IRCloggy #git 2011-04-14

Provider of IRC logs since 2005.

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

	2011-04-14
	→ RobertLaptop joined	00:01
	← javaanse_jongens left	00:01
	→ Swimming_bird joined	00:01
	→ javaanse_jongens joined	00:02
	← luxurymode left	00:02
	→ jrmuizel_ joined	00:04
	← cjz left	00:05
	← devth left	00:06
	← sako left	00:06
	→ derek_ joined	00:06
	← javaanse_jongens left	00:06
	← derek_ left	00:06
	→ vanmik joined	00:07
	← logik-bomb\|away left	00:08
	← Anubisss left	00:08
	← sftp left	00:09
	← jaeckel left	00:09
	→ Anubisss joined	00:09
	← NetHawk left	00:10
	→ javaanse_jongens joined	00:10
	← mattikus left	00:10
	→ sftp joined	00:12
	→ Predominant joined	00:12
	← jrdavid_ left	00:14
	→ jrdavid joined	00:14
	→ rchavik joined	00:15
	→ variable joined	00:15
	← kenichi left	00:15
	→ wo9x joined	00:15
	→ rondon joined	00:17
	→ sftp_ joined	00:18
	← sftp left	00:18
	→ NetHawk joined	00:19
	← smolyn left	00:21
	← wo9x left	00:22
	← drake01 left	00:22
	→ drake01 joined	00:23
	← wilmoore_ left	00:25
	← nb left	00:25
	← linusoleander left	00:26
	← sontek left	00:27
	← rphillips left	00:27
	← bottiger left	00:27
	← SimonP86 left	00:29
deitarion	Where do the Git docs explain the difference (as far as how they get set) between the author and committer fields in `git log --format=fuller`?	00:29
	← systemaddict left	00:30
	← javaanse_jongens left	00:31
Arrowmaster	probably in git commit man page	00:31
	→ javaanse_jongens joined	00:32
deitarion	Ahh. --author. Thanks.	00:32
	← beyondcreed left	00:34
	→ sparrW joined	00:34
	← javaanse_jongens left	00:36
juliohm	what means ..HEAD notation?	00:37
	git format-patch ..HEAD	00:37
	→ javaanse_jongens joined	00:37
	← beatak left	00:40
	→ SethRobertson joined	00:40
	← ramshacklecpu left	00:41
	← NetHawk left	00:43
sparrW	given a git://foo.git URL, how do i check out a copy of it?	00:44
	← larsl left	00:44
	← rendar left	00:46
	← dharrigan left	00:46
	→ dharrigan joined	00:48
	← dharrigan left	00:48
	→ dharrigan joined	00:48
juliohm	sparrW, just do `git clone git:foo.git localDir`	00:49
	← aef left	00:50
	← Swimming_bird left	00:50
	→ Fandekasp joined	00:50
	→ aef joined	00:51
	← hyperair left	00:55
sparrW	thanks juliohm	00:57
	→ jamescarr joined	00:58
	← jamescarr left	00:58
	→ jamescarr joined	00:58
jamescarr	hey when I fork someone's repository is it possible for me to pull new changes they make into my fork?	00:58
	→ phantomcircuit joined	01:00
	→ rayners joined	01:02
	→ xiangfu joined	01:02
	← deitarion left	01:03
	← cbz left	01:05
	← towski left	01:05
	← GodEater left	01:07
	→ bottiger joined	01:07
	→ cbz joined	01:08
	→ GodEater joined	01:09
	← GodEater left	01:09
	→ GodEater joined	01:09
	← Anubisss left	01:12
	→ Anubisss joined	01:13
	← drake01 left	01:14
	→ poseidon joined	01:15
	→ Buglouse joined	01:15
	← orafu left	01:16
	← Evious left	01:16
	→ sontek joined	01:16
	→ orafu joined	01:17
	← Chaser left	01:18
	← javaanse_jongens left	01:19
rchavik	jamescarr, add their repo as another remote, and you can pull from it	01:19
jamescarr	I just did a fetch upstream	01:20
	→ mattikus joined	01:20
	← jfkw left	01:20
	→ javaanse_jongens joined	01:21
rchavik	git pull upstream/master will merge it into your current branch	01:21
	← cloudpopup left	01:21
rchavik	or git rebase upstream/master if you want to fast forward	01:21
	someone please correct me if i'm wrong :)	01:22
	← panike left	01:24
	← gigamonk` left	01:25
	← juliohm left	01:25
	← jamescarr left	01:26
	← spaceonline left	01:26
	→ Remixman joined	01:26
	→ seivan joined	01:27
	← mdel_ left	01:27
	→ amondracorp joined	01:28
	← javaanse_jongens left	01:31
	← Vampire0 left	01:31
	← hachiya left	01:32
	← oc80z left	01:32
	→ hachiya joined	01:32
	← squiggy left	01:32
	→ javaanse_jongens joined	01:32
	→ oc80z joined	01:33
	→ Vampire0 joined	01:33
	→ nb_ joined	01:34
	→ rphillips joined	01:35
	← leavittx left	01:38
	→ Fuzzy joined	01:39
	← Fuzzy left	01:40
	→ shiba_yu36 joined	01:40
	→ CloudySunshine joined	01:40
	eyda\|mon → eydaimoon	01:41
	ChanServ set mode: +v	01:41
CloudySunshine	Is it possible to cluster git servers in an active/active method where they locally mirror each other perfectly?	01:41
	← SethRobertson left	01:43
	offby1 scratches head	01:45
	→ Vampire0_ joined	01:45
offby1	I assume you could give each of them a post-commit hook that pushes to the other fellow	01:45
	→ alberto56 joined	01:46
	→ ChUrP joined	01:46
CloudySunshine	that sounds really easy	01:46
	→ SethRobertson joined	01:47
	← Vampire0 left	01:48
bremner	CloudySunshine: well, one master many slaves is easy	01:48
	offby1 enjoys that arrangement at home	01:49
	← xiangfu left	01:49
	← theoretical left	01:51
	← Remixman left	01:56
	→ drake01 joined	01:57
	← Fandekasp left	01:57
	← javaanse_jongens left	01:58
	→ DarthGandalf joined	01:58
	tizzo-afk → tizzo	01:58
	→ xiangfu joined	01:58
	→ Fandekasp joined	01:58
Fandekasp	hello	01:59
	I'm having some confusions with my git server	01:59
	→ javaanse_jongens joined	01:59
offby1	!repro	02:00
jast	please paste (e.g. using http://git.pastebin.com/) a transcript of your terminal session in which the problem occurred so that we can troubleshoot more easily. [automatic message]	02:00
	← amerine_ left	02:00
	← orafu left	02:00
	← sechrist left	02:01
Fandekasp	I installed gitolite, and after reading this http://sitaramc.github.com/gitolite/doc/1-INSTALL.html I can now git clone an empty directory called gitolite-admin from my client, but I don't understand by reading the doc what is the use of this project, and I'd like my own project to be git clonable like this. Do someone know about gitolite and git servers ?	02:01
bremner	yes	02:01
Fandekasp	cool :)	02:01
	← erichammond left	02:01
bremner	it sounds like you need to read more of the docs	02:02
	like adding users and repos	02:02
Fandekasp	in the admin doc	02:02
bremner	also, gitolite-admin should not be empty	02:02
	← avdg left	02:02
Fandekasp	arf does that mean I failed somewhere during the installation?	02:02
	→ intripoon_ joined	02:03
bremner	perhaps.	02:03
	you might also read the "is gitolite overkill" document	02:03
	→ orafu joined	02:04
Fandekasp	I have a .git directory in the gitolite-admin folder, sorry	02:04
	ok for the doc, thanks	02:05
	→ brianpWins joined	02:05
	← intripoon left	02:05
	→ jamescarr joined	02:06
Fandekasp	ah bremner the gl-setup didn't work the first time, and now my gitolite-admin has a conf and keydir :)	02:07
bremner	better	02:07
	tizzo → tizzo-afk	02:08
	→ cutmail joined	02:09
	← Anubisss left	02:10
	← bitkiller left	02:10
	← L42y left	02:10
	← boombatower left	02:14
	→ amerine joined	02:21
	→ jfkw joined	02:21
	→ mosno joined	02:22
mosno	how can i see the difference between changes on the same line between two commits?	02:23
	ie. i'm having to visually scan the line for changes	02:23
	→ petdance joined	02:26
	→ Yuuhi` joined	02:26
	→ sebersole_ joined	02:27
	← sebersole left	02:28
	← Yuuhi left	02:28
	← Elfe left	02:28
	→ Elfe_ joined	02:29
	→ stliu joined	02:30
offby1	mosno: use a better diff viewer.	02:30
	for example, the one built into emacs highlights those changes within the line.	02:31
	I would have thought that simply typing "git diff" would do some fancy highlighting like that.	02:31
	hmm, apparently not. Pity	02:31
	→ luxurymode joined	02:32
	← aaronlevy left	02:32
	← tjholowaychuk left	02:33
	← hvgotcodes left	02:33
	← pedrobelo left	02:34
	→ robotarmy joined	02:34
	← lu_zero left	02:36
Fandekasp	bremner I've added a user and a project in my conf/gitolite.conf, then ci and push, then git clone the new project, and it works. But on the server, I'd like to symlink this new project to the existing one, and I can't find where are the sources for the new project. There is a /home/git/repositories/project.git but do you know where is my code ?	02:37
	→ towski joined	02:37
	← javaanse_jongens left	02:39
	→ MikhX joined	02:39
	← alberto56 left	02:39
	→ javaanse_jongens joined	02:41
	→ Swimming_bird joined	02:41
offby1	Fandekasp: it's there, but since your repositories are "bare", they're not in a convenient or obvious form.	02:42
	They're all in subdirectories named .git/objects, and they're compressed	02:42
Fandekasp	erf :(	02:43
offby1	why is that a problem?	02:43
Fandekasp	Because the git repository is used by track in another location	02:43
	← sebersole_ left	02:43
offby1	I don't understand what that means	02:43
Fandekasp	so I'd like a symlink somewhere	02:43
	Trac sorry,	02:43
	→ hyperair joined	02:43
offby1	does trac require a working tree with all the files?	02:43
Fandekasp	That's how I've configured it, but I'm not an expert, maybe there is a better way to do	02:44
offby1	I don't know enough about trac to advise you.	02:44
	← javaanse_jongens left	02:45
	→ ak0rn\|2 joined	02:45
Fandekasp	So I'd need to download my project in the client, then update the gitoline project and push it. And then modify my trac config to read gitoline instead of the current git project, am I right ?	02:45
	→ sebersole joined	02:46
	→ javaanse_jongens joined	02:47
	→ azazul joined	02:48
azazul	hi	02:48
	→ linoj joined	02:48
	← ak0rn left	02:48
linoj	hiya, how would i do a git pull of a different branch than i presently have?	02:48
frogonwheels	linoj: a git pull is git fetch + merge. in the git fetch part, all the branches are pulled.	02:49
azazul	i'm looking for some software that could pull autimatically updates from multiple gite repositories to my local file system	02:49
frogonwheels	linoj: do a git branch -r to see the remote branches.	02:49
azazul	( on windows )	02:49
	← hcl2_ left	02:49
azazul	i there any such tool , or will i have to make something like that on my own ?	02:50
frogonwheels	linoj: if say one is called 'remotes/source/feature' you can do git checkout feature and it will create a new branch 'feature' that has remotes/source/feature as it's upstream remote-tracking branch	02:50
	← javaanse_jongens left	02:50
linoj	thx	02:50
	→ hobodave joined	02:51
frogonwheels	azazul: batch file?	02:51
	→ javaanse_jongens joined	02:52
frogonwheels	azazul: have a batch file that reads a list of remotes / target directories and put it into the command scheduler. not too hard.	02:52
azazul	yeah , i guess	02:52
	i was just hoping that there was a small app that did it for me	02:53
frogonwheels	azazul: I have one (on windows) that pulls my repos to a backup drive.	02:53
azazul	well , shell scripting is easy , that isnt a problem	02:53
frogonwheels	azazul: I guess it's a unix thing. why bother creating an app that does it specifically when you have batch files and a task scheduler?	02:53
azazul	one thing : what should i google for , if i want the script to automatically create a log for "last updates" ?	02:54
	as in , list of files that have been changed	02:55
	← javaanse_jongens left	02:55
azazul	or , is was there a built in functionalit and i dont have to worry about that	02:55
	*functionality	02:55
	← kpreid left	02:55
	← jamescarr left	02:55
frogonwheels	azazul: well how about you do a fetch, then do a log from origin/branch to branch then do a merge of origin/branch into branch ..	02:56
	← glitterfang left	02:56
infid	how do you get 'git gui here' to show up in nautilus (gnome's gui file explorer)?	02:56
	→ javaanse_jongens joined	02:57
	← dreiss left	02:57
azazul	hmm .. i better do this in the morning	02:58
	← petdance left	02:59
	→ Openfree` joined	03:00
Openfree`	is there any option to disable create remote git branch on my server?	03:00
	→ intripoon joined	03:00
azazul	i gues i will get back to this idea a bit later	03:01
	at least i know that it will be possible	03:01
	← azazul left	03:01
Openfree`	It's quite annoy when people type something wrong, a new branch created, I just need this option to give them a warning!	03:02
	← intripoon_ left	03:03
infid	is msysgit better than tortoisegit	03:04
frogonwheels	infid: tortoisegit uses msysgit afaik	03:05
	infid: tortoisegit is a bit too 'tortoisesvn' for my liking. (ie git concepts shoe-horned into svn concepts). It really depends on your usage pattern.	03:05
	infid: personally, I don't actually use explorer to visualise my work tree.	03:06
	infid: I prefer to see 'what's changed in my project' rather than 'what's the status of this file'	03:06
	→ wilmoore joined	03:07
frogonwheels	infid: so I tried tortoise but discarded it pretty quickly.	03:07
Openfree`	seems NO that option, at least I'm looking at the code...	03:07
infid	frogonwheels: well i installed msysgit then i installed tortoisegit and i'm not seeing any tortoisegit options showing up at all in explorer	03:10
	← texel left	03:11
infid	and i want to keep the 'git bash here' option, so if they cant play well together i'm going to choose msysgit	03:11
mosno	offby1: cool thanks	03:11
	next question... i have a commit in repo A that i want to apply in repo B... how?	03:12
offby1	depends	03:12
	does the commit have ancestors in repo B?	03:12
eydaimoon	how can I remove a file from the repo without removing it locally?	03:12
offby1	eydaimoon: "git rm --cached path/to/file"	03:13
	then "git commit"	03:13
eydaimoon	offby1: thank you	03:13
Captain`Ramen	infid: have you tried gitextensions?	03:13
infid	no	03:13
offby1	eydaimoon: of course you could also back it up, then do a simple "git rm path/to/file; git commit", then restore. But this is more elegant.	03:13
Captain`Ramen	give it a spin.	03:14
mosno	offby1: yes	03:14
	offby1: but i only want to apply changes from a single commit, not all the commits leading up to it	03:14
	offby1: as if i scp a single file from host A to host B, then do a commit in host B	03:15
	→ t0rc joined	03:15
	→ petdance joined	03:15
offby1	mosno: ah, I think you want to fetch from A to B, but then "git cherry-pick"	03:15
	it's really two separate issues:	03:15
	1) how do you get that commit's "patch" onto some other branch	03:16
	2) how to get stuff from a branch on one box to another box	03:16
	← Error404NotFound left	03:16
eydaimoon	offby1: yeah, I like this better, and it's easier. Also I should have figured it was cached... been reading the progit book a bit :)	03:17
	offby1: thanks :)	03:17
	← javaanse_jongens left	03:18
mosno	offby1: thanks, i will look into cherry-pick. i've done no work with remote repos yet, so i have some other learning to do	03:19
	→ javaanse_jongens joined	03:19
	← kukks left	03:25
Fandekasp	What are git control files ?	03:25
	← zivester_ left	03:26
	→ zivester_ joined	03:26
	→ cenuij joined	03:26
	← cenuij left	03:26
	→ cenuij joined	03:26
infid	i thought git rm --cached only takes things out of the staging area, you're saying it deletes it out of the repo?	03:27
	ilteris_ → ilteris	03:28
Fandekasp	When I try to configure trac, the hidden git folder in the project repository is read, but the git folder with compressed data isn't : http://paste.pocoo.org/show/371335/. So in my /home/git/repositories/servicepanel.git/, what would be the Git control files according to you ?	03:28
	→ majonto joined	03:29
	← majonto left	03:29
	→ majonto joined	03:30
	← majonto left	03:30
	→ jon_at_dh joined	03:31
offby1	shrug	03:31
Fandekasp	:(	03:34
	← MikhX left	03:35
	→ MikhX joined	03:36
	← linoj left	03:36
	← mattikus left	03:38
	← seivan left	03:39
	Predominant → hotcalichick6969	03:39
	→ seivan joined	03:40
	→ jaeckel joined	03:41
	→ sechrist joined	03:42
	→ Remixman joined	03:47
	← MikhX left	03:51
	→ MikhX joined	03:54
	→ cakehero joined	03:54
cakehero	Two mathematicians were outside a bar and observed two people go in, and later three people left. One mathematician says to the other "if one more person goes in it will be empty!"	03:55
	Wait - wrong room	03:55
	← sechrist left	03:55
nevyn	I laughed..	03:55
cakehero	git reset --hard	03:55
	← orafu left	03:55
cakehero	There, that's better	03:55
	→ orafu joined	03:56
	→ Error404NotFound joined	03:56
	← Error404NotFound left	03:56
	→ Error404NotFound joined	03:56
	← MikhX left	03:56
	→ Goliath\|2 joined	03:56
	← phantomcircuit left	03:56
	← iAmerikan left	03:58
	← Goliath left	03:59
offby1	ha! that's a good oe	04:00
	"git reset --hard" ... hyuk	04:00
	offby1 slaps thigh	04:00
infid	how do 'export' a git working dir so that it doesn't contain the .git? do i have to just make a copy then delete the .git? (wanting to script this)	04:00
	→ sechrist joined	04:00
	→ iAmerikan joined	04:00
offby1	infid: man git-archive	04:00
jast	infid: the 'git-archive' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-archive.html [automatic message]	04:00
	→ mattikus joined	04:00
offby1	ooh, ooh, I win	04:00
	offby1 dances like an idiot in the end zone	04:00
infid	doesnt git archive force it to be a compressed thing though	04:00
	i want it to be uncompressed	04:00
	→ Dave^\| joined	04:01
	← sechrist left	04:02
offby1	dunno what made you think that	04:02
	read the man page	04:02
	→ sechrist joined	04:02
offby1	oh, I guess if you ask for a "zip" archive, it indeed compresses, seeing as how that's what zips are all about.	04:03
	But if you ask for tar, e.g., it doesn't compress.	04:03
infid	the man page did make me think that	04:03
offby1	so: git archive --prefix=whatever HEAD \| (cd .. ; tar xf -)	04:03
	that might not be exactly right, but it's close.	04:03
infid	Format of the resulting archive: tar or zip. If this option is not given, and the output file is specified, the format is inferred from the filename if possible (e.g. writing to "foo.zip" makes the output to be in the zip format). Otherwise the output format is tar.	04:03
	→ Dave^\|\| joined	04:03
Arrowmaster	tar is not a compressed format	04:04
offby1	I dunno, _I_ don't get "compression" from reading that, but maybe it's just m,e.	04:04
	← Dave^\|\| left	04:04
	← OOPMan left	04:04
	← hobodave left	04:04
infid	i dont want it compressed or otherwised reformatted, but i guess i can make do with untar'ing it back	04:04
	→ hobodave joined	04:04
Arrowmaster	you can pipe the output of git archive directly to tar in that case	04:05
	← Barry-Nichols left	04:05
	← seivan left	04:05
	← orion left	04:06
	offby1 suspects infid is confused about something, but cannot think of exactly what	04:06
infid	i want the equivalent of cp -r foo/ bar/ ; rm -rf bar/.git	04:09
	→ sunaku joined	04:09
	hotcalichick6969 → Predominant	04:10
	← jaeckel left	04:11
	→ orion joined	04:12
	→ avinashhm joined	04:14
	← brianpWins left	04:14
	QshelTier → Bombe	04:15
	→ phantomcircuit joined	04:19
	← Finalflash50 left	04:20
	→ Finalflash50 joined	04:21
	← jon_at_dh left	04:21
	← Finalflash50 left	04:21
	→ OOPMan joined	04:23
	→ jon_at_dh joined	04:26
	→ dr0id joined	04:26
	← AkumaStreak left	04:28
offby1	did we not just give you the equivalent? Am I misunderstanding something?	04:29
	Have you actually _tried_ git-archive?	04:29
	← hackez left	04:30
	← cutmail left	04:30
	← Linex left	04:30
	← fisted left	04:31
	→ bindaas joined	04:32
	→ xerora joined	04:32
	← cakehero left	04:33
	→ cutmail joined	04:33
	← t0rc left	04:34
	← cutmail left	04:34
	← [intra]lanman left	04:35
infid	nah i got it thanks	04:35
	← sebersole left	04:37
infid	so git revert doesn't work on uncommitted files, like svn?	04:37
	some googling is saying i have to do 'git checkout -- files' to revert local changes, is that right?	04:38
	what the heck is "--"	04:38
	→ hackez joined	04:38
wereHamster	infid: to separate options from the rest of the arguments	04:40
	→ Yuffster joined	04:41
wereHamster	actually in this case to tell git to treat files as a pathspec and not a branch	04:41
offby1	"git revert" is very different from "svn revert".	04:42
	→ roop joined	04:42
offby1	It means "make a new commit whose changes counteract those from some other commit".	04:42
	"negate" ... "cancel out" ... whatever term you like	04:42
	→ cutmail joined	04:42
offby1	"nuke"	04:42
	pew pew	04:42
	offby1 takes a chill pill	04:42
	← drake01 left	04:43
	← rayners left	04:43
	← dr_lepper left	04:44
infid	wereHamster: where can i read about all the contexts of -- ?	04:44
	← Yuffster left	04:45
wereHamster	infid: in man git-checkout	04:46
jast	infid: the 'git-checkout' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-checkout.html [automatic message]	04:46
	← dharrigan left	04:46
infid	thanks	04:46
	→ dharrigan joined	04:47
	→ CJack joined	04:48
	← jrmuizel_ left	04:48
infid	actually that man page doesnt explain the contexts	04:49
	← oc80z left	04:49
	→ oc80z joined	04:50
	→ freimatz joined	04:50
	ChanServ set mode: +v	04:50
	← petdance left	04:50
infid	and googling for this doesnt work: git " -- "	04:50
CJack	ok .. do i have a voice now ?	04:51
	yay	04:51
dr0id	man pages are pretty confusing until you get a good grasp over git by reading books like progit or book.git-scm.com :)	04:51
CJack	Hi all	04:51
dr0id	CJack: give me voice :P	04:51
	← ISF left	04:51
	← juangiordana left	04:51
CJack	drOid ... love ProGit .. a major help to learning	04:51
	→ ISF joined	04:51
	→ fisted joined	04:52
	← robotarmy left	04:52
CJack	dr0id: voice as in this channel's anti-spam process	04:52
dr0id	I do know what voice is, heh :P	04:53
CJack	infid: what was the question ?? .. i got here late	04:53
	dr0id: voice as in Mode +v	04:53
	→ dr_lepper joined	04:54
	← EricInBNE_ left	04:55
wereHamster	in the double-dash is used in pretty much any gnu commandline app	04:56
Arrowmaster	dr0id: read the topic	04:56
wereHamster	infid: and the man page doesn't explain what it means specifically, but it gives you an example where to use it	04:57
dr0id	Arrowmaster: ?	04:57
	← ISF left	04:57
	→ Linex joined	04:57
Arrowmaster	the -- means stop processing the comandline for options, for git checkout its like a seperator between hash and filenames so it wont try to parse the filename as a hash	04:58
	→ EricInBNE joined	04:59
	← zcopley left	04:59
	← cxadams left	04:59
	← ceej left	05:01
	→ ThomasLocke joined	05:04
	← meder left	05:05
	← kanru left	05:05
	→ meder joined	05:05
	→ endra` joined	05:05
	→ cordoval_ joined	05:05
	→ kanru joined	05:05
	→ bwwww joined	05:05
	← endra left	05:06
	→ ExtraSpice joined	05:06
	← VladDrac left	05:08
	← cutmail left	05:08
	→ VladDrac joined	05:08
	→ ph^ joined	05:08
	→ cutmail joined	05:11
	→ krosenvold joined	05:12
	← sftp_ left	05:15
	← variable left	05:15
	→ Jake__ joined	05:16
	→ jutaro joined	05:16
	→ robotarmy joined	05:18
	→ adaptee joined	05:18
	← Spockz left	05:19
	→ int3__ joined	05:20
	← davey left	05:21
	← javaanse_jongens left	05:22
	→ seivan joined	05:22
	→ javaanse_jongens joined	05:23
	→ jeffwheeler joined	05:24
	← jutaro left	05:25
	← javaanse_jongens left	05:26
	← jeffwheeler left	05:27
	→ davey joined	05:28
	davey → Guest60131	05:28
	← brodo left	05:28
infid	thanks	05:29
	← ketas left	05:30
	→ javaanse_jongens joined	05:30
	← ChUrP left	05:35
	← madewokherd left	05:36
CJack	Arrowmaster: yes, 2 dashes with spaces on each side is saying: "Here is the end of the options, the next item will be a filename/path"	05:38
	← rideh left	05:38
	← sechrist left	05:38
	→ slonopotamus joined	05:39
	← amerine left	05:39
	← Zenith77 left	05:39
	→ airborn joined	05:41
	→ tyteen4a03 joined	05:46
	→ linusoleander joined	05:50
	← xiong left	05:51
	← cenuij left	05:52
	→ cenuij joined	05:52
	← cenuij left	05:52
	→ cenuij joined	05:52
	→ rideh joined	05:53
	→ Garen_ joined	05:55
	→ TrineLocke joined	05:57
	→ thiago joined	05:57
	→ ketas joined	05:58
	← Garen left	06:00
	← slonopotamus left	06:00
	← linusoleander left	06:01
	← phaedrix left	06:01
	→ slonopotamus joined	06:02
	← sunaku left	06:04
	← CloudySunshine left	06:04
	← bwwww left	06:05
	← bauruine left	06:06
	← Fissure left	06:08
	→ guardian joined	06:10
	→ cbreak_work joined	06:11
	← hobodave left	06:12
	← seivan left	06:13
	← towski left	06:14
	← jense_ left	06:14
	← jfkw left	06:15
	← KKA left	06:15
	→ mari00sh joined	06:15
	→ KKA joined	06:16
	→ leo2007 joined	06:17
	→ Fissure joined	06:21
	← guardian left	06:21
	← rubydiamond left	06:22
	← slonopotamus left	06:23
	→ CloudySunshine joined	06:23
	→ psoo joined	06:23
	← Kudos left	06:24
	← karihre left	06:24
	← cordoval_ left	06:24
	→ cordoval_ joined	06:25
	→ seivan joined	06:25
	→ karihre joined	06:26
	→ frsk joined	06:26
	← LouisJB left	06:27
	→ rubydiamond joined	06:27
	→ towski joined	06:28
	→ amerine joined	06:29
	← amerine left	06:32
	→ Garen joined	06:32
	→ simplechat joined	06:33
	→ profynet joined	06:34
	→ larsmoil joined	06:35
	← ak0rn\|2 left	06:35
	← Garen_ left	06:35
	→ ak0rn\|2 joined	06:35
	→ joel_falcou joined	06:36
	→ Kudos joined	06:36
	→ michael_lf joined	06:37
	→ namemon joined	06:37
	→ d0k joined	06:38
	→ dermoth\|home joined	06:38
	← dermoth\|home left	06:38
	← profynet left	06:40
	← alip left	06:41
	→ LouisJB joined	06:42
	← javaanse_jongens left	06:42
	→ saschagehlich joined	06:43
	→ aminpy joined	06:44
	→ vdv joined	06:44
	← TrineLocke left	06:45
	→ alip joined	06:45
	→ javaanse_jongens joined	06:45
	→ TrineLocke joined	06:45
	→ tyfighter joined	06:46
	→ singlow joined	06:47
	→ Deesl joined	06:48
	→ brianpWins joined	06:48
wereHamster	CJack: in git it's not always used in this exact sense. Rather it's a generic delimiter, it doesn't need to be options/filename but can also be tree-ish/pathspec etc	06:48
	→ steph021 joined	06:49
	← ph^ left	06:49
	← steph021 left	06:49
	→ steph021 joined	06:49
	← LouisJB left	06:49
	← vdv left	06:50
	← walf443 left	06:50
	→ walf443 joined	06:52
	← airborn left	06:53
	→ giallu joined	06:56
	→ jceb joined	06:56
	→ pdelgallego joined	06:57
	→ soc42 joined	06:59
	← orion left	06:59
	← sigh left	07:00
	← brianpWins left	07:02
	→ wo9x joined	07:03
	→ orion joined	07:04
	← orion left	07:04
	→ orion joined	07:04
	← Garen left	07:05
	→ Garen joined	07:05
	← CJack left	07:05
	→ brianpWins joined	07:05
	→ workmad3 joined	07:07
	→ linusoleander joined	07:08
	← orafu left	07:08
	→ orafu joined	07:08
	← gullden left	07:09
	→ Yuuhi`` joined	07:09
	← saschagehlich left	07:09
	→ bauruine joined	07:10
	→ guardian joined	07:10
	← Yuuhi` left	07:10
	→ saschagehlich joined	07:11
	→ xxtjaxx joined	07:11
	← javaanse_jongens left	07:11
	→ LouisJB joined	07:11
	← int3__ left	07:12
	→ ph^ joined	07:12
	→ troessner joined	07:13
	→ javaanse_jongens joined	07:14
	→ saschagehlich_ joined	07:15
	← saschagehlich left	07:16
	← ph^ left	07:17
	→ jaeckel joined	07:18
	→ nevyn^ joined	07:18
	→ henrikbjorn joined	07:18
	← dr0id left	07:19
	→ mastroDani joined	07:19
	→ pratz joined	07:20
	← aminpy left	07:20
	→ yota joined	07:20
	← tty1 left	07:21
	→ uwjweq joined	07:21
	← yota left	07:23
	→ MikhX joined	07:24
	← cordoval_ left	07:25
	→ cordoval_ joined	07:25
	← sycofly_ left	07:25
	→ eijk joined	07:25
	← wo9x left	07:26
	→ _pingu joined	07:27
	→ dc5ala joined	07:28
	→ ningia joined	07:28
	← saschagehlich_ left	07:29
	→ TiBook joined	07:29
	← nevyn^ left	07:30
	← steph021 left	07:30
	→ cilly joined	07:31
	→ mikewintermute joined	07:31
	→ aminpy joined	07:32
	← LouisJB left	07:33
	← adaptee left	07:34
	← Adys left	07:36
	→ borwar joined	07:38
	→ georgedove joined	07:38
	← xerora left	07:39
georgedove	Hello. I have a webserver for development. All changes in source files which occur there are made through the ftp. After changes have been made I have to connect through ssh and to make the commit from console. (It's not possible to have the local copy of the all source files and then just pull changes to remote server). Question: Is there web oriented app for making commits on remote server, or maybe some GUI tool that can do it through the ssh?	07:39
	← bindaas left	07:40
thiago	not that I know of	07:40
	why isn't it possible to have a local copy of everything?	07:40
	→ bindaas joined	07:40
cbreak_work	georgedove: git is a command line app	07:41
	you can invoke it from scripts too	07:41
	→ bentob0x joined	07:41
	→ dpino_ joined	07:41
cbreak_work	but I think it's better to have a local work dir	07:41
georgedove	cbreak, are you suggest to write scripts	07:41
	→ bkozal joined	07:42
	← simplechat left	07:43
georgedove	there are a lot of services on the server that i cant run on my windows machine	07:43
	→ drake01 joined	07:43
georgedove	so i cant test the code on local machine	07:43
	→ optikfluffel joined	07:47
	← bauruine left	07:47
	← MikhX left	07:50
	← drake01 left	07:50
	→ azariah4_ joined	07:50
	ChanServ set mode: +v	07:52
	← tyfighter left	07:52
azariah4_	How do I undo 'git add foo.c' ?	07:52
joschi	azariah4_: `git reset foo.c`	07:54
	← mosno left	07:54
	→ sako joined	07:55
azariah4_	joschi: thanks	07:55
	→ bauruine joined	07:57
thiago	azariah4_: git status tells you what to do	07:57
	← codespectator left	07:58
	← fs-sylr left	07:58
	← linusoleander left	07:58
	→ fs-sylr joined	07:58
	← javaanse_jongens left	07:59
	→ davzie joined	08:00
	→ javaanse_jongens joined	08:01
	→ babbio joined	08:01
	→ roop_ joined	08:01
	← optikfluffel left	08:01
	→ roop__ joined	08:03
	← roop left	08:04
	→ eschnou joined	08:04
	→ ph^ joined	08:06
	← roop_ left	08:06
	← brianpWins left	08:08
	← ak0rn\|2 left	08:09
KolakCC	https://github.com/mindstormer/Basilmobile/commit/b6546a8d8a9050c285336646b85b1817c6e36ec5 < Can somebody tell me what happened to the 3rd file?	08:09
	→ ak0rn\|2 joined	08:09
KolakCC	Nothing changed but all the lines are red and then readded with green.	08:09
	← ph^ left	08:10
selckin	line endings?	08:10
	→ saschagehlich joined	08:10
KolakCC	Bleh. How can I fix that?	08:10
	← javaanse_jongens left	08:12
	→ Spockz joined	08:12
	← roop__ left	08:12
	→ noonoooo joined	08:12
	← theoros left	08:12
	→ javaanse_jongens joined	08:13
	→ dr0id joined	08:13
	→ leavittx joined	08:15
	← javaanse_jongens left	08:16
	← _sh3 left	08:17
KolakCC	selckin?	08:17
	→ javaanse_jongens joined	08:17
selckin	don't change them?	08:17
	→ codespectator joined	08:18
KolakCC	I didn't change them on purpose.	08:19
selckin	use a good editor	08:20
	→ linusoleander joined	08:20
	→ fga joined	08:21
	← Goliath\|2 left	08:21
azariah4_	I'm using repo with git, and did a git commit then repo upload. Now I want todo the exact same commit but with a different commit message, but git add does not seem to add anything.	08:23
frerich	KoalkCC: Maybe you have the core.autocrlf configuration switch enabled, so that git 'fixed up' line endings?	08:23
Fandekasp	hi	08:24
KolakCC	frerich: Okay, i'll try disabling that, thanks	08:24
	→ PigeonFriend joined	08:25
	→ lu_zero joined	08:25
	← cordoval_ left	08:25
Fandekasp	I struggle with this error : http://paste.pocoo.org/show/371419/ . I use gitolite, and I'm sure that my config is fine and my ssh public key is correctly placed in the keydir folder, with the same name as specified in the conf. I really don't understand why I keep getting this permission error (and a few hours ago, it worked :/ ) . Any idea 6	08:25
	→ cordoval_ joined	08:26
Fandekasp	-6+?	08:26
	→ optikfluffel joined	08:26
	→ Adys joined	08:26
	← mojavy left	08:26
	→ rendar joined	08:27
jast	wrong username in SSH URL perhaps?	08:28
	(remember, it needs to be the user gitolite is installed under)	08:28
	← krosenvold left	08:29
Fandekasp	no it's fine	08:30
	I think I found out	08:30
	I can download another project	08:30
	so that mean this one is corrupted or something ?	08:30
	→ cutmail_ joined	08:30
Fandekasp	But I have no clue how to fix it	08:30
	→ ph^ joined	08:30
jast	the files are probably fine, it's just a filesystem permissions problem	08:31
	← bkozal left	08:31
jast	perhaps someone accidentally pushed to the project bypassing gitolite (e.g. as a different user)	08:31
	← ph^ left	08:31
Fandekasp	hum	08:32
	← cutmail left	08:33
Fandekasp	I pulled the same project from another user in the server, but didn't push	08:33
KolakCC	What would be the reccomended core.autocrlf configuration for 2 windows computers?	08:33
selckin	false	08:34
	← cutmail_ left	08:34
jast	I was going to say true	08:34
	:)	08:34
	← CloudySunshine left	08:35
selckin	you don't want to git to mess with it, else you jus get suprises, its an option to cover up for reall ybad editors that always change the line editngs	08:35
	→ empity joined	08:35
selckin	don't use those editors.	08:35
jast	the new official way to do it is by using core.eol and attributes, anyway	08:35
	slightly more work but perfectly safe once the attributes are set properly	08:36
KolakCC	git-config core.autocrlf false	08:36
	would be the command, righT?	08:36
jast	if you don't anticipate any non-windows people ever working with the repository, sure	08:36
	(nor any cygwin users)	08:37
	← linusoleander left	08:37
selckin	how does that mather?	08:37
jast	they'd be likely to start introducing LF line endings	08:38
	and once you've got mixed line endings in the repository, things get ugly	08:38
selckin	doesn't mather	08:39
	→ cutmail joined	08:40
jast	but you'll start getting a lot of whitespace damage in diffs very soon	08:40
	← fga left	08:40
	→ roop__ joined	08:40
jast	which means merges are going to be painful	08:40
robinbowes	I have a fork of this: https://github.com/loggly/puppet-modules here: https://github.com/robinbowes/puppet-modules	08:42
	→ ptr_uzl joined	08:42
	→ roop_ joined	08:42
robinbowes	I've cloned my fork on my workstation and am trying to pull in an upstream change (a new top-level dir was added)	08:43
	→ NetHawk joined	08:43
robinbowes	But, I can't get it to show up in my local clone	08:43
	Doh!	08:44
	I just noticed, I added my own repo as the upstream!	08:44
	→ mags joined	08:44
	← namemon left	08:44
	← robotarmy left	08:45
	← roop__ left	08:45
	adamholt → adamholt_away	08:46
	adamholt_away → adamholt	08:46
	→ tatsuya joined	08:47
	← dharrigan left	08:47
	→ dharrigan joined	08:48
	← dharrigan left	08:48
	→ dharrigan joined	08:48
	← Spockz left	08:48
	→ Spockz joined	08:49
	← tgunr_ left	08:49
	← cutmail left	08:50
	→ rgr joined	08:50
	→ cutmail joined	08:50
	← rgr left	08:50
	← ak0rn\|2 left	08:50
	→ Dr4g joined	08:50
	← PigeonFriend left	08:52
	→ mattt0 joined	08:53
	→ rgr joined	08:53
	← Spockz left	08:53
	→ ak0rn joined	08:54
	ChanServ set mode: +v	08:54
mattt0	I am not sure what difference it makes to make a --bare init. When is it proper to use this instead of a normal init?	08:55
	→ PigeonFriend joined	08:55
	→ RaceCondition joined	08:55
	→ mklappstuhl joined	08:55
	→ drake01 joined	08:56
	→ metin joined	08:56
	→ Elfe joined	08:56
thiago	when you want a bare repository	08:57
Fandekasp	Does someone has an idea how to fix this error : http://paste.pocoo.org/show/371429/ ? Thanks in advance	08:57
thiago	check that the user has the permission to read all files in the repository and write to all directories	08:58
	← metin left	08:58
mattt0	thiago: Was that for me?	08:58
Fandekasp	check done, git:git everywhere	08:58
mattt0	:/	08:58
	← Elfe_ left	08:59
Fandekasp	ah	08:59
	no my fault	08:59
	there is a apache:apache for a few file	08:59
	← PigeonFriend left	09:00
Fandekasp	mattt0, that was that ... so stupid -_-' thank you	09:00
jast	mattt0: for hosting, mostly. bare repositories can be freely pushed to, unlike non-bare repositories where you don't want to push to the checked out branch.	09:01
	← rgr left	09:01
	→ rgr joined	09:01
jast	bare = there is no working copy of the files, just the database of commits	09:01
	← leavittx left	09:02
mattt0	jast: Can you explain the part of not wanting to push to the checked out branch? What do you mean by this?	09:03
thiago	mattt0: yes	09:04
	→ PigeonFriend joined	09:04
	→ LouisJB joined	09:04
	→ achadwick joined	09:05
	→ jophish joined	09:05
	← ak0rn left	09:07
	→ rmAdfinis joined	09:07
mattt0	thiago: To check permissions? How does that go along with what jast said? Is there more to it, than checking permissions?	09:08
thiago	mattt0: sorry, to you it was just the "use --bare when you want a bare repository"	09:08
	the permissions was to Fandekasp	09:08
mattt0	lol thought you said yes to me :)	09:09
	Or, you actually did :P	09:09
	→ larsl joined	09:09
	→ woozie joined	09:09
mattt0	I dont feel the git man is explanatory enough for beginners	09:10
	← bn_ left	09:10
	← ki__ left	09:11
	→ Titosemi joined	09:12
	← roop_ left	09:14
mattt0	thiago: Can you confirm if I have understood this correctly? The point of a bare repository is to allow free pushes for instance, as it is not set up to a specific working tree, as this would be a problem if you have multiple users pushing at any point?	09:16
	← seivan left	09:16
	← jceb left	09:16
	← rendar left	09:16
	← rmAdfinis left	09:17
thiago	mattt0: the point of a bare repository is a repository without a working dir. It's just the Git database.	09:17
	← Predominant left	09:17
thiago	mattt0: and you should not push to a non-bare repository.	09:18
	→ Predominant joined	09:18
	→ snearch joined	09:18
mattt0	How come thiago? I think there is something im not aware of there	09:18
jast	simple	09:19
	if you push to a checked out branch (bare repositories don't have checked out branches because there's nowhere to check them out to), git updates the commit history but not the checked out files	09:19
	because that opens a huge can of worms: what if the checked out files have changes? what if someone is editing them right now? etc.	09:19
	so, after pushing to a checked out branch, the checked out files (plus the index) no longer match the history of the repository	09:20
	this usually causes a lot of confusion so we forbid it by default	09:20
	← gebi left	09:21
	← dc5ala left	09:22
hyperbor1ean	anyone has any examples of using git-svn tag ? I've been trying exactly like this git svn tag <url> and I get a weird error about svn_path_join	09:22
	← kassah left	09:23
	→ rendar joined	09:23
	→ nevyn^ joined	09:23
mattt0	jast: Thanks for the explanation. That hepled	09:24
dr0id	wait, I am confused now :P	09:24
	← TrineLocke left	09:24
	→ gebi joined	09:25
dr0id	bare repo means an empty repo, right ?	09:25
	← stliu left	09:25
	→ TrineLocke joined	09:25
	← cordoval_ left	09:25
mjt	bare = without .git subdir	09:26
	→ cordoval_ joined	09:26
jast	no	09:26
	→ trupheenix joined	09:26
jast	bare = without working tree, i.e. only the stuff that's usually in .git/	09:26
	← TrineLocke left	09:27
mjt	that's another way to say the same, sorta :)	09:27
dr0id	ohh, without any changes or anything. basically when git status is "nothing to commit" :P	09:27
BinGOs	hyperbor1ean: git svn tag <tag>	09:27
mjt	dr0id: no!	09:27
	→ TrineLocke joined	09:27
	← javaanse_jongens left	09:27
dr0id	then ? :S	09:28
	→ curtana joined	09:28
dr0id	"and you should not push to a non-bare repository." is confusing me buddy	09:28
mjt	that's wrong	09:29
	→ simplechat joined	09:29
dr0id	uggh	09:29
	← simplechat left	09:29
	→ simplechat joined	09:29
dr0id	jast: is there some confusion here ?	09:29
	← d0k left	09:29
	→ rmAdfinis joined	09:29
	→ javaanse_jongens joined	09:31
	→ jasper joined	09:32
	← gebi left	09:32
	→ whyking joined	09:34
	→ opalka joined	09:34
	← optikfluffel left	09:36
	→ aidalgol joined	09:38
	→ starcoder2 joined	09:38
	→ seivan joined	09:39
sgronblo	Hey can I list files that have existed in a directory at some point?	09:42
	← cutmail left	09:42
jast	dr0id: http://git.or.cz/man/gitglossary	09:42
dr0id	read that, makes lil sense really :S	09:43
	manual is definitely confusing for newb	09:43
	s	09:43
jast	non-bare = files with your code in them + .git subdir. bare = only the .git dir, directly at the top level	09:43
hyperbor1ean	BinGOs: where <tag> is the whole svn url with /tags/my_tag or just the my_tag thing? it doesn't work for either of those	09:43
	BinGOs: thanks	09:43
dr0id	jast: so you should not push to repos where there is .git + code files ?	09:44
jast	dr0id: usually not	09:45
dr0id	sounds weird, people push everyday on github.	09:45
jast	yes, but the repositories on github don't have physical copies of your code files	09:45
	they just contain the contents of your files somewhere deep inside the git object database	09:45
	github uses bare repositories on their servers	09:46
dr0id	hmm	09:46
	I see	09:46
jast	you know what, I suggest you try it yourself	09:46
	create a bare repository, clone from it locally, edit/add/commit some stuff in the clone and push it	09:46
	then look at the two repositories	09:46
	that will probably make it a lot clearer to you	09:46
dr0id	I have done it yesterday	09:47
	→ gebi joined	09:47
	← javaanse_jongens left	09:47
dr0id	on my server and it was just fine, maybe I am missing something	09:47
jast	have done what?	09:48
dr0id	there were files on the server, did git init->add . -> commit	09:48
	then git clone on my dev box	09:49
	modified files, added new	09:49
	→ javaanse_jongens joined	09:49
dr0id	git add . -> got commit on my box and then git push	09:49
jast	and you got no warnings/errors?	09:49
	→ fivetwentysix joined	09:50
dr0id	no	09:50
	→ mikelangelo joined	09:50
	← fivetwentysix left	09:50
	← yaotti_aw__ left	09:50
jast	what version of git are you running on the server?	09:50
dr0id	I was learning git last month and had done git push and got this error	09:50
	I asked here, and someone said me the same thing	09:50
	I started re-learning git on Sunday and yesterday did my first git push everything was fine :)	09:51
	I will check one sec	09:51
	should be 1.6.3.3	09:51
	← blast_hardcheese left	09:51
dr0id	actually, I wish try git push again right now :)	09:51
	1.6.4	09:51
	it is	09:51
jast	okay, we already had the warning (but not the error) in 1.6.x	09:52
	starting from 1.6.2 I think	09:52
hyperbor1ean	git svn branch -t <tag> is for creating new tags in svn	09:52
	← gebi left	09:52
dr0id	also last month when I got that error, I was using some 1.5.3 or something	09:53
	but the server was same though :/	09:53
jast	the error message is generated on the server	09:53
BinGOs	'tag' probably got added as a shortcut at some point.	09:53
dr0id	weird then, not sure	09:53
mattt0	Having a server for my remote git repositories which is gonna have multple repos depending on project, I am going to have a folder structure named git/project-name where each project-name folder is a --bare init ... This seems right, doesnt it? Or is there some other more proper way to organize multiple repos	09:55
	← engla left	09:55
workmad3	mattt0: you could look at gitolite for managing the server	09:56
jast	mattt0: by convention, bare repositories live in directories like projectname.git; it doesn't matter how you structure the whole thing though	09:56
mattt0	Ok thanks both of you workmad3 and jast	09:56
dr0id	jast: so basically github don't hold any files ? when we see the files from web browser they are fetching from .git and showing ?	09:57
jast	dr0id: yeah, though they might be using some fancy in-house caching system on top of that. who knows.	09:57
dr0id	true	09:57
selckin	.git caontains just files	09:57
jast	yeah, but special files :}	09:58
	magical, magical files	09:58
dr0id	jast: is there some command that can tell whether the repo is bare or not ?	09:58
	or do we just consider it bare if no files/folders except .git is present	09:58
jast	dr0id: in properly configured repositories: git config core.bare	09:58
	only extremely old or manually modified repositories will fail to give you a result	09:59
dr0id	on my server it says false :D	09:59
	→ drl joined	10:01
	→ marcob joined	10:01
	→ kpreid joined	10:04
dr0id	jast: buddy, the value is always false :P	10:05
	if the files are not there and you run git status then it will say that those files are deleted	10:06
	→ yaotti_aw joined	10:06
jast	just deleting the files doesn't make the repository bare	10:06
	that's why I said: create your own fresh bare repository to see the difference	10:06
dr0id	you mean just git init and nothing else ? :)	10:07
	even then git config core.bare is false	10:07
	← snearch left	10:09
	← seivan left	10:10
xelister	in git how to do like svn info	10:10
	to see gir URL, current version etc	10:10
FauxFaux	git remote -v	10:11
	git fetch && gitk --all	10:11
xelister	this is the version of current locally checked out version?	10:11
	→ gs joined	10:11
FauxFaux	How about you run them and find out.	10:11
xelister	again, how to see the version name of currently locally checked out version? possibly with indication if it is modified (liek M versions in svn)	10:11
FauxFaux	git status; git rev-parse HEAD	10:12
xelister	I dont have gitk	10:12
	→ seivan joined	10:13
	Vampire0_ → Vampire0	10:13
FauxFaux	It's in nearly every distro's package manager; I'd guess you're using a debian clone if you have git but no gitk, sudo apt-get install gitk	10:13
	xelister touches FauxFaux's gf boobs honk* honk ; dont CTCP version me ;)*	10:13
xelister	its a debian	10:13
	ubuntu actually	10:13
FauxFaux	It's quicker than asking what OS you're on, and normally more accurate than asking, even including people like me.	10:14
	← mklappstuhl left	10:15
	← sitaram left	10:16
	→ ph^ joined	10:20
	→ roop_ joined	10:22
	← marcob left	10:23
	→ namemon joined	10:24
	← jasper left	10:24
	→ winterweaver joined	10:24
	→ marcob joined	10:25
	hyperbor1ean → hyperboreean	10:25
	← aidalgol left	10:25
	→ roop__ joined	10:26
	← marcob left	10:26
	← cordoval_ left	10:26
	→ marcob joined	10:26
	→ cordoval_ joined	10:26
	← _pingu left	10:26
	← roop_ left	10:28
	← sraue left	10:28
	← cenuij left	10:28
	→ jadu joined	10:29
	← drake01 left	10:29
	→ _pingu joined	10:30
	→ lanthan joined	10:31
	→ sitaram joined	10:32
	← sitaram left	10:32
	→ sitaram joined	10:32
	← lanthan left	10:32
	← JEEB left	10:32
	→ sraue joined	10:33
	← seivan left	10:34
	→ lanthan joined	10:34
	← shiba_yu36 left	10:35
	→ cenuij joined	10:36
	← cenuij left	10:36
	→ cenuij joined	10:36
	← winterweaver left	10:36
	← whyking left	10:36
	← namemon left	10:37
	→ whyking joined	10:37
	← Adys left	10:38
	← javaanse_jongens left	10:39
	→ Adys joined	10:40
	→ javaanse_jongens joined	10:41
	→ icwiener joined	10:42
	→ Kjellski joined	10:42
	→ avdg joined	10:42
Kjellski	Is there a way to find out which server a repository is saved to?	10:43
	← _pingu left	10:43
	← nevyn^ left	10:43
	→ nevyn^ joined	10:43
FauxFaux	You mean, who's pulled your repository? Or what git remote -v prints?	10:43
Kjellski	FauxFaux: I can see the local repository and don't know where it was backed up to... where it was pushed to?	10:44
	← larsmoil left	10:44
	← RaceCondition left	10:45
FauxFaux	Given a push of a repo is basically a file copy, no?	10:45
	→ larsmoil joined	10:46
Kjellski	FauxFaux: I don't know. I'm pretty new to git, so I just wanted to know whether there is a way to find out where a repo was pushed to and where that remote side is or not...	10:46
	← trupheenix left	10:48
Kjellski	FauxFaux: And the git remote -v shows just one line with origin.	10:48
	→ DarkUranium joined	10:49
DarkUranium	hey	10:49
	I've a question -- I've got two branches, master and console	10:49
	how can I merge a single file from console into master?	10:49
	now, I know I can simply copy over the file, but what would be the "proper" way of doing it?	10:50
	→ SRabbelier\|Lappy joined	10:50
DarkUranium	"git checkout console fname.ext"?	10:51
	← javaanse_jongens left	10:52
	← Fandekasp left	10:52
albel727	to copy it over and git add it is the best way. and that checkout does exactly "copying it over" part. I doubt you want to perform even a tricky merge here (like git merge --no-commit, unstage all but frame.ext, git-commit), cause there are implications in having a merge info in history.	10:52
DarkUranium	albel727: okay, thanks =)	10:53
	→ JEEB joined	10:53
	→ javaanse_jongens joined	10:55
	→ linusoleander joined	10:55
	→ etnt` joined	10:56
	← linusoleander left	10:56
	→ gebi joined	10:58
	→ JohnFlux joined	10:58
JohnFlux	How do I branch and track "origin/master" to a new branch name "fixes" ?	10:58
	← etnt left	10:58
thiago	git branch fixes origin/master	10:58
dr0id	git checkout -b fixes origin/master	10:58
	eww :P	10:59
JohnFlux	thiago: and that will track?	10:59
	← etnt` left	10:59
dr0id	yes JohnFlux	10:59
JohnFlux	cool	10:59
dr0id	the second version will create and checkout	10:59
	→ etnt joined	11:00
	← saschagehlich left	11:00
etnt	it seem that 'git tag -l' also list branch names, how can I list tags only?	11:01
dr0id	just git tag ?	11:01
	← pdelgallego left	11:01
	→ mabrand joined	11:02
teuf	etnt: git tag -l only gives tags here	11:02
	← albel727 left	11:04
	← cenuij left	11:04
	← javaanse_jongens left	11:04
etnt	hm...I got loads of old branchnames in .git/refs/remotes/origin/ , leftovers from an old svn conversion	11:04
	they all show up with git tag -l	11:04
	→ alberto56 joined	11:06
	← amondracorp left	11:06
	← dpino_ left	11:07
	← kpreid left	11:08
	→ javaanse_jongens joined	11:09
	← cilly left	11:10
	→ RaceCondition joined	11:11
	→ elyob joined	11:15
	→ sftp joined	11:15
	← Ruudjah left	11:15
	→ juliohm joined	11:17
	← antagomi1 left	11:18
juliohm	Hi, what means the ..HEAD notation in git?	11:18
	→ nimrod10 joined	11:18
juliohm	this is an interval notation, right?	11:18
	→ dpino_ joined	11:19
	→ cilly joined	11:20
	← nimrod10` left	11:21
	→ glitterfang joined	11:21
	← aminpy left	11:21
	→ mariusso joined	11:22
mariusso	So, how do I get back to merge conflict? Say I resolved a conflict incorrectly, in only one file, how do I get that one file back to the conflicted state so I can re-resolve it?	11:23
	→ cenuij joined	11:24
	← cenuij left	11:24
	→ cenuij joined	11:24
	← davzie left	11:26
	→ Jake_ joined	11:26
	← cordoval_ left	11:26
	→ linusoleander joined	11:26
	→ cordoval_ joined	11:26
	→ davzie joined	11:27
	→ sebersole joined	11:27
	← cenuij left	11:28
mariusso	'git checkout -m <file>' seems to be the magic formula	11:28
juliohm	mariusso, do you know what means the origin/master..HEAD notation?	11:28
	← mags left	11:29
	← Jake__ left	11:30
	→ iband joined	11:30
xelister	git rev-parse HEAD	11:30
	how to get the shorter format of the git version?	11:30
rane_	--short	11:31
	→ namemon joined	11:31
xelister	thanks	11:31
	→ Jygga joined	11:31
	→ aminpy joined	11:31
	→ spaceonline joined	11:32
dr0id	juliohm: that's used to see the commits that are available in HEAD but not in origin/master	11:32
	git show origin/master..HEAD	11:32
	git show or git diff , don't remember exactly :S	11:32
	← iband left	11:33
	→ iband joined	11:33
	← iband left	11:33
	→ iband joined	11:34
	→ gfto joined	11:35
etnt	hm...correction I got unannotated tags in the output of 'git tag -l', but where are they stored?	11:35
dr0id	.git/refs/tags ?	11:35
etnt	dr0id: nope, only 3 annotated ones there	11:36
	← mabrand left	11:36
dr0id	I am not any good with tags :P still to read some tuts on it, but I guess unannotated tags are known as lightweight tags :D	11:36
	← drev1 left	11:37
etnt	dr0id: yes, but I'm confused because I don't understand where they are store, find .git -name <tag> doesn't give me anything	11:37
bremner	there is also the possibility of packed refs	11:38
	← hyperair left	11:38
dr0id	yup :)	11:38
	← woozie left	11:38
bremner	so do git show-ref "mystery tag"	11:39
etnt	aha, that could explain it, didn't know about that	11:39
	gives me: 977b164b9b8f30f95fc4db26aeb419fe34f0dcdb refs/tags/pre_merge_1_4-0	11:40
	but it doesn't exist in that dir though	11:40
bremner	right. its packed	11:41
etnt	ok, thx	11:41
	→ wo9x joined	11:42
	ChanServ set mode: +v	11:42
	→ bitkiller joined	11:43
	→ urkud joined	11:46
	→ pdelgallego joined	11:46
	→ ereslibre joined	11:46
	← ereslibre left	11:46
	→ ereslibre joined	11:46
	→ saleh joined	11:47
	→ keystr0k joined	11:48
	← larsmoil left	11:49
evelyette	hi, I setup the git post-receive hook to update my working dir automatically ... but the problem is that the working dir is owned by root and the git (gitolite) doesn't have permission to update it ... I get this error: remote: error: cannot open .git/FETCH_HEAD: Permission denied	11:50
	← pratz left	11:50
wereHamster	evelyette: and your question is ... ?	11:50
evelyette	I could change the permissions , which would work, but then the other service doesn't work, because it has to be run as root	11:50
	so I'm asking, is there any elegant solution to this ... I would create a new group	11:50
wereHamster	evelyette: faq non-bare	11:51
jast	evelyette: Pushing to non-bare repositories is discouraged: please see https://git.wiki.kernel.org/index.php/GitFaq#non-bare [automatic message]	11:51
evelyette	but is there a way that gitolite can still have permission to update even though the working dir is owned by root	11:51
	→ albel727 joined	11:51
	← troessner left	11:52
	→ hendeby joined	11:52
	→ kpreid joined	11:52
	→ kotique joined	11:52
kotique	If you run "git branch" at this point, you'll see that git has temporarily moved you to a new branch named "bisect".	11:52
	bash-3.2$ git branch	11:52
	* (no branch)	11:52
evelyette	wereHamster, well, the git push .. is the only one that pushes to the working dir ... there are no local changes, so this isn't a problem	11:52
kotique	how do i git archive from this bisected state?	11:53
wereHamster	evelyette: it doens't matter. YOu shouldn't push to a non-bare repo	11:53
	kotique: git archive	11:53
kotique	wereHamster, what's the tree spec?	11:53
wereHamster	HEAD ?	11:53
kotique	are you sure?	11:53
wereHamster	HEAD is where you are at	11:54
kotique	did you do bisects previously?	11:54
evelyette	wereHamster, well this wasn't the question ... I want to know if the git can have access to root working dir ?	11:54
	← robinbowes left	11:54
wereHamster	evelyette: no	11:54
	ChanServ set mode: +v	11:54
wereHamster	kotique: if you hate HEAD, then you can also do git branch oh-I-hate-HEAD; git archive oh-I-hate-HEAD; git branch -d oh-I-hate-HEAD	11:55
freimatz	hi, how can i comment out a line in config	11:55
kotique	that will reset bisect state probably	11:55
wereHamster	kotique: no	11:55
	creating a brnach will not reset anything	11:55
	freimatz: have you tried the standard under unix?	11:56
	← albel727 left	11:56
freimatz	what is the stadard under unix?	11:56
	← fairuz left	11:56
freimatz	unix is 20 years ago for me :-)	11:57
	→ albel727 joined	11:57
wereHamster	#	11:57
freimatz	thx	11:57
	first machine i worked on with unix had a Motorola 68000 CPU	11:57
	→ avdg1 joined	11:58
	→ jsick3 joined	11:58
	← urkud left	11:59
	← avdg left	11:59
	← avdg1 left	11:59
	← javaanse_jongens left	11:59
	→ avdg joined	11:59
	← walf443 left	11:59
	→ mab01 joined	12:00
	→ walf443 joined	12:00
	← kpreid left	12:01
	→ javaanse_jongens joined	12:01
	← linusoleander left	12:01
	→ mklappstuhl joined	12:02
bremner	newb	12:04
	→ Spockz joined	12:04
dr0id	:P	12:05
	← lanthan left	12:05
	→ larsmoil joined	12:06
	← javaanse_jongens left	12:06
	← alberto56 left	12:06
	← iband left	12:06
	→ javaanse_jongens joined	12:07
	← ereslibre left	12:08
juliohm	dr0id, thank you :-)	12:09
	→ ereslibre joined	12:09
	← ereslibre left	12:09
	→ ereslibre joined	12:09
dr0id	juliohm: so what was it git show ?	12:09
	← juliohm left	12:10
	→ shiba_yu36 joined	12:10
	← javaanse_jongens left	12:10
	← rgr left	12:10
	→ jfkw joined	12:11
	← glitterfang left	12:11
	→ javaanse_jongens joined	12:11
	→ fairuz joined	12:12
	→ juangiordana joined	12:14
	→ linusoleander joined	12:15
	→ foocraft joined	12:17
	← ymasory left	12:17
	→ urkud joined	12:18
	→ dajhorn joined	12:18
	→ hcl2_ joined	12:23
	← mklappstuhl left	12:23
mattt0	Is checkout only used to change branch?	12:23
	→ nixness joined	12:24
	← dr0id left	12:25
	← roop__ left	12:26
dr_lepper	checkout is used to retrieve file or files from the repository into the working copy	12:26
	branch changing is just a side-effect	12:26
	← cordoval_ left	12:26
	→ SimonP86 joined	12:27
	→ cordoval_ joined	12:27
dr_lepper	you can, for example, checkout an older commit; or a single file from an older commit	12:27
	→ ziro` joined	12:27
ziro`	hi, is it possible to make it clearer what branch you're working on with git?	12:27
mattt0	So how can branch changing be a side-effect?	12:27
	dr_lepper	12:27
patrikf	it's not a side effect, it's a different usage	12:28
	← foocraft left	12:28
dr_lepper	ziro`, you can add branch name to PS1	12:28
	← Predominant left	12:28
	→ [intra]lanman joined	12:28
	← [intra]lanman left	12:28
	→ [intra]lanman joined	12:28
	→ PabloM joined	12:28
dr_lepper	mattt0, when during checkout you reference a commit by a branch name, git switches current branch	12:28
	→ Predominant joined	12:28
patrikf	dr_lepper: no.	12:29
dr_lepper	patrikf, very argumented :)	12:29
	← ziro` left	12:30
	← Predominant left	12:30
	→ ziro` joined	12:30
	← ziro` left	12:30
mattt0	:P	12:31
	→ roop joined	12:32
	→ Predominant joined	12:32
	→ ericbarnes joined	12:32
mattt0	dr_lepper: You gave an example that you could checkout an older commit... Is there any other use of checkout than that or changing branch, sideeffect or not? I need a clear definition of checkout and its usage	12:35
	← Predominant left	12:35
selckin	checkout can do a lot of things	12:35
	→ _pingu joined	12:35
dr_lepper	mattt0, basically it just retrieves given set of files from the repository. I guess you'd want to see man git-checkout for details	12:37
jast	mattt0: the 'git-checkout' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-checkout.html [automatic message]	12:37
	→ rgr joined	12:37
mattt0	ok thanks	12:38
	← icwiener left	12:38
	← _pingu left	12:40
	← NetHawk left	12:40
	→ robinbowes joined	12:41
	→ bartek joined	12:41
	→ _pingu joined	12:43
	← dharrigan left	12:43
	← cordoval_ left	12:44
	→ Predominant joined	12:45
	→ gorekee joined	12:48
	← svg left	12:48
	→ dharrigan joined	12:48
	← Predominant left	12:49
	← gebi left	12:49
	→ Predominant joined	12:49
	← nixness left	12:49
	→ owst joined	12:50
	→ fr0sty\|mobile joined	12:50
	← Predominant left	12:50
	→ Predominant joined	12:51
	← Kjellski left	12:51
	← Predominant left	12:52
	→ Predominant joined	12:52
	← shiba_yu36 left	12:53
	→ sigh joined	12:55
	← rondon left	12:55
KolakCC	I need some help with my .gitignore file. I have this line: */config/config.php	12:56
	← kotique left	12:56
KolakCC	However, when I edit that file, it still wants to commit the change	12:56
jacobat	KolakCC: Git only ignores files it doesn't know about	12:57
	← RaceCondition left	12:57
jacobat	KolakCC: You will need to tell git to remove the file if you don't want it listed	12:57
KolakCC	jacobat: With git rm file?	12:57
	→ wo9x_ joined	12:57
	← crashanddie left	12:57
jacobat	KolakCC: Yes - copy it somewhere first though	12:57
	→ ereslibre_laptop joined	12:58
	→ shiba_yu36 joined	12:59
	← ereslibre left	12:59
	→ dv310p3r joined	13:00
KolakCC	jacobat: But now it wants to commit that I deleted the file, then the file will be deleted for everybody, right?	13:00
	← wo9x left	13:00
KolakCC	jacobat: But the config file shouldn't be deleted, just ignored	13:01
	→ crashanddie joined	13:02
	← SimonP86 left	13:03
fr0sty	KolakCC: people will be able to check out the old version of the file if they need to.	13:03
	→ drubies\|mbp joined	13:03
	→ deekayen joined	13:04
	← linusoleander left	13:04
	→ pratz joined	13:05
fr0sty	and you will only have this problem once (when going across the deletion).	13:05
KolakCC	Okay.	13:05
	It still lists the file in "unstaged changes", can I remove it from there?	13:05
fr0sty	you might want to add and commit a 'template' config file for people to copy and modify.	13:05
KolakCC	It doesn't want to commit it though.	13:05
	← deekayen left	13:05
fr0sty	git rm --cached <filename>	13:05
	→ theoretical joined	13:06
fr0sty	that will leave your local copy.	13:06
	← mab01 left	13:06
	→ deekayen joined	13:06
	← aminpy left	13:06
	→ hvgotcodes joined	13:07
	→ linusoleander joined	13:07
	← logankoester left	13:07
	← _pingu left	13:09
	← jadu left	13:09
	→ thierryp joined	13:10
	← cilly left	13:11
tokkee	.oO( git config --global alias.unadd 'rm --cached' )	13:11
	← roop left	13:11
selckin	git status tells you how	13:11
	→ dermoth\|home joined	13:12
avar	where would I put a server git hook to deny pushes of certain tag names (because some people here did an unholy push of foreign tags into another repo)	13:12
KolakCC	In git gui for windows, can I remove untracked files from unstaged changes?	13:13
	→ kukks joined	13:13
rane_	i hope not	13:13
	← jaeckel left	13:13
	← wo9x_ left	13:14
	→ ben42 joined	13:15
ben42	hello	13:16
	← deekayen left	13:16
fr0sty	avar: the update hook is probably the place	13:16
ben42	can git diff ignore entire added or removed empty lines ?	13:16
selckin	-w will ignore all whitespace	13:17
	→ deekayen joined	13:17
fr0sty	if you want to be heavy handed pre-receive will deny the entire push if it contains unallowed tags. the update hook will only deny the specific tags...	13:17
ben42	it doesn't seem, according to the man but maybe i missed it	13:17
	selckin, it keeps telling me about added or removed empty lines	13:17
	→ Finalflash50 joined	13:18
	→ logankoester joined	13:18
	→ saschagehlich joined	13:18
fr0sty	\n is not whitespace	13:19
	for the purposes of "ignore whitespace changes", that is...	13:19
	→ panike joined	13:19
	← saschagehlich left	13:20
	→ robert_david joined	13:20
	← borwar left	13:20
avar	fr0sty: right, hacking that..	13:20
	→ jatt joined	13:21
	→ linoj joined	13:21
	← linoj left	13:21
	← javaanse_jongens left	13:21
	→ roop joined	13:21
mattt0	Is it considered to be good practice to do a checkout after a commit/push?	13:21
fr0sty	for what purpose?	13:22
	→ chrisdev joined	13:22
Pieter	no	13:22
	→ javaanse_jongens joined	13:23
	← urkud left	13:23
	← chrisdev left	13:24
selckin	why is it always the people with +v that ask the most silly questions	13:24
crashanddie	to make sure you're paying attention	13:25
	← elb0w left	13:25
selckin	does it voice all webchat people maybe, how odd	13:25
	→ Tabstar joined	13:25
	← javaanse_jongens left	13:26
	← silfreed left	13:27
	→ javaanse_jongens joined	13:27
	→ Fuzzy joined	13:27
	← fs-sylr left	13:27
	← Soliton left	13:27
	← gorekee left	13:28
	→ silfreed joined	13:28
	→ Soliton joined	13:29
	← javaanse_jongens left	13:30
	→ koo6 joined	13:30
	→ svg joined	13:30
	→ rem joined	13:31
	→ javaanse_jongens joined	13:31
	→ jonatasnona joined	13:32
	← rem left	13:33
	→ candybar joined	13:33
	→ nembutal joined	13:33
	→ ziro` joined	13:33
	← nembutal left	13:34
	→ nembutal joined	13:35
	→ Barry-Nichols joined	13:36
	← Barry-Nichols left	13:36
	→ Barry-Nichols joined	13:36
	→ pprkut joined	13:36
	→ Ryback_ joined	13:37
	← Ryback_ left	13:37
	→ Ryback_ joined	13:37
	← nembutal left	13:37
	← Swimming_bird left	13:37
	→ tewecske joined	13:38
	← whyking left	13:38
	← rgr left	13:39
	→ q0rban joined	13:40
	← ph^ left	13:41
	← q0rban left	13:41
	→ chrisdev joined	13:41
	→ foocraft joined	13:42
	→ urkud joined	13:42
	← jelmer left	13:43
JohnFlux	This works: git push -f origin "HEAD:refs/merge-requests/1734"	13:43
	how do I make that an alias?	13:43
	I've tried; bug = push -f origin "HEAD:refs/merge-requests/$1"	13:43
	Claviceps → _Clavicep	13:43
JohnFlux	but it says: error: src refspec 1734 does not match any.	13:43
fr0sty	JohnFlux: you can't use $1	13:43
	→ rayners joined	13:43
	_Clavicep → Claviceps	13:44
	← DarkUranium left	13:44
	← drl left	13:44
fr0sty	not like that, at least.	13:44
	← tatsuya left	13:44
selckin	JohnFlux: probably need to prepend with !git	13:44
	→ q0rban joined	13:45
	→ saschagehlich joined	13:46
	→ DaveIngram joined	13:47
fr0sty	yeah, bug = !git push -f origin "HEAD:/refs/merge-requests/$1"	13:47
JohnFlux	I can't seem to get that to work	13:47
	bug = !git push -f origin "HEAD:refs/merge-requests/$1"	13:47
	oh	13:47
	missing /	13:47
	nm, doesn't matter	13:47
	git bug 1734	13:48
	fatal: remote part of refspec is not a valid name in HEAD:refs/merge-requests/	13:48
	← rchavik left	13:48
fr0sty	"!sh -c git ..." maybe...	13:48
JohnFlux	no change	13:48
	bug = !sh -c 'git push -f origin "HEAD:refs/merge-requests/$1"'	13:48
fr0sty	lose the double quotes?	13:49
	(might be killing the expansion)	13:49
	bug = !sh -c 'git push -f origin HEAD:refs/merge-requests/$1'	13:50
	← henrikbjorn left	13:51
JohnFlux	same error. I've used " instead ' before, like:	13:51
	· locate = !sh -c 'git ls-files \| grep --color=auto "$1"' -	13:51
	this works	13:51
	→ jelmer joined	13:51
	→ fga joined	13:51
	→ nembutal joined	13:51
	→ cordoval_ joined	13:52
selckin	the - ?	13:52
	→ engla joined	13:52
mattt0	An easy way to publish from a remote repo to the web?	13:52
	← engla left	13:52
	→ engla joined	13:52
	← blind_oracle left	13:52
	← ziro` left	13:52
	→ xiong joined	13:53
JohnFlux	selckin: hmmmmm, that - seems to be very important	13:53
	→ Myth17 joined	13:53
	← Myth17 left	13:53
	→ Myth17 joined	13:53
	← koo6 left	13:53
	→ hyperair joined	13:53
JohnFlux	This works: bug = !sh -c 'git push -f origin "HEAD:refs/merge-requests/$1"' -	13:53
	→ ziro` joined	13:53
JohnFlux	I wonder what that - means	13:53
	→ koo6 joined	13:54
fr0sty	The final dash is so that arguments start with $1, not with $0.	13:54
	^ from man githooks	13:54
jast	the 'githooks' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/githooks.html [automatic message]	13:54
	→ Ruudjah joined	13:55
	← owst left	13:55
	→ owst joined	13:55
	→ mrx1_ joined	13:55
	mrx1_ → mrx1	13:56
	← cordoval_ left	13:56
	← roop left	13:56
	→ cordoval_ joined	13:56
	→ drl joined	13:57
	→ BlindOracle joined	13:57
mrx1	test.	13:57
	ChanServ set mode: +v	13:59
fr0sty\|mobile	mrx1: just ask your question...	13:59
selckin	why would they start at 0	14:00
mrx1	fr0sty: no, some guy on another channel told me he can't speak here	14:00
	← saleh left	14:00
mrx1	so i tested if i can :)	14:01
	← Typh left	14:01
selckin	needs to be registered i think	14:01
	← larsmoil left	14:01
fr0sty	or voiced	14:01
nembutal	do u see me now?	14:01
fr0sty	yes...	14:01
nembutal	great	14:01
	can someone help me with gitignore pattern?	14:02
	→ cilly joined	14:02
nembutal	i want to exclude all paths that contain "obj" directory	14:03
	← opalka left	14:03
nembutal	how to do that?	14:03
fr0sty	do you have multiple obj/ directories?	14:03
nembutal	yes	14:04
	→ jaeckel joined	14:04
	→ nitish joined	14:04
fr0sty	and you don't want to track anything in them (or their subdirectories)?	14:04
nembutal	no, nothing	14:04
fr0sty	what have you tried so far?	14:04
	→ dim joined	14:04
nembutal	/obj/	14:05
fr0sty	obj/	14:05
	that ^ should work	14:06
	→ ceej joined	14:06
	→ v0n joined	14:06
nembutal	it works but I wonder why	14:07
	← Myth17 left	14:07
	→ nitish_mythology joined	14:07
nembutal	counterintuitive	14:08
fr0sty	how so?	14:08
	→ cannonball joined	14:08
	← psoo left	14:09
nembutal	i would expect it just excludes top level obj dir	14:09
fr0sty	why would you expect that?	14:09
	you need a leading slash for that to be the case...	14:09
	"A leading slash matches the beginning of the pathname. For example, "/*.c" matches "cat-file.c" but not "mozilla-sha1/sha1.c". <-- (from man gitignore)	14:09
jast	the 'gitignore' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/gitignore.html [automatic message]	14:09
nembutal	it doesnt contain "*" so i wouldnt assume it matches obj subdir	14:10
	jast: i've read that man page	14:10
	and came here nevertheless	14:10
fr0sty	not very well...	14:10
	fr0sty wanders off for coffee...	14:11
	← nitish left	14:11
nembutal	anyway thanks for help	14:11
	← gs left	14:12
	← sigh left	14:12
	← michael_lf left	14:12
	→ roop joined	14:13
	← nembutal left	14:14
jast	it seems awfully difficult to read a small piece of bold text	14:14
	→ Yuffster joined	14:14
	→ saleh joined	14:14
	→ SimonP86 joined	14:14
	→ cwillu_also_at_w joined	14:14
	← cwillu_also_at_w left	14:14
	→ AkumaStreak joined	14:16
	← nitish_mythology left	14:16
	← mrx1 left	14:16
	→ ph^ joined	14:17
	→ jceb joined	14:19
	← ph^ left	14:20
	→ ph^_ joined	14:20
	→ michael_lf joined	14:20
fr0sty	jast: in all fairness the formatting is client-specific...	14:20
	← koo6 left	14:20
fr0sty	but still...	14:20
	← javaanse_jongens left	14:20
jast	perhaps I should move it to the front	14:21
fr0sty	I wonder what sort of life people presume you to lead where all you do is lurk, waiting to pounce on people's suggestions to read the manpages	14:21
	← jfkw left	14:21
	← picasso left	14:22
jast	!jast @	14:22
	[automatic message] jast runs a script that provides all kinds of information, but rumour has it that he's also human. Sometimes.	14:22
	→ aminpy joined	14:22
jast	let's see if that works better	14:23
bremner	fr0sty: but I do that, it's jast that just is always faster	14:23
	← Remixman left	14:23
jast	bremner: try again :}	14:23
	→ mikewintermute_ joined	14:23
fr0sty	jast: probablility of success .4	14:23
	→ nixness joined	14:24
	← mikewintermute left	14:24
	mikewintermute_ → mikewintermute	14:24
jast	well, my away log will keep collecting data for me	14:24
	→ rondon joined	14:25
	← Finalflash50 left	14:25
	→ xerora joined	14:25
	→ ymasory joined	14:26
	← mattt0 left	14:26
	→ picasso joined	14:26
	← shiba_yu36 left	14:27
	← foocraft left	14:27
	← pratz left	14:27
	q0rban → q0rban\|otp	14:27
	→ hobodave joined	14:28
	→ pratz joined	14:28
	→ airborn joined	14:30
	→ beatak joined	14:30
	← gfto left	14:30
	← mattly left	14:31
	← namemon left	14:32
	→ mattly joined	14:33
	→ Swimming_bird joined	14:33
	← jceb left	14:33
	← georgedove left	14:34
	← hobodave left	14:35
	← mikelangelo left	14:35
	→ dnorman joined	14:35
	← roop left	14:36
	→ RaceCondition joined	14:38
	→ AAA_awright_ joined	14:38
	→ gs joined	14:38
	→ Myth17 joined	14:38
	← deekayen left	14:38
	→ soc42_ joined	14:39
	← AAA_awright left	14:39
	← soc42 left	14:40
	→ Remixman joined	14:41
	→ peper joined	14:41
	→ drake01 joined	14:42
	← dnorman left	14:42
	→ deekayen joined	14:43
	→ georgedove joined	14:46
	← georgedove left	14:46
	← drake01 left	14:46
	← Barry-Nichols left	14:46
	← urkud left	14:47
	← Deesl left	14:48
	→ kumbayo joined	14:49
	→ argor joined	14:50
	→ jfkw joined	14:51
	← argor left	14:51
	→ webPragmatist joined	14:53
	← TiBook left	14:53
	← kumbayo left	14:53
	→ whatdoesitwant joined	14:53
	→ reds joined	14:54
	→ drev1 joined	14:54
	→ crocket joined	14:54
crocket	How can I download git://slackbuilds.org/slackbuilds/perl/perl-xml-twig ?	14:54
	git://slackbuilds.org/slackbuilds is the original git repository.	14:54
	elliottcable → ec\|detached	14:56
	→ BinaryMuse joined	14:57
cmn	crocket: man git-clone	14:57
jast	crocket: [automatic message] the 'git-clone' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-clone.html	14:57
	← avinashhm left	14:57
	← EricInBNE left	14:58
	→ icwiener joined	14:58
	← thierryp left	14:59
	→ texel joined	14:59
	← Croms left	15:00
crocket	meh~~~~	15:02
	→ bn_ joined	15:02
	→ alvaro_o joined	15:03
	→ NetHawk joined	15:03
	→ Croms joined	15:03
	→ icwiener_ joined	15:03
	← nixness left	15:04
	→ deuterium joined	15:04
	→ avinashhm joined	15:04
	← icwiener left	15:04
	hendeby → hendeby\|afk	15:05
	→ kumbayo joined	15:06
	← fga left	15:07
	← ShooterMG left	15:07
	→ helvete joined	15:08
	→ ShooterMG joined	15:08
	← Myth17 left	15:08
	← wilmoore left	15:08
	→ alberto56 joined	15:10
	→ gusnan joined	15:11
	← Remixman left	15:11
	→ Myth17 joined	15:12
	← Myth17 left	15:12
	→ Myth17 joined	15:12
	→ dreamcode joined	15:12
	← dreamcode left	15:12
	→ dreamcode joined	15:12
	→ nitish joined	15:14
	← nitish left	15:14
	→ nitish joined	15:14
	← dim left	15:14
	→ dim_ joined	15:14
	← cilly left	15:14
	← eschnou left	15:15
crocket	According to "man git-clone", I can't clone only a subdirectory of a git repository.	15:15
jast	[automatic message] the 'git-clone' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-clone.html	15:15
	→ _psychic_ joined	15:15
crocket	I have to clone an entire repository.	15:15
Bombe	Right.	15:16
	← Guest60131 left	15:16
	← teukka left	15:16
	→ Davey joined	15:16
	← aminpy left	15:16
	← cbreak_work left	15:17
	← Myth17 left	15:17
	→ teukka joined	15:17
	→ aminpy joined	15:18
	← bentob0x left	15:19
	→ foocraft joined	15:20
	← leo2007 left	15:20
	← angelsl left	15:21
	→ tjholowaychuk joined	15:21
	→ henrikbjorn joined	15:21
	→ leo2007 joined	15:21
	→ pj_1 joined	15:22
	← etnt left	15:22
	ChanServ set mode: +v	15:23
	→ webPragmatist1 joined	15:23
	← webPragmatist left	15:23
pj_1	I created a branch, switched to a it, made changes, committed the changes and did a git push	15:23
	but I don't see that branch on github.com	15:23
	anything I am doing wrong?	15:23
cmn	git push doesn't push branches that don't exist upstream	15:24
	git push origin yourbranch	15:24
	← iravid left	15:24
cmn	I should say "that don't exist in the remote" instead of "upstream"	15:24
	→ bkozal joined	15:25
	→ iravid joined	15:25
pj_1	oh yeah. I forgot . thanks	15:25
	← linusoleander left	15:25
	→ warthog9 joined	15:26
	→ verto joined	15:26
	→ nevyn^_ joined	15:29
	← nevyn^ left	15:29
	nevyn^_ → nevyn^	15:29
	← nevyn^ left	15:29
	← sako left	15:30
	→ dnorman joined	15:31
	→ slonopotamus joined	15:31
	← bauruine left	15:32
	← rmAdfinis left	15:32
darkskiez	if i've resolved conflicts during a git merge, and later want to rebase, why cant the rebase use the merge resolution that i used for the merge, or is that was rerere is for.	15:32
	← deekayen left	15:33
	→ envi joined	15:34
	← avinashhm left	15:34
	← crocket left	15:34
cmn	darkskiez: that is exactly the reason rerere exists	15:34
	if you want to do that king of thing, you should activate it	15:34
	s/king/kind	15:34
	← michael_lf left	15:35
	→ Anubisss joined	15:35
	← dnorman left	15:36
darkskiez	but can rebase not see the merge commit that also changes files as a patch rather than just dropping it	15:37
	← Drakevr left	15:37
	→ deekayen joined	15:37
	→ fs-sylr joined	15:37
cmn	rebase can't know you already resolved conflicts	15:38
	unless it uses the rerere machinery	15:38
	← giallu left	15:39
	← alvaro_o left	15:39
darkskiez	but it can see from the merge commit what was changed ?	15:39
	← hackez left	15:40
	← thiago left	15:40
	← SRabbelier\|Lappy left	15:41
	← AkumaStreak left	15:41
	→ AkumaStreak joined	15:42
	q0rban\|otp → q0rban	15:42
	→ d0k joined	15:42
darkskiez	instead of just deleting the merges from the history, treat them as patches	15:42
	← singlow left	15:42
	→ robotarmy joined	15:42
	← jonatasnona left	15:43
mariusso	darkskiez: Did you try to use the -p option with rebase, which tries to recreate merges?	15:43
	→ mklappstuhl joined	15:43
	→ Heimidal joined	15:43
darkskiez	AAAAAAAAAaaahhhh	15:44
	i had a blind spot looking at the man page	15:44
	→ kipras joined	15:45
	← freimatz left	15:45
darkskiez	magic	15:46
xelister	darkskiez: =)	15:47
	→ coachz joined	15:48
	→ Finalflash50 joined	15:48
	→ rmAdfinis joined	15:48
	→ __name__ joined	15:50
	← rmAdfinis left	15:52
infid	to delete a file from a repo / locally, do you do: 'rm file; git rm file' ?	15:52
karstensrage	just git rm file	15:53
coachz	do you want to remove it from your disk also ?	15:53
	→ cilly joined	15:53
coachz	or just form the staging area	15:53
	← reds left	15:54
infid	both	15:54
coachz	git rm path/file	15:54
infid	from the repo, staging and disk	15:54
	← kumbayo left	15:54
coachz	if you ever manually delete it from disk, you can sync your staging area up with what the disk is using 'git add -A'	15:55
	→ devth joined	15:56
	← fr0sty\|mobile left	15:56
	→ fr0sty\|mobile joined	15:56
infid	what if i rm a file locally but now want it back locally	15:56
	eg i do 'rm file' from disk, but now want to restore it from the head of the repo	15:57
mariusso	git checkout HEAD -- path/file	15:57
PerlJam	infid: git status will tell you how to restore it	15:57
	→ Drakevr joined	15:57
infid	thanks	15:57
coachz	how do i delete a tracking branch ?	15:57
	← tyteen4a03 left	15:57
mariusso	git branch -D <branch>	15:58
	← jason237 left	15:58
	← JohnFlux left	15:58
	→ oriba joined	15:58
devth	coachz: do you mean the remote brach?	15:58
coachz	thx	15:58
	← Spockz left	15:58
coachz	local	15:58
	that works	15:58
devth	k, yep	15:58
	← acelent left	15:58
	→ amerine joined	15:58
	→ acelent joined	15:59
	→ jason237 joined	15:59
	← j0ran left	15:59
	→ wilmoore_ joined	16:00
	← vanmik left	16:03
infid	i deleted a file with 'git rm file ; git commit -a -mm' but when i do 'git ls-files -d' it doesn't show it	16:05
	← webPragmatist1 left	16:05
coachz	you are trying to restore a file from a previous commit ?	16:06
fr0sty	infid: if you committed the delete then the file is gone from both the index and working tree.	16:06
	if you did ls-files before the commit it would have shown you the delete...	16:06
infid	fr0sty: so git ls-files -d is only if it's deleted from the index?	16:06
	ok thanks	16:06
	but so would have 'git status'	16:06
	i guess i just misunderstood what git ls-files is for	16:06
fr0sty	git-ls-files - Show information about files in the index and the working tree	16:07
	← RaceCondition left	16:07
infid	yeah it's the 'and the working tree' that through me off	16:07
	← joel_falcou left	16:07
infid	is there another way to do what i'm wanting	16:07
	like ls the history to only show me deleted files	16:07
fr0sty	what do you want?	16:07
	← q0rban left	16:08
fr0sty	git log --name-status --diff-filter=D? <-- just a guess...	16:08
	→ q0rban joined	16:08
infid	wow, nice thanks	16:09
	→ vanmik joined	16:10
	q0rban → q0rban\|lunch	16:10
cmn	r	16:10
	← pj_1 left	16:10
cmn	damn, wrong screen	16:10
	→ gitRMhelp joined	16:10
mariusso	fr0sty: That was a cool and powerful options combo :) nice	16:11
	← kylehayes left	16:11
	← mklappstuhl left	16:11
	← nitish left	16:11
	→ nitish joined	16:12
	→ kumbayo joined	16:12
	← babbio left	16:12
coachz	how do a create my local branch and make it a tracking branch from origin/dev that i pulled down ?	16:12
	→ kenichi joined	16:13
mariusso	coachz: git checkout -b <local branch> origin/dev	16:13
coachz	thanks	16:13
mariusso	coachz: That both creates the (tracking) branch and checks it out	16:13
coachz	worked perfectly, thank man	16:14
	→ hackez joined	16:14
	mariusso embarks on a rebase of 620+ patches, with several merges in the branch..	16:14
eydaimoon	how do I delete a branch on a remote repo on something like github for example? Is it possible through git?	16:15
	BinaryMuse → BinaryMuse[away]	16:15
	← AkumaStreak left	16:15
rane_	git push <remote> :<branch>	16:15
spb	eydaimoon: man git-push /delete	16:16
jast	eydaimoon: [automatic message] the 'git-push' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-push.html	16:16
eydaimoon	thanks	16:17
	← drl left	16:17
	→ migrantgeek joined	16:17
	→ leadgy joined	16:17
leadgy	hi guys	16:17
	anyone with EGit experience?	16:18
	← crashanddie left	16:18
	← uwjweq left	16:19
	← achadwick left	16:19
	← flaguy48 left	16:20
fr0sty	mariusso: git rebase with merges may not do what you expect, so be aware...	16:20
	← bn_ left	16:20
fr0sty	leadgy: just ask the question...	16:21
	→ bn_ joined	16:21
leadgy	I'm pushing the current branch(master) via EGit with Push to Upstream, but in history I don't see the origin to change	16:21
	however, if I do a "git push" from console, the history in Eclipse gets updated	16:21
	hi fr0sty, thanks for replying	16:22
	← Finalflash50 left	16:22
rendar	damnit, i've wrong the last commit's comment, can i chage that now?	16:22
devth	git commit --amend	16:23
rendar	ok	16:23
	← TrineLocke left	16:23
leadgy	fr0sty: but if in Eclipse I do a Fetch after the push, the history gets updated and it shows correctly that the origin(remote branch) was updated	16:23
	← mikewintermute left	16:24
mariusso	fr0sty: I know. Ideally I want to get rid of all those merges, and rebase it properly. I'm giving it a go both with and without -p though..	16:25
	q0rban\|lunch → q0rban	16:25
	→ drake01 joined	16:26
	→ bonhoffer joined	16:26
bonhoffer	what does it mean if "Your branch is ahead of 'origin/master' by 1 commit."	16:28
	i get that from a git status	16:28
spb	it means what it says	16:28
PerlJam	it means you've got a commit you haven't pushed.	16:28
fr0sty	bonhoffer: it means exactly what it says...	16:28
coachz	bonhoffer, you know that origin/master is a local copy of master on your server	16:28
bonhoffer	PerlJam, so i just thought i pushed it via git push origin/master	16:28
PerlJam	bonhoffer: it means you have one more commit in your local branch than you do in the branch its tracking, origin/master	16:29
coachz	when you run 'git remote update' it pulls down the latest origin/master and others	16:29
fr0sty	leadgy: it is possible Egit is not using the correct upstream repository (or you perhaps don't have one defined).	16:29
bonhoffer	am i on a local branch? i thought i was on master	16:29
coachz	run 'git branch -a' and you will see all your branches AND the remotes you pulled down	16:29
fr0sty	when you did it from the console did you just do 'git push'?	16:29
	← oriba left	16:29
coachz	you do NOT push origin/master or any of the remotes	16:29
	you push your local branches	16:29
bonhoffer	oh	16:29
	git push origin/master	16:30
coachz	no	16:30
	→ drl joined	16:30
coachz	git push origin master	16:30
bonhoffer	yes, that	16:30
coachz	in my example origin is found in .git/config file	16:30
	and pushes to that server	16:30
bonhoffer	i get "everything up to date"	16:30
coachz	so you are even now	16:30
bonhoffer	true, ok	16:30
coachz	all is well	16:30
leadgy	fr0sty: but if I do a "git push" from the local folder everything gets pushed correctly and then "git branch -v -v" shows the remote version was updated	16:31
	→ RaceCondition joined	16:31
coachz	but origin/master is a local cached copy of the server	16:31
	← cordoval_ left	16:31
bonhoffer	coachz, makes sense	16:31
coachz	and you get the latest version of that with 'git remote update'	16:31
leadgy	doing a "git branch -v -v" after pushing from EGit shows [1 step behind]	16:31
FauxFaux	leadgy: Ooh, didn't know about git branch -vv, nice.	16:31
fr0sty	leadgy: which is behind?	16:31
	← gusnan left	16:32
	← kenichi left	16:32
	→ kenichi joined	16:33
	← ziro` left	16:34
leadgy	fr0sty: the origin/master remains behind	16:35
	with 1 commit	16:35
fr0sty	so the local branch is one commit ahead?	16:35
	does 'git fetch origin' fix the problem?	16:36
	← nitish left	16:37
	→ nitish joined	16:37
	→ tyfighter joined	16:38
	← codespectator left	16:38
	← simplechat left	16:39
leadgy	fr0sty: yes, it fixes the issue	16:39
	← towski left	16:39
leadgy	but I wanted to obtain that directly from Eclipse/EGit when I was doing the push to upstream	16:39
fr0sty	then the 'git push' Egit uses is at fault.	16:39
leadgy	why do I have to fetch after push?	16:39
devth	best option is to use the CLI instead of egit	16:40
	← tyfighter left	16:40
fr0sty	it is pushing "git push origin mybranch:theirbranch" probably, which doesn't update origin/theirbranch.	16:40
	(just a guess)	16:40
	→ Zephyrus joined	16:40
leadgy	do you use EGit, maybe you could try my test case and tell me if you get same results	16:40
devth	no, i avoid it. coworker uses it and says its buggy	16:41
	(i also avoid eclipse :)	16:41
leadgy	:)	16:41
	fr0sty: how can I test if the case you told me above is what happens in reality?	16:42
	← xiangfu left	16:42
	BinaryMuse[away] → BinaryMuse	16:42
fr0sty	I think you already did, in essence. you push a commit to the remote repository but the update isn't reflected in refs/remotes/origin/master	16:42
	Egit is a re-implementation of git in Java (I believe) so it isn't doing any native 'git' operations at all.	16:43
	← jatt left	16:43
	→ mando joined	16:43
	→ namemon joined	16:44
	← drl left	16:45
leadgy	fr0sty: this is my EGit Push configuration: http://min.us/mbf9XlXUGcrOZC#1	16:45
	← robinbowes left	16:45
leadgy	fr0sty: I didn't specify anything for RefSpec - maybe I should?	16:46
	it says it pushes refs/heads/:refs/heads/	16:46
	don't know if this is correct	16:47
	← dharrigan left	16:47
	← helvete left	16:47
fr0sty	that's the default, should be fine...	16:47
	→ dharrigan joined	16:48
	← dharrigan left	16:48
	→ dharrigan joined	16:48
leadgy	fr0sty: well then, it's not working - probably buggy as you said..	16:48
	→ vol7ron joined	16:48
leadgy	I hope someone that uses EGit could test this behavior	16:48
	→ trupheenix joined	16:51
	← teukka left	16:51
	→ teukka joined	16:51
	ptr_uzl → ptr_uzl_away	16:51
	→ Chillance joined	16:52
	→ smolyn joined	16:52
	→ linoj joined	16:52
	→ zekus joined	16:52
	→ LiamH joined	16:52
	→ gebbione joined	16:53
	→ saschpe joined	16:53
	→ metcalfc joined	16:54
	→ sako joined	16:54
	← LouisJB left	16:55
	← davzie left	16:55
	→ yunosh joined	16:56
	← PigeonFriend left	16:56
	→ drl joined	16:56
	→ krosenvold joined	16:57
	← curtana left	16:57
	← vol7ron left	16:57
yrlnry	Suppose commit X has caused a test failure, and then I make commit Y, with parent X, to fix the test failure. What's a good way to mention X in the commit message for Y? Is it enough to just say "Fix test failure in preceding commit"? Or it it better to include the abbreviated SHA1 of X?	16:58
	The former is fragile if Y gets cherry-picked elsewhere. But the latter is not transparent and forces the reader to run a command to figure out that the commit referred to is actually Y^.	16:58
	→ IslandUsurper joined	16:58
	← SimonP86 left	16:58
	→ robinbowes joined	16:58
	→ SimonP86 joined	16:59
	→ Textmode joined	17:00
fr0sty	quote the header of the previous commit? : e.g. "Fix bug introduced by 'Add frotz to bar-dongle in baz widget'"	17:00
yunosh	hi, for some reason commit ids of two partially merged branches don't line up anymore (probably some accidental rebase or something). when i try to merge those branches, i get conflicts of course. none of the merge strategies seem to able to cope with that. how can i tell the merger to actually look out for the same changes in both branches, or how can i tell git to skip a numer of commits when doing the merging?	17:00
	→ cschneid joined	17:00
	adamholt → adamholt_away	17:01
	adamholt_away → adamholt	17:01
	← k0rnel_ left	17:01
	→ k0rnel joined	17:01
fr0sty	yrlnry: a unique enough string makes 'git log --grep="..."' somewhat easy.	17:02
	← acelent left	17:03
	← dharrigan left	17:03
fr0sty	yunosh: "partially merged branches"?	17:04
	← linoj left	17:04
	← HrznDefeated left	17:04
	→ eletuchy joined	17:05
	← migrantgeek left	17:05
	← mattly left	17:05
	→ tvw joined	17:05
	→ dreiss joined	17:06
	→ mattly joined	17:07
	→ TrineLocke joined	17:07
yunosh	fr0sty: yes, they are merged, up to certain point in the past, but git doesn't seem to recognize it, i guess because for the most recent merged commits, the two branches have different commit ids. now one of the branches has even more commits that i want to merge into the other branch. but git not only tries to merge those new commits, but also all the old commits from the point on where the commit ids diverge	17:07
fr0sty	yunosh: the branches are merged, or they are not...	17:08
	does 'git merge-base A B give the last merge commit between A and B ?	17:08
yunosh	for git they are not. for the code they are. let me check	17:08
	← ningia left	17:08
	→ t0rc joined	17:08
cbreak	if there's no merge commit, then they have not been merged	17:09
	→ elementz joined	17:09
elementz	hi all.	17:09
cbreak	there's nothing like "git doesn't recognize it"	17:09
	because git KNOWS what was merged and what was not. there's nothing to recognize	17:10
fr0sty	yunosh: "kinda having the same changes" is not the same as "merged"...	17:10
cbreak	when you merge two branches in git, it will make a new commit that has both branch heads as parent	17:11
	and merge all changes since the branches diverged	17:11
elementz	i just tried to use the following .gitignore rules https://github.com/github/gitignore as a global setting for my git installation. added a java project, and ended up with the .class files being tracked	17:11
	i used this line to set global ignores git config --global core.excludesfile ~/.global_ignore	17:12
	← slonopotamus left	17:12
cbreak	elementz: all of them?	17:12
fr0sty	elementz: what was in ~/.global_ignore?	17:12
elementz	cbreak: you mean if i used all of those ignorefiles? yep. i just checked out the repo	17:12
	→ towski joined	17:12
cbreak	elementz: so, you know that you used none of them, right?	17:13
yunosh	yes, i know but somebody messed up this tree, so that git somehow lost track about that last merging	17:13
elementz	cbreak: yeah, i guess... but how come?	17:13
fr0sty	yunosh: there was no merge, then...	17:13
cbreak	because you didn't tell git to use an ignore file	17:13
	→ gp5st joined	17:13
cbreak	besides .global_ignore	17:13
fr0sty	which was likely empty...	17:13
	if it even exists...	17:14
yunosh	i KNOW the difference between what git considers a merge and what happened in the work tree	17:14
gp5st	is there anythign similar to git commit -v but that comments the patch lines?	17:14
	← spaceonline left	17:14
cbreak	yunosh: what happens in a work tree is irelevant	17:14
	→ chrischambers joined	17:14
fr0sty	yunosh: can you pastie: 'git log --oneline --graph --decorate A B' ?	17:15
	← Zephyrus left	17:16
	→ ysulaiman joined	17:16
	← Yuffster left	17:16
	→ cakehero joined	17:16
elementz	fr0sty: cbreak hm. .global_ignore is full of foo.gitignore files. and shouldn't the line above add all those files to the global ignore rules?	17:16
fr0sty	or git log --graph --format="%h %d" A B if you are paranoid...	17:16
gp5st	also, is there anyway to do commits while doing an interactive add?	17:16
fr0sty	elementz: the argument is "excludesfile" not "excludesdirectory"	17:17
elementz	ah ok.	17:17
	elementz blushes	17:17
	← leo2007 left	17:18
cbreak	elementz: it will read the .global_ignore file and ignore everything in it	17:18
ysulaiman	Hi. Anyone using git to track LaTeX files on Windows?	17:18
	← dpino_ left	17:19
	← eletuchy left	17:20
	← nhance left	17:20
	→ Zephyrus joined	17:20
	→ eletuchy joined	17:20
cbreak	I used git to track LaTeX files on Mac OS	17:20
	→ nhance joined	17:20
elementz	cbreak: so there is no simple way to actually add all those .gitignore files? do i understand it right, that i would have to add them to a custom .global_ignore file?	17:20
cbreak	elementz: cat fileA fileB >> .global_ignore	17:21
ysulaiman	cbreak: Did you use the vc bundle?	17:21
	→ pdelgallego_ joined	17:21
devth	elementz: you prob only want to use those files on a project-by-project basis instead of globally	17:21
cbreak	ysulaiman: no bundle	17:21
	→ squiggy joined	17:21
ysulaiman	cbreak: I see.	17:21
fr0sty	yunosh: you lose interest?	17:23
	← mando left	17:24
yunosh	no, but i figured i don't want to pastie that few year's worth of 20000 commits list. i'm trying right to reset the tree to the last known stable state and re-merge	17:24
	← nitish left	17:24
	← ShooterMG left	17:24
	← ptl left	17:24
	→ Myth17 joined	17:25
	← Myth17 left	17:25
	→ Myth17 joined	17:25
	← robert_david left	17:25
	→ ptl joined	17:25
yunosh	and that actually seemed to have worked fine. i guess i have a rewritten history now, but i don't care at this point	17:26
	← whaley left	17:26
	→ fission6 joined	17:27
	← jonanin left	17:27
fission6	if i have a merge conflict how do i take mine?	17:27
	← alex__c2022 left	17:28
	→ alex__c2022 joined	17:28
	→ LouisJB joined	17:28
	← frogonwheels left	17:28
	→ Bass10 joined	17:28
	→ frogonwheels joined	17:29
	→ nitish joined	17:29
	← nitish left	17:29
	→ nitish joined	17:29
	← t0rc left	17:29
	← Bass10 left	17:29
	← Myth17 left	17:30
	→ jdummy joined	17:30
	→ Bass10 joined	17:30
jdummy	Hi all. Any idea why git status shows all files in the project as modified, even though they haven't been modified?	17:31
	← Jippi_mac left	17:31
ysulaiman	jdummy: What does git diff show?	17:31
gp5st	jdummy: permission or mtime changes?	17:32
	jdummy: diff it like ysulaiman said	17:32
cehteh	lineendings trouble?	17:32
	→ beyondcreed joined	17:32
jdummy	git diff shows everything as well	17:32
fission6	"error: Your local changes to 'settings.py' would be overwritten by merge. Aborting." how do i take my local copy	17:32
	← ricky left	17:33
cehteh	jdummy: then look closer you quite certainly did something very stupod	17:33
yunosh	eh, that doesn't work of course. how do i revert all commits, back to a certain point, without having to specify all commits individually? iiuc git revert only reverts individual commits, while git reset rewrites the history, right?	17:33
gp5st	fission6: commit or revert it (revert discards your changes	17:34
	→ friskd joined	17:34
gebbione	hi i m reading the kernel.org tutorial on GIT	17:34
gp5st	alt, you could stash it then stash pop	17:34
cehteh	yunosh: no ..	17:34
	← chrizz left	17:34
jdummy	cehteh: you are correct :) the files have been chmodded. Thanks ysulaiman, gp5st cehteh	17:34
gebbione	I am not really following it completely i would like to understand the logical structure of GIT, any better tutorial?	17:35
cehteh	git revert adds a reverse commit on top, git reset resets history to some (earlier) point but doesnt rewrite it	17:35
devth	gebbione: http://progit.org/book/	17:35
gebbione	devth: thanks	17:36
gp5st	gebbione: the git-book	17:36
	→ spaceonline joined	17:36
	← gitRMhelp left	17:36
yunosh	cehteh: hm, but if i reset, my tree is behind origin's tree. or do i need to specify some other reset option?	17:36
	← dreamcode left	17:37
yunosh	the problem is those change i want to revert are already in origin, not only locally	17:37
	→ slonopotamus joined	17:37
cehteh	sure .. it chops commits off, you really loose then	17:37
	if you then commit again you divert from the origin	17:37
	yes then use git revert	17:38
	← coachz left	17:38
cehteh	origin is yours, not some upstream?	17:38
	← jelmer left	17:38
fission6	gp5st, okay i commited still have unmerged issues, i just want to take "mine"	17:38
yunosh	which takes me back to the original question, can i tell revert to revert more than one commit? we're talking about a few hundred commits	17:38
	orgin is upstream but under my control	17:38
cehteh	outch .. iirc revert has a -n option for no commit (RTFM) .. that lets you revert a lot commits and then commit this reversals in one big commit	17:39
	→ robert_david joined	17:39
	→ alex__c2022_ joined	17:39
cehteh	but if you want to revert more than a fre commit you prolly want to do something else	17:40
	→ siprbaum joined	17:40
cehteh	starting a new branch and declare that 'official'	17:40
yunosh	meh	17:40
gp5st	fission6: then fix up the merge	17:40
cehteh	or reset --hard and tell everyone that he has to follow you :P	17:40
fr0sty	yunosh: git revert can take a range of commits...	17:41
gp5st	if it's a file that's different on each computer fission6 it prob shouldn't be under scm	17:41
fission6	how do i use "mine"	17:41
cehteh	ah i didnt know :)	17:41
	← cilly left	17:42
	→ cillybabe joined	17:42
devth	fission6: git checkout --ours <file>	17:42
fission6	then commit that?	17:42
	cillybabe → cilly	17:42
devth	yep	17:42
fission6	ok	17:42
	← alex__c2022 left	17:42
	alex__c2022_ → alex__c2022	17:42
	→ gigamonkey joined	17:42
fission6	thanks	17:42
	→ mithridates joined	17:42
	→ sechrist joined	17:43
	← zekus left	17:43
mithridates	does anyone know how git store information in sha-1 and how it extracts it?	17:44
	can someone tell me what algorithm is used there?	17:44
	→ gebi joined	17:44
cbreak	the source can	17:44
PerlJam	"store information"?	17:44
	→ reds joined	17:44
mithridates	yes, date,...	17:44
gp5st	mithridates: you can't extract info from a sha, that's the point. it's a one-way/trap-door function	17:45
mithridates	date, size ,.. I call these things infromation	17:45
sgronblo	yeah you cant store anything in a hash, its a one-way transformation	17:45
fr0sty	http://repo.or.cz/w/git.git?a=tree;f=Documentation/technical;hb=HEAD	17:45
	that's technical documentation, including file formats ^^	17:45
gp5st	mithridates: i'm not 100% sure what is being sha'd at any given time, but it's basically the state of that file/tree. if the file/tree changes, so does the sha and a difference can be detected	17:45
mithridates	what do u mean about one-way?	17:45
Nugget	sha1 is one-way.	17:46
	you can't un-sha something	17:46
	→ zekus joined	17:46
fr0sty	http://en.wikipedia.org/wiki/Cryptographic_hash_function	17:46
PerlJam	mithridates: think of it as a black hole. You put stuff in, and it never comes back out.	17:46
	→ booginga joined	17:46
	→ ricky joined	17:46
	→ joel_falcou joined	17:46
mithridates	so what's the point of puting stuff into it?	17:46
PerlJam	mithridates: you can check if two black holes have been constructed the same, but you have to make them first.	17:46
Nugget	to comapre things and spot changes	17:46
mithridates	you mean it's like md5 sum-check?	17:46
gp5st	mithridates: yes	17:47
	→ chrizz joined	17:47
Nugget	yes, but more secure than md5	17:47
gp5st	but for the state of the object at that time	17:47
PerlJam	mithridates: the point is really "uniqueness"	17:47
sgronblo	well i wouldnt compare it to a black hole	17:47
	its a function, so you do get something out of it	17:47
gp5st	yeah, it's an (should very well be) unique value to describe that point in time	17:47
mithridates	well, if uniqness was the point there is some sha collision problems...	17:47
	→ thiago_home joined	17:48
PerlJam	mithridates: because ... ?	17:48
	← saleh left	17:48
gp5st	mithridates: that's a very rare occurance	17:48
yunosh	fr0sty: i'm trying to grasp git-rev-parse(1) but i can't figure out how to specify all commits from a tag up to HEAD for git revert. this is rocket science	17:48
gp5st	mithridates: for better hash functions it's almost impossible to create a collision by hand. and in general, two states _wont_ have the same hash unless they are the same	17:49
fr0sty	tag..HEAD ?	17:49
mithridates	so you guys all sure 100% that sha-1's purpose is only comparison like md5 checksum	17:49
	← hackez left	17:49
	← GrowLinux left	17:49
cbreak	not comparison	17:49
	it's used as identifier	17:49
fr0sty	identification	17:49
	→ madewokherd joined	17:49
cbreak	_unique_ identifier	17:49
	→ codespectator joined	17:49
	BinaryMuse → BinaryMuse[omg	17:49
	BinaryMuse[omg → BinaryMuse[omgfo	17:49
sgronblo	pseudo-unique or something	17:49
	BinaryMuse[omgfo → BinaryMuse[screw	17:49
mithridates	ok I gochya	17:49
	→ gastaldi joined	17:50
cbreak	it's also a content-derived specifier	17:50
mithridates	thanks	17:50
yunosh	fr0sty: nope	17:50
devth	mithridates: read up in http://progit.org/book/ch9-2.html	17:50
gastaldi	hey	17:50
cbreak	so it's reproducible	17:50
	← leadgy left	17:50
gastaldi	I added some files on my GIT project and "git reset --hard" did not delete them	17:50
	is that right ?	17:50
sgronblo	gastaldi: only adding them won't put them in the repository yet	17:50
gastaldi	I simply copied the files, no GIT ADD were performed	17:51
cbreak	if you want to delete files, git rm is the way	17:51
sgronblo	first you add, then you commit	17:51
fr0sty	yunosh: why not? what is missing? (or added)?	17:51
	→ jhoepken joined	17:51
gastaldi	My intention is to revert to the last commit	17:51
devth	gastaldi: that's right. use `git clean` to remove untracked files	17:51
	gastaldi: man git clean	17:51
jast	gastaldi: [automatic message] the 'git' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git.html	17:51
	← nitish left	17:51
yunosh	fr0sty: git cannot even find the commits: "fatal: Cannot find 'before-update-20110408..HEAD'"	17:51
devth	gastaldi: man git-clean	17:52
jast	gastaldi: [automatic message] the 'git-clean' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-clean.html	17:52
	→ sraue_ joined	17:52
	← sraue_ left	17:52
gastaldi	git clean does not removes the added folder	17:52
	← Croms left	17:52
	→ Croms_ joined	17:52
devth	gastaldi: read the options	17:52
gastaldi	Ah ! It worked with -d -f	17:52
	But why git reset --hard does not do that ?	17:53
	← Yuuhi`` left	17:53
gp5st	mithridates: in fact, it's "so good" that often only the first 7 or 8 chars of the sha can be used and have it be unique in the project. sometimes there will be another one or two that share those, but never the full sha	17:53
devth	gastaldi: because those files aren't in the repo yet	17:53
gastaldi	devth: You mean that if I run git add it will work ?	17:53
	← walf443 left	17:53
fission6	what do i do to address this, ! [rejected] master -> master (non-fast forward)	17:53
	→ phaedrix joined	17:54
PerlJam	fission6: nothing?	17:54
devth	gastaldi: yes	17:54
gastaldi	fission6: git push origin master --force	17:54
PerlJam	fission6: pull first?	17:54
mithridates	gp5st: it was the reason that I became suspicious that may be there is way that it extracts data from it	17:54
	→ Yuuhi`` joined	17:54
gp5st	gastaldi: fission6 that's dangeours	17:54
PerlJam	fission6: you haven't given enough information to get a useful response :)	17:54
	fission6: but don't listen to gastaldi	17:54
gp5st	fission6: that means that the version on the server is newer than the last version you pulled from the server	17:55
fission6	ah got it thanks	17:55
gp5st	i'd recommend to pull from the server	17:55
	→ rns joined	17:55
gp5st	merge	17:55
	then push	17:55
fission6	yea that worked	17:55
gp5st	mithridates: what caused you to be suspicious ? and no, you can't extract data from it	17:55
	← gastaldi left	17:56
	← codespectator left	17:56
	← kumbayo left	17:57
	→ walf443 joined	17:57
mithridates	gp5st: because a portion of hash doesn't change if you make a small change, I know my reason is not technical but it was just a simple question in my mind	17:57
gp5st	mithridates: no, a small changes causes the whole hash to change	17:57
	→ gasull joined	17:57
gp5st	normally the ones that conflict in the first couple chars are random	17:57
	→ jonatasnona joined	17:57
mithridates	gp5st: For example, changing dog to cog produces a hash with different values for 81 of the 160 bits (wikipedia)	17:58
	→ hemmecke joined	17:58
mithridates	but I think you are right	17:58
	it says it changes the whole hash	17:58
gp5st	mithridates: it's a good question, don't get me wrong. it's just misguided since small changes will, more oftent han not, change the whole hash dramatically	17:59
	← owst left	17:59
mithridates	I gochya	17:59
	tnx	17:59
gp5st	np:)	17:59
fr0sty	yunosh: where did you get that tag name?	17:59
gp5st	like i said, good question and i'm sure you learned something about hashes	17:59
	← Jake_ left	17:59
	← cilly left	18:00
yunosh	fr0sty: i set it earlier myself	18:00
sgronblo	mithridates: the hash algorithms arent that complicated really, and when you realise how the bits are flying around you'll probably realise that it's basically impossible to extract any information back from the end result	18:00
yunosh	and git tag still shows it too	18:00
	← sebersole left	18:00
fr0sty	does "git show before-update-20110408" give you what you expect.	18:00
	→ logik-bomb joined	18:00
	← logik-bomb left	18:00
	→ logik-bomb joined	18:00
fr0sty	?	18:00
	→ sebersole joined	18:01
	→ leo2007 joined	18:01
yunosh	fr0sty: yes	18:01
fr0sty	and 'git rev-parse before-update-20110408..HEAD' ?	18:01
gasull	Hi. Why do I get this error? http://dpaste.com/532055/ Thanks.	18:02
fr0sty	gasull: what are you trying to accomplish?	18:02
	→ opalka joined	18:02
yunosh	fr0sty: that gets me some SHAs	18:03
	→ Plouj joined	18:03
Plouj	hi	18:03
gasull	fr0sty: I want to grab the latest version in origin of media/css/screen.css and overwrite what I have in local.	18:03
Plouj	is there a way to find _all_ tags reachable from a given commit?	18:03
fr0sty	yunosh: but it doesn't work with 'git revert' ?	18:03
yunosh	fr0sty exactly	18:03
fr0sty	git --version ?	18:03
gasull	fr0sty: git version 1.7.2.2	18:04
	→ Evious joined	18:04
	← eydaimoon left	18:04
fr0sty	gasull: fetch is not what you want. you need checkout	18:04
	→ flaguy48 joined	18:04
yunosh	fr0sty 1.7.1	18:04
	← thiago_home left	18:05
	← flexable777 left	18:05
gasull	fr0sty: error: path 'media/css/screen.css' is unmerged	18:05
	fr0sty: using checkout	18:05
	← robotarmy left	18:05
fr0sty	yunosh: git revert only takes single commits in 1.7.1	18:06
	adamholt → adamholt_away	18:06
	adamholt_away → adamholt	18:06
fr0sty	gasull: are you in the middle of a merge?	18:06
	→ woozie joined	18:06
gasull	fr0sty: yes	18:06
	← bla left	18:06
	→ bla joined	18:06
fr0sty	git checkout --theirs media/css/screen.css	18:06
	← Plouj left	18:06
	→ pedrobelo joined	18:07
yunosh	fr0sty: oh. an i thought that would be a recent version	18:07
gasull	fr0sty: how can I tell git to forget everything from the last commit?	18:08
	→ eydaimoon joined	18:08
	← ysulaiman left	18:08
	← robert_david left	18:09
	← panike left	18:09
fr0sty	yunosh: 1.7.4.4 is the latest	18:09
	→ flexable777 joined	18:09
fr0sty	gasull: do you want to abort your merge?	18:10
gasull	fr0sty: yes	18:10
	→ cordoval_ joined	18:10
gp5st	is there anythign similar to git commit -v but that comments the patch lines? also, is there anyway to do commits while doing an interactive add?	18:10
	← marcob left	18:10
gasull	fr0sty: git reset —hard HEAD?	18:10
	← slonopotamus left	18:10
	→ bgy joined	18:10
fr0sty	yes, that will abort the merge.	18:11
	← dreiss left	18:11
gasull	fr0sty: thanks	18:11
patrikf	gp5st: everything after the first comment line is ignored.	18:11
elementz	ok, i am getting really confused now. i added the following to my core.excludesfile http://pastebin.com/aASeZxKA , but still .project .classpath and .zip files are being tracked by git	18:11
patrikf	gp5st: so with commit -v, the patch is not part of the commit message	18:11
fr0sty	elementz: ignore/exclude does not work on tracked files...	18:11
elementz	fr0sty: i removed my .git dir in the respective project, and then did a new git init && git add . - to no avail	18:12
	→ acelent joined	18:12
	→ irc2samus joined	18:12
	← acelent left	18:12
	← trupheenix left	18:12
	← bguery left	18:13
bnovc	I had a submodule in foo/bar and when I did a merge it had a conflict in .gitmodules. I remoevd it from gitmodules and just want to remove it entirely now	18:13
	but when I do a `submodule update --init`, it says that there is no entry in .gitmodules for it	18:13
fr0sty	elementz: are you sure the excludesfile name is correct?	18:13
gp5st	patrikf: i thought it was, since there are no #s prefixing the lins	18:13
	← eydaimoon left	18:14
elementz	fr0sty: how can i check?	18:14
bnovc	is there a better way to fix this than searching through .git to remove whatever stored that	18:14
	→ theoros joined	18:14
	← sontek left	18:14
	→ sontek joined	18:14
	→ p3rg2z joined	18:15
	→ eydaimoon joined	18:15
elementz	fr0sty: git config -l gives me core.excludesfile=.global_ignore	18:15
gp5st	patrikf: i guess you're right?:-p now is there a way to commit within an interactive add?	18:15
elementz	fr0sty: should .global_ignore reside under ~ ?	18:16
	→ saleh joined	18:16
	← ricky left	18:16
	← eydaimoon left	18:17
	→ eydaimoon joined	18:17
Captain`Ramen	something strange happened in git-svn today. someone here committed to both trunk and a branch at the same time using svn	18:17
	he probably checked out /	18:18
	→ whaley joined	18:18
	← saschpe left	18:18
Captain`Ramen	so now, when git svn fetch[ing], i have two different commits with the same git-svn-id. will this break git svn?	18:19
	→ dyer joined	18:19
	← dyer left	18:19
	→ dyer joined	18:19
fr0sty	elementz: use the full path...	18:20
	→ saschpe joined	18:20
elementz	fr0sty: yep. that seemed to be the problem. thank you	18:21
	← jdummy left	18:21
	→ HG` joined	18:21
	→ dreamcode joined	18:21
	← dreamcode left	18:21
	→ dreamcode joined	18:21
	← bindaas left	18:22
	← eydaimoon left	18:22
	→ mehwork_ joined	18:23
	→ eydaimoon joined	18:23
eydaimoon	test	18:23
	finally. ok, how can I stash just one file in a change? is it possible?	18:23
	eydaimoon → eydaimon	18:23
	← hemmecke left	18:24
gp5st	can multiple commands be aliased into a single command in git?	18:24
	like if i wanted to run add -p and then commit -v	18:24
FauxFaux	You can use ! to make it a shell command, git config --global alias.foo '!git add -p && git commit -v' etc.	18:25
	← workmad3 left	18:25
	← Pyker left	18:28
	← phantomcircuit left	18:28
gp5st	FauxFaux: thanks:)	18:28
	→ phantomcircuit joined	18:29
	→ Pyker joined	18:29
	Pyker → Guest76411	18:29
	→ dreiss joined	18:30
	← soc42_ left	18:30
	← LouisJB left	18:30
	← mattly left	18:31
	← ph^_ left	18:31
	← ceej left	18:32
	→ dannas joined	18:32
	→ codespectator joined	18:33
	→ mattly joined	18:33
	← codespectator left	18:33
	← rendar left	18:36
	← eletuchy left	18:36
	← ereslibre_laptop left	18:37
	← cakehero left	18:38
	← guardian left	18:39
	→ hackez joined	18:40
	→ robotarmy joined	18:40
	← cordoval_ left	18:41
	← noonoooo left	18:43
	→ icwiener joined	18:45
	← elementz left	18:45
	← icwiener_ left	18:45
	→ Myth17 joined	18:45
	→ voidx joined	18:45
	→ hemmecke joined	18:46
	← irc2samus left	18:46
	→ elementz joined	18:47
	← elementz left	18:47
	→ elementz joined	18:47
	← hemmecke left	18:48
	→ nitish joined	18:48
	→ LouisJB joined	18:49
	→ kingfishr_ joined	18:50
	← drake01 left	18:50
	← deekayen left	18:51
	← xiong left	18:51
	→ drake01 joined	18:51
kingfishr_	generally speaking, are git operations slowed down more by the raw size of the repository or the number of commits saved in the repo?	18:51
	→ deekayen joined	18:51
	← Myth17 left	18:52
	→ levifig joined	18:52
fr0sty	kingfishr_: which operations?	18:52
yrlnry	kingfishr_: depends a lot on the operation.	18:52
	← nitish left	18:53
yrlnry	kingfishr_: something like git log PATH has to search through the commits, but the repository size probably won't affect it much.	18:53
	→ sycofly joined	18:54
kingfishr_	yrlnry, git status would probably be most affected by the size, right?	18:54
wereHamster	kingfishr_: number of files in the working tree	18:54
kingfishr_	ah good point	18:54
PerlJam	kingfishr_: why are you worrying about the speed of git operations anyway?	18:54
wereHamster	kingfishr_: clone, however, is pretty much linearly dependent on repo size	18:55
	→ toabctl joined	18:55
	→ dl9pf joined	18:55
	→ L42y joined	18:55
	→ rson joined	18:55
	← jason237 left	18:56
kingfishr_	wereHamster, this occurred to me. Can you think of other operations which would be negatively impacted by, say, committing a 10GB binary to your repo?	18:56
	← voidx left	18:56
rson	If I pull changes from a remote repo that has poorly formatted commit messages, it's bad to rebase those commits against my master, is that correct?	18:56
	← khaije\|minder left	18:57
wereHamster	kingfishr_: all of them. Don't do it	18:57
yrlnry	kingfishr_: push/fetch	18:57
	→ robert_david joined	18:57
kingfishr_	PerlJam, just collecting information because we're thinking of splitting our repo up into multiple repos.	18:58
wereHamster	kingfishr_: while your'e at it, don't put files which are more than a couple megabytes into the git repo	18:58
PerlJam	kingfishr_: because it's already slow?	18:58
	→ eletuchy joined	18:59
kingfishr_	wereHamster, I'm aware of best practice here; I'm just wondering what the real consequences are of not following it	18:59
wereHamster	kingfishr_: pain. Lots of pain.	19:00
	← namemon left	19:00
kingfishr_	PerlJam, not too slow, but enough that we have to turn off automatic gc because it can ruin your day	19:00
	→ Heimidal_ joined	19:01
	← Heimidal left	19:02
	→ troessner joined	19:03
fr0sty	dare I ask the size of .git?	19:03
	→ nyingen joined	19:04
kingfishr_	4 gigs	19:04
	← sycofly left	19:04
	← troessner left	19:04
Captain`Ramen	...	19:04
	texel → texas	19:04
	texas → texel	19:04
fr0sty	what sort of crap is in there?	19:05
bremner	"assets"	19:05
	→ erichammond joined	19:05
bremner	we call crap "assets" now	19:05
fr0sty	bremner: no doublt...	19:05
	→ sycofly joined	19:05
kingfishr_	the thing is, most of the unnecessary size is from people who checked in vendored code and binaries long time ago for some reason. So, we can easily pull that out and then have a <100MB repo	19:05
	← rson left	19:05
wereHamster	fr0sty: maybe you are right with 'crap' :P	19:05
	kingfishr_: then do it!	19:05
kingfishr_	and by 'pull that out' I mean make a new repo	19:05
	but i'm trying to push for further splitting up (by projects) and i'm attempting to find pros/cons for that	19:06
	→ dunpeal joined	19:06
wereHamster	you have multiple projects in a single repo?!?	19:06
kingfishr_	indeed	19:06
dunpeal	Hey. I'm squashing a bunch of commits with `rebase -i`, and getting some conflicts. In a bunch of files, I now have "<<< HEAD" and ">>> 123 foo commit"	19:06
wereHamster	kingfishr_: that's stupid.	19:07
dunpeal	Is there an easy way to tell Git to resolve all of them in favor of HEAD?	19:07
wereHamster	one git repo = one project. End of story	19:07
	→ nembutal joined	19:07
wereHamster	dunpeal: checkout --ours -- . ?	19:07
dunpeal	`checkout --ours/--theirs` doesn't work.	19:07
kingfishr_	wereHamster, I agree with you, but I can't come up with super compelling arguments. The best I've got is that it's annoying to see so many unrelated commits.	19:08
dunpeal	wereHamster: `fatal: --ours/--theirs is incompatible with switching branches`	19:08
	→ mklappstuhl joined	19:08
fr0sty	wereHamster: that does 'take their version' not 'resolve conflicts in their favor' IIRC.	19:08
	ChanServ set mode: +v	19:08
yrlnry	fr0sty: it checks out the stage-2 version from the index.	19:08
wereHamster	kingfishr_: the argument is that one git repo = one project. There is nothing to discuss	19:08
yrlnry	fr0sty: then you can git-add the checked-out version, which is the one you wanted.	19:08
fr0sty	rly, yrlnry?	19:08
	→ lanthan joined	19:09
	← robotarmy left	19:09
kingfishr_	wereHamster, yep, tha'ts definitely not an argument. That's an unsupported statement.	19:09
yrlnry	fr0sty: see man page	19:09
	fr0sty didn't know that...	19:09
wereHamster	kingfishr_: did you use svn before?	19:09
kingfishr_	again, i'm not disagreeing with you, i'm just looking for better reasons. A link would be great...	19:09
fr0sty	yrlnry: when did all this 'stage' stuff start showing up in the docs?	19:09
kingfishr_	wereHamster, yes	19:09
yrlnry	fr0sty: a long long time ago.	19:09
	← lanthan left	19:10
fr0sty	I've only heard people yammer about it in the last few months...	19:10
yrlnry	fr0sty: as far as I know the index has always had stage 0,1,2,3 to record the different vrsions inc ase of a conflict.	19:10
	← eijk left	19:10
	→ eijk joined	19:10
	← levifig left	19:10
	→ mt3 joined	19:11
wereHamster	kingfishr_: what arguments do your coworkers have for keeping one repo?	19:11
	→ lanthan joined	19:11
yrlnry	fr0sty: git-checkout got --ours in Aug 2008	19:11
kingfishr_	mostly just that that's how it currently is	19:11
	→ voidx joined	19:11
fr0sty	kingfishr_: having one repository requires multiple 'master' branches for each repository to prevent mingled commits.	19:11
wereHamster	that's no argument either	19:11
	← DaveIngram left	19:12
	← gebbione left	19:12
fr0sty	it invites people to commit changes to the wrong branch, or comingled changes in the same commit.	19:12
yrlnry	fr0sty: and git-rev-parse supported the :[0-3] notation for the stage number in the index since Oct 2006.	19:12
	→ mattp_ joined	19:12
fr0sty	and makes browsing history potentially difficult...	19:12
mattp_	say i make a fork on github, how do i rebase said fork to sync to upstream?	19:12
	sorry if this is a dumb question	19:13
yrlnry	We have multiple repos here at my workplace, one per project. I think it's a mixed bag.	19:13
fr0sty	yrlnry: thanks.	19:13
wereHamster	kingfishr_: another argument: separate repos allow you to work on each project separately. You won't have to clone the big repo if you just want to work on one small project	19:13
yrlnry	On the down side, it sometimes happens that project A incorporates project B, and then someone needs to make commits to both A and B that are interdependent; it would be good if the history recorded that, but it doesn't.	19:14
	→ jonanin joined	19:14
	← rondon left	19:14
fr0sty	submodules...	19:15
yrlnry	Similarly, when working on A, and you realize it needs a concomitant change to B, you need to take extra care to make sure that someone checking out A doesn't get confused by the test failures they will see if they fail to run the tests against the right version of B. A mingled repository would avoid the problem completely,	19:15
fr0sty	but that brings its own barrel of fun...	19:15
yrlnry	Yes, but I consider "submodules" a "con" argument in one word. :)	19:15
	← mt3 left	19:16
yrlnry	Another one is that if you are looking for some component, just doing git-grep or whatever will not find it, you just have to magically know which repo to look in.	19:16
PerlJam	If there's tight coupling between projects, I tend to put them in the same repo rather than use submodules.	19:16
	There are too many update headaches with submodules.	19:16
yrlnry	PerlJam: But even if there's not tight coupling, you get the same problems. For example, A is an application that uses module B, which is developed independently of A.	19:17
PerlJam	right.	19:17
yrlnry	It has often happened that the project I am working on will call library function foo, and I can't find foo, and have to ask someone, and oh, it turns out to bee in the blarf.git repo, which I never even cloned.	19:17
	→ linoj joined	19:17
dunpeal	wereHamster: thanks, and I used the wrong syntax before.	19:18
yrlnry	I am not so well set up to argue against a large single repository, since we don't do that here and so I can't report on the non-obvious problems.	19:18
dunpeal	wereHamster: `git checkout --ours` does work, user just didn't specify any paths.	19:18
	← empity left	19:20
	← linoj left	19:20
	→ gebbione joined	19:20
	← zivester left	19:20
	→ levifig joined	19:21
mattp_	anyone?	19:21
	← whaley left	19:21
	→ olle_ joined	19:22
fr0sty	mattp_: add upstream as a remote, fetch it; git checkout <yourbranch>; git rebase upstream/branch; git push origin <yourbranch>	19:22
mattp_	fr0sty: got it. cheers	19:22
	← jrdavid left	19:23
	← toabctl left	19:23
olle_	Hi, when I do a git ls-remotes I have one listed that I want to get rid of. The output is at https://gist.github.com/920264	19:24
wereHamster	delete it	19:24
olle_	How would I go about deleting refs/head/master?	19:24
wereHamster	olle_: do you know how to delete brnaches from remote repos?	19:24
	← pielgrzym left	19:24
olle_	no but if thats what I need to do I can read up on that.	19:25
	← theoros left	19:25
	→ theoros joined	19:25
	← nembutal left	19:26
	← gebi left	19:27
	← gp5st left	19:27
	→ whaley joined	19:28
	← hendeby\|afk left	19:28
	→ zivester joined	19:29
olle_	wereHamster: thanks, worked out!	19:30
	← HG` left	19:31
	← Dr4g left	19:32
	← makubi left	19:33
	→ tgunr_ joined	19:33
	→ makubi joined	19:34
	→ AkumaStreak joined	19:34
	→ gebi joined	19:35
	← leo2007 left	19:36
	← RaceCondition left	19:36
	← Textmode left	19:37
	→ theoros_ joined	19:37
	← theoros left	19:37
	→ ricky joined	19:38
	← saschagehlich left	19:38
	← gasull left	19:39
	← drl left	19:39
	← gs left	19:41
	→ pantsman joined	19:42
	→ Textmode joined	19:42
	← squiggy left	19:43
	→ saschagehlich joined	19:44
	← tjholowaychuk left	19:44
	← olle_ left	19:44
	← gebbione left	19:44
	BinaryMuse[screw → BinaryMuse	19:46
	→ dl9pf_ joined	19:46
	← theoros_ left	19:46
	← dl9pf left	19:46
	← saleh left	19:49
	← pantsman left	19:49
	← bartek left	19:50
	← eletuchy left	19:52
	← dreamcode left	19:52
	← aminpy left	19:52
	→ bartek joined	19:52
	→ drl joined	19:53
	← xelister left	19:53
	← OOPMan left	19:53
	← kenichi left	19:53
	→ warz joined	19:53
warz	l	19:53
	→ kenichi joined	19:54
	← warz left	19:54
	→ squiggy joined	19:54
	→ sbell joined	19:55
	→ warz joined	19:56
	← drake01 left	19:56
	← warz left	19:57
	→ LongBeach joined	19:58
	→ tyfighter joined	19:58
	← mmattice left	19:59
	→ codebeaker joined	20:00
	← envi left	20:00
	→ comptekki joined	20:01
	← mklappstuhl left	20:01
	← joel_falcou left	20:01
	← saschagehlich left	20:02
	← drubies\|mbp left	20:02
	→ vdv joined	20:02
	→ drubies\|mbp joined	20:02
	← Dave^\| left	20:02
	→ olle_ joined	20:02
	→ theoros joined	20:03
	← Heimidal_ left	20:04
	→ Heimidal joined	20:05
	→ drizzd joined	20:05
	→ tjholowaychuk joined	20:06
	← ptl left	20:07
	→ mklappstuhl joined	20:07
	← kynlem left	20:08
	→ OOPMan joined	20:08
	← woozie left	20:08
	← whatdoesitwant left	20:08
	→ Barry-Nichols joined	20:08
	← Textmode left	20:09
	→ saleh joined	20:09
	← keystr0k left	20:10
	← mastroDani left	20:11
	→ ptl joined	20:13
	→ RaceCondition joined	20:14
	→ KaffeeJunky123 joined	20:15
	← jonatasnona left	20:15
	→ drake01 joined	20:15
	← deuterium left	20:16
KaffeeJunky123	I'm having trouble in getting rid of a local branch, I'm trying to delete the branch with git branch -D branchname, and then I'm creating a new branch with the same name, but for some reason it got the same patches applied to it as the branch I deleted, so it isn't an exact copy of master :/	20:16
	→ cordoval_ joined	20:16
	→ mastroDani joined	20:16
	→ Textmode joined	20:17
	→ Sonderblade joined	20:17
KaffeeJunky123	any Ideas what to do?	20:17
	← saleh left	20:17
fr0sty\|mobile	How are you creating it?	20:17
KaffeeJunky123	git branch -f name	20:18
	I tried without the -f too	20:18
	← Titosemi left	20:18
fr0sty\|mobile	What is checked out right now?	20:18
KaffeeJunky123	master	20:19
fr0sty\|mobile	Sure?	20:19
KaffeeJunky123	I know that I cannot delete the branch I'm working on	20:19
	yeah sure	20:19
fr0sty\|mobile	Does "git show master" show you the commit you want for branch foo?	20:19
KaffeeJunky123	yes	20:20
fr0sty\|mobile	git branch foo master then...	20:20
	← adreghiciu left	20:20
KaffeeJunky123	okay I did that and now I did git ckeckout foo	20:21
	→ adreghiciu joined	20:21
KaffeeJunky123	and it does report changes to some files	20:21
	→ glitterfang joined	20:21
	→ kpreid joined	20:21
	← dannas left	20:21
KaffeeJunky123	exactly those files that I don't want patched, because the patch I applied is outdated for the current commit	20:21
fr0sty	'changes'?	20:22
	you have uncommitted changes in your working copy?	20:22
KaffeeJunky123	I seem to have uncommited changes in the foo branch	20:23
	the one I just created from master :/	20:23
fr0sty	uncommitted changes do not belong to a branch...	20:23
	they are in your working copy	20:23
	they float along as you change branches.	20:23
	do you want to keep those changes? (or apply them elsewhere?)	20:24
	← metcalfc left	20:25
KaffeeJunky123	okay thanks	20:25
	a checkout -f in my master branch got rid of it	20:25
	← LongBeach left	20:26
	← fr0sty left	20:27
	→ Mattias joined	20:27
	← reds left	20:28
	← phantomcircuit left	20:28
sako	hey guys so does git diff <commit> head path/to/file give me the difference between head and that commit?	20:28
	→ gasull joined	20:29
sako	there was a bunch of lines deleted in a commit we got the sha id	20:29
	i want to show what was deleted in a patch style diff	20:30
	anyone know if thats correct?	20:31
	→ mmattice joined	20:31
	→ slonopotamus joined	20:31
	→ ptl_ joined	20:32
yrlnry	sako: git diff <commit> head -- path/to/file	20:33
	← verto left	20:33
	← ptl left	20:33
	← dajhorn left	20:33
yrlnry	sako: git diff <commit> HEAD -- path/to/file	20:33
	or you could just git diff <commit>:path/to/file	20:33
	← henrikbjorn left	20:33
	→ reds joined	20:35
	→ mando joined	20:36
	→ saleh joined	20:37
	← cordoval_ left	20:37
	q0rban → q0rban\|afk	20:38
	← dunpeal left	20:39
	→ codespectator joined	20:40
	← phishy left	20:42
	→ Luke joined	20:42
	Luke → Guest59163	20:43
	← Mattias left	20:43
	← KaffeeJunky123 left	20:43
	→ SRabbelier1 joined	20:43
	← SRabbelier left	20:43
	← SRabbelier1 left	20:43
	← ptl_ left	20:43
	→ SRabbelier joined	20:44
sako	thanks	20:45
	← saschpe left	20:46
yrlnry	sure	20:46
	→ panike joined	20:48
	← Guest59163 left	20:48
	→ workmad3 joined	20:48
	→ linusoleander joined	20:49
	→ LukeScott joined	20:49
	→ phantomcircuit joined	20:49
	← drl left	20:49
	→ Zenith77 joined	20:50
	→ ptl joined	20:50
	← p3rg2z left	20:50
	← jsick3 left	20:50
	← LukeScott left	20:51
	→ ShadowBelmolve joined	20:51
	→ namemon joined	20:52
	→ lukescott joined	20:52
	q0rban\|afk → q0rban	20:54
	← AkumaStreak left	20:55
	← lukescott left	20:56
	→ lukescott joined	20:58
lukescott	Does anyone know the difference between $ git submodule add URL local/path when "path" exists or doesn't exist? What if "path" is a symbolic link? Is that a "bad" thing? I've read a ton on submodules but I'm still a bit confused. I don't know if that is the "right" thing to do. I'm trying to avoid duplicated files. (I have several apps pointing to the same dependency)	20:59
	← PabloM left	21:00
	← tgunr left	21:00
	→ jelmer joined	21:00
	→ jason237 joined	21:00
	← cschneid left	21:00
	→ ki__ joined	21:00
	→ tgunr joined	21:00
	→ drl joined	21:02
	← IslandUsurper left	21:04
	→ robotarmy joined	21:04
	← ericbarnes left	21:04
	→ jutaro joined	21:05
	AAA_awright_ → AAA_awright	21:05
	→ apocalyptiq joined	21:05
	← bartek left	21:05
	← olle_ left	21:06
	← albel727 left	21:07
	→ Finalflash50 joined	21:07
mariusso	lukescott: I don't think that's wise. Each submodule is it's own repo, with its own state, since the top-level repo syncs the submodule to whatever the "submodule SHA1" points to for the current HEAD in the top-level repo. This means that each "identical" sub-module needs to be able to synced at various SHA1s at the same time.	21:07
	→ dnorman joined	21:07
mariusso	lukescott: With a symlink for each "identical" submodule to the same physical repo on disk, you loose that	21:08
mithridates	I'm trying to clone kernel with git, if I disconnect or kill the process and run the command again will it continue or restart downloading	21:08
	?	21:08
	← deekayen left	21:08
mariusso	it'll restart	21:08
	← comptekki left	21:09
	→ eletuchy joined	21:09
sgronblo	it will?	21:09
mithridates	mariusso: anyway to handle that issue?	21:09
mariusso	mithridates: The sever creates a unique pack file for your clone, which it needs to recreate the next time, (and which might be different) so the clone restarts	21:09
	→ jin28 joined	21:10
sgronblo	will one pack file be created for the whole repo?	21:10
mariusso	mithridates: If the repo is exposed through http, you can grab the whole repo (with pack files) using a downloader which can resume	21:10
	← Ryback_ left	21:10
mithridates	but kernel is huge and if I let it to keep downloading it in one time my room mates will rape me for internet speed!!!!!!	21:10
	→ paja joined	21:10
mariusso	sgronblo: it depends, normally it repacks to one large one containing all the objects you don't have	21:11
	→ helvete joined	21:11
mariusso	mithridates: Good, then you have something to do while it's downloading ;-)	21:11
sgronblo	wouldn't it make more sense to split it up a bit? so you can clone part of it and if it fails and you have to start over you would at least have gotten somewhere	21:11
	→ verto joined	21:11
lukescott	mariusso: I'm not entirely clear what that means. I basically need my application repo to see the dependency (framework) as readonly and manage that framework in it's own repo. The only reason why I'm using submodules is because at some point Beanstalk (the service I use) will support submodules for deployment. I just want to pull the latest version from the specified branch. Does that make sense? Is what I'm doing not suited	21:12
mithridates	mariusso: so after wget -c http://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git what should I do to make it on my local machine?	21:12
mariusso	sgronblo: clone resume is being discussed	21:12
sgronblo	mariusso: i see	21:12
	← jason237 left	21:12
mariusso	sgronblo: http://article.gmane.org/gmane.comp.version-control.git/168906/match=clone+resume	21:13
	ChanServ set mode: +v	21:13
mithridates	mariusso: any idea that how I can make it a local project after downloading it?	21:13
lukescott	mariusso: When I make the directory a softlink git seems to ignore the sub module completely, besides having a reference in .submodules. I think that might work for what I'm trying to do. Normally I would just use symlinks, but I need Beanstalk to pull the framework during deployment.	21:14
	← drubies\|mbp left	21:14
mariusso	mithridates: You can clone with the original URL, and use the local download as a reference. That way it'll pick all the objects you have locally first, and download the missing ones	21:14
	← paja left	21:15
mithridates	how can I clone with the original URL without downloading it and then refer it to that http manually downloaded pack	21:15
	→ ineiros joined	21:16
	← rayners left	21:18
jin28	test to see if I can talk. I need some basic help but I think I'm getting the terminology wrong	21:18
	I downloaded a repository that's read only, had some problems with it made fixes. I want to send a list of just the changes to the author whats the git command to create the changefile ?	21:18
workmad3	jin28: you want to generate a patch file then, I believe	21:18
jin28	yeah	21:18
mariusso	mithridates: You can also just clone locally, and afterwards point the origin.url to the remote repo, so it fetches there on the next 'git pull' or 'git fetch'	21:19
workmad3	jin28: iirc, you can do that with git diff	21:20
mithridates	mariusso: you mean I download this file for ex. http://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git then use this command: 'git pull linux-2.6.git' ?	21:20
	→ guardian joined	21:21
	← [intra]lanman left	21:21
workmad3	mithridates: you'd do 'git clone <repo-url>' which would create a local clone	21:21
	← panike left	21:21
mithridates	workmad3: well I did that but it started to download it , I don't want it because kernel is huge and I can't keep my connection for a long time	21:22
workmad3	mithridates: so you only want a shallow clone of it	21:23
	mithridates: you can do 'git clone <repo-url> --depth 5' to only get the latest 5 commits	21:24
mithridates	I don't know what you call it, I want to download it manually with wget, then make it as if what git clone does	21:24
jin28	<workmad3> thanks, that will work as a quick work around	21:24
mithridates	workmad3: but if I do that I can't search through old stuff	21:24
	← jutaro left	21:24
workmad3	mithridates: oh, you do want the full repo...	21:25
	mithridates: well, then you're looking at git clone then... and look for the git:// url for the repo, not the http one	21:25
mithridates	workmad3: I have it, it's here: git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git	21:25
	→ woozie joined	21:26
	→ gebbione joined	21:26
workmad3	mithridates: clone that one then... it'll be the most efficient you'll get	21:26
	← booginga left	21:26
	← TrineLocke left	21:27
	→ bdiu joined	21:27
	→ mzero joined	21:27
mithridates	workmad3: yes I know that, but the problem is that it's HUGE and I can't keep my connection, and git is not resumeable ! so I wanted to get it with wget -c and --limit-rate option.........	21:27
	← L42y left	21:27
	← v0n left	21:27
	← larsl left	21:27
lukescott	What would be the appropriate way to manage external dependencies that you want to be read-only? (As in you don't plan to make changes directly to that dependency, or submodule). I would like to make changes to the dependency as a dedicated repo/clone outside of the other repo. Reason being is I have several "applications" that need the "framework" in the private/library folder. These applications should use the latest branch	21:28
workmad3	mithridates: afaik, there isn't a way to generate such a file	21:28
mithridates	does git have --limit-rate option for downloading ?	21:28
mzero	I've got "git status" reporting that there is a change to be committed (staged), but "git diff --cached" gives no output at all!	21:28
	→ larsl joined	21:28
mzero	any idea what i should check?	21:29
ayust	git update-index --refresh	21:29
mzero	nope	21:30
	← fr0sty\|mobile left	21:31
	→ bauruine joined	21:31
	→ fr0sty\|mobile joined	21:32
mzero	oy - it was bad interaction with less (my pager)	21:32
	← RaceCondition left	21:33
	→ bousquet joined	21:35
	← woozie left	21:35
	← logik-bomb left	21:35
	← Zephyrus left	21:36
	← workmad3 left	21:37
	→ amcsi__ joined	21:37
	← robert_david left	21:37
lahwran	how do I list current remotes?	21:39
ayust	git remote -v	21:39
lahwran	ah, no `show`.	21:39
ayust	alternatively if you just want the remote names, you can leave off the -v	21:39
lahwran	...oh...	21:39
	I guess I don't have any remotes	21:39
	:P	21:39
	← amcsi_ left	21:40
	→ BiggFREE joined	21:40
	→ pantsman joined	21:42
	← pantsman left	21:42
	→ pantsman joined	21:42
mithridates	I didn't get any answer :( how can I download a huge project with bandwidth limit or -c feature like wget	21:42
	→ jackjennings joined	21:42
FauxFaux	mithridates: There's basically no support for anything like that; if you're cloning over http (or http is available) you can download the large pack-files by hand with whatever.	21:43
mithridates	FauxFaux: then what? git pull file ?	21:43
FauxFaux	No, just put them in .git and pray.	21:43
mithridates	tnx	21:44
bremner	what about man git-bundle?	21:44
jast	[automatic message] the 'git-bundle' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-bundle.html	21:44
bremner	then wget the bundle	21:44
	or heck, make a tarball/zip/file and wget that.	21:45
	→ robert_david joined	21:45
mariusso	mithridates: Maybe you could wget this http://www.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git/ and then when you're done, 'git clone <path to wget repo> linux && cd linux && git remote set-url origin git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git && git pull'?	21:45
	q0rban → ericduran	21:46
	ericduran → q0rban	21:46
mariusso	mithridates: The download URL can also be changed to something closer to you, like www.us.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git/ for norway etc..	21:46
	Eh, that was the US	21:47
	http://www.no.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git/ for Norway	21:47
	You get the drift	21:47
	← elementz left	21:47
mariusso	mithridates: In total, you'll be looking at downloading 904MB	21:48
	← sbell left	21:48
mithridates	mariusso: I gochya, tnx, oh is it totally 904MB?	21:48
	so it shouldn't be too much even with git clone	21:48
	← q0rban left	21:49
	← vanmik left	21:49
mariusso	mithridates: nah, the bottleneck is really the server..	21:50
	← wanders_ left	21:51
	→ wanders joined	21:52
	← mzero left	21:52
	→ kylehayes joined	21:53
	← jackjennings left	21:53
	mariusso just downloaded the whole thing himself	21:53
kylehayes	Can gc be run against bare repos too?	21:54
mariusso	yes, gc doesn't touch the working directory at all	21:54
	→ AdmiralSausage joined	21:54
mariusso	kylehayes: You often should run gc to compress a repos size on servers too, since they get 'fragmented' on push	21:55
lukescott	mariusso: Would you know the answer to my question?	21:55
kylehayes	mariusso: ok good to know	21:55
	thank you	21:55
	→ EricInBNE joined	21:56
mariusso	mithridates: Actually, doing a 'git gc --aggressive' on the linux kernel after downloading it, it's only 480MB in /.git	21:56
	.git/	21:56
	mithridates: So I think I gave you too high a number earlier	21:57
	← uau left	21:57
mithridates	mariusso: I'm trying to get those files but the http ones don't seem to be available	21:57
mariusso	lukescott: Sorry, didn't see your question	21:57
	→ hobodave joined	21:57
lukescott	mariusso: (hopefully this makes sense) What would be the appropriate way to manage external dependencies that you want to be read-only? (As in you don't plan to make changes directly to that dependency, or submodule). I would like to make changes to the dependency as a dedicated repo/clone outside of the other repo. Reason being is I have several "applications" that need the "framework" in the private/library folder. These ap	21:57
	← whaley left	21:57
	→ uau joined	21:58
	→ drizzd_ joined	21:58
mithridates	what does gc --agressive do? show the size?	21:59
mariusso	lukescott: With plain git that use-case is not plainly supported. However, there's a Python framework, which supports that, kinda like a substitute for git submodules	21:59
	mithridates: Garbage collects the repo, and compresses it into larger pack files	21:59
	← dv310p3r left	22:00
	BinaryMuse → BinaryMuse\|away	22:00
mariusso	lukescott: Trying to remember what it's called, but now it escapes me	22:01
jast	"compresses into larger files" – you don't read something like that every day ;)	22:01
	← kipras left	22:01
mariusso	jast: lol :-D	22:01
lukescott	mariusso: I could theoretically use relative symlinks though, right? As long as I check out both repos in the same place. My concern is being able to deploy the complete project using Beanstalk. They're going to be adding submodule support soon.	22:01
mithridates	I just couldn't understand it	22:01
	← drizzd left	22:01
lukescott	mariusso: So what I tried doing is adding the submodule on a symlink that already exists. Git seems to ignore the submodule and just add a reference into .submodules. But you said there was a draw back to doing that? I didn't fully understand.	22:02
mariusso	jast: Yeah, normally when you commit, you create many small (object) files in the .git/ repo parts. That's suboptimal for git, since a larger pack file with all the objects better compressed is much better for performance and storage space :)	22:02
	← ymasory left	22:02
	← airborn left	22:03
mariusso	lukescott: The state of the submodule is recorded in the top-level repo. So when you switch branches, or move to another commit, the submodule should be updated to whatever SHA1 for the submodule was recorded for that commit. That means that each repo which uses that submodule must support that this submodule can be a different SHA1s. You cant do that if they all share the same physical repo on disk	22:04
	← opalka left	22:04
mithridates	mariusso: it's what I get from git gc --agressive git://<link>.git	22:04
	fatal: Not a git repository (or any of the parent directories): .git	22:04
lukescott	mariusso: So what you're saying is an earlier version would link to the newer version of a dependency, not the version of the dependency that it was at the time. Wouldn't that also be the case with regular symlinks?	22:05
	← drl left	22:06
	← theoretical left	22:06
mithridates	mariusso: I just didn't understand what you told me! how can I get that in 480 MB ?	22:06
	← iAmerikan left	22:06
lukescott	mariusso: (I mean regular as in if you just used symlinks without submodule reference)	22:06
mariusso	mithridates: What I meant to say is that it probably was only ~480MB in the first place :)	22:06
mithridates	mariusso: that's good but now I should download 900 MB ?	22:07
	← vdv left	22:07
mariusso	lukescott: With the symlink, git will just record the symlink, nothing more. So a sub-module with only a symlink makes no sense	22:08
	mithridates: No, 900 was with both the repo, and the working directory you get checked out afterwards. It was my bad when I reported the size the first time	22:09
	BinaryMuse\|away → BinaryMuse	22:09
	← robert_david left	22:09
mithridates	mariusso: ok, to get everything clear, now I started cloning with git clone git:\\... how can I check the size? what's the size?	22:10
	→ ereslibre joined	22:10
	← ereslibre left	22:10
	→ ereslibre joined	22:10
	← mklappstuhl left	22:11
mariusso	mithridates: it won't report the total size, but it'll show you how much you have cloned upto now, and how many percent	22:11
mithridates	mariusso: it's funny because they could make a field for total size with a simple calculation	22:11
lukescott	mariusso: So with submodules the project is not actually storing the dependency. It's actually storing the SHA1 hash of that dependency (the commit). And if I use a symlink in-place of that dependency then I'm not recording the SHA1 hash I'm recording a symlink. Is that right? So basically the snapshot of my app has no idea what commit is being used, which is why it's not a good idea.	22:12
mariusso	mithridates: Worst case, point wget to marius.storm-olsen.com/sources/linux-2.6.tar ;-)	22:13
	→ patmaddox joined	22:13
	mariusso has to head off..	22:13
	← jin28 left	22:13
mithridates	mariusso: thanks man for your help	22:13
sako	hey guys so i am using git to version control my vim configs and also using pathogen...	22:14
	vim/bundle/snipmate is a git submodule... why does it keep telling me the above message?	22:14
	I get this when i do git status modified: vim/bundle/snipmate (untracked content)	22:14
	← jaeckel left	22:14
	→ GodEater_ joined	22:15
	→ Weiss joined	22:15
	→ mklappstuhl joined	22:15
	← ShadowBelmolve left	22:16
	← yunosh left	22:16
	→ tpope joined	22:16
	→ jaeckel joined	22:16
	← GodEater left	22:18
	→ mattalexx joined	22:18
	← apocalyptiq left	22:18
	← bkozal left	22:19
	→ agile joined	22:19
	→ drl joined	22:19
	→ woozie joined	22:20
	← sebersole left	22:20
	← d0k left	22:20
	← dermoth\|work left	22:20
	← verto left	22:21
	← AdmiralSausage left	22:22
	← dnorman left	22:23
	→ [intra]lanman joined	22:23
	← tmz left	22:23
	← ThomasLocke left	22:24
	← ptl left	22:24
	← drl left	22:24
	→ ptl joined	22:24
	← beatak left	22:25
	→ whyking joined	22:26
	← danlentz_ left	22:29
	→ ymasory joined	22:29
eydaimon	what does it mean when I have two branches with the same name, like here? https://gist.github.com/67d225c8ca4057377507	22:29
mithridates	why git doesn't show the total size?	22:30
gebbione	hi	22:31
	anybody in the house?	22:31
	← alester left	22:32
gebbione	yo yo	22:32
jast	gebbione: the other one has a different name, it's literally called "mordor (2)"	22:33
	← hobodave left	22:33
	→ uroboros joined	22:33
jast	though actually current versions should make that invalid	22:33
	what does this say: git show-ref	22:34
gebbione	jast ... i m just starting with GIT not sure that answer was for me	22:35
mithridates	is it dummy to have a total size in cloning (progress.c) ? or is it a dummy idea or there is a reason that there is no total size?	22:35
gebbione	I am looking for some answer to my lame questions	22:35
	→ AdmiralSausage joined	22:35
	← guardian left	22:35
gebbione	i keep all my projects in a specific folder and i would like to use a separate folder as my GIT repository	22:36
jast	sorry, I'm not capable of coherent thought right now :) I need sleep	22:36
	← nb_ left	22:36
gebbione	for instance myGitRepo	22:37
	→ drl joined	22:37
	→ danlentz joined	22:37
	← danlentz left	22:37
	→ danlentz joined	22:37
	← danlentz left	22:37
	← mortenmj left	22:37
gebbione	can i add each folders of my projects to this single repo or do i need to create a separate repo for each folder/project	22:37
	→ danlentz joined	22:38
	← danlentz left	22:38
jast	we recommend creating a separate repository for each independent project	22:38
	→ danlentz joined	22:38
	← tgunr left	22:38
	→ tgunr joined	22:38
	→ ISF joined	22:38
jast	usually this is not a big bother because it's easy to create 'em and they don't really have a lot of overhead	22:38
	← amerine left	22:39
	← eletuchy left	22:39
	← tgunr left	22:40
	→ tgunr joined	22:40
	← woozie left	22:41
	← codebeaker left	22:41
	← mastroDani left	22:41
	→ nb_ joined	22:42
	← chrisdev left	22:42
	← drl left	22:43
	→ chrisdev joined	22:44
	→ tgunr__ joined	22:44
eydaimon	jast: I've not created anything literally called "mordor (2)"	22:44
	← tgunr left	22:44
eydaimon	jast: one sec on show-ref	22:44
	jast: https://gist.github.com/5298b46b5d93e3aa2e18	22:45
lukescott	What is the "correct way" to update submodules within a super project?	22:45
	As in pull new versions of each submodule	22:45
jast	eydaimon: well, something created a branch with that name, though with recent versions it's pretty difficult to create that kind of name	22:45
	← reds left	22:45
eydaimon	jast: checkig it out gives me a detached state	22:45
jast	lukescott: if someone has properly committed updates to the submodules, git submodule update --recursive	22:46
eydaimon	jast: doesn't that suggest that it put me on the remote branch?	22:46
jast	after pulling the changes in the supermodule that contain the submodule changes	22:46
	eydaimon: since git doesn't currently consider it a valid branch name, who knows what it does with that	22:46
	the 'git branch' tool, that is	22:46
	→ angelsl joined	22:47
	← Sonderblade left	22:47
eydaimon	jast: the history makes no mention of the creation of such a branch	22:47
jast	branch creation is not versioned	22:47
	→ soidexe joined	22:48
gebbione	can directories be added to a GIT repo?	22:48
	→ cakehero joined	22:48
jast	at least not by git	22:48
lukescott	jast: That isn't working. I need to pull the newer versions. Right now the submodules are stuck on an older commit.	22:48
jast	gebbione: yes, but empty ones are ignored	22:48
soidexe	hey there! how do I remove a file from list to be committed?	22:48
eydaimon	well, I deleted it	22:48
	← alex__c2022 left	22:48
gebbione	i just tried to add something	22:48
jast	lukescott: probably the submodule references weren't updated to a newer version (update submodule, use "git add path/to/submodule" in main repo, commit, push)	22:49
gebbione	git add .././myfolder/*	22:49
	← chrisdev left	22:49
eydaimon	jast: must have been a git joke, last commit was on April 1st :P	22:49
gebbione	but it gave me the error ../../myfolder/subfolder is outside repository	22:49
jast	soidexe: git status mentions virtually all things you can do to change what's currently planned for a file	22:49
soidexe	please help! I added wrong file to be commited. will "git reset" reset the changes from that file?	22:49
lukescott	jast: The super project isn't seeing that it's new. Do I need to go into each submodule and do "git pull origin master" first?	22:50
jast	lukescott: well, yes	22:50
	→ workmad3 joined	22:50
jast	otherwise how are you going to tell the superproject which different version to record :)	22:50
soidexe	jast, git status shows me files that I don't want to commit so how do I remove them from it?	22:50
lukescott	jast: First time using Git. Wasn't sure how this was supposed to work.	22:50
jast	soidexe: 'git status' shows several sections. at the top of each section, it gives you commands to change things. if it says "use git reset to undo", do that, etc. :)	22:51
lukescott	jast: So I can do git submodule foreach git pull origin master, then git submodule update —recursive ?	22:51
	→ alex__c2022 joined	22:51
jast	lukescott: no, then the second part will just undo the first	22:51
	lukescott: a superproject internally remembers which version of each submodule to use, but only when you tell it (which you do by 'git add'ing the submodule after you have checked out the right version of it, then committing)	22:52
soidexe	jast, okay, seems like working	22:52
	← AdmiralSausage left	22:52
	← c00kiemon5ter left	22:53
	→ c00kiemon5ter joined	22:53
	← c00kiemon5ter left	22:53
	→ c00kiemon5ter joined	22:53
lukescott	jast: So after "git submodule foreach git pull origin master" do "git add ." ? Where does "git submodule update" fall in the equation? I think I get it. I think I only do git submodule update if I need to pull down what someone else did.	22:54
jast	lukescott: yep. someone else updates the records for the submodules -> you do git submodule update	22:54
	you change submodules yourself -> you don't need git submodule update	22:54
lahwran	how do I revert a file to the latest commited state?	22:55
jast	lahwran: git checkout HEAD -- file	22:55
	← aces1up left	22:55
lukescott	jast: Ah I get it now. Do you know of any good clients (on Mac) that work with submodules, or am I just stuck on the command line?	22:55
lahwran	git checkout HEAD -- file1 file2 #this valid?	22:55
jast	lahwran: sure	22:55
	lukescott: a) I don't have OSX, b) I haven't really ever felt the need to look for any	22:55
	→ drl joined	22:55
	→ fr0sty joined	22:55
lahwran	I wouldn't say you're "stuck" on the commandline ...	22:56
gebbione	git add .././myfolder/* gave me the error ../../myfolder/subfolder is outside repository ... but the folder is not empty	22:56
	is this something not allowed?	22:56
	← Amorphous left	22:56
jast	gebbione: when you do 'git init', you can only add stuff to that repository that's in the same directory where you did the init, or subdirectories	22:56
	← mari00sh left	22:56
jast	otherwise you'd probably end up adding your whole hard drive to your project repository ;)	22:57
gebbione	right so i cannot keep all my repos in a separate folder	22:57
	lets say repos	22:57
lahwran	how do I show what revision and such I'm on?	22:57
gebbione	and then in repos/subfolders put the actual repositories	22:58
jast	if your projects are in different folders, you just create a repository in each one	22:58
	→ eletuchy joined	22:58
jast	lahwran: git log shows the current one at the top, for instance	22:58
lahwran	that works	22:58
jast	there are dozens of ways to get at that kind of info	22:58
	git log answers many questions simultaneously and is easy to type ;)	22:58
	← DarthGandalf left	22:59
jast	and now, good night. I'm pretty darn tired.	22:59
gebbione	jast: thanks for the confirmation, i would have liked to have a separate folder for the repositories but i guess thats not how this has been designed like	23:00
	→ DarthGandalf joined	23:03
	→ maguey_ joined	23:06
	← ISF left	23:07
fr0sty\|mobile	gebbione: why would you want the .git directory somewhere else?	23:07
gebbione	to have a central folder with all the sub repos	23:08
	separate from the working files	23:08
fr0sty\|mobile	Sub-repos?	23:08
gebbione	myfolder/repo1 ... myfolder/repo2	23:08
fr0sty\|mobile	gebbione: you can always create a symlink...	23:09
	→ ShooterMG joined	23:09
fr0sty\|mobile	But I'm probably missing your point...	23:09
	← lukescott left	23:10
gebbione	i guess your might be a solution but i still dont want to be restricted to be able to add files from my current folder only	23:11
	← LouisJB left	23:11
	→ reds_ joined	23:11
	→ amerine joined	23:11
	← saleh left	23:11
gebbione	for instance at the moment i am creating the repo under the folder of the project as i could not add a folder outside	23:11
	when i upload this repo to a remote server how will it work	23:12
	→ albel727 joined	23:12
gebbione	will it copy the full folder?	23:12
fr0sty\|mobile	You have nested repositories?	23:12
gebbione	no	23:12
	i have my eclipse workspace/project1	23:12
	so at the moment i have to create repos under each project	23:13
	and then i will want to upload the repositories to my remote server	23:13
fr0sty\|mobile	As it should be	23:13
	One repo per project...	23:13
	→ Amorphous joined	23:13
fr0sty\|mobile	Git push...	23:13
	→ AdmiralSausage joined	23:14
gebbione	ok so on the server i will get only the files under .git	23:14
	on a folder i decide to push them too?	23:14
	just asking because i dont want to overwrite files or create problems on the remote server	23:15
	→ LouisJB joined	23:15
	← bousquet left	23:15
	→ bousquet joined	23:16
	← cakehero left	23:16
fr0sty\|mobile	You can push wherever you want, but faq non-bare	23:17
jast	Pushing to non-bare repositories is discouraged: please see https://git.wiki.kernel.org/index.php/GitFaq#non-bare [automatic message]	23:17
	→ nimrod10` joined	23:18
	← kermit left	23:19
	← nimrod10 left	23:20
	→ Nikoli_ joined	23:22
	← Nikoli left	23:22
	→ _psychic__ joined	23:22
	← gigamonkey left	23:23
	← maguey_ left	23:24
	← tewecske left	23:24
	← namemon left	23:25
	← _psychic_ left	23:25
	← mattalexx left	23:26
	Tabstar → Tabmow	23:26
	← eijk left	23:26
	→ dlam joined	23:26
	← bitkiller left	23:27
	→ sigh joined	23:27
	← _psychic__ left	23:27
dlam	mmm, is there a way to do 'git diff b3358a11 <revision before b3358a11> some.file'?	23:27
	trying to find an equivalent to 'svn diff -c 55690'	23:28
fr0sty\|mobile	Maan gitrevisions	23:29
	← __name__ left	23:29
dlam	wait i found it... git diff b3358a11^..b3358a11	23:29
	← ExtraSpice left	23:30
cbreak	look at git shnow	23:30
	show	23:30
fr0sty\|mobile	Git show hash would give you the diff and the commit information	23:30
	→ phantomcircuit_ joined	23:30
	← Heimidal left	23:30
	adamholt → adamholt_away	23:31
	adamholt_away → adamholt	23:31
	→ Wes joined	23:31
dlam	oooo git show seems easier, thanks!	23:31
Wes	I'd like to pull a single change from a remote repository, without any of it's descendents. Is this possible, or should I just use diff/patch?	23:32
	→ pyroscope joined	23:32
	← c00kiemon5ter left	23:32
lahwran	what's that thing called that shows current git repo on ps1?	23:32
	→ tmz joined	23:32
	← phantomcircuit left	23:34
	← icwiener left	23:34
Wes	git cherry-pick changeset	23:34
	oops	23:34
	← LiamH left	23:35
	→ c00kiemon5ter joined	23:36
	← c00kiemon5ter left	23:36
	→ c00kiemon5ter joined	23:36
	← alip left	23:40
	← drl left	23:40
	← ereslibre left	23:41
	→ alip joined	23:41
	← rettub left	23:41
	→ ISF joined	23:42
	← BinaryMuse left	23:42
	← nyingen left	23:43
	← Chillance left	23:44
	→ cenuij joined	23:44
	← cenuij left	23:44
	→ cenuij joined	23:44
	→ verto joined	23:45
fr0sty\|mobile	lahwran: git bash-completions	23:45
	→ rettub joined	23:46
	← slonopotamus left	23:46
	→ Titosemi joined	23:49
	→ seivan joined	23:50
	→ drl joined	23:52
	→ robotarm_ joined	23:52
	← robotarmy left	23:55
	← uroboros left	23:56
	→ rchavik joined	23:58
	← mithridates left	23:58
	→ coldboot\|home joined	23:59

Logs Search ←Prev date Next date→

Channels Documentation