats what ppl said about CVS => SVN back then	01:17
ab3	yep, [email@hidden.address] is in git remote show origin	01:17
rapha	and also about RCS => CVS _way_ back then lol	01:17
PHLAK	hmmm...	01:17
rapha	anyway off to bed and thx for the link again!	01:18
PHLAK	welcome	01:18
	→ ereslibre joined	01:18
PHLAK	ab3: and you've generated and added your ssh key?	01:18
	← kennethlove left	01:19
ab3	yes, it worked fine untill I fetched and dcommited my work to the original svn trunk	01:19
	but I can't even do that now...	01:19
	wang_ → wang	01:19
	← voker57 left	01:20
PHLAK	try creating a new folder/repo	01:20
	pulling that repo down	01:20
	make a small change and commit/push that	01:20
	→ voker57 joined	01:21
ab3	ok but what about my current changes?	01:21
PHLAK	thinking...	01:22
	if you KNOW what you have locally is good	01:22
	→ crack_head joined	01:22
PHLAK	you can allways do "git push origin master -f" to force an overwrite of the remote repo with your local repo	01:23
	← voker57 left	01:25
	→ ryoma joined	01:25
	→ voker57 joined	01:25
ryoma	how come sha-1 are printed in base 16 instead of, say, ascii85?	01:25
	← scroogey left	01:25
	← spaceonline left	01:26
	→ EricInBNE joined	01:26
	← chris2 left	01:26
	→ Evanlec joined	01:27
	→ veiz joined	01:27
ab3	adding the -f flag doesn't change anything...	01:29
	← voker57 left	01:29
	← bradly left	01:29
	← ericindc left	01:30
	← wshimmy left	01:31
	→ mrlemao joined	01:32
	← aspotashev\|eeepc left	01:32
	→ voker57 joined	01:33
	→ mm_202 joined	01:33
	← Cromulent left	01:34
	→ wshimmy joined	01:35
	→ _bsy_ joined	01:36
	← _bsy_ left	01:36
	← voker57 left	01:37
	→ Adlai joined	01:38
	→ voker57 joined	01:39
	→ tongueroo joined	01:40
	← voker57 left	01:40
	→ _bsy_ joined	01:41
	→ sinsun joined	01:41
	sinsun Free Chinese Class is still welcome Linux users and friends. Join #sinsun-says	01:42
	→ voker57 joined	01:42
	← ereslibre left	01:42
	→ levifig joined	01:43
	← levifig left	01:45
	→ levifig joined	01:46
	← voker57 left	01:46
	→ voker57 joined	01:47
	← uau left	01:47
	← tongueroo left	01:48
	← voker57 left	01:50
	← mrlemao left	01:50
	← gonzojive_ left	01:50
	→ voker57 joined	01:50
	→ uau joined	01:50
	→ aspotashev\|eeepc joined	01:51
	→ danferranti joined	01:53
	← _bsy_ left	01:53
	→ Krynnotaur joined	01:53
	← voker57 left	01:54
	← Krynnotaur left	01:54
	← sinsun left	01:55
	← danferranti left	01:55
	→ Krynnotaur joined	01:55
	← levifig left	01:56
	→ gonzojive_ joined	01:56
	→ voker57 joined	01:57
	← AkumaStreak left	01:58
	← voker57 left	02:01
	← hzlocky left	02:02
	→ voker57 joined	02:02
	← lohrun left	02:02
	← frede left	02:03
	→ EiNZTEiN joined	02:05
	→ danferranti joined	02:05
	← Theravadan left	02:06
	→ scarabx joined	02:07
	→ gonzojive__ joined	02:07
	→ ctusar joined	02:07
	← voker57 left	02:08
	→ voker57 joined	02:09
	← Fissure left	02:10
	← spearce left	02:12
	→ mbrumbelow joined	02:13
	→ alezandro joined	02:13
	→ _bsy_ joined	02:14
	← voker57 left	02:14
	← PHLAK left	02:14
	← aspotashev\|eeepc left	02:14
	→ voker57 joined	02:15
	← benn_d left	02:15
	→ AkumaStreak joined	02:16
	← danferranti left	02:16
	← EiNZTEiN left	02:17
	→ EiNZTEiN joined	02:18
	← voker57 left	02:19
	→ voker57 joined	02:19
	← murbank left	02:20
	→ cytrinox_ joined	02:21
	← cytrinox left	02:21
	cytrinox_ → cytrinox	02:21
	← fualo left	02:21
	→ qmx joined	02:22
	← WALoeIII left	02:23
	→ DrAk0 joined	02:23
	← voker57 left	02:23
	→ voker57 joined	02:24
	← gonzojive_ left	02:24
	← _bsy_ left	02:28
	→ benn_d joined	02:28
	← voker57 left	02:28
	→ voker57 joined	02:30
	→ qrush joined	02:30
	→ benn_d_ joined	02:33
	← heo left	02:33
	← richardcavell left	02:34
	← EiNZTEiN left	02:34
	← voker57 left	02:34
	→ voker57 joined	02:35
	← benn_d left	02:36
	→ zehrique_ joined	02:36
	← panpainter left	02:37
	← zehrique left	02:38
	zehrique_ → zehrique	02:38
	← voker57 left	02:39
	→ qrush_ joined	02:39
chuck	If I have a brand new, uninitialised remote repo and two local branches (master, new-wiki-list-abstract), how would I go about making new-wiki-list-abstract be the new remote master branch and the local master branch be the remote old-config branch?	02:39
	← breitz left	02:39
	← joshsdc left	02:39
	← bcardarella left	02:39
	→ jinzhu joined	02:43
	→ atreyu joined	02:44
ab3	is it normal that when you do git svn fetch blabla that is fetches every previous revision?	02:45
	← atreyu left	02:45
	← SjB left	02:45
	← sergiodj left	02:45
	→ SjB joined	02:46
	← mm_202 left	02:46
offby1	of course	02:49
	assuming you didn't already _have_ those revisions.	02:49
	← DrAk0 left	02:49
offby1	If you _did_ already have 'em, then that's a whopping huge bug.	02:49
	→ voker57 joined	02:52
	→ tjafk2 joined	02:52
	← memiux left	02:53
	← qmx left	02:53
	→ atreyu joined	02:54
	← atreyu left	02:55
	→ panpainter joined	02:56
	← voker57 left	02:58
	→ voker57 joined	02:58
	← qrush left	02:59
	← tobias__ left	03:01
	→ eletuchy joined	03:02
	← voker57 left	03:02
	→ voker57 joined	03:02
	→ atreyu joined	03:04
	← atreyu left	03:05
	← voker57 left	03:06
	→ voker57 joined	03:07
	← segher left	03:07
	→ Theravadan joined	03:08
	← tjafk1 left	03:09
	← akitada left	03:09
	← jinzhu left	03:10
	← voker57 left	03:11
	→ voker57 joined	03:11
	← eletuchy left	03:14
	← zehrique left	03:15
	← voker57 left	03:15
	→ voker57 joined	03:16
	← voker57 left	03:20
	→ voker57 joined	03:21
ab3	abe@higg:pygame-mirror$ git svn dcommit	03:23
	Committing to svn://abe@seul.org/svn/pygame/trunk ...	03:23
	A test_git	03:23
	Item already exists in filesystem: File already exists: filesystem '/home/svn/repos/svn/pygame/db', transaction '2639-1', path '/trunk/test_git' at /sw/lib/git/git-svn line 492	03:23
	→ joshsdc joined	03:23
ab3	any idea what goes wrong here	03:23
	I don't find the test_git file and have done a lot more changes that the test_git file	03:24
	← gonzojive__ left	03:26
	← MisterN left	03:26
	← juanjoc left	03:27
	→ kwatz joined	03:28
	→ girishr joined	03:29
	← voker57 left	03:29
	→ _bsy_ joined	03:29
	→ dreiss joined	03:30
	→ voker57 joined	03:30
	← pheaver left	03:31
	→ pheaver joined	03:32
	← pheaver left	03:32
	→ escrooge1 joined	03:34
	← voker57 left	03:34
	← panpainter left	03:34
	→ gonzojive_ joined	03:34
	← faithfulgeek left	03:34
	→ faithfulgeek joined	03:34
	← qrush_ left	03:38
	→ kmamykin joined	03:41
	← jjuran left	03:41
	→ heo joined	03:41
	→ voker57 joined	03:41
	← ab3 left	03:42
	← Theravadan left	03:45
	→ drio joined	03:46
drio	I have a project in a box... I go to another box and I clone the original repository...	03:46
	I then make some changes and commit to the local copy..	03:46
	then I can either:	03:47
	push to the origin..	03:47
	or pull from the remote in the origin..	03:47
	if I do this last one.. nothing gets pull..	03:47
	if I push to the origin..	03:47
	the changes go to the index....	03:47
	so I still have to commit...	03:47
	← joshsdc left	03:47
drio	but the log changes are already there..	03:47
Ademan	what do you mean nothing gets pulled? are your changes in a different branch?	03:47
	→ dschulz joined	03:47
drio	origin has master	03:48
	let's call the orgin machine fred	03:48
	← AkumaStreak left	03:48
drio	and the machine I clone and develope: milnouse	03:48
	sorry, milhouse	03:48
	so I do this:	03:48
	→ qrush joined	03:48
drio	1. create my git project in fred	03:48
	2. go to milhouse and clone the project in fred	03:48
	← voker57 left	03:48
drio	3. I add some changes in fred.... and I commit.. to the master branch	03:49
	then I just want to push those changes to fred...	03:49
	I can either pull from milhouse or push to fred...	03:49
	if I push...	03:49
	everything goes to the master branch in fred .. but to the index... it doesn't really get merged...	03:49
	if I pull...	03:49
	This is what I get..	03:49
	← wagle_home left	03:50
drio	http://pastie.org/591241	03:50
	→ squisher joined	03:51
	← girishr left	03:51
madewokherd	I don't know why pull is recommended	03:52
	I always find it simpler just to fetch and do things with the remote branches	03:52
	→ kastner joined	03:52
	← johnnyg1 left	03:53
kastner	Alright, I know how annoying git-svn questions are so I'll be brief ;), I created my repo with git svn clone -s $SVN/site, I now have /htdocs in my root (this is right)... if I do get checkout -b erik remote/erik it will check out the branch erik ($SVN/site/branches/erik)	03:53
	← dschulz left	03:53
kastner	the problem is that INSIDE of that directory are all my SVN branches.. so if I were to merge from erik/test1 to master, I'm going to have a big problem	03:54
	so I either need to re-import all the branches somehow (with their parent structure), OR I need a way to say "ignore the parent when merging"	03:54
	→ voker57 joined	03:55
	→ wagle_home joined	03:55
	→ cloudhead joined	03:55
	← jfkw left	03:55
	← rudi_s left	03:55
kastner	humm: http://stackoverflow.com/questions/258590/how-do-i-import-svn-branches-rooted-in-different-directories-into-git-using-git-s	03:56
	→ ericindc joined	03:56
	← kmamykin__ left	03:57
	← ctusar left	03:59
	→ srid joined	04:01
	→ richardcavell joined	04:03
	→ AkumaStreak joined	04:03
	← AkumaStreak left	04:03
	← qrush left	04:03
	← enherit left	04:06
	→ aussiegeek joined	04:06
	← voker57 left	04:07
	→ voker57 joined	04:07
	← chuck left	04:11
	→ mrlemao joined	04:12
	← greboides left	04:12
	← mrlemao left	04:13
	← voker57 left	04:13
	→ voker57 joined	04:13
	→ ben_h joined	04:15
	← kastner left	04:16
	← voker57 left	04:18
	← loincloth left	04:18
	→ relix joined	04:18
	→ chuck joined	04:19
	→ voker57 joined	04:19
	→ pheaver joined	04:20
	← cloudhead left	04:22
	← voker57 left	04:23
	→ voker57 joined	04:23
	← ben_h left	04:26
	← gonzojive_ left	04:26
	→ gonzojive_ joined	04:27
	→ fualo joined	04:27
	→ ben_h joined	04:27
	← relix left	04:28
	← escrooge1 left	04:28
	← tfarina left	04:29
	← kwatz left	04:30
	→ fualo_ joined	04:31
	→ Rich_Morin joined	04:31
	← markelikalderon left	04:31
	← Rich_Morin left	04:32
	← voker57 left	04:34
	← marcelo_ left	04:36
	→ voker57 joined	04:38
	→ porcnick joined	04:39
	→ RayMorgan joined	04:39
	→ hydrogen joined	04:40
	← voker57 left	04:40
	← RayMorgan left	04:40
	→ RayMorgan joined	04:40
	→ voker57 joined	04:40
	→ codeswing joined	04:40
	→ aspotashev\|eeepc joined	04:41
	← voker57 left	04:44
	→ voker57 joined	04:45
	← fualo left	04:46
	← doener left	04:46
	→ doener joined	04:46
	← faithfulgeek left	04:47
	→ faithfulgeek joined	04:47
	← voker57 left	04:49
	→ segher joined	04:49
	→ voker57 joined	04:50
	← Bass10 left	04:52
	← gonzojive_ left	04:53
	← voker57 left	04:54
	→ voker57 joined	04:56
	→ bdowning joined	04:59
	← voker57 left	05:00
	→ voker57 joined	05:00
	← scarabx left	05:02
	← voker57 left	05:04
	→ voker57 joined	05:04
	→ meanburrito920_ joined	05:05
	← meanburrito920_ left	05:07
	→ srbaker joined	05:08
srbaker	folks	05:08
	in my repo i have a dir called foo and one called bar	05:08
	is it possible to git clone and only get stuff below foo?	05:09
aspotashev\|eeepc	no	05:09
	git is not svn	05:09
srbaker	didn't think so, but wanted to be sure before i moved everything	05:09
	thanks	05:09
	← srbaker left	05:09
	→ omenking joined	05:10
omenking	HELPEPEEPL	05:10
	← stuffcorpse left	05:11
	← voker57 left	05:12
	→ voker57 joined	05:12
	← richardcavell left	05:16
omenking	I have a branch	05:16
	and somehow I lost the commit	05:16
	I want to merge my branch into my master	05:16
	but the commit is gone and it says my master is up to date with my branch which isn't the case	05:16
	how do I force my master to accept everything in my branch?	05:17
	→ Altreus_ joined	05:17
	← voker57 left	05:17
	→ reprore joined	05:17
	← Altreus left	05:18
	→ voker57 joined	05:18
	← dodo_the_last left	05:20
	← voker57 left	05:22
	→ DavidKlein joined	05:23
	← hydrogen left	05:24
	→ voker57 joined	05:26
	→ ProperNoun joined	05:29
	→ Theravadan joined	05:31
	→ mpolesky joined	05:34
	← mpolesky left	05:35
	→ WALoeIII joined	05:36
	← voker57 left	05:36
	← altrux left	05:37
	→ voker57 joined	05:37
	← WALoeIII left	05:38
	← reprore left	05:39
	← orafu left	05:40
	→ orafu joined	05:41
	→ WALoeIII joined	05:41
	← voker57 left	05:41
	→ dpino joined	05:42
	← ben_h left	05:44
	→ dirk2 joined	05:45
	→ voker57 joined	05:46
	← Theravadan left	05:48
	← WALoeIII left	05:49
	→ WALoeIII joined	05:49
	← codeswing left	05:50
	← voker57 left	05:51
	→ voker57 joined	05:51
	← pheaver left	05:52
	→ scientes joined	05:54
	← voker57 left	05:54
	→ stuffcorpse joined	05:57
	→ reprore joined	05:58
	→ voker57 joined	05:58
	← voker57 left	06:02
	→ voker57 joined	06:03
	→ panpainter joined	06:03
dirk2	Can anybody help with a hopefully easy newbie rebase question? If this is a FAQ, a link with a good description would already help ;)	06:04
	http://pastebin.com/m1fa325e8	06:04
	→ pheaver joined	06:04
dirk2	I haven't figured out how to push rebase to an upstream remote repository	06:05
crab	dirk: push -f	06:06
dirk2	crab: Yes, push -f does work	06:06
	← voker57 left	06:07
dirk2	crab: but is this the intended way?	06:07
	crab: it seems to me that people pulling from the remote afterwards have issues	06:07
	← DavidKlein left	06:07
	→ sitaram joined	06:08
crab	yes, they will. you should never rebase things other people can see.	06:08
	→ voker57 joined	06:08
crab	(or you should announce that you're going to rebase branch x so they should be prepared for it)	06:08
	but basically, avoid rebasing commits that have been published to other people.	06:08
	→ danferranti joined	06:08
dirk2	crab: prepared for it: So they have to throw away their old stuff and do a new clone?	06:09
crab	(the people who are having problems after pulling can do "git reset --hard origin/work" to recover and set their work branch to what you push -f'ed)	06:09
	dirk: oh no, no need to re-clone. just reset --hard.	06:10
dirk2	crab: ah, ok :)	06:10
	crab: never rebase: so instead, merging should be done?	06:10
	← voker57 left	06:12
crab	dirk2: it depends on what you're trying to do. sometimes a branch that everyone agrees can be rebased is the right thing to do (but given enough people, someone is sure to make a mistake at some point and you'll have to put in effort to sort it out)	06:12
	→ Deradner joined	06:12
	→ voker57 joined	06:12
	→ bugfux joined	06:13
crab	merging work into master every now and then avoids that problem, and may or may not be more appropriate for you. i don't know.	06:13
	← crack_head left	06:13
dirk2	crab: Well, basically I have X - Y - Z on top of linux-stable (e.g. 2.6.30.4). Now, if 2.6.30.5 comes, I like to have X - Y - Z on top of 2.6.30.5	06:13
	← danferranti left	06:14
dirk2	... and provide the result to everybody	06:14
	← Adlai left	06:15
	→ Adlai joined	06:15
	← __Joker left	06:16
	← voker57 left	06:16
	→ voker57 joined	06:17
	joshthecoder → joshthecoder_afk	06:20
	← WALoeIII left	06:21
	→ WALoeIII joined	06:21
	← voker57 left	06:23
	→ voker57 joined	06:24
	← Derander left	06:24
milluminu	ls	06:26
	← voker57 left	06:28
	→ mithro joined	06:31
	→ voker57 joined	06:33
	← khmarbaise_ left	06:34
	← squisher left	06:39
	→ WALoeIII_ joined	06:39
	← voker57 left	06:39
	→ voker57 joined	06:40
	← pheaver left	06:41
dirk2	Back again ;) So what's the best way to maintain X - Y - Z on top of e.g. linux stable like described above and provide the result to everybody remote/upstream? If not using rebase and push -f ?	06:42
	→ karmi joined	06:43
	→ s34n_ joined	06:44
	← voker57 left	06:44
	→ voker57 joined	06:44
s34n_	I would like to add a file from a different directory than the one that contains .git/	06:45
	the file is in the /etc directory	06:45
	but I'd really like to include it in my git repo	06:46
	s34n_ → s34n	06:48
	← voker57 left	06:48
s34n	I would like to add a file from a different directory than the one that contains .git/	06:49
	the file is in the /etc directory	06:49
	but I'd really like to include it in my git repo	06:49
	→ voker57 joined	06:49
	← _bsy_ left	06:49
	→ _bsy_ joined	06:49
	← _bsy_ left	06:50
	← porcnick left	06:51
	← SjB left	06:51
	← voker57 left	06:52
	→ voker57 joined	06:52
	← WALoeIII left	06:52
	WALoeIII_ → WALoeIII	06:52
	→ panpaint-1 joined	06:53
	← panpaint-1 left	06:53
	← dreiss left	06:56
	→ panpaint-1 joined	06:57
	← panpaint-1 left	06:57
	← voker57 left	06:59
	→ tongueroo joined	06:59
	→ voker57 joined	06:59
	→ thiago_home joined	07:00
	→ jinzhu joined	07:00
	← panpainter left	07:03
	→ ben_h joined	07:03
	→ bcardarella joined	07:03
	Deradner → Derander	07:06
	→ d0k joined	07:08
	→ porcnick joined	07:08
	← bcardarella left	07:09
	→ raoul joined	07:09
	→ jceb joined	07:14
	← tongueroo left	07:15
	Krynnotaur → Divestiture	07:16
	← ben_h left	07:17
	→ xyz joined	07:18
	→ Guest80557 joined	07:18
	Guest80557 → ben_h	07:19
	→ dreiss joined	07:19
	← srid left	07:24
	← juan_arandaalvar left	07:27
	← veiz left	07:28
	→ priidu joined	07:28
	← Dashkal left	07:29
	← mithro left	07:31
	→ RaceCondition joined	07:32
dirk2	bbl. If anybody has a hint regarding the rebase topic discussed above, I will read the logs ;) Thanks!	07:35
	← aspotashev\|eeepc left	07:35
	→ aspotashev\|eeepc joined	07:35
	← dirk2 left	07:36
	→ harinath joined	07:36
	← porcnick left	07:37
	← RayMorgan left	07:39
	→ gonzojive_ joined	07:40
	← gonzojive_ left	07:40
	→ ADmad joined	07:41
ADmad	hello, is it possible to get syntax highlighting for php files under cgit ?	07:42
	→ smerrell joined	07:42
	← xyz left	07:42
milluminu	thats silly	07:44
	← s34n left	07:44
	→ vals_ joined	07:44
ADmad	milluminu: was that for me ?	07:44
	davegb3\| → davegb3	07:45
milluminu	because git has nothing to do with any particulr language	07:45
	but yeah technically it is	07:45
	← tango_ left	07:45
	vals_ → tango_	07:45
	→ Modius_ joined	07:46
	→ onigiri joined	07:46
milluminu	you could just pipe it	07:46
ADmad	milluminu: i guess you didnt read my question properly.. i am talking about cgit the web interface for git	07:46
doener	milluminu: cgit is a web-interface which can show the blobs to the user. Asking for syntax highlighting in that context isn't really stupid	07:46
milluminu	git show master:dir/file \| some prettifier	07:46
	← jceb left	07:46
milluminu	sorry i missed that admad	07:46
	← onigiri left	07:47
	→ onigiri joined	07:48
	← Modius_ left	07:48
	→ Modius joined	07:48
	← omenking left	07:49
Aides	nice idea, btw	07:50
doener	ADmad: there's some script in filters/ that uses "highlight" for the highlighting	07:52
	ADmad: http://hjemli.net/git/cgit/tree/filters/syntax-highlighting.sh?id=6445a3ad0987ba66eef555c9caa3fc378f99ee22	07:52
	ADmad: according to the command's home page (apparently in german only), it also supports php, so just adding another entry to the case switch should do	07:53
ADmad	doener: yes i found there's a "source-filter::" option in config, so need to figure out what value to set for it	07:53
doener	ADmad: should have the path to the filter script AFAICT	07:54
	ADmad: see the comments in http://hjemli.net/git/cgit/tree/filters/syntax-highlighting.sh?id=6445a3ad0987ba66eef555c9caa3fc378f99ee22	07:54
ADmad	doener: yes i am checking it out, thank you	07:55
doener	oh, heh, the highlight command actually also has english docs: http://www.andre-simon.de/doku/highlight/en/highlight.html	07:56
ADmad	ah thats saves me using google translate :)	07:56
	where can i check the logs for this channel?	07:57
doener	see topic	07:57
	though I usually just google "git irc log", first hit	07:58
	→ SeySayux_ joined	07:58
SeySayux_	Hello, how do I import from SVN?	07:58
ADmad	doener: got it. thx	07:58
	← harinath left	07:59
\|jon\|	SeySayux_: look at git-svn	07:59
SeySayux_	Okay, thanks	07:59
	→ harinath joined	08:01
	← voker57 left	08:05
	→ voker57 joined	08:06
	← smerrell left	08:09
	← onigiri left	08:09
	← voker57 left	08:10
	→ voker57 joined	08:10
	← dreiss left	08:10
	← releod left	08:12
	→ patrikf joined	08:12
	← voker57 left	08:14
	← RaceCondition left	08:14
	→ voker57 joined	08:15
	→ ianmcorvidae\|alt joined	08:17
	→ onigiri joined	08:17
	→ poige joined	08:18
poige	Hi! Suppose, I've commited ver 2.0, then 3.0 and 4.0. Now I want to have commited ver 1.0 and rebase it to be the first commit (instead of ver 2.0). How can I?	08:19
	→ bcardarella joined	08:20
	← onigiri left	08:21
	← voker57 left	08:21
	← Derander left	08:21
	→ onigiri joined	08:21
	→ voker57 joined	08:21
	← segher left	08:22
	← Divestiture left	08:22
	→ myabc joined	08:23
	→ hzlocky joined	08:24
sitaram	doener: http://pastebin.com/d56010bf (finding dropped/lost stashes again!)	08:25
	doener: does that look good?	08:26
doener	sitaram: http://pastebin.com/m1dbacb8b	08:28
poige	Hi! Suppose, I've commited ver 2.0, then 3.0 and 4.0. Now I want to have commited ver 1.0 and rebase it to be the first commit (instead of ver 2.0). How can I?	08:28
doener	sitaram: and calling log multiple times is bad	08:28
	sitaram: if the pager is run for each one, you have to quit each one	08:28
	→ mbroeker joined	08:29
doener	sitaram: and if, for example, the first two don't trigger the pager, but the third one does, you likely won't see the first two at all	08:29
	sitaram: or rather "if less decides that paging is not required for the first two, but for the third one"	08:30
sitaram	doener: ok wait wait; I mainly wanted confirmation of the operative "[[...]]" line...	08:30
doener	sitaram: heh, didn't check that for syntax or so, I fail at proof-reading bash scripts	08:31
sitaram	doener: not shell syntax; the ^ == ^2^ bit	08:31
	does that describe a triangular commit properly (seems it does, from your paste)	08:31
	← bcardarella left	08:31
SeySayux_	Git was made by Linus Torvalds, right?	08:31
doener	that confirms the triagle, but ignores all commits except for A B Y	08:32
	← voker57 left	08:32
	→ voker57 joined	08:32
sitaram	doener: that's ok; typically I only want to see B	08:32
doener	sitaram: well, it ignores X, which must not exist for a possible stash entry	08:32
	sitaram: the shape is uncommon for git, (as B would not happen, could have fast-forwarded from A to Y), but common for svn imports	08:34
	← ankit9_ left	08:34
sitaram	doener: does that shape come up often (I guess my branches dont merge back and forth immediately so I never saw it, mine have more commits in between when I do that)	08:34
	doener: svn does not exist in my world ;-)	08:34
doener	sitaram: may also happen in environments where people prefer --no-ff merges	08:35
	← ben_h left	08:35
doener	sitaram: e.g. devA gets his brach up to date by merging, then upstream pulls from him with --no-ff	08:35
	ADmad → ADmad\|away	08:36
sitaram	doener: hmm ok; but if I wanted to help a newbie who lost a stash, this will at least give him a good set of candidates, and the messages can be used to further (mentally) filter	08:36
doener	sitaram: how much better than the --grep is that?	08:37
	← voker57 left	08:37
	→ voker57 joined	08:37
sitaram	doener: dont know; depends whether things like --no-ff are more indicative of an expert of using a message on a stash	08:38
	a/of/or	08:38
	s/of/or (damn!)	08:38
doener	s/of/or/2 I guess? (sed syntax)	08:38
sitaram	:) thanks!	08:38
doener	sitaram: and the --grep _does_ capture stash entries with custom messages	08:39
	← poige left	08:39
	→ ben_h joined	08:39
sitaram	doener: hmm? because of the "on branch"?	08:39
doener	sitaram: no the "(on branch)" is when you are on a detached head	08:40
	sitaram: non-custom message: WIP on $name: $commit_message_of_$name	08:40
	sitaram: custom message: On $name: $custom_message	08:41
sitaram	aah	08:41
doener	sitaram: so it's the "$WIP on \\|On $" part that matters	08:41
sitaram	doener: followed by the non-space\+:	08:41
	ok	08:42
	well never mind then, scratch all this... :(	08:42
doener	that's required in any case, the non-space replaces "(no branch)" when you were on some branch when you stashed	08:42
	← voker57 left	08:42
crab	hi.	08:43
doener	hey crab	08:43
	→ voker57 joined	08:43
doener	sitaram: http://git.pastebin.com/m6087a64	08:46
	sitaram: the --merges thing is also helpful. While regular commits might say "On foo: fjksfjsdlf", merge commits are very very unlikely to do that	08:47
sitaram	doener: ok	08:47
	← voker57 left	08:48
	→ voker57 joined	08:49
	← reprore left	08:50
	→ ankit9_ joined	08:52
	← krawek left	08:52
	← Evanlec left	08:53
	← hbunny left	08:53
	→ reprore joined	08:53
	← reprore left	08:53
	← voker57 left	08:57
	→ voker57 joined	08:57
	← onigiri left	08:57
	→ ph^ joined	08:58
	→ lohrun joined	08:58
	→ HG` joined	08:59
	→ onigiri joined	08:59
	→ ben_h_ joined	09:02
	← voker57 left	09:03
	← ben_h left	09:03
	→ voker57 joined	09:03
	→ Sigma joined	09:05
	→ Chillance joined	09:06
	← fualo_ left	09:07
	→ JNR___ joined	09:11
	← voker57 left	09:11
	→ voker57 joined	09:11
	← patcito left	09:12
	← ryoma left	09:12
	← WALoeIII left	09:12
	→ Tuomas joined	09:14
	← voker57 left	09:15
	→ voker57 joined	09:16
	← heo left	09:19
	→ roberthorrey joined	09:22
roberthorrey	so i committed something to my branch with the wrong message. any way to change it retroactively?	09:22
wereHamster	roberthorrey: git commit --amend	09:23
roberthorrey	ah, perfect, thanks	09:23
	← _Vi left	09:28
	← rubyist left	09:38
	→ rubyist joined	09:40
	← ToxicFrog left	09:40
	→ ToxicFrog joined	09:41
	→ SuperRoach joined	09:41
SuperRoach	lol, love the topic	09:41
roberthorrey	the google summer of code project looks neat	09:45
	it's going to make my scm of choice even better :)	09:45
	← HG` left	09:45
	→ robinr joined	09:47
	→ __iron joined	09:51
	← tokkee left	09:54
	← Altreus_ left	09:57
	→ frede joined	10:00
	← lu_zero left	10:01
	→ Evanlec joined	10:08
	← voker57 left	10:09
	← Strogg left	10:09
	→ voker57 joined	10:09
	→ therrg joined	10:09
	→ Strogg joined	10:10
	→ arachnist joined	10:13
	→ sitaram_ joined	10:13
arachnist	hi	10:13
	i've probably got a lame question, but, oh well	10:13
	→ tbuehlmann joined	10:15
	→ HG` joined	10:15
doener	arachnist: so far, there's no question at all	10:15
arachnist	i've added a remote for a git repo, but it doesn't seem to work as intended. "git pull sparc" (where sparc is the second remote) gives me this: http://dpaste.com/83928/	10:15
doener	arachnist: well, you didn't tell it what to merge	10:16
arachnist	but i already have those configuration stuff	10:16
	remote.sparc.url=arachnist@khas:scm/arbor	10:16
	remote.sparc.fetch=+refs/heads/master:refs/remotes/sparc/master	10:16
doener	arachnist: maybe you meant "git fetch sparc"?	10:16
arachnist	hmm	10:16
doener	arachnist: pull is fetch + merge/rebase	10:16
arachnist	well, i want to merge all the changes from "sparc"	10:17
doener	arachnist: "sparc" is a remote, which has possibly multiple branch heads, so you need to tell pull which one you want to merge	10:18
	→ PaNtH3R joined	10:18
doener	arachnist: e.g. "git pull sparc master"	10:18
arachnist	oh	10:18
	← sitaram left	10:18
PaNtH3R	hi ... all.. I have one noob question	10:18
doener	(this won't update the remote tracking branch, as it by-passes the configured default fetch refspec)	10:18
PaNtH3R	Is it possible to auto pull/fetch submodules when pulling/fetching master?	10:19
arachnist	doener: thanks, "git pull sparc master" works. essentially, "git pull sparc '+refs/heads/master:refs/remotes/sparc/master'" also did, but was too long to remember	10:20
doener	arachnist: the difference is that the latter also updates the remote tracking branch	10:21
arachnist	good to know	10:21
doener	arachnist: if you set branch.<name>.{remote,merge} to sparc and refs/heads/master respectively, just "git pull" will also work, and update the remote tracking branches	10:22
	← JNR___ left	10:22
	→ sitaram joined	10:22
	→ Weasel[DK] joined	10:24
	← HG` left	10:25
	→ cao joined	10:32
	← rubyist left	10:32
	cao → Guest94814	10:33
	→ rubyist joined	10:33
	← sitaram_ left	10:35
	→ pielgrzym joined	10:37
	← bugfux left	10:39
	← rubyist left	10:39
pielgrzym	hi there :) I'm quite new to git :) I set up a remote bare repo etc. created proper git configs and .gitignore files. In .gitignore I specified .vbk which is a folder holding vim swapfiles etc. it's nearly in each of the projects folders. Is there a way to tell git to remove all instances of .vbk directory recursively all accross the project?	10:39
	→ rubyist joined	10:39
	← myabc left	10:40
pielgrzym	I mean the file is ignored, but I wanna get rid of the preexisting .vbk folders in my remote git repo :)	10:40
davegb3	pielgrzym: put .vbk/ in the .gitignore	10:40
crab	why are there .vbk files in your remote git repo at all?	10:40
Aides	pielgrzym, you want git to delete them?	10:40
	physically?	10:40
pielgrzym	davegb3: already got it there, need to delete it	10:40
Aides	or from the repository?	10:41
pielgrzym	Aides: yes :)	10:41
	err	10:41
	← PaNtH3R left	10:41
pielgrzym	from the repo	10:41
	I only need them on dev machine where I actually use vim :) and they can be created now since they will be ignored by gitignore. I only want to get rid of the old ones	10:41
	→ HG` joined	10:42
	← sitaram left	10:42
Aides	git rm `find . -type d -name .vbk`	10:42
pielgrzym	Aides: wow :) hah, this is a bashfull solution :) thanks! :) is there a way to dry-run it?	10:43
Aides	echo git rm `find . -type d -name .vbk`	10:43
	Aides feels bashy today	10:43
	→ froschi joined	10:43
pielgrzym	;)	10:43
Aides	git rm -n `find . -type d -name .vbk`	10:44
	that would dry-run it more gitty	10:44
pielgrzym	Aides: hah, worked flawlessly :) thanks again :)	10:45
doener	git rm -n --ignore-unmatch .vbk */.vbk	10:45
	hm, no, actually not	10:45
	→ segher joined	10:46
Aides	doener, that's bashy too, and for single-level only	10:46
	← HG` left	10:46
doener	Aides: actually I meant the * to be interpreted by git, should have escaped it	10:47
	→ sitaram joined	10:47
doener	Aides: but the "* matches /" assumption is only true when there's no / in the pattern	10:47
	Aides: e.g. "git rm \*.txt" deletes .txt files, regardless of the directory they're in	10:48
	→ mlpug joined	10:48
	→ sitaram_ joined	10:50
	← Guest13774 left	10:50
Aides	ah, okay	10:51
	← jinzhu left	10:51
pielgrzym	ah git is fabulous	10:52
	← patmaddox left	10:52
pielgrzym	how those damn .svn files pissed me off	10:52
	← sitaram_ left	10:52
	→ psynaptic joined	10:54
	→ markelikalderon joined	10:55
pielgrzym	one other thing. I've got an entry in my .gitignore: media/pictures/* but I can't add the pictures folder itself to git. I only want git to ignore it contents	10:55
	→ juanjoc joined	10:55
doener	pielgrzym: git doesn't track directories	10:55
pielgrzym	ah, indeed :) it tracks the content itself :)	10:55
	doener: so no way to achieve this?	10:56
doener	pielgrzym: a common way is to have a .gitignore file in such a directory, probably containing	10:56
	just a * (instead of having the rule in the top-level .gitignore)	10:56
pielgrzym	hm good idea :) thanks :)	10:57
doener	pielgrzym: the other common wisdom is: let your build system/installer/software create required directories	10:57
	→ malumalu joined	10:57
pielgrzym	doener: I guess I'll stick to that one since putting gitignore would be troublesome (I've got one django engine and several fcgi runfiles each having it's own media dir, it would be a pain in the neck to remember putting .gitignore file each time I tinker there	10:58
	→ scarabx joined	11:01
	→ LimCore joined	11:03
	→ dodo_the_last joined	11:03
	→ kumbayo joined	11:05
	Aides looks for perl syntax highlighting library	11:05
	← voker57 left	11:08
	→ voker57 joined	11:08
SeySayux_	How do I allow other people to pull from my repo? Do I need to set up my firewall or something?	11:12
LimCore	SeySayux_: public server are not very secure, how about some free services like github etc (esp. for open source projects)	11:13
SeySayux_	I'm using SF.net	11:14
	But as I understand it, I can allow other people to pull from my repo, I can pull from other people's repo, etc? Otherwise it's just the same as svn, except that I need to use an extra 'git push' command	11:15
	← arachnist left	11:16
	← kmamykin left	11:16
thiago_home	SeySayux_: you can run git-daemon	11:16
	or you can give people ssh accounts	11:16
	← sitaram left	11:16
	→ kmamykin joined	11:16
SeySayux_	ssh accounts?	11:17
doener	SeySayux_: fetching directly from your working repo isn't really recommended, as that means that your local history becomes public, so all history rewriting is "forbidden"	11:17
	SeySayux_ starts twitching	11:17
SeySayux_	Okay... so where should they fetch from instead?	11:17
doener	SeySayux_: your repo at sf is your "distribution point", you push there whatever you deem ready for public consumption	11:17
	SeySayux_: others can fetch from there, to get your stuff into their local repos	11:18
	← markelikalderon left	11:18
SeySayux_	Okay, okay... I think I got it. Now how is this different from svn?	11:18
doener	SeySayux_: then they can create patches to be send by email, or have their own public repo (e.g. at sf or somewhere else) so you can fetch from there	11:18
thiago_home	it's not	11:18
doener	SeySayux_: with svn you have a single repo, with git, everyone has one or more repos	11:19
SeySayux_	Isn't that just the same as copying your source tree on your own hdd, and commit when you think it's ready?	11:19
	(in svn)	11:20
doener	SeySayux_: not really. You don't have any history locally with svn. You may not even have commit rights on the one central repo	11:20
SeySayux_	Well, it isn't so much different from SVN. Explain me again why SVN users are 'dumb', 'ugly', 'morons', 'idiots', etc?	11:21
doener	SeySayux_: With git, I could clone your public repo, do 20 commits, rewrite them a few times to clean them up, and then push them to _my_ public repo and tell you "Look, I cloned your repo and made some commits that do $GREAT_STUFF", you can fetch from there and merge my stuff	11:22
	SeySayux_: with svn, I can give you one big diff and say "deal with it"	11:22
SeySayux_	[...] you can fetch from there [...]. Where? Your local computer?	11:23
doener	SeySayux_: no, from my public repo	11:23
mlpug	SeySayux_, wikipedia git and "distributed revision control" articles describe very well the charactiristics of distributed vs central and towards what kind of scenarios git has been optimised	11:23
SeySayux_	so, in fact, everyone needs a public repo somewhere on github or sf.net ? Or am I wrong?	11:23
doener	SeySayux_: you might have your public repo at sf.net, I could clone it, and push my stuff to github, or repo.or.cz, or wherever	11:24
	SeySayux_: of course I could also have that public repo on my local box and allow fetching from there (e.g. via git-daemon)	11:24
SeySayux_	Okay, so let's see if I got it...	11:25
doener	SeySayux_: but then I'd have at least two repos locally, one for me to work on, and one with stuff ready for public consumption (and that's what I'd export via git-daemon)	11:25
	← rubyist left	11:26
	→ rubyist joined	11:26
SeySayux_	I have my own local work repo on which I can commit as often as I like. When I want to share it with the other devs, I put it up on github. They can fetch it from there, modify it in their own repos and commit as often as they like, and put it on their own github repos, from which I can fetch trhem again. When I like some piece of code, I put it up in the central repo on sf.net. Right?	11:27
doener	SeySayux_: for example, just that the "central repo" thing only exists by convention. The repo on sf.net (to take your choice) is only special because you say so	11:28
	SeySayux_: If I think that Joe Random's repo at $HOSTING_SERVICE is better, I can take that as my upstream	11:28
	→ vashism523 joined	11:29
vashism523	ey if you guys are bored check out facebook zombies - http://apps.facebook.com/zombies/links.php?r=719927515&nref=st	11:29
	← vashism523 left	11:29
doener	SeySayux_: e.g for wine, there's the "official" repo, and there's wine-hacks. The latter incorporates some patches that upstream doesn't have (yet)	11:30
SeySayux_	I appointed the repo at sf.net now as the project's main repo, where everybody should commit 'approved' code... I understand that nothing stops somebody from putting their code elsewhere, just like nothing stops somebody from forking my code...	11:30
doener	SeySayux_: with git, each side can fetch from the other to incorporate their commits (merging, cherry-picking, rebasing, whatever)	11:31
	SeySayux_: with svn, you'd have two independent repos and some pain to carry over change from one to the other	11:31
SeySayux_	Okay, I get it... The whole concept of Git is pretty new to me...	11:32
	→ cao joined	11:32
	cao → Guest94211	11:32
doener	SeySayux_: with git, you always commit locally, you only push to the other repo. Of course you can allow multiple devs to push to the main repo at sf.net, it's not that everybody has to have his own public repo	11:33
	→ markelikalderon joined	11:33
SeySayux_	Because with svn it's much simpler: you commit, someone else checks out, and it's done... But simpler is not always better ;)	11:33
doener	SeySayux_: but it's _possible_ to do so, so contributors can exploit the power of version control, even if they may not push to your main repo at all	11:33
	(or commit to the central repo, with svn)	11:33
	with svn, commit permission = permission to use version control. No commits, no version control.	11:34
	← ankit9_ left	11:34
SeySayux_	Okay, thank you very very much for the explaination ;) off_t o_implement, I guess	11:34
doener	:-)	11:34
	→ ToxicFrog` joined	11:35
doener	SeySayux_: if you're looking for some interesting reading material, try "The Git Parable", "Pro Git", "Git For Computer Scientists" and "Git From The Bottom Up" (ordered from high-level to low-level)	11:35
	(the last one being way more low level than you'll need during day to day work, but possibly interesting to understand the data model)	11:36
	→ brizly1 joined	11:37
	→ RaceCondition joined	11:41
	→ MisterN joined	11:41
	→ shenie joined	11:42
	→ sunn joined	11:42
	← RaceCondition left	11:43
	→ RaceCondition joined	11:43
	→ Jinzhu joined	11:44
	→ qrush joined	11:45
	→ kollektiv joined	11:47
	→ adeodatus joined	11:47
kollektiv	hi	11:48
	why does this not work http://img195.imageshack.us/i/errordzx.png/ ?	11:48
	← ToxicFrog left	11:48
	← therrg left	11:48
Aides	you need --hard to change your working tree	11:49
	← Guest94814 left	11:49
kollektiv	but that would reset the entire thing	11:50
	i just want to unstage one file	11:50
Aides	then use git checkout	11:50
zomg	kollektiv: you already unstaged it with reset	11:50
Aides	ah, unstage	11:51
	i thought you want to undo your changes	11:51
doener	kollektiv: the "locally modified" message means that there are difference between the index and working tree for that files	11:51
	s/s$//	11:52
	kollektiv: i.e. you already successfully unstaged it	11:52
kollektiv	oh	11:52
	ok i thought it was an error msg	11:52
	→ lu_zero joined	11:52
	← markelikalderon left	11:53
	← brizly left	11:53
kollektiv	if i do 'git -rm --cached file' but the file is already in the index does it just ignore the --cached option	11:57
doener	kollektiv: hm? --cached only makes sense if the file is in the index	11:57
	kollektiv: --cached means "delete from the index, but not the working tree"	11:57
	(for rm, that is)	11:57
kollektiv	o well i mixed up working tree and index	11:57
	it seems like it just deletes it from the working tree	12:00
	if it was already added	12:00
	and committed previously	12:00
doener	hm? git rm never deletes from the working tree only AFAIK	12:00
	either from the index and working tree, or from the index only (--cached)	12:01
	← karmi left	12:02
	→ tokkee joined	12:03
	→ chris2 joined	12:04
kollektiv	whats in the index? staged files?	12:05
MisterN	yes	12:07
	kollektiv: but the version that was staged, and NOT the version in the working directory	12:07
	kollektiv: when you stage a file, it gets copied into the index	12:08
doener	depends on how you define "staged". Some people would interprete that as "only the files I ran 'git add' on, since the last commit"	12:08
thiago_home	you may interpret it that way	12:08
	but that's not how git interprets it	12:08
	so you should interpret it the same way the tool does, to avoid misunderstandings :-)	12:08
	the last commit is staged	12:09
	← adeodatus left	12:09
thiago_home	plus everything you added, minus everything you removed	12:09
kollektiv	i see	12:09
doener	the index is really like an index of files that should be in the next commit. And as commits are snapshots and not deltas, this needs to also include files that you didn't change since the last commit	12:12
	ADmad\|away → ADmad	12:12
doener	so "git rm --cached foo" really removes the index entry for "foo", it doesn't add a "delete foo" entry or so	12:12
	everything deals with snapshots, diffs etc. are created on the fly	12:13
davegb3	... so if i add & commit file X, then change it, then add & commit it again, git physically stores 2 full copies of the file?	12:14
doener	davegb3: yep, but one of it will be unreachable, and thus prunable	12:14
cbreak	why unreachable?	12:15
	you can of course reach past commits	12:15
davegb3	cbreak: you beat me to it!	12:15
doener	you committed only once	12:15
cbreak	that's the whole point of an SCMS	12:15
davegb3	no, i committed twice	12:15
doener	oh wait, you didn't	12:15
	ok, so it's not unreachable and not prunable	12:16
	and there are two blobs, one for each version of the contents of X	12:16
	those were created when you did "git add X", as loose objects (possibly compressed, but still the whole contents)	12:17
	only when things get packed, git uses delta compression to save space	12:17
davegb3	is that git gc that packs the blobs then? (just interested in what goes on underneath the hood)	12:18
doener	gc is the highest-level tool for that purpose and does packing as well as pruning	12:18
	repack is slightly more low level and does just packing (thought it can drop some unreachable objects in its progress, it's not really pruning)	12:19
	and at the lowest level, there's pack-objects which does the real work	12:19
	in general, gc is what you should use	12:19
davegb3	ok... thanks, it helps to understand what's going on.. sorry to butt in on someone else's question!	12:20
	→ sitaram joined	12:21
	→ ryah joined	12:21
	→ donny joined	12:21
kollektiv	am i suppose to run git gc myself	12:21
cbreak	no	12:21
	you can	12:21
kollektiv	or is there some compression thing that happens automatically	12:21
	→ codeswing joined	12:22
kollektiv	nvm	12:23
	it was answered above	12:23
	→ teolicy joined	12:23
	← joevano left	12:24
	donny → joevano	12:24
	← scarabx left	12:24
doener	there's "gc --auto", but I'm not really sure when/by which commands it's triggered	12:24
	← hzlocky left	12:25
ryah	I want to edit the commit message of a commit which is not the most recent	12:25
doener	rebase -i	12:25
teolicy	Hi. There's something I don't understand about git push. I have a repo, whose master is at ref 'C'. It came from origin, whose master (origin/master) is at 'A', which is about 18 commits behind 'C'. In my repo, 'C'^17 is 'B', an unnamed ref whose parent is 'A' (so the graph is: A->B->...->C).	12:25
	I want to push just B unto origin, and keep C and all the refs leading up to it locally.	12:25
	I thought 'git checkout B ; git push origin HEAD:master' is what I want, but my push is rejected (non fast forward).	12:26
	git pull --rebase says there's nothing new to fetch.	12:26
	What am I missing?	12:26
	→ marvil07 joined	12:26
	← psynaptic left	12:27
ryah	doener: can i do that while keepin gthe commit in the same place?	12:27
doener	teolicy: I'd do "git push origin B:master" instead of doing the checkout, but that shouldn't affect the outcome	12:27
	ryah: define "in the same place" please	12:27
	ryah: changing the commit message creates a new commit anyway, so of course you get a new history	12:28
sitaram	teolicy: what does "git log --oneline --left-right B...origin/master" say	12:28
	teolicy: note 3 dots not 2 in there	12:28
doener	teolicy: do a "git fetch origin" before that log	12:28
sitaram	aah! that's got to be it!	12:28
	someone else moved the origin beyond A	12:29
ryah	doener: nevermind. i forgot to rebase --continue	12:29
doener	sitaram: hm, why three dots? The commits that are in B but not in origin/master don't really matter, do they?	12:29
teolicy	"git log --pretty=oneline --left-right 0f010b...origin/master" (note I did --pretty=oneline, I think it's what you meant) shows me three commits (which I agree is odd): B, A, A^.	12:30
	sitaram, doener: ^^^^	12:30
doener	teolicy: --oneline is supported in recent git versions ;-)	12:30
teolicy	doener: Don't get cocky :)	12:31
doener	teolicy: what's the exact pull command you used? Just "git pull --rebase" shouldn't work on a detached HEAD	12:31
	→ ereslibre joined	12:31
teolicy	When I did the pull --rebase I was at master (my local master, aka 'C').	12:31
doener	teolicy: ok, and that has remote set to origin and merge set to refs/heads/master?	12:32
	← mlpug left	12:32
kollektiv	what would i use git reset --soft for? the description makes it seem like it does nothing 'Does not touch the index file nor the working tree at all, but requires them to be in a good order'	12:32
teolicy	I /think/ so, but I'm not sure I can say that with confidence. How can I verify?	12:32
doener	kollektiv: git commit -m tmp; git commit -m tmp; git commit -m tmp; git reset --soft HEAD~3; git commit -m "All three tmps squashed together"	12:32
sitaram	teolicy: doesnt the output show any "<" or ">" signs? are there any ">" or only "<"?	12:33
doener	kollektiv: i.e. the soft reset still resets the branch head	12:33
teolicy	('git remote' prints 'origin')	12:33
	The output of the git log showed: <B, >A, >A^.	12:33
	(using three dots)	12:33
	(using two dots it said just: >A, >A^)	12:33
doener	teolicy: git remote show origin	12:34
	→ britishseapower joined	12:34
doener	teolicy: should show "Local branches configured for 'git pull'" and "master" should be in that section	12:34
britishseapower	whats a pack header ?	12:34
sitaram	teolicy: this means (to me) that both A and A^ are not contained in B, so your A->B->...->C is not true	12:34
britishseapower	fatal: protocol error: bad pack header	12:34
	← RaceCondition left	12:34
britishseapower	just got that when doing git pull	12:34
	used to work	12:34
teolicy	doener: git remote show origin shows: http://dpaste.com/83952/	12:35
kollektiv	doener: what do you mean by squashed?	12:35
teolicy	sitaram: Hmm. I don't know what to say, I'm pretty sure they are, I thought I was beyond that point with git... Let me re-verify please. One moment.	12:35
britishseapower	i got it.	12:36
	i had a file open	12:36
teolicy	pfft. OK, I'm a moron. I used a ref to refer to B, and it was the wrong ref (the one from before the git pull --rebase). I plead nearly one bottle of nice wine to my shabby defence, sorry for wasting your time.	12:37
doener	kollektiv: http://git.pastebin.com/m2713185b	12:37
sitaram	teolicy: add a "--boundary" to that and you'll see the latest common commit; you definitely have a divergence, going by the "<" and ">"	12:37
teolicy	sitaram, doener: ^^^	12:37
sitaram	teolicy: plead one bottle? or give one bottle? ;-)	12:37
	→ veiz joined	12:38
teolicy	You pay for the shipment, I pay for the bottle. I'm in Israel by the way. :)	12:38
sitaram	:)	12:38
	sitaram is strictly rum and coke -- has forsaken all other forms of drink :)	12:39
sitaram	(except the occasional Baileys)	12:39
kollektiv	doener: where does D,E, and F go in that example	12:39
	like can you get back to them	12:39
doener	kollektiv: they became unreachable, at least through "master", but they'll still be in the reflogs	12:40
	kollektiv: that's why I kept them in the graph, they still exist, they're just not in master's history anymore	12:40
	← mbrumbelow left	12:41
kollektiv	ah k	12:41
teolicy	sitaram: 32 degreees C; rum and coke / Baileys would kill me. Ice cold white whine is the only drinkable substance. (and even that vaporizes if you don't drink it fast enough :)	12:41
	doener prefers a good Malt, but isn't opposed to a good wine, especially if it goes along with a good steak :-)	12:41
teolicy	(also, I think the friggin' bastards at yahoo weather are lying, I want that meteorologist who claims its just 32C to come over and have a stroll in the sun)	12:42
sitaram	teolicy: I'm in India. 32C? Pfft!	12:42
	:D	12:42
teolicy	Exceedingly good point, I must admit. :)	12:42
	doener just decided to have a good Single Malt this evening :-)	12:42
	→ rlb3 joined	12:43
teolicy	Anyway, enough yakking, thanks a lot for your help, I'll be off now, back to coding.	12:43
	teolicy tips hat	12:43
	← teolicy left	12:43
ADmad	sitaram: you have the wrong nick for enjoying such drinks ;)	12:43
sitaram	ADmad: ??	12:46
	ADmad: you Indian by any chance? and referring to the religious significance of my name? again, to use teolicy's words: Pffft! ;-)	12:46
	← dodo_the_last left	12:46
ADmad	sitaram: yes sitaram drinking rum doesnt sound too nice :P	12:46
sitaram	ADmad: and it's not a random nick; it's my name..	12:47
	→ woototo joined	12:47
	sitaram would love to know what most of the nicks mean and how they were chosen	12:47
ADmad	sitaram: haha	12:47
doener	go to the office, get a new one. I suggest "Max Power"! :-)	12:47
woototo	hi. Is there any place were i can download the git's own repository as a bundle?	12:47
sitaram	doener: what does doener mean, if it means anything (always wanted to ask; please ignore if you wish!)	12:48
	← rlb3 left	12:49
	sitaram → Max_Power	12:49
	Adlai uses his real name...	12:50
Max_Power	thank doener	12:50
	thanks	12:50
Adlai	"Handles are for tools"	12:50
rapha	Max_Power: it's a food ... turkish food ... some kind of meat; here in germany they put it into kind of a bun together with veggies	12:50
doener	I got that nick in school, after several modifications to my earlier nicks. It's actually the (short)name of a fast-food meal. The full name is "Dönerkebab", is turkish and means as much as "rotating meat" or so)	12:50
	Max_Power → sitaram	12:50
sitaram	:)	12:50
doener	it was way better than the previous iteration, and as I don't take myself serious anyway, I stuck to it	12:51
ADmad	sitaram: how about SitaramOnRum :)	12:51
	→ ab3 joined	12:51
sitaram	Adlai: it's quite common in India for people to have nicknames that the family and friends use, which may be quite different from the real name. In effect, it becomes an alternate name	12:52
	Adlai: SitaRum maybe better :)	12:52
	minor spelling change	12:52
Adlai	heh	12:52
	btw, I think you intended that for ADmad ?	12:52
sitaram	Adlai: yes -- sorry! (dang tab completion!)	12:53
woototo	it would be a not-so-bad-idea to add support for bundles in the git-scm.org "Other Download Options"	12:53
Adlai	hm, can I have different push and fetch URLs for the same remote?	12:54
doener	woototo: why a bundle?	12:54
woototo	because of no direct connection	12:54
doener	Adlai: with 1.6.4, you can	12:54
ADmad	sitaram: that was my first choice but i could sound a bit offensive to the religious minded hence didnt mention :)	12:54
	→ reprore_ joined	12:54
qrush	anyone know how to make custom messages appear on the client side during a git push?	12:54
doener	woototo: hm? If you can get the bundle, you should be able to clone, too. And then you can create a bundle yourself	12:54
qrush	doener: you gotta know something about this :)	12:55
Adlai	doener: yep, I have 1.6.4. When I do `git config remote.origin.<TAB>', though, I only see one url option	12:55
doener	qrush: output from hooks is sent to the client	12:55
woototo	doener, i will get it from windows, transfer to linux via usb and then unbundle	12:55
qrush	doener: Not from what I saw...which hook? pre-receive?	12:55
doener	Adlai: probably not yet in the bash completion, check man git-config	12:55
woototo	anyway, it is an alternative download option :)	12:55
doener	woototo: also, bundles are usually custom made to include just the required stuff	12:56
Adlai	doener: I use zsh! zsh tab completion is the last word!	12:57
doener	woototo: if you trust me, I can create a bundle for you	12:57
	Adlai consults `man git-config'	12:57
woototo	i trust you, but only if you don't have more important things to do :)	12:57
sitaram	doener: why trust? after unbundling he can check the top SHA to exist in the git repo via gitweb or so can't he?	12:57
qrush	doener: i'm trying out the scripts here for pre/post receive and not seeing anything http://book.git-scm.com/5_git_hooks.html	12:58
sitaram	at worst, it won't be the top SHA there if Junio pushed something else out in those few seconds/minutes	12:58
doener	woototo: so, which stuff to you already have? "git for-each-ref refs/remotes/origin" or whatever	13:01
woototo	nothing. just a binary installation of git	13:03
	→ karmi joined	13:03
Adlai	wow, git-rerere is cool.	13:05
doener	woototo: http://dot-dash.at/stuff/git.bdl -- sits on DSL	13:05
	→ Bass10 joined	13:06
woototo	doener is the dude	13:06
Adlai	er, probably a very stupid question, but to what should I set rerere.enabled to enable rerere?	13:07
	`git config rerere.enabled true' ?	13:07
doener	sitaram: I _could_ know about a bundle related exploit.	13:09
	← froschi left	13:10
qrush	doener: perhaps there's some config value I have to set on the bare repo to have its hooks send messages back to the client side? sorta stumped here :/	13:10
	→ froschi joined	13:11
	← ia left	13:12
ADmad	doener: thanks for the help regarding cgit highlighting earlier... after struggling for quite sometime finally got it running... problem was the latest package cgit-0.8.2.1.tar.gz available doesnt support the "source-filter" config option... once i released that got latest code through git everythign went smoothly	13:12
doener	ADmad: ah!	13:12
	qrush: just a moment, my git install seems fucked up	13:12
	→ ia joined	13:12
	← isnull left	13:13
sitaram	doener: an exploit which attacked his git via a mal-formed bundle? hmmm...	13:13
doener	sitaram: yep. I don't know of any, but still he has to trust me that my bundle won't cause e.g. a buffer-overflow and runs some shell code, doing bad stuff	13:14
sitaram	true	13:14
	woototo: this leads me to ask: why cant you install git on the windows machine that does have internet, grab the repo the normal way, and make your own bundle?	13:14
woototo	because i can't install new stuff	13:15
	← ryah left	13:15
woototo	don't ask :)	13:15
doener	woototo: btw, that bundle contains my remote tracking branches, i.e. refs/remotes/origin*, so use the right refspec for fetch	13:15
qrush	or if anyone know this :) i'm trying to figure out how to get the git daemon to print out messages on push	13:15
	on the client side, that is	13:16
Ilari	qrush: You are trying to stuff them to stdout or stderr?	13:16
qrush	Ilari: stdout hopefully, basically after a `git push` i want to see some messages coming from the server. heroku does this, i can't find anywhere that tells how to do this	13:17
	→ prusnak joined	13:17
sitaram	qrush: they just come; at least I've done then on he post-receive hook	13:17
	qrush: is your hook even executing?	13:18
doener	qrush: I get stdout as well as stderr just fine	13:18
woototo	umph, `git bundle unbundle git.bdl` doesn't do anything	13:18
qrush	I'm printing out on stdout on post-receive, and i see that fine on the server side. on the client side i want to see those messages as well	13:18
doener	qrush: http://git.pastebin.com/m75407103	13:18
sitaram	woototo: gitfetch	13:18
qrush	sitaram: yes, definitely	13:18
sitaram	git fetch	13:19
	woototo: unbundle is really plumbing, per the docs	13:19
Adlai	sorry to nag, but does `git config rerere.enabled' => 'true' mean that I've properly enabled rerere?	13:19
sitaram	Adlai: yes I think so	13:19
Adlai	cool.	13:19
doener	qrush: same for post-receive, also prints stuff just fine	13:19
	→ botanicus joined	13:20
qrush	doener: Still only seeing the messages on the verbose log of the git daemon	13:20
sitaram	Adlai: although I just noticed a "rerere.autoupdate" that I dont remember seeing before; and it defaults to false :(	13:20
	← EricInBNE left	13:20
qrush	doener: they're not getting transmitted to the client	13:20
doener	woototo: "git fetch ./git.bdl +refs/remotes/origin/:refs/remotes/origin/"	13:20
qrush	doener: even with a simple echo :)	13:20
roberthorrey	man, git is so awesome. it's one of those tools i want to use even in places it shouldn't be used. like embedded systems that run my thermostat, etc.	13:21
woototo	sitaram: tried that: `git fetch git.bdl` -> Unable to chdir or not a git archive	13:21
Adlai	sitaram: yes, rerere.autoupdate controls whether the index gets updated automatically after rerere kicks in.	13:21
doener	woototo: needs the ./ prefix to make git look at the file	13:21
Adlai	it defaults to false so that you get a chance to do a "sanity check"	13:21
	← botanicus left	13:21
qrush	doener: so yeah, short of recompiling git to print out custom stuff is there anything that can be done?	13:21
	→ botanicus joined	13:21
Ilari	qrush: What if you stuff them to stderr?	13:22
doener	qrush: so you're pushing through git-daemon?	13:22
qrush	doener: yes	13:22
	doener: http://gist.github.com/172779	13:22
	seeing that on the log	13:23
sitaram	Adlai: aah! makes sense; I just looked at the code and saw the message changes "Resolved" to "Staged" if this is true	13:23
doener	qrush: that's probably the reason then, daemon capturing the output without relaying it.	13:23
qrush	so is there another way to offer push/fetch anonymously?	13:23
	that will actually transmit the data?	13:24
sitaram	qrush: you're using git daemon for push?	13:24
	wow...	13:24
qrush	sitaram: this is sort of a different application for pushing :) we're trying to get away with not doing ssh key auth	13:25
	→ cloudhead joined	13:25
	→ tvw joined	13:26
woototo	no luck. please note that i don't have any git history. i.e. the unbundling should also create the ".git" dir etc.	13:27
doener	woototo: it won't.	13:28
woototo	ah. so i have to init-db	13:28
doener	woototo: mkdir git; cd git; git init; git fetch /path/to/bundle +refs/remotes/origin/:refs/remotes/origin/; git checkout -t origin/master	13:28
	woototo: init-db is really really old ;-	13:28
	)	13:28
woototo	yay. 100% success. Thank's doener and sitaram	13:30
	→ squentin joined	13:32
	← veiz left	13:37
	→ HG` joined	13:43
	← ruskie left	13:43
	← patmaddo_ left	13:46
	← marvil07 left	13:51
	← codeswing left	13:53
	→ johnnyg joined	13:54
	→ eMyller_ joined	13:57
	→ pstuteville joined	13:58
	← pstuteville left	14:00
	→ pstuteville joined	14:00
	← woototo left	14:02
	→ loincloth joined	14:03
	→ veiz joined	14:05
	→ RaceCondition joined	14:06
	← cloudhead left	14:07
	← gonzojive left	14:10
	← LimCore left	14:11
	← qrush left	14:12
	→ dodo_the_last joined	14:17
rleigh	Hi folks. When doing an "add -i" to interactively add hunks, it's possible to split hunks up to commit part. However, if you have two adjacent lines, it's not possible to split them up into the individual lines. Is there any way to add just one of those lines to the index?	14:18
	← britishseapower left	14:18
	→ onigiri_ joined	14:18
doener	rleigh: (e)dit allows you to edit the patch in your editor	14:18
	← onigiri left	14:18
doener	rleigh: and "git gui" can stage individual lines	14:19
Tommy[D]	Whats the easiest way to merge some changes from another git tree into your own one?	14:19
	→ jfkw joined	14:19
rleigh	doener: Cool, I'll give them a try, thanks!	14:19
	eMyller_ → eMyller	14:25
	→ hzlocky joined	14:25
	→ Sho_ joined	14:27
	← shenie left	14:28
	← roberthorrey left	14:32
	← Jinzhu left	14:32
davegb3	can i just sanity check this: i have a tracking branch tracking someone else's repo on github. i've branched from that (for a fix) locally and made some commits. as the original branch has made some more commits in the mean time, if i pull those in, what's the best way to merge those in to my "fix" branch, so it looks like it's branched from the tip of the original branch. git rebase ..something?	14:33
Ilari	davegb3: 'git rebase <remote>/<branch>'.	14:34
	davegb3: (after fetching <remote>)	14:34
davegb3	ok, and that leaves my commits on my branch, just altered so they come after the new commits on the remote branch?	14:34
	→ SjB joined	14:35
wereHamster	davegb3: yes	14:35
ab3	hello	14:35
davegb3	i presume git rebase myTrackingBranch would do the same, if myTrackingBranch is tracking <remote>/<branch>	14:36
	(and myTrackingBranch has been updated)	14:36
ab3	I have a strange problem when I try to dcommit changes to svn with git svn dcommit	14:36
	http://dpaste.com/83982/	14:36
	I don't have a file named test_git...	14:37
	any idea how I can make dcommit my changes to svn?	14:37
	→ almostautomated joined	14:38
	← Akufen left	14:42
	→ khmarbaise joined	14:43
	→ cubuanic joined	14:48
	→ hydrogen joined	14:50
	→ danferranti joined	14:51
offby1	hmm, no :-\|	14:52
	you're _sure_ you don't have a file of that name?	14:52
sunn	I want to host a git repository that is only available to a few specific people -- is http authentication my only option, or is there some other way that I'n not aware of? (I don't have root access on the server)	14:53
offby1	sunn: you can use gitosis	14:53
	it uses ssh, but you don't need root to set it up	14:53
	→ laney_ joined	14:53
ab3	offby1: yes	14:54
	← danferranti left	14:54
sunn	Okay, I will look into that, thanks	14:54
	davegb3 → davegb3\|	14:55
Ilari	sunn: It is bit nasty to setup without root access (assuming you can't get admins to create an extra account for you and to grant sudo privs to that).	14:55
offby1	oh yeah	14:55
	offby1 forgot about the setting-up-the-new-account part :-\|	14:55
	← bdowning left	14:56
ab3	git rm test_git	14:56
	fatal: pathspec 'test_git' did not match any files	14:56
	→ _grepper joined	14:56
sunn	Well, I won't be able to get another account, and definately no sudo access :/	14:57
	→ bdowning joined	14:57
	→ Fullmoon joined	14:58
Ilari	sunn: For setting up gitosis without extra user account you need TWO keypairs (you probably already have one).	15:00
	→ enherit joined	15:01
sunn	I see -- is that a problem beyond simply running ssh-key-gen again though?	15:01
	ssh-keygen rather	15:02
	← Fullmoon left	15:02
	→ Fullmoon joined	15:02
	← SjB left	15:02
	← ADmad left	15:03
Ilari	sunn: There are also some pitfalls. On both client and server.	15:04
	→ SjB joined	15:05
	→ mm_202 joined	15:06
Ilari	sunn: On client, it revolves around offering the right key. On server, the key used for gitosis access MUST NOT be the same key that grants shell access.	15:06
Tommy[D]	Whats the easiest way to merge some changes from another git tree into your own one? Or isnt there such an option in git?	15:06
Ilari	Tommy[D]: Pull (fetch + merge)?	15:07
	← dodo_the_last left	15:07
	SeySayux_ → SeySayux	15:08
	← syd_ left	15:09
Ilari	sunn: Also, AFAIK, the second key for gitosis can't be passphrase-protected unless you want to type it out each time.	15:10
	← univate left	15:11
	← _grepper left	15:12
	→ univate joined	15:12
	→ cloudhead joined	15:13
Tommy[D]	Ilari: i am talking about a different tree, not the one i am pulling from	15:13
Ilari	Tommy[D]: What does "tree" mean? Repository?	15:14
sitaram	sunn: the progit book shows a manual way to do what gitosis does -- basically manage ~/.ssh/authorized_keys -- if you want to try it. Doing the whole thing while having only one account on the server is tricky but not impossible	15:14
	sunn: make sure you have password access to the server so if you screw up the keys in some way you can move aside your local priv key and login to fix things	15:15
	→ gaveen joined	15:16
sunn	I guess I'll have a look at that aswell then	15:17
sitaram	sunn: it doesn't address doing everything as one user. In order to retain control to go in and fix things, as well as restrict your "guests" from doing whatever they want, you will end up replicating a lot of gitosis functionality, especially the stuff inside "authorized_keys". Play with gitosis on the side to get a feel for all the ssh tricks it uses	15:18
	← khmarbaise left	15:18
sitaram	sunn: http://sitaramc.github.com/0-installing/9-gitosis-server-install.html tries to explain most of it	15:18
Tommy[D]	Ilari: i have 1 repo created with git-svn and one normal git repo and now i want to merge the changes from git-svn repo into my other one	15:20
Ilari	If one knows the tricks gitosis plays with SSH, it shouldn't be too hard to to hack something crude together (assuming 'echo $VARIABLE' is safe).	15:20
	Tommy[D]: Sounds like fetch + merge (a.k.a. pull).	15:21
	→ g3d joined	15:22
Ilari	'echo $VARIABLE' being "safe" means that it doesn't do any weird stuff no matter what the environment variable contains.	15:22
	→ juan_arandaalvar joined	15:23
sitaram	Ilari: half of the stuff gitosis does is in managing the keydir and gitosis.conf and setting up authorized_keys on a push to gitosis-admin; if that can be done manually it's fairly straightforward	15:23
	→ gonzojive joined	15:23
	← juan_arandaalvar left	15:23
	davegb3\| → davegb3	15:24
	← SjB left	15:24
Ilari	At least I can't find character that would do anything special when echoing a env.variable	15:24
	→ markelikalderon joined	15:25
	→ davi_ joined	15:25
davi_	hi all. currently I use SVN on a system which my bash user is my bug tracking account and my VCS account. so, to login on the dev server, login on the bug tracking system and commit, I use my username "davi".	15:26
Ilari	davi_: SVN?	15:26
davi_	how can I associate my git commits to my username without rely on git config user.name?	15:26
	davegb3 → davegb3\|	15:27
Ilari	davi_: GIT_{AUTHOR,COMMITTER}_{NAME,EMAIL}	15:27
Tommy[D]	Ilari: eh, fetch+merge? How should that work, if origin is different for both?	15:27
	→ almostautomated_ joined	15:27
Ilari	Tommy[D]: If those repos don't share history, merge is probably going to go horribly wrong...	15:28
davi_	Ilari: sorry, I didn't understand. could you clarify or point me to the right keywords to search? currently, I'm googling for 'git relate commit bash user'	15:28
Ilari	davi_: Those are environment variables.	15:28
davi_	Ilari: from the commiter, right?	15:28
	← veiz left	15:29
AAA_awright	GIT_AUTHOR_NAME="username" GIT_AUTHOR_EMAIL="..." git commit	15:29
	← aspotashev\|eeepc left	15:29
AAA_awright	Right?	15:29
Ilari	AAA_awright: AFAIK, ESYNTAX	15:29
	→ veiz joined	15:29
Tommy[D]	Ilari: they do, but currenlty, i created a branch on the second one, copied the latest version of the first one to that branch, git-commit that, then merge this commit into the master branch, but that does not preserve the history	15:29
davi_	but, using env vars, I can do a commit using something like GIT_AUTHOR_NAME="Ilari" git commit, can't I?	15:30
	→ SjB joined	15:30
wereHamster	AAA_awright: if you only want to set the author, you can also use git commit --author="Your Name [email@hidden.address]	15:30
Ilari	I have Bash 3.2.48 and that sort of specifying two env.vars doesn't seem to work.	15:30
	davi_: That kind of stuff should work.	15:31
AAA_awright	Ah, COMMITTER is the default one, AUTHOR is by default undefined right?	15:32
davi_	Ilari: and there isn't a way to prevent this? I work with 7 more developers, and I don't want to leave this "role" to one of them commit using someone else "username"	15:32
AAA_awright	No, it is the same as committer	15:32
	← kpreid_ left	15:32
	→ kpreid_ joined	15:33
Ilari	davi_: If you have full blown shell account with direct write access to repo, you can do nastier stuff than committing with wrong usernames...	15:34
	joshthecoder_afk → joshthecoder	15:34
davi_	Ilari: no one have access to the repository :)	15:34
patrikf	davi_: don't develop with people you don't trust?	15:34
	← MisterN left	15:34
davi_	the repository is on another server	15:34
	patrikf: if I was the boss, I would fire those people... :P	15:35
patrikf	ah, I see	15:35
	← malumalu left	15:37
	← Fullmoon left	15:38
davi_	Ilari: sorry, I'll try to clarify. I've 2 servers: the repo (currently, SVN) server and the devel server (Apache, MySQL, bla, bla, bla). all developers have (SSH) access to the devel server. but we only have access to the repo server when (1)logging in the bug tracking system or (2)commiting to repository. since the bugtracking system is fully integrated with SVN, I want to migrate from SVN to Git. And, no, I don't want git-svn. I want to	15:38
	get rid of SVN	15:38
Ilari	davi_: Committing is local operation. And pushing is too late to check the committer/author info.	15:40
	davi_: Sometimes it is possible to log the pusher.	15:40
	← RaceCondition left	15:42
davi_	Ilari: what do you recommend me to do/search?	15:42
	← almostautomated left	15:43
Ilari	davi_: Would "(2) committing to repository" mean "pushing to repository" with git?	15:44
davi_	Ilari: yeah	15:45
Ilari	davi_: Presumably one could tell the users apart by their accounts or keys. Note that gitosis needs one-line modification to enable logging the pushes in hooks.	15:46
	← alezandro left	15:48
	← gonzojive left	15:49
	← almostautomated_ left	15:49
davi_	Ilari: ok. thank you very much for your time and patience. two more questions: what sould I search for? 'git hook log push' / some time ago, I read that gitosis was discontinued, what do you think about gitosis?	15:50
offby1	.oO("discontinued"?)	15:50
	→ bcardarella joined	15:50
davi_	offby1: sorry, I can't remember which words were used. :( English isn't my mother language	15:51
offby1	as far as I know, gitosis works fine	15:51
davi_	offby1: thanks	15:52
	← voker57 left	15:53
	→ voker57 joined	15:53
	→ almostautomated joined	15:54
	← bcardarella left	15:54
	→ omenking joined	15:54
	→ almostautomated_ joined	15:58
	→ danferranti joined	16:00
	← SjB left	16:02
	→ Aikawa joined	16:02
	← hydrogen left	16:03
	← Aikawa left	16:03
	← HG` left	16:03
	← dpino left	16:04
	→ Aikawa joined	16:04
	← ProperNoun left	16:04
	→ krawek joined	16:06
	← laney_ left	16:06
	← mbroeker left	16:07
	← danferranti left	16:10
	← juanjoc left	16:13
	← bgerber left	16:14
sitaram	davi_: where did you read that gitosis was discontinued?	16:15
offby1	perhaps it's merely heavily discounted	16:16
sitaram	davi_: I've been suspecting it myself but hadn't actually read it anywhere	16:16
	← almostautomated left	16:16
sitaram	offby1: yes 100% discount of your money back :)	16:16
davi_	sitaram: I don't remember... probably on one of Gentoo changelogs...	16:16
	sitaram: well, for Gentoo, if a package didn't has any updates in 2 days, it's set as abandoned... :P	16:17
sitaram	davi_: yes there's a gentoo guy who'se modified it rather heavily! almost unrecognisable actually :)	16:17
	davi_: which means for somethings they expect an update by the time the previous one compiles?	16:17
	sitaram ducks and runs	16:17
davi_	haha	16:18
Aides	;)	16:18
	→ mbroeker joined	16:18
	sitaram is working on gitosis-lite. Pure perl, no installation, no extra users, no "admin by push", no "gitweb/daemon" management, but includes branch R/RW like Documentation/howto/update	16:18
	Aides loves how no one outside of gentoo community has found any other reason to rant upon it	16:18
sitaram	Aides: joke <> rant	16:19
	Aides: I admire those guys; almost envy. But that life is not for me	16:19
Aides	yeah, meant that, sorry	16:19
	→ bgerber joined	16:20
sitaram	its like dave barry saying: ubuntu is an ancient african word meaning "can't install debian" :) What would he say about gentoo, I wonder?	16:21
	→ Derander joined	16:22
Aides	gentoo is a modern english word meaning "a very fast and nice penguin"	16:25
	→ khmarbaise joined	16:26
	→ JasonWoof joined	16:26
	→ xanonus joined	16:27
doener	Aides: gentoo is a fictional word meaning "a penguin for which you only get the dna, which babbles a lot of gibberish while you grow it" (but you can choose to grow it with fast legs, but without a head!)	16:29
	SCNR	16:29
	→ _grepper joined	16:29
Aides	don't follow the gibberish part, but quite accurate, yes ;)	16:30
doener	Aides: well, others describe gentoo as "watching shit fly by"	16:30
wereHamster	gibberish = compiler output while you build the system ?	16:31
doener	Aides: or is the building more quiet nowadays? It's been a while since I used gentoo	16:31
	→ dodo_the_last joined	16:31
Aides	afaik, it is quiet by default	16:31
doener	in 2003 (or so) it was quiet verbose, but ok	16:32
	← magcius left	16:34
	sitaram has to go... :(	16:36
	← sitaram left	16:36
	← reprore_ left	16:36
crab	"SCNR" when it comes to strange experiments with penguin DNA is just asking for trouble.	16:37
	→ reprore_ joined	16:37
	→ zacharyp joined	16:38
	← ben_h_ left	16:38
	← zacharyp left	16:38
	→ repror___ joined	16:38
	davegb3\| → davegb3	16:39
	→ WinterWeaver joined	16:39
	← reprore_ left	16:41
	← botanicus left	16:42
	→ marcelo_ joined	16:44
	→ djgera joined	16:46
	→ Theravadan joined	16:46
	→ ctusar joined	16:49
	→ pheaver joined	16:51
	→ altrux joined	16:51
	← eMyller left	16:53
	→ Akufen joined	16:54
	← JasonWoof left	16:55
	→ magcius joined	16:55
	← Sigma left	16:58
	← ricky left	16:58
	← pstuteville left	16:58
	← djgera left	16:59
	← scientes left	17:00
	→ crack_head joined	17:00
	→ offby1` joined	17:00
	→ MisterN joined	17:02
	← veiz left	17:02
	→ spearce joined	17:04
	← offby1 left	17:05
Nugget	gentoo is a modern english word that means "I love the idea of FreeBSD but I'm looking for something that's less stable."	17:06
	→ aspotashev\|eeepc joined	17:06
	← gaveen left	17:07
	→ HG` joined	17:07
	← Evanlec left	17:07
	→ fualo joined	17:09
	→ hydrogen joined	17:09
	→ Husio joined	17:11
Husio	hello, how to force to apply changes from stash?	17:11
	→ WALoeIII joined	17:11
	offby1` → offby1	17:12
hyperair	what do you mean force apply?	17:13
	← spearce left	17:14
	→ spearce joined	17:14
Husio	if i type git stash show @{0} it shows me that 2 files has changed, git shash list shows nothing and git shash apply @{1} returns 1	17:15
	→ javatexan joined	17:15
	→ rnaveiras joined	17:18
	← krawek left	17:20
	→ fualo_ joined	17:21
	→ gaveen joined	17:21
	→ Michael__ joined	17:21
	→ ShadeHawk joined	17:21
	← Michael__ left	17:22
	← Chillance left	17:22
Husio	it's not shash, it's future commit :)	17:24
	→ Evanlec joined	17:26
	→ Krynnotaur joined	17:27
	← javatexan left	17:27
	→ pstuteville joined	17:28
	→ netkevin joined	17:35
	→ developernotes joined	17:36
	← fualo left	17:36
	→ jceb joined	17:37
cubuanic	hello, guys. i need to completely delete last commit on remote origin on master branch. how i can do this?	17:43
	← Akufen left	17:43
	→ muammar joined	17:43
	→ selckin joined	17:43
thiago_home	push HEAD~ to it	17:43
	you'll need to force it	17:43
	remember you'll screw up everyone who has downloaded that commit already	17:44
cubuanic	this is my personal repo, so i'm only one who can download, so it's ok	17:45
	← _grepper left	17:46
	← onigiri_ left	17:46
	← ctusar left	17:47
	→ Cromulent joined	17:48
	→ dreiss joined	17:48
netkevin	thiago: wouldn't it be HEAD^?	17:49
	← SuperRoach left	17:49
netkevin	or HEAD~1	17:49
doener	netkevin: which is the same as HEAD~	17:49
	netkevin: HEAD~ = HEAD~1 = HEAD^1 = HEAD^	17:49
netkevin	I thought ~ needed a number. Huh.	17:49
doener	netkevin: but HEAD~2 == HEAD^^ != HEAD^2	17:49
thiago_home	^ used to need a number	17:49
	oh, wait, ~ used to need a number	17:49
	but that changed a long time ago	17:50
netkevin	"long time ago" in git time.	17:50
	→ karmi_ joined	17:50
netkevin	It moves pretty quickly.	17:50
thiago_home	yes, something like 1 year	17:50
doener	~ means the same as ~1 since 1.5.4.5~11	17:53
	→ Thamster joined	17:54
	← SeySayux left	17:57
	→ relix joined	17:57
	→ ben_h joined	17:59
	← mcdebugger left	18:00
	← karmi left	18:00
	→ hobodave joined	18:02
	→ SjB joined	18:02
	← Theravadan left	18:02
	← crack_head left	18:03
	→ LiamH joined	18:03
	→ crack_head joined	18:05
	← ereslibre left	18:11
	← harinath left	18:12
	← jceb left	18:14
	← aspotashev\|eeepc left	18:15
	→ mcdebugger joined	18:16
	→ rgl joined	18:17
	← Cromulent left	18:20
	→ ph^_ joined	18:21
	→ benn_d joined	18:22
	→ myabc joined	18:25
	→ lucs joined	18:26
	→ RayMorgan joined	18:26
	→ boto joined	18:30
	← myabc left	18:30
	← pstuteville left	18:32
	← mcdebugger left	18:33
	← Thamster left	18:33
	→ onigiri joined	18:33
	← ShadeHawk left	18:34
	← onigiri left	18:36
	← crack_head left	18:37
	→ juanjoc joined	18:38
	← benn_d_ left	18:39
	→ Grahack joined	18:41
	← Grahack left	18:41
	→ offby1` joined	18:45
	→ ricky joined	18:46
	← offby1 left	18:46
	offby1` → offby1	18:46
	→ dodo_the_last1 joined	18:46
	← boto left	18:47
	← muammar left	18:48
	← ph^_ left	18:48
	← johnnyg left	18:48
	← gaveen left	18:48
	→ dnyy joined	18:49
dnyy	When I create a new branch, shouldn't it be empty? It's showing files from the master branch. ;o	18:49
	→ \ask joined	18:50
	→ thomd joined	18:50
thiago_home	dnyy: where did you create the branch from?	18:50
	← HG` left	18:51
dnyy	thiago_home: well, was inside the master branch when i did it. ;o	18:51
	→ crack_head joined	18:51
	→ Akufen joined	18:52
bdrewery	no, it shouldnt be empty	18:52
	you are branching from the exact spot you branched from	18:52
	so it should match exactly	18:52
thiago_home	then you created the branch from exactly the same commit you were in	18:52
	it's a copy of that branch	18:52
bdrewery	I think you're thinking of creating a new project	18:52
dnyy	well, for some reason, i thought that last time i created a branch it was empty	18:53
	→ locellcount joined	18:53
dnyy	welll, the branch is for gh-pagse, so i guess i should remove files not pretaining to that from this branch?	18:53
	← ab3 left	18:53
bdrewery	ah no	18:53
locellcount	How do you roll back a pull? I've tried git reset and git reset --hard	18:54
bdrewery	dnyy: you want a branch with no parent	18:54
dnyy	bdrewery: aaah, how do i do that?	18:54
bdrewery	dnyy: http://madduck.net/blog/2007.07.11:creating-a-git-branch-without-ancestry/	18:56
dnyy	much thanks :D	18:57
bdrewery	it's mainly	18:57
	git symbolic-ref HEAD refs/heads/newbranch	18:57
	rm .git/index	18:57
dnyy	err, well echo ref etc broke something lol	18:58
	← developernotes left	19:00
	davi_ → davividal	19:00
	← aussiegeek left	19:01
dnyy	there _has_ to be an easier way to do this ;-;	19:01
madduck	22 20:57 < bdrewery> git symbolic-ref HEAD refs/heads/newbranch	19:02
	22 20:57 < bdrewery> rm .git/index	19:02
	and no, there is no easier way	19:02
	→ Cromulent joined	19:02
	← dodo_the_last left	19:02
madduck	at least noone ever told me of an easier way, and this post up there has been one of the most popular pages on my site for two years	19:02
	→ relix_ joined	19:03
bdrewery	oh nice haha	19:03
madduck	~35 unique hits/day if memory serves right	19:03
bdrewery	yeah that's the way I've always made them	19:03
	it's really not that hard	19:03
	and most git people will frown on it	19:03
	even though git.git uses the method	19:03
madduck	if you understand what's going on, then there's nothing to say against it	19:03
dnyy	i'm not saying anything against it :p	19:03
madduck	if you don't understand each step and why it's necessary, then you are out of shallow waters	19:03
dnyy	i was just surprised there wasn't a way to do it built into git itself	19:04
	← relix left	19:04
dnyy	just a flag for "no ancestory"	19:04
madduck	dnyy: it's not exactly what you want with a DVCS	19:04
	use a separate repo if you don't want shared ancestry	19:04
	i think the way git.git tracks manpages is confusing and wrong	19:04
netkevin	locellcount: a pull includes a few different things. It updates the remote tracking branch and merges into local branches.	19:05
	locellcount: I thing reset will do what you need if you give it the right reference and commit.	19:05
	→ kumbayo_ joined	19:05
netkevin	But it depends on if you are trying to un-update the remote tracking branch or the local branch	19:06
dnyy	Hrm, well im still getting a "pathspec 'gh-pages' did not match any file(s) know to git."	19:07
netkevin	(commonly there would be no reason to reset the remote tracking branch, btw.)	19:07
	→ relix joined	19:07
	← kumbayo left	19:07
	← juanjoc left	19:07
	→ dpino joined	19:08
locellcount	netkevin: lol what what in the what now? It's ok, I just shrugged my shoulders and decided to live with it, it's only a local dev db anyway. Thanks for the pointers though, just goes to highlight how much I still have to learn about this scm... for which you don't have my thanks :D	19:10
	← ben_h left	19:11
	→ herenbdy joined	19:11
	→ porcnick joined	19:14
	→ khmarbaise_ joined	19:14
netkevin	locellcount: If you're confused about remote tracking vs. local branches, the first thing to learn is 'git branch' and 'git branch -r' (which will show local and remote-tracking branches, respectively, so you can see the difference.)	19:14
	At least that's a starting place.	19:15
	→ JNR___ joined	19:16
	← Cromulent left	19:16
locellcount	netkevin: yeah, good advice	19:17
	→ nessundorma joined	19:18
	← dodo_the_last1 left	19:21
	← relix_ left	19:22
	→ dbell joined	19:23
herenbdy	Why do I recieve this error message when I attempt to push (using gitorious): fatal: The remote end hung up unexpectedly?	19:23
	→ spaceonline joined	19:24
Weiss	herenbdy: are you trying to push using the native Git protocol rather than something like SSH?	19:24
thiago_home	herenbdy: server down	19:24
	herenbdy: what URL are you pushing to?	19:25
herenbdy	"git push [email@hidden.address] master" is what i'm attempting	19:25
	← khmarbaise_ left	19:25
herenbdy	the site instructs me to do so, but there's something more I need to do it seems :/	19:26
thiago_home	the server is probably experiencing problems	19:27
	that has been happening to gitorious every now and then	19:27
	→ Luke-Jr joined	19:28
Luke-Jr	Is there a way to restrict a SSH connection to only do push to a specific git repository	19:28
	?	19:28
	← rnaveiras left	19:28
	← dnyy left	19:29
Ilari	Luke-Jr: Restrict on what side?	19:29
thiago_home	Luke-Jr: yes	19:29
	what are you using on the server side to handle the incoming SSH commands?	19:30
Luke-Jr	OpenSSH	19:30
herenbdy	thiago_home: nothing on my side i'm screwing up then, like my SSH keys?	19:30
thiago_home	after OpenSSH	19:30
	what shell?	19:30
	← spaceonline left	19:30
thiago_home	herenbdy: no, not on your side	19:30
Luke-Jr	I was planning to force a script on the connection of some sort, not allow a shell	19:30
thiago_home	then do that	19:31
Luke-Jr	and restrict git access how? :/	19:32
	← khmarbaise left	19:32
thiago_home	by using a special shell	19:32
herenbdy	hmm, well the beauty of this is that I can still make commits :)	19:32
thiago_home	like git-shell	19:32
charon	Luke-Jr: gitosis?	19:32
herenbdy	just started migrating from svn to git today	19:32
	→ danferranti joined	19:33
Luke-Jr	oh, neat	19:33
	← crack_head left	19:33
	← \ask left	19:34
	→ rudi_s joined	19:36
	← robinr left	19:36
	→ [Po]lentino joined	19:38
	→ djszapi joined	19:40
djszapi	Hello!	19:40
	Can i get notification mail about a git comment, as a member of that project ?	19:40
Aides	a git commit you mean?	19:41
thiago_home	yes	19:41
	install the post-receive hook that will send you the notification	19:41
djszapi	Aides: sorry, commit.	19:42
	← WALoeIII left	19:43
	→ onigiri joined	19:43
	→ developernotes joined	19:43
developernotes	Is anyone aware of a broadcasting feature for git similar to ZeroconfExtension for hg? (http://mercurial.selenic.com/wiki/ZeroconfExtension)	19:44
	← JNR___ left	19:45
	← cubuanic left	19:45
	← danferranti left	19:47
	← jmspeex left	19:48
herenbdy	thiago_home: why is it that I see lots of other people pushing/committing fine in gitorious' feed?	19:48
thiago_home	herenbdy: I can't connect either.	19:48
	→ seanmccann joined	19:48
thiago_home	it's not you.	19:48
herenbdy	oh	19:48
thiago_home	$ git ls-remote [email@hidden.address]	19:49
	Connection closed by 87.238.52.168	19:49
	→ spaceonline joined	19:49
	→ Cromulent joined	19:51
	→ eMyller joined	19:56
	← Krynnotaur left	19:56
	→ khmarbaise joined	19:57
	← khmarbaise left	19:58
	← kyxzme left	19:58
	← dbell left	19:59
	← spaceonline left	20:01
	→ jjuran joined	20:02
	← developernotes left	20:03
	← UukGoblin left	20:03
	→ UukGoblin joined	20:03
	← Cromulent left	20:08
	← netkevin left	20:08
Luke-Jr	How can I tell Git where to find the SSH key? ☺	20:11
doener	you don't. You tell ssh where to find it	20:11
	see man ssh_config	20:11
bdrewery	Luke-Jr: ~/.ssh/config	20:13
	→ relix_ joined	20:13
Luke-Jr	I tell SSH where to find it with command line args	20:13
	I'd rather not have my git config split up	20:13
bdrewery	Luke-Jr: yes, most other apps which use ssh have a param to pass settings to ssh, but git does not	20:14
Luke-Jr	not even a config value?	20:14
djszapi	lvh: one more question	20:14
bdrewery	10 lines of code probably to support it, but nope	20:15
charon	you can easily wrap with a shellscript and set GIT_SSH to that	20:16
djszapi	How can i say the prefix to python setup.py script ?	20:16
	like in case of ./configure --prefix=/usr	20:17
bdrewery	charon: ah cool	20:20
	→ call911 joined	20:20
call911	i have a question: in a merge point which of the two parents points to the original branch?	20:20
	← relix left	20:21
call911	i mean, is it always the first, or the last, or is it random?	20:21
	← madewokherd left	20:23
offby1	I suspect it's not random	20:23
	I also suspect there's an easy way to find out, which I happen to not know	20:23
	I further suspect that at least one of [doener, sitaram, Ilari] knows	20:23
call911	I looked some gitwebs but got confused	20:24
	← onigiri left	20:25
doener	call911: a "merge point"? You mean a merge commit?	20:26
call911	yes	20:26
doener	call911: at least "git merge foo" (assuming that it creates a merge) creates a merge commit with first parent = HEAD, second parent = foo	20:26
call911	thanks!	20:27
	→ sergiodj joined	20:28
	→ onigiri joined	20:28
	→ robinr joined	20:28
	→ johnnyg joined	20:28
	← __iron left	20:31
	→ crack_head joined	20:32
	→ __iron joined	20:33
	← sergiodj left	20:35
	← call911 left	20:35
	→ gonzojive joined	20:40
	→ AkumaStreak joined	20:40
	← kumbayo_ left	20:42
	→ davidw joined	20:43
davidw	is there a way to apply git patches without using git?	20:43
herenbdy	#join Svn	20:44
Adlai	Hi, can I make it so that emerge runs in emacsclient rather than starting a whole new instance?	20:44
herenbdy	woops	20:44
Adlai	s/,/, how/	20:44
charon	davidw: unless they were produced with git-specific options (such as binary) they are 'patch' compatible, so you can usually 'patch -p1 < some.diff'	20:44
Adlai	I've noticed that there's a mergetool.<tool>.cmd config option, but it doesn't seem to have any default value, and it looks as though that's for customized mergetools... it'd help to be able to see what the current command for emerge was, so I could modify that	20:45
Ilari	davidw: Also, if you have git installed, git apply (doesn't need a repo).	20:47
	← thomd left	20:53
	← KiBi_ left	20:55
	→ KiBi joined	20:55
	← eMyller left	20:58
	→ qhoxie joined	21:01
	← frede left	21:02
	← rphillips left	21:04
	→ Sigma joined	21:04
	→ rphillips joined	21:04
	← crack_head left	21:05
	← qhoxie left	21:06
	→ malumalu joined	21:07
almostautomated_	I can see the output on stdout yet when adding the send-email pipe this break...	21:08
	git format-patch -M -s --cover-letter --subject-prefix='PATCH-v2/RFC' --in-reply-to='1249428804.2774.52.camel@GWPortableVCS' --stdout -6 \| git send-email --identity tbf_git --to [email@hidden.address]	21:08
	any ideas?	21:08
	→ jcape joined	21:13
Aides	git send-email apparently does not read stdin by default	21:13
	→ patcito joined	21:13
almostautomated_	well, I'll be a monkey's uncle!	21:13
Aides	if i'm not mistaken, you can pass all your git-format-patch arguments directly to git-send-email	21:14
almostautomated_	I didn't see --cover-letter on there	21:14
	almostautomated_ re-checks	21:14
	→ corni joined	21:14
almostautomated_	nope don't see it in the auto-completion	21:15
	almostautomated_ checks the man	21:15
Aides	these commands usually invoke each other and pass arguments	21:15
	yeah, it works	21:16
almostautomated_	ahh cool	21:16
	I'll try doing the whole thing as a send-email then	21:16
	→ beni- joined	21:17
Aides	err, someone screwed my muttrc	21:17
almostautomated_	ya know... that just sounds wrong!	21:18
Aides	screwing muttrcs? yeah...	21:18
	Aides wonders if git-send-email can send nice pull requests	21:20
	→ spaceonline joined	21:21
almostautomated_	Aides: looks like that is working, thanks for the help.	21:21
Aides	you're very welcome ;)	21:22
	→ diegoviola joined	21:24
davidw	charon, Ilari: thanks!	21:25
	→ machride1 joined	21:28
	← machrider left	21:28
	→ crack_head joined	21:29
	machride1 → machrider	21:30
	← crack_head left	21:31
	← mbroeker left	21:32
	→ mbroeker joined	21:34
	← altrux left	21:35
	→ ctusar joined	21:36
	← djszapi left	21:41
	← Sigma left	21:41
	← rphillips left	21:43
	← beni- left	21:46
	→ altrux joined	21:47
	joshthecoder → joshthecoder_afk	21:47
	← hobodave left	21:50
	→ rphillips joined	21:50
	→ Michael_ joined	21:51
	→ internet joined	21:51
	← MisterN left	21:51
	→ diego joined	21:53
	← hydrogen left	21:53
	→ reidab joined	21:53
	← lohrun left	21:53
	diego → Guest42916	21:53
	← reidab left	21:54
	→ reidab joined	21:54
	→ scientes joined	21:55
	→ adam_0 joined	21:55
	← [Po]lentino left	21:56
	← markelikalderon left	21:57
adam_0	is there an easy way to git mv an entire directory and subdirectories?	21:57
thiago_home	aside from using git mv?	21:57
adam_0	well I can't figure out how to git mv it	21:57
	← ctusar left	21:58
adam_0	heh	21:59
	→ markelikalderon joined	21:59
adam_0	just did it and it worked	21:59
	for some reason it was giving me trouble earlier	21:59
	→ Cromulent joined	21:59
	← adam_0 left	21:59
	← dpino left	22:00
	← kibombo left	22:00
	← internet left	22:03
	← blankthemuffin left	22:04
	→ syd_ joined	22:04
	← xanonus left	22:04
	← dreiss left	22:04
	← hzlocky left	22:05
	← SjB left	22:05
	→ SjB joined	22:05
	← Cromulent left	22:06
	→ archis joined	22:06
	← archis left	22:06
	→ thomd joined	22:07
	← relix_ left	22:07
	← Michael_ left	22:09
	← davidw left	22:09
	← thomd left	22:11
	← spaceonline left	22:13
	→ nadavoid joined	22:13
	← thiago_home left	22:18
	→ robin joined	22:21
	← rudi_s left	22:21
	joshthecoder_afk → joshthecoder	22:23
	→ rudi_s joined	22:24
nadavoid	If I want to push my repository (the .git dir) and the files that it is managing to a server that doesn't have git installed on it, what are my options?	22:24
	I want to do something like "git push" or "git upload"	22:24
Ilari	nadavoid: HTTP push? Or push to local repo and upload that somehow? Sshfs (no idea if that works).	22:25
kollektiv	is there any way to make 'git log' word wrap. it goes off the screen in cygwin	22:25
nadavoid	Ilari: I might look into HTTP push, but I think that's doubtful, since (I think) the HTTP server needs to be set up special to receive it.	22:27
	→ devnotes joined	22:27
nadavoid	SSHFS sounds promising.	22:27
patrikf	if you have ssh then you can install git as well...	22:27
nadavoid	patrikf: even if I don't have root access?	22:27
patrikf	sure, you can always install it inside your home directory	22:28
	(the simplest option being to kindly ask your admin to install git :-))	22:28
	← devnotes left	22:28
nadavoid	patrikf: would it normally install a custom compile on the server?	22:28
	I'm thinking of things like shared hosts.	22:28
herenbdy	anyone else having problemson gitorious?	22:28
patrikf	nadavoid: I think shared hosts that offer SSH access but aren't vservers are rare	22:29
	← d0k left	22:29
patrikf	nadavoid: but yes, you can place the binaries into ~/bin or something similar (and adjust your path)	22:29
nadavoid	patrikf: OK, thanks, that's helpful.	22:30
Ilari	nadavoid: Just remember that $PATH for interactive and non-interactive logins may not be the same (git uses non-interactive).	22:30
	→ developernotes joined	22:30
nadavoid	ok	22:30
	← developernotes left	22:31
	← fualo_ left	22:32
	← reidab left	22:32
nadavoid	Another +1 for the #git channel. Thanks everyone.	22:32
	→ fualo joined	22:32
	← corni left	22:32
	→ reidab joined	22:32
	← robinr left	22:34
xenoterracide	is there a tutorial somewhere on how I would rewrite all commit authors? I've got commit's from root@ and username@ I want to change those to what I put in .gitconfig	22:35
	→ corni joined	22:36
herenbdy	anyone else having problems on gitorious?	22:36
almostautomated_	xenoterracide: man filter-branch should help	22:36
	down in the examples section	22:36
	→ EricInBNE joined	22:37
	→ EricInBNE_ joined	22:37
	← EricInBNE_ left	22:37
xenoterracide	almostautomated_: thanks	22:38
almostautomated_	Is it normal, when applying a series of posted patches to need to remove [0/xx] from the mbox before doing git am?	22:38
	since it doesn't have anything in it I'd assume most people don't bother putting it in the mbox to get applied	22:39
	But I just want to be sure	22:39
	← corni left	22:40
almostautomated_	and do people generally find it easier to work with shallow or deep chained series?	22:44
	nm, on the first question... just tested three different setups and each required removal of the cover letter... so that is a 'yes'	22:46
	← tvw left	22:47
	← Aikawa left	22:48
	→ ereslibre joined	22:48
	← robin left	22:50
	← repror___ left	22:56
	← WinterWeaver left	22:58
	→ diegovio1a joined	22:58
	→ diego joined	22:59
	diego → Guest75424	23:00
	→ Altreus joined	23:01
Altreus	Does anyone know svn well enough to know whether my .svn directories contain enough data for me to create a git repository from a subversion working copy?	23:01
	→ WALoeIII joined	23:01
Altreus	Seeing how I have lost and re-lost the svn repo pretty thoroughly	23:02
	← Guest42916 left	23:02
	← cloudhead left	23:02
Altreus	oh I tell a lie it's right here	23:03
	← Altreus left	23:03
cbreak	do subversion trees even contain the history?	23:04
herenbdy	I don't think so	23:04
	← jspiros left	23:05
	→ madewokherd joined	23:05
	→ diegovio2a joined	23:07
	← WALoeIII left	23:09
	→ eletuchy joined	23:11
	→ an0maly joined	23:11
	→ panpainter joined	23:14
	← diegoviola left	23:14
	← diegovio1a left	23:14
	→ vashism69 joined	23:14
vashism69	ey if you guys are bored try facebook zombies :D - http://apps.facebook.com/zombies/links.php?r=719927515&nref=st	23:15
	← vashism69 left	23:15
Ilari	No, they don't.	23:17
	→ ctusar joined	23:18
	← karmi_ left	23:19
	→ suitdetony joined	23:19
	← mbroeker left	23:20
	→ WALoeIII joined	23:20
	→ jspiros joined	23:20
	← diegovio2a left	23:20
suitdetony	cuando nací me dijo la partera: hay mijo cuando estés grande pobre de la chava que te de las nalgas si que va a gritar en la primera noche que se meta contigo ya que pareces burro jajajaja	23:21
	→ diegoviola joined	23:22
	← chris2 left	23:26
	← priidu left	23:27
	← g3d left	23:28
	← warthog9 left	23:28
	← ricky left	23:31
	← diegoviola left	23:33
	→ eMyller joined	23:37
	← malumalu left	23:40
Ademan	is it possible to diff your working copy with the latest commit? If i understand correctly, one could accomplish this by adding the working copy, and somehow referencing these staged objects... but I don't know how I'd do that...	23:43
	→ hzlocky joined	23:46
teuf	Ademan: are you looking for git diff ?	23:46
	→ benn_d_ joined	23:47
	← ph^ left	23:53
	← reidab left	23:53
	→ ianmcorvidae joined	23:54
	← onigiri left	23:56
	← ianmcorvidae\|alt left	23:58
	← rudi_s left	23:58

Logs Search ←Prev date Next date→

Channels Documentation