IRCloggy #git 2013-11-18

Logs Search ←Prev date Next date→ Channels Documentation

Provider of IRC logs since 2005.
WARNING: As Freenode became unjoinable and lost all warnings in topics, we cannot log channels on Freenode anymore.

2013-11-18

→ oldtopman joined00:00
← teran left00:00
→ Goplat joined00:01
oldtopman I just changed the port for ssh on my gitolite server - how do I use git and specify a port number?00:01
→ cagedwisdom joined00:02
← sam113101 left00:02
bremner oldtopman: you can specify it in the client ssh configuration00:02
or, see man git-push URL00:03
gitinfo the git-push manpage is available at http://jk.gs/git-push.html00:03
→ carpii joined00:03
oldtopman bremner: git remote set-url origin is of no use then?00:03
→ GT4066 joined00:03
bremner sure, that should work too.00:03
→ undecim joined00:03
oldtopman I have to use that anyway, since the computer shuffles between local network and outside.00:04
carpii i want to step through some commits in series, to test each stage to see where a bug was introduced. So I git checkout HEAD~20 and test. Is there now an easy way to advance to the next commit ?00:04
← mdiers left00:04
bremner carpii: is there some reason you are not using git bisect?00:06
→ sam113101 joined00:06
carpii yes00:06
bremner carpii: maybe use git-rev-list in advance?00:07
carpii that looks useful, thankyou00:07
← hiroyuki left00:08
bremner I did something similar with rebase: http://paste.debian.net/66345/00:08
← threesome left00:08
oldtopman bremner: How does git://host.xz[:port]/path/to/repo.git/ work when I specify the url as git://host.xz:repo.git00:08
bremner oldtopman: the latter is not correct syntax00:09
← Kayra left00:09
bremner and anyway git:// has nothing to do with ssh00:09
→ frojnd joined00:09
frojnd Hi.00:09
← mmcdaris left00:11
frojnd I try to push changes to branch and I get: ![rejected] mybranch -> mybranch (non-fast-forward) and then error: failed to push some refs... updates were rejected because the tip of your current branch is behind its remote counterpart. Integrate the remote changes e.g. git pull before pushing again00:11
I tried git pull and it says up-to date00:11
← raonyguimaraes left00:12
carpii what does git status say ?00:12
← snuffeluffegus left00:12
← freestyl3r left00:12
→ freestyl3r joined00:13
→ mfeltner joined00:13
frojnd carpii: Your branch is ahead of 'origin/master' by 1 commit.00:13
carpii: and: use "git push" to publsh your local commits00:13
← maestrojed left00:14
← freestyl3r left00:14
← m8 left00:14
EugeneKay That's for the master branch00:14
→ freestyl3r joined00:14
EugeneKay `git checkout mybranch; git status`00:14
→ hahuang65 joined00:15
EugeneKay And the error message is pretty self-explanatory ;-)00:15
← freestyl3r left00:15
frojnd EugeneKay: Your branch is ahead of 'origin/master' by 1 commit.00:15
and (use "git push" to publish your local commits)00:15
EugeneKay Well either mybranch isn't the name of your branch, or you're doing something wrong00:15
→ freestyl3r joined00:16
EugeneKay does not troubleshoot theoretical issues00:16
frojnd EugeneKay: I did as you asked except I changed mybranch with actual name of my branch00:17
EugeneKay !secret !repro00:17
gitinfo [!topsecret] If your situation is so secret that you can't tell us how to !reproduce it, there's really not much we can do. Perhaps you need to find or buy support you can trust? There are people with clearances even on #git but they may not want to identify themselves. However, you are protected by the biggest secret of all: We just don't care.00:17
Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session -- or, even better for complex issues, design a minimal case in which your problem can be reproduced, and share it with us. This will help immensely with troubleshooting.00:17
← mfulz left00:18
← petey__ left00:18
frojnd EugeneKay: http://sprunge.us/IcGf00:19
oldtopman bremner: Thanks for the help - managed to figure it out with the link you gave me.00:20
→ john2x joined00:20
EugeneKay And that still doesn't show you what the output of `git status` is. That's why we look for /transcripts/ instead of the output of a single command.00:20
← tombee left00:20
← Spyro left00:21
frojnd EugeneKay: that's the git status output. Did I missed something?00:21
EugeneKay s/doesn't/only/00:21
The failed `git push`, your `git checkout`, etc00:21
frojnd eh ok00:21
← starsinmypockets left00:22
aartist How I can find out when the file was first added into repository?00:23
frojnd EugeneKay: git push http://sprunge.us/IMai00:24
EugeneKay: and git checkcout just gives me: Your branch is ahead of 'origin/master' by 1 commit.00:26
→ Davey joined00:26
← wooy left00:26
frojnd EugeneKay: and (use "git push" to publish your local commits)00:26
← Balliad left00:26
→ wooy joined00:27
→ rjhunter joined00:27
← BSaboia_ left00:27
→ johnkevinmbasco joined00:28
frojnd EugeneKay: any ideas?00:28
→ oojacoboo joined00:28
← eletuchy left00:28
← Vivekananda left00:29
→ eletuchy joined00:29
← echo-area left00:29
frojnd can I undo last commit on my branch? and then git pull, do changes and then commit and push?00:30
→ macabre joined00:30
frojnd EugeneKay: do you need anymore info?00:32
carpii you seem to be trying to push a local branch to origin master00:33
→ BSaboia_ joined00:33
carpii just mere your branch into local master and then push master00:33
merge00:33
← hyperair left00:34
frojnd carpii: on local I have two branches. If I do git branch I can see I'm in mybranch currently00:34
anyways, how do I branch into local master?00:35
← SocalNick left00:35
thiago what do you mean by "branch into" ?00:35
carpii how do you switch to local master branch?00:35
git checkout master00:35
frojnd wait why would I merge into local master? I need mybranch :)00:35
carpii oh i think you really need to just read a good git guide00:35
no disrespect but i dont think you understand how to use git even on a basic level :/00:36
→ dyng joined00:36
← freestyl3r left00:36
→ freestyl3r joined00:37
aartist I have a list of file and I like to know when they were first added into git repo. How I can find that out?00:37
→ Kayra joined00:37
frojnd carpii: no problem. I thought If I have second branch I can do changes in this second branch and then pushing changes for this second branch.00:38
← freestyl3r left00:38
carpii frojnd: you typically create a branch, work on that feature, then when you want to push to remote, you merge it into master and push that00:38
alternativwly, if you want to push to a remote branch so another user can work on it00:38
you can git push origin remote_branch_name00:39
rjhunter aartist: you could start with `git log path/to/myfilename` and look at the earliest commit00:39
← oldtopman left00:39
← endiruna left00:39
rjhunter aartist: `git log --folow` if you're interested following renames00:39
frojnd carpii: yeah I tried to use this second feature. git push origin remote_branch_name (which in my case is the same as my local) and I get these errors00:40
carpii aartist: this might help too -- git log --pretty=format:'%Cred%h%Creset -%C(yellow)%d%Creset %s %Cgreen(%cr) %C(cyan)<%an>%Creset' --abbrev-commit public/index.html | tail -n100:41
just ensures its on one line00:41
tho the colors are unnecessary, i just copied it from an alias :)00:42
rjhunter aartist: git log also has `--reverse --date-order` options00:43
carpii nice00:44
← nnyk_ left00:44
← ulkesh left00:44
→ nnyk_ joined00:45
→ macrover joined00:45
← ndngvr left00:45
→ ndngvr joined00:45
→ HumanSky joined00:47
← MattMaker left00:47
→ ewnd9 joined00:48
→ juvenal joined00:49
← ntio left00:49
← Vile left00:49
← BSaboia_ left00:49
frojnd carpii: don't know why but merging to master works00:49
but when trying to push to other branch it wont work00:50
gonna have to read few things00:50
carpii thats a great idea :)00:50
→ juls joined00:50
→ Evelynn joined00:51
→ mribeirodantas joined00:52
← ewnd9 left00:54
→ ewnd9 joined00:54
← Evelynn left00:55
→ raonyguimaraes joined00:56
→ rchavik joined00:57
← Kayra left00:57
→ Kayra joined00:58
← harish_ left01:00
→ vipzrx joined01:00
← ewnd9 left01:02
← nwertman left01:02
→ Spyro joined01:02
→ ewnd9 joined01:02
← Ange_blond left01:04
→ kukks joined01:05
→ JEEB joined01:06
→ Balliad joined01:06
← HumanSky left01:08
← ewnd9 left01:09
→ ewnd9 joined01:09
← rjhunter left01:09
→ hyperair joined01:10
← Davey left01:11
← PeterBooker left01:11
→ lvella joined01:11
← shruggar left01:11
→ freestyl3r joined01:13
← penguin42 left01:13
← Marbug left01:15
→ boombatower joined01:18
→ weiyang joined01:18
← musicmatze left01:22
← lvella left01:22
→ PDX-rmjack joined01:23
← pfFredd left01:24
← mizu_no_oto left01:25
← nnyk_ left01:26
→ musicmatze joined01:28
→ echo-area joined01:30
→ daito joined01:31
← carpii left01:33
← huli_ left01:34
← ewnd9 left01:35
→ SiliconDon joined01:35
← freestyl3r left01:36
← daito left01:36
← King_Hual left01:36
→ freestyl3r joined01:37
← SiliconDon left01:39
← rchavik left01:39
← BiggFREE left01:42
← rafaqueque left01:42
→ luc_ joined01:42
← bgerber left01:42
→ bgerber joined01:43
→ dhruvasagar joined01:44
← mribeirodantas left01:44
← kazi left01:46
→ bambams joined01:46
← bambams left01:46
→ bambams joined01:46
→ harish joined01:47
← DonRichie left01:48
→ DonRichie joined01:49
→ Chwing joined01:49
← causasui left01:51
← griotspeak left01:52
← pedda left01:52
→ lavaman joined01:53
← Guest__ left01:53
juvenal → juvenal_away01:54
→ hwrd|work joined01:59
→ bwreilly joined01:59
← asQuirreL left02:00
→ Marbug joined02:00
→ saorge joined02:02
→ kazi joined02:03
← bgerber left02:05
→ bgerber joined02:05
← bambams left02:05
→ bambams joined02:05
← saorge_ left02:06
→ daito joined02:06
→ sagittarian joined02:06
→ tmcmahon joined02:07
← Balliad left02:07
← lavaman left02:07
← harish left02:08
← Kayra left02:08
→ pure joined02:09
pure Would there be an easy way to auto increment the @version tag in javadoc when a file is comitted?02:09
bremner pure: see "$Id$" in man gitattributes02:11
gitinfo pure: the gitattributes manpage is available at http://jk.gs/gitattributes.html02:11
EugeneKay The short answer is man git-describe02:11
gitinfo the git-describe manpage is available at http://jk.gs/git-describe.html02:11
bremner well, yeah, that would be better.02:11
← whitby left02:11
→ MattMaker joined02:13
pure So, would $Id$ be the only way to do it automatically?02:15
← jday_ left02:15
pure Oh, or I could just use a post commit hook and do some stuff.02:16
→ causasui joined02:16
← causasui left02:16
→ causasui joined02:18
← RaceCondition left02:18
← mitz_ left02:20
→ mitz_ joined02:22
→ chuckharmston joined02:24
← MattMaker left02:26
← GT4066 left02:26
→ kcj joined02:29
← chuckharmston left02:29
→ chuckharmston joined02:30
→ dsdeiz joined02:31
→ Vampire0 joined02:31
→ maslowbeer_ joined02:31
← abelsiqueira left02:31
← Vampire0_ left02:32
→ rjhunter joined02:32
← ozette left02:33
→ chuckhar_ joined02:33
→ Hexorg_ joined02:34
→ griotspeak joined02:34
← chuckharmston left02:35
← Watabou left02:35
→ Watabou joined02:35
Hexorg_ Hey guys... I'm *really* new to git, and version control... Feel bad - very useful (thanks devs!), I should have learnt it earlier. Anyway, can I use git as an archive for my code? I usually write code at home and at work computers, so I want to transfer all the files to work, and then want to be able to transfer stuff from work back.02:36
offby1 yep02:38
it's quite good at that02:38
granted, if that's _all_ you want, Dropbox or rsync might be even better02:38
but chances are, you're gonna hack at home, then go to work and forget to sync, and start hacking; soon enough, you'll have created (without even intending to) "diverging branches", and git makes it super-easy to "merge" those back together.02:39
→ harish joined02:39
← Naoe-Kanno left02:39
offby1 Hexorg_: so, use it. Also sign up for github, and use it too; it's super-slick02:39
← thelorax123 left02:39
offby1 is not paid to endorse github, nor the Hair Club for Men02:39
← tmcmahon left02:39
Hexorg_ offby1, yeah! I started a project in eclipse and I've been commiting changes. Is there a special way I should transfer now? Or can I just copy the whole project folder?02:40
← bambams left02:40
← PDX-rmjack left02:41
→ bambams joined02:41
← bambams left02:41
→ bambams joined02:41
offby1 the way to "transfer" is to get _some_ git repository somewhere that's accessible via the Internet (like I say, github is a good choice). Then you "add" that repository as a "remote" to your eclipse repository. Then you "push".02:41
Then you can "clone" that repository onto other machines.02:41
→ thelorax123 joined02:41
← griotspeak left02:41
Hexorg_ I see. Awesome! Thanks!02:41
offby1 for what it's worth, I found git to be quite confusing and prickly even after _months_ of using it. But I ultimately got to the point where I wouldn't give it up for anything02:41
granted this was back in the Stone Age, when you had to create commits with pointed sticks and chipped-stone tools. But still.02:42
→ parcs joined02:44
Hexorg_ I see gitHub wants a paid membership for private repos. I don't think I'm allowed to post this code for public. Is there a way to "push" into file?02:45
offby1 stares blankly02:45
offby1 Hexorg_: I think bitbucket offers private repositories for free.02:45
I have no idea what '"push" into file' means.02:46
Hexorg_ offby1, to create something like a tarball that will have all of my files AND commits02:46
← kpreid left02:46
→ w0ng joined02:47
offby1 Hexorg_: you could do that by just tarring up your enitre repository. But why would you want to do that?02:47
→ kpreid joined02:47
offby1 *entire02:47
deryni I believe git-bundle may do what you want, sort of.02:47
offby1 let's figure out what he/she wants first02:47
Hexorg_ Well I'm transfering the code over a flash drive. So I just wanted to have a tarball02:48
→ Gu_______ joined02:48
offby1 well, if you're using git, it's quite unusual to transfer the entire tree.02:48
typically you just transfer the minimum "delta" needed to bring one repository in sync with another.02:49
"git push" and "git pull" do that for you.02:49
so you could put a repository on the flash drive, and push to it, or pull from it.02:49
so, you'd have (at least) three repositories: one on each machine on which you hack, and another on the flash drive.02:50
Hexorg_ offby1, oh ok, so I don't have to use an online resource to push/pull? It looked like I HAD to use an online repo02:50
offby1 You'd occasionally connect the flash drive to this machine or that, bringing each in sync with the other.02:50
deryni Putting the "online repo" on the flash drive would work.02:50
offby1 you don't _have_ to use an online resource; a flash drive would be fine. It's just that "online resources" (I call them "hosted git repositories") are more common02:50
and arguably more convenient, but Your Mileage May Vary™.02:51
deryni And no, the only thing that makes "online" a requirement is the "getting to it from both home and work" part.02:51
offby1 nod02:51
although I hear they make flash drives small enough to fit in the passenger seat of a car these days :)02:51
a reasonably healthy fellow can lift carry it into and out of the car himself!02:52
← yogin left02:52
Hexorg_ Yeah, I'm a university student, and by "work" I ment my lab haha02:52
That's also why I don't want to use a paid account - I don't have much money.02:52
offby1 I'd be afraid I'd lose the flash drive.02:52
well, github is only like $7/month for a few private repositories. I think that's all I'm paying.02:53
I bet your tuition is a tad more than that.02:53
Hexorg_ I have an ssh access to my machine in the lab, I think I'll keep the main tree there.02:53
deryni Does your lab have internet accessible services? They could host the repo themselves.02:53
offby1 oh sure; if you can get to the box via ssh, then it'll do just fine.02:53
Hexorg_ Yeah :D thanks guys02:53
offby1 You can have two repositories on the lab machine: one sort of common one, and one that you hack in.02:53
← Gu_______ left02:53
offby1 each hacking repository will exchange data with that common one.02:53
(You should really make the common repository be a "bare" repository, by the way.)02:54
Hexorg_ Hmm. ok02:55
← Hexorg_ left02:57
offby1 given that you can "talk to" the lab machine from home via ssh, you might be tempted to have just _two_ repositories: one on each machine. That can be made to work, but I advise against it; having _three_ repositories -- one central "bare" one and two for hackin' -- is conceptually easier.02:59
← Pyrogerg left02:59
offby1 nods gravely02:59
→ Skrypter joined02:59
← CheckDavid left03:00
← Skrypter left03:00
← Bombe left03:01
← macrover left03:01
→ conan_ joined03:01
→ Pyrogerg joined03:01
→ silviof1 joined03:01
→ blahdodo joined03:01
← conan left03:02
→ Bombe joined03:04
→ jmg joined03:04
← jmg left03:04
← silviof left03:04
← chuckhar_ left03:04
← CKyle left03:04
← TGiFallen_ left03:05
← Sonderblade left03:05
→ abelsiqueira joined03:07
→ ThinkRusty87 joined03:09
← ThinkRusty87 left03:09
→ madewokherd joined03:09
← Bombe left03:10
→ leeN joined03:11
← hwrd|work left03:13
zacts can I easily convert an rcs repo over to git, while retaining rcs history?03:13
← raonyguimaraes left03:14
juvenal_away → juvenal03:16
→ rchavik joined03:17
canoon i didn't know people used rcs03:17
zacts, google seems to work http://stackoverflow.com/questions/1619737/migrate-project-from-rcs-to-git03:18
← shinnya left03:19
← kukks left03:19
deryni If you can fake it as a cvs repo there are tools you can use. I don't if any of them do rcs by itself though since I didn't think rcs really had a concept of "repos".03:19
→ Bombe joined03:20
→ PDX-rmjack joined03:21
← dyng left03:21
→ sarkis joined03:22
offby1 people still use rcs?03:24
← oojacoboo left03:24
echo-area Hmm. I thought tailor might be able to do such things, but tailor's homepage had already gone down03:24
→ isleshocky77 joined03:26
→ jmaya joined03:27
isleshocky77 I've been using git-svn for the last year or so. I'm getting ready to switch to pure git; however, I'm not ready yet. As part of this preparation I wanted to clean up the author names, emails. I did this through git filter-branch. That seems to work fine. I then ran "git update-ref -d refs/original/refs/heads/master" afterwards, cause I thought that was to clean up the back up made after the branch-filter. However, now when I try to ru03:29
deryni Cut off after 'However,'03:29
isleshocky77 deryni: haha. Great. The meat and potatoes of it.03:30
'm not ready yet. As part of this preparation I wanted to clean up the author names, emails. I did this through git filter-branch. That seems to work fine. I then ran "git update-ref -d refs/original/refs/heads/master" afterwards, cause I thought that was to clean up the back up made after the branch-filter. However, now when I try to run git svn rebase or git svn dcommit, I'm getting: "BUnable to determine upstream SVN information fro03:30
← jmaya left03:30
rjhunter isleshocky77: try pasting one sentence at a time :-)03:30
deryni Cut off again. IRC message length limits. Just paste from 'However'.03:30
isleshocky77 however, I'm not ready yet. As part of this preparation I wanted to clean up the author names, emails. I did this through git filter-branch. That seems to work fine.03:31
← Vinnie_win left03:31
isleshocky77 I then ran "git update-ref -d refs/original/refs/heads/master" afterwards, cause I thought that was to clean up the back up made after the branch-filter.03:31
However, now when I try to run git svn rebase or git svn dcommit, I'm getting: "BUnable to determine upstream SVN information from HEAD history. Perhaps the repository is empty. at /usr/lib/git-core/git-svn line 780.". Any thoughts on fixing this?03:31
→ beautifulmind joined03:32
isleshocky77 rjhunter: Thanks.03:33
→ Vinnie_win joined03:33
rjhunter isleshocky77: does the error message literally start with "BUnable"?03:33
← thelorax123 left03:33
isleshocky77 From what i've read, it looks like "git update-ref refs/remotes/git-svn refs/remotes/origin/master" May help me, but honestly, I have no idea what that command is doing - or which refs to have it use.03:33
No. The "B" was a mistake03:34
rjhunter isleshocky77: if it were really there i'd have suggested re-installing git :-)03:34
→ MattMaker joined03:34
← dhruvasagar left03:35
deryni Does the current refs/remotes/git-svn ref point to what refs/original/refs/heads/master pointed to before you deleted it?03:35
isleshocky77 rjhunter: Got it. yeah. I think I had a key board key hit while I hadn't expected the original command to take so long. It takes a bit of time to figure out it can't do it.03:35
→ thelorax123 joined03:35
isleshocky77 deryni: I'm not sure.. .So that's the thing, I'm not sure if I have the same structure as the person saying to use that command. The first thing I know is that I have it under refs/remotes/svn, not "git-svn". But I don't see a refs/remotes/origin/master. I'm guessing I don't have exactly the same scenario as the person saying to use that.03:36
← thelorax123 left03:36
isleshocky77 But basically I need to tell git-svn what git commit matches up to the svn remote commit.03:36
juvenal → juvenal_away03:36
isleshocky77 I've made one commit which isn't on the remote svn03:36
→ oojacoboo joined03:37
deryni Did update-ref spit out the sha of the ref it deleted?03:37
isleshocky77 deryni: I no longer have the output of it. I didn't know it had messed anything up until an hour later when I tried dcommitting.03:38
→ thelorax123 joined03:38
deryni Hm... reflog might still have it.03:39
→ Guest99863 joined03:39
→ kukks joined03:39
← Guest99863 left03:39
→ War2 joined03:39
← War2 left03:39
→ War2 joined03:39
→ diegoviola joined03:39
juvenal_away → juvenal03:40
→ Celelibi joined03:40
Celelibi Hello.03:40
How can I see the url I pull from?03:40
isleshocky77 HEAD@{1} is the filter-branche: rewrite, and HEAD@{0} is the latest commit which hasn't been pushed03:40
rjhunter Celelibi: git remote -v03:41
Celelibi thanks.03:41
damn -v...03:41
deryni isleshocky77: HEAD@{1} is the rewritten history head? And HEAD@{2} is?03:41
isleshocky77 HEAD@{2}: checkout: moving from master to master03:41
it looks like part of my previous dcommit03:42
← Dougie187 left03:42
deryni isleshocky77: 'git update-ref refs/remotes/svn refs/heads/master' *might* fix things for you but I'm not at all sure about that since I don't really know what git-svn is expecting to find there. Make sure you keep track of what is currently under refs/remotes/svn just in case.03:42
isleshocky77 deryni: That's the exact command I had waiting on the console that I was going to try, but decided to consult here first. I'm going to make a backup of the directory and try running it. I'm just wondering how that will affect the non-dcommitted comit03:44
zacts cannon: oh cool03:44
rjhunter is scanning through the git-svn source to see exactly how it figures out what has been dcommitted.03:44
deryni isleshocky77: Should be ok I think but you can use 'master~' if you want to try that.03:44
isleshocky77 rjhunter: Thank you. I was scanning through the docs to see if there was an obvious git svn SET COMMIT or something dumb. No such luck03:45
deryni Yeah. I took a quick scan through git-svn but perl isn't my thing and I wasn't giving it enough attention.03:45
rjhunter isleshocky77: If you're worried you might lose that commit, `git format-patch` can save it to a file.03:45
→ macabre^ joined03:45
zacts cannon: I am thinking that it might be useful for old UNIX tools that used rcs.03:45
isleshocky77 rjhunter: Good point. I can pull a patch out of the backup now.03:45
← psyl0n left03:45
dmamills → dmamills_zz03:45
→ corygibbons joined03:45
ketas- → ketas03:46
isleshocky77 deryni: rjhunter: error: there are still refs under 'refs/remotes/svn' \n fatal: Cannot lock the ref 'refs/remotes/svn'.03:46
→ AWinterman joined03:46
→ ging joined03:47
← abelsiqueira left03:47
→ CKyle joined03:47
← macabre left03:47
→ eletuchy_ joined03:48
parcs hello. why does git fetch take a lot of cpu time on a large repository?03:48
← sarkis left03:48
← eletuchy left03:48
← Celelibi left03:50
ging can someone help me out with using the format patch command? i am trying to get it to output a single patch to go to go from the latest commit to a specific older commit, i can only seem to get it to do hundreds of little patches03:50
isleshocky77 ging: look at --stdout03:50
ging well the patch is to update the source from an older commit to the latest03:50
→ sohail joined03:50
isleshocky77 ging: You can direct that into a single file.03:51
→ Northmance joined03:51
← Northmance left03:51
← eletuchy_ left03:53
← Vinnie_win left03:54
→ Vinnie_win joined03:55
parcs looks like git fetch performs a memcpy+malloc of each file in the tree. i wonder why03:56
juvenal → juvenal_away03:56
parcs filename, that is03:56
← macabre^ left03:57
→ joelrobichaud joined03:57
rjhunter isleshocky77: it looks like git-svn looks up the last dcommitted push by 1) parsing log comments from `git rev-list --first-parent --pretty=medium` and 2) looking the svn info up in the "rev_map"03:58
isleshocky77 rjhunter: Thanks for looking that up. Any idea what I would do to fix this? (sorry to be a pain)03:59
ging isleshocky77: thanks, but that's not really what i am looking for, that is basicly stiching them all together what i want is a single patch03:59
i can get the diff and make my own patch without a problem but i really want to try and do it using the format patch command04:00
→ ezekielnoob joined04:00
juvenal_away → juvenal04:01
canoon ging if you want to get it as one04:01
you should do a rebase to squash them all04:01
then format-patch04:01
isleshocky77 ging: The description of format-patch is "Prepare each commit with its patch in one file per commit, formatted to resemble UNIX mailbox format." I think you would have to do what canoon just said.04:02
rjhunter isleshocky77: which is to say, if you've got a git commit message that ends with "git-svn-id: http://example.com/svn/trunk@3775 8092fc87-c524-0410-9efc-e669fe64eaf9", then git svn will look for .git/svn/refs/remotes/trunk/.rev_map-8092fc87-c524-0410-9efc-e669fe64eaf904:02
deryni ging: Why do you want to use format-patch for this specifically?04:02
← PDX-rmjack left04:03
→ PDX-rmjack joined04:03
← PDX-rmjack left04:03
rjhunter isleshocky77: how to fix it? i dunno yet but hopefully talking about it and looking depeer will lead to a solution :-)04:03
isleshocky77 rjhunter: So I should be manually creating that file?04:03
.../trunk is a file, not a directory.04:03
rjhunter: it has one line which is a hash, but doesn't seem to correspond to anything in git-log04:04
SamB isleshocky77: you sound pretty hosed04:05
I mean, trunk is a file? WTH?04:05
← War2 left04:05
SamB this may bear little relation to your current problem though04:05
ging deryni: i'm doing something which needs to be easily repeatable to go into some documentation instructions, i don't want to have to manually add all the patch header details to each diff04:07
isleshocky77 It's just the end of the ref where I think it's supposed to point to the commit it's on.04:07
canoon isleshocky77, my recommendation would be to git svn clone it again and just copy your commit over and dcommit it from the new one04:07
isleshocky77 SamB: I'm looking at some other git-svn checkouts and they all seem to have something similar at the end of the line.04:08
canoon and do the filter branch when you are actually going to convert04:08
deryni I'm with canoon on this one I think.04:08
Then I think you are left with the rebase option. Which you can do in a throw-away branch if you want to keep the split commits in the real branch.04:09
isleshocky77 canoon: Yeah, that's what I was thinking originally as well. But I'm really trying to leave that as the last last resort. It's a huge project with lots externals and 4000 commits and 20's of branches and like 75 tags. It was a huge pain to do it the first time and get it right. I want to smack myself for messing it up after 1 and a half over author names.04:09
→ Cervantes joined04:09
SamB isleshocky77: or, wait, you meant that there was a file named .git/svn/refs/remotes/trunk ? I dunno about that ...04:09
isleshocky77 SamB: Yeah.04:09
deryni You don't need to get all the branches and tags if you don't need git to work with them I believe.04:10
isleshocky77 SamB: It looks like when I did the original git svn clone I did it funky. All my other git svn clones have refs/remotes/git-svn as the file pointing to the commit. This has refs/remotes/svn/trunk04:10
SamB strange04:10
→ rurufufuss joined04:11
→ tang^m joined04:11
rjhunter a comment in the Git/SVN.pm source says the rev_map files "are disposable unless noMetadata or useSvmProps is set", and says they are pairs of SVN revision number + git commit SHAs04:12
(stored in binary -- 4 bytes + 20 bytes)04:12
isleshocky77: the name "trunk" or "git-svn" depends on options to "git svn clone"04:13
→ preflex_ joined04:13
← Vinnie_win left04:13
ging canoon i'm trying to do what you suggest but not quite sure how to make it work04:14
isleshocky77 rjhunter: Got it. And I think at the time the tutorial I was following recommended putting it a level deeper for some reasons. Hence the remotes/svn/trunk04:14
→ dyng joined04:14
canoon ging, git rebase -i04:14
← preflex left04:14
preflex_ → preflex04:14
canoon then just squash everything except for one04:15
← trusktr left04:15
→ Vinnie_win joined04:15
canoon ging, the real problem actually is the format patch includes the subject from one commit04:15
so your really only getting one subject04:15
← Octalot left04:17
← undecim left04:18
→ eletuchy joined04:18
isleshocky77 Has anyone used svn2git lately? Does it do a better job of handling externals? I remember that was a huge pia. I have 30 of them.04:19
← eletuchy left04:19
→ eletuchy joined04:19
deryni What are you expecting to happen to them?04:21
→ Quazil joined04:22
← aartist left04:23
Quazil Can you force the creation of an empty 'master' branch on a bare repo?04:23
rjhunter isleshocky77: just to clarify your repo's position, what do you see when you run `find -type f .git/svn`?04:23
Quazil: you'll need to create a commit with an empty tree, then set "master" to that commit04:24
isleshocky77 394 lines of output.04:24
rjhunter isleshocky77: are they all rev_map files?04:25
Quazil rjhunter: Can you do that on the bare repo directly? Or do you have to clone it first?04:25
→ mizu_no_oto joined04:25
→ eruMelkor joined04:25
isleshocky77 rjhunter: one .rev_map.*** and one unhandled.log for every tag and branch04:25
→ Katafalkas joined04:27
rjhunter Quazil: it's easiest to create commits from a non-bare clone, although you could use git plumbing commands to hand-craft a commit into a bare repository04:27
Quazil rjhunter: I'd be doing this from grit right after I created the repo, can you point me in the right direction to look up the plumbing?04:28
juvenal → juvenal_away04:29
rjhunter Quazil: It's likely that grit exposes everything you need. You'd need to create an empty "tree" object, create a "commit" object that refers to that empty tree, then finally set the master "reference" to that commit04:30
Quazil ty04:30
→ zastern joined04:31
isleshocky77 deryni: rjhunter: Thanks for trying to help me out. I'm trying a new clone with svn2git. The new directory is up to 50M, the old one was 6.5G :'( I love bloated projects.04:31
→ arjun_ joined04:32
deryni You shouldn't need to clone the whole thing. I assume most of that is in tags and branches you don't care about.04:32
← daito left04:33
← tang^m left04:34
← nerdfiles left04:34
← geomyidae left04:34
← sgambino left04:35
isleshocky77 deryni: I'm not sure what i need yet and what I don't though04:35
← axfcampos left04:36
Quazil rjhunter: Sweet, it worked - thanks!04:37
rjhunter Quazil: :-)04:37
deryni isleshocky77: Depends what you are trying to do I suppose.04:38
→ geomyidae joined04:39
← chamunks left04:39
BombStrike Let's say I have 2 branches, "A" and "B", then I start a branch "C" using "B" as a base, later I merge "B" in "A", how do I rebase "C" to the HEAD of "A"? git rebase A C ? git rebase --onto A B C ?04:40
ging ah i've given up and gone for using diffs04:40
← sohail left04:41
ging i would have though it would be easy to make patches from any point to any point with the format patch command04:41
isleshocky77 deryni: eff this. I just realized I have an old copy of this directory on an old machine I stopped using. I should just be able to run git svn rebase and get it up to speed.04:42
→ Vinnie_win_n joined04:43
← |Freeman| left04:43
→ szarubin joined04:44
deryni ging: It should be. Just not as one patch.04:44
juvenal_away → juvenal04:44
rjhunter ging: If you want to be binary-safe, make sure you use `git diff --binary`04:44
← Vinnie_win left04:44
rjhunter ging: (and apply with `git apply`)04:44
juvenal → juvenal_away04:45
← mizu_no_oto left04:45
← CKyle left04:46
← Goplat left04:46
ging deryni: the problem with doing all patches, certain things that get changed a lot get a huge list of patches, mostly undoing each other04:46
← yacks left04:47
→ tang^m joined04:47
canoon ging, doesn't matter04:47
→ undecim joined04:47
→ Goplat joined04:47
← kcm1700 left04:49
← zarubin left04:49
← karupanerura left04:49
← w0ng left04:49
← Tabmow left04:49
← wrp6 left04:49
→ wrp6 joined04:49
→ w0ng joined04:49
→ Tabmow joined04:49
→ kcm1700 joined04:49
→ zz_karupanerura joined04:49
deryni Sure, but that doesn't change the fact that format patch does support from X to Y easily, just not in the way you want. I'm still not sure what about the manual diff isn't what you want exactly. Unless you meant you want the overall diff but each commit message.04:49
zz_karupanerura → karupanerura04:50
← tang^m left04:51
← kukks left04:51
ging deryni: well yeah what i want to do might not really always work, what i wanted was the over all diff and the latest commit message, but that could turn out to be the wrong one04:52
canoon ging, whats it for?04:55
← mfeltner left04:57
→ sqlnoob joined05:01
← eruMelkor left05:01
→ hwrd|work joined05:01
ging canoon: creating patches to go into a debain package05:02
→ chuckharmston joined05:04
→ maleknet joined05:04
→ jaseemabid joined05:05
← Vinnie_win_n left05:08
→ Vinnie_win joined05:10
→ irqq joined05:11
→ LAAJH3A joined05:12
→ guest242 joined05:15
→ aartist joined05:16
← diegoviola left05:17
→ trusktr joined05:18
→ daito joined05:18
← etcetera left05:21
→ etcetera joined05:23
→ dhruvasagar joined05:24
→ CKyle joined05:24
→ nptl joined05:24
→ War2 joined05:24
→ Pengoo joined05:24
→ mfeltner joined05:28
← dhruvasagar left05:29
→ dhruvasagar joined05:30
← mfeltner left05:32
→ a-l-e joined05:33
← dhruvasagar left05:33
→ maestrojed joined05:33
← thelorax123 left05:33
→ el3ctron joined05:33
→ Vile joined05:34
← Vile left05:34
← jaseemabid left05:35
← MattMaker left05:35
→ Guest25023 joined05:35
→ runeg joined05:35
→ thelorax123 joined05:35
→ dhruvasagar joined05:35
← Guest25023 left05:35
→ Guest25023 joined05:35
Guest25023 → diegoviola05:36
→ MattMaker joined05:36
← Spyro left05:37
→ oddjobsman joined05:38
← maestrojed left05:38
→ Spyro joined05:40
→ Tommy[D]_ joined05:40
← bgerber left05:42
→ bgerber joined05:43
→ maveneagle joined05:43
← maveneagle left05:43
← warbaque left05:43
← Pengoo left05:43
← dhruvasagar left05:43
→ dhruvasagar joined05:44
← Tommy[D] left05:44
→ timthelion joined05:45
← nptl left05:45
← sagittarian left05:46
← Quazil left05:46
← chuckharmston left05:52
← dhruvasagar left05:53
→ dhruvasagar joined05:54
← guest242 left05:55
→ guest2421 joined05:55
→ warbaque joined05:55
→ maestrojed joined05:55
← maestrojed left05:56
← CKyle left06:01
→ hoschi joined06:02
← dhruvasagar left06:02
→ npnp joined06:03
← npnp left06:03
→ npnp joined06:03
← npnp left06:03
→ dhruvasagar joined06:03
→ npnp joined06:03
← npnp left06:04
→ jaseemabid joined06:04
← oddjobsman left06:05
→ npnp joined06:05
→ Skrypter joined06:06
← trusktr left06:06
← irqq left06:07
→ Pengoo joined06:07
→ blackwc joined06:08
→ nnyk_ joined06:08
blackwc hello. I have a message: leftover conflict marker.06:08
← oojacoboo left06:08
blackwc I've done no merging or anything like that.06:08
how I can fix this? It gives me the file and the line of the file where the conflict marker is06:09
canoon blackwc, remove it?06:10
skorgon edit the file and resolve the conflict06:10
blackwc but I'm not sure why it is conflicting06:10
I don't know what it wants me to do with it06:10
canoon, it's a necessary README06:10
canoon no i mean remove the conflict marker06:10
open the file there should be >>>> or <<< there06:11
delete the part you don't want06:11
→ kbart joined06:11
canoon and git add the file06:11
blackwc open which file?06:11
canoon git status06:11
blackwc I opened the README it says it conflicting and there is no arrows06:11
← john2x left06:12
blackwc is conflicting**06:12
canoon what does git status tell you?06:12
blackwc canoon, it gives me a list of files ready for commit. I can't commit b/c of conflict06:12
the conflicting file is "add new" because I removed it and added it again thinking it would resolve it.06:13
← PaulCapestany left06:13
blackwc er "new file"*06:13
← m0viefreak left06:13
← tpollard left06:13
canoon is everything staged for commit?06:14
← canoon left06:14
blackwc some modified files are not staged.06:14
→ guest242 joined06:14
blackwc nothing that would seem to cause this problem though.06:14
→ saurabh joined06:15
moritz just 'git add README'06:15
← guest2421 left06:15
blackwc i did06:15
saurabh → Guest8697106:15
blackwc scripts/README.md:2: leftover conflict marker06:16
that's what it says when I try to commit06:16
← AWinterman left06:16
→ teran joined06:17
← maurycy left06:18
→ cff joined06:18
→ k_sze[work] joined06:18
k_sze[work] Is `git submodule update` supposed to fail if the submodule is dirty?06:19
→ JeremyWei joined06:19
skorgon blackwc: i think git by itself would let you commit conflict markers. is there some hook running that scans for some pattern and is causing the issue?06:21
← daito left06:21
silviof1 → silviof06:21
blackwc I think I know the problem06:21
← DonRichie left06:21
blackwc it was a markdown file and I used "---------" under a subtitle in the document06:22
at the exact location it said there was a conflict marker06:22
I changed it to "## subtitle" and it worked06:22
← nnyk_ left06:22
→ nnyk_ joined06:23
→ ThomasLocke joined06:24
← Kabaka left06:25
mattcen Hi all. I've got a really messed up merge attempt that I can't explain; it's trying to change files it shouldn't. Is anybody able to have a look if I pastebin some output? I can't make the repo accessible, but can paste pretty much any output that'll help diagnose the issue.06:25
→ dreiss joined06:25
← hahuang65 left06:26
→ hahuang65 joined06:26
→ DonRichie joined06:26
← npnp left06:28
→ mfeltner joined06:29
→ ayaz joined06:29
→ daito joined06:29
← dreiss left06:29
milki mattcen: just pastebin it first06:30
→ sagittarian joined06:30
← Curtisbeef_ left06:30
→ matanya joined06:32
→ lavaman joined06:33
← mfeltner left06:33
← jaseemabid left06:34
mattcen Sure: http://paste.debian.net/66384/06:34
← madewokherd left06:34
← guest242 left06:35
→ jaseemabid joined06:35
→ Guest95138 joined06:35
← the_cybersphinx left06:35
← leeN left06:35
mattcen I can't see why it is that despite only a handful of files being updated by the 22869-new-kernel branch, I get so many changes made when I try to merge that in, and *most* of the changes are wrong; the ones it staged on non-interactively mostly *revert* changes to the 'wtf' branch that I want to keep06:35
← Guest95138 left06:35
→ Guest95138 joined06:35
← diegoviola left06:36
Guest95138 → diegoviola06:37
← hwrd|work left06:38
→ acl__ joined06:38
← acl__ left06:39
→ b_rare joined06:40
← acl_ left06:40
← blackwc left06:43
← nnyk_ left06:45
← Gamecubic left06:45
← dhruvasagar left06:45
→ dhruvasagar joined06:45
→ AWinterman joined06:47
→ guest242 joined06:47
→ maveneagle joined06:48
← maveneagle left06:48
← Eryn_1983_FL left06:48
→ Raging_Hog joined06:48
→ rob_w joined06:49
→ maveneagle joined06:49
← maveneagle left06:49
← AWinterman left06:51
← a-l-e left06:51
← matanya left06:53
← gonzzor left06:53
← dhruvasagar left06:54
→ diego1 joined06:55
→ freimatz joined06:55
→ dhruvasagar joined06:55
← diego1 left06:56
→ diego1 joined06:56
diegoviola → Guest9887006:56
diego1 → diegoviola06:56
← heroux left06:56
← Guest98870 left06:56
→ abrkn joined06:57
← War2 left06:58
← Goplat left06:59
← MattMaker left07:00
→ MattMaker joined07:00
→ shoshin joined07:01
→ lite_ joined07:02
← jaseemabid left07:02
← rjhunter left07:02
→ pantsman joined07:04
← MattMaker left07:04
← h0rrorvacui left07:05
← dhruvasagar left07:05
→ zapb joined07:06
→ m0ltar joined07:07
juls → master_j07:07
← echo-area left07:09
→ lb1a joined07:11
→ dhruvasagar joined07:11
→ bttf joined07:11
bttf i'm trying to write an automated process to bootstrap a machine with some apps, git included ... i have the scripts on github. is there a way i can wget a tarball of the source?07:12
→ echo-area joined07:12
→ chuckharmston joined07:12
→ cbreak-work joined07:14
lb1a bttf, yes07:15
← bwreilly left07:15
arand Nut sure about tarball, zip file of any arbitrary commit though, yes.07:15
→ TomyLobo joined07:16
Cork bttf: https://github.com/git/git/releases/tag/v1.8.5-rc207:16
https://github.com/git/git/archive/v1.8.5-rc2.tar.gz07:16
lb1a bttf, https://www.google.com/search?q=github+get+tarball07:16
← teran left07:16
→ teran joined07:17
← galaris left07:17
→ AWinterman joined07:17
arand Oh, nevermind, they have tarballs of arbitrary commits as well, just replace the suffix in the zip link.07:17
← dhruvasagar left07:18
← chuckharmston left07:18
arand (or if you have a tag, use that, as per above)07:18
← teran left07:18
← diegoviola left07:19
→ dhruvasagar joined07:19
bttf thx guys for some reason my previous google searches did not unearth this07:20
← kazi left07:20
→ MattMaker joined07:21
← graphitemaster left07:21
← guest242 left07:22
← s0beit left07:22
→ guest242 joined07:22
→ s0beit joined07:22
→ graphitemaster joined07:23
← MattMaker left07:24
→ rath joined07:24
→ MattMaker joined07:24
← joelrobichaud left07:25
← b_rare left07:25
karupanerura → zz_karupanerura07:26
← Cervantes left07:26
← bttf left07:27
→ b_rare joined07:27
→ takamichi joined07:28
→ Zarthus joined07:28
← AWinterman left07:28
→ user258467 joined07:28
← dhruvasagar left07:29
← szarubin left07:29
→ dhruvasagar joined07:29
→ mfeltner joined07:29
← pantsman left07:30
→ longbeach joined07:30
← ayaz left07:30
→ halu joined07:31
← arand left07:31
→ teran joined07:32
→ marthinal joined07:33
→ thierryp joined07:33
← mfeltner left07:34
← kermit left07:34
→ yargy joined07:34
→ papegaaij joined07:35
→ psoo joined07:35
← MattMaker left07:38
→ Josh1986 joined07:39
→ MattMaker joined07:39
freimatz hi, what git server should we choose for windows?07:39
← dhruvasagar left07:39
→ h0rrorvacui joined07:39
← lavaman left07:40
← Josh1986 left07:40
→ diegoviola joined07:40
→ dhruvasagar joined07:40
← _5kg left07:40
→ Josh1986 joined07:40
→ _5kg joined07:41
← diegoviola left07:41
→ diegoviola joined07:41
_ikke_ Some suggestions found on google: http://bonobogitserver.com/ http://gitstack.com/07:41
→ soee joined07:41
Josh1986 cant authorize via ssh, win7 - authentication failed07:41
→ j0ran1 joined07:42
Josh1986 what can it be, I generated and added public key?07:42
_ikke_ Josh1986: What are you using07:43
freimatz _ikke_: thx07:43
← daito left07:44
← hans_henrik left07:45
→ daito joined07:45
freimatz _ikke_: atr that possible candidates or really suggestions? ;-)07:45
← el3ctron left07:45
→ hans_henrik_ joined07:47
_ikke_ candidates07:47
→ ciampix joined07:47
← h12o left07:47
→ Search12 joined07:47
→ h12o joined07:48
← Norm_ left07:48
→ GeertV joined07:49
← daito left07:49
→ akiress joined07:49
thiago freimatz: git should be enough for Windows07:49
freimatz: are you having problems with that?07:49
Josh1986 _ikke_: i was installed gitlab, exactly like in installation guide07:50
_ikke_: and now i want push changes, but cant07:50
thiago why not?07:51
what's the error?07:51
→ ace_me joined07:51
ace_me I have a checkout repo,I did some modif and I want to revert to the checked out version07:52
thiago ace_me: have you committed your changes?07:52
Josh1986 fatal: Could not read from remote repository07:52
thiago Josh1986: and before?07:52
← beautifulmind left07:52
Josh1986 it works fine via http07:52
thiago Josh1986: pastebin the entire output, please07:53
_ikke_ ace_me: So all changes you made can be deleted?07:53
Josh1986 yes, I`m testing now07:53
→ netiul joined07:54
→ AWinterman joined07:55
→ zarubin joined07:56
← TomyLobo left07:57
→ arand joined07:57
→ daito joined07:58
→ npnp joined07:59
→ eijk joined07:59
← h0rrorvacui left07:59
→ _Qman joined08:00
Josh1986 thiago: what exactly you want in p&b?08:00
← freimatz left08:00
→ bvi joined08:00
→ mishok13 joined08:01
← j0ran1 left08:01
→ jnavila joined08:01
→ harobed joined08:01
thiago Josh1986: from the moment the command prompt was displayed to the next08:02
→ h_kon joined08:02
ace_me git reset --hard will also delete new created files ?08:02
thiago ace_me: did you git add them?08:03
← Skrypter left08:03
→ holderbaum joined08:03
→ nnyk_ joined08:03
→ rethus joined08:03
← eijk left08:03
← zastern left08:04
→ juril joined08:04
→ Davey joined08:04
← bvi left08:05
hans_henrik_ → hans_henrik08:05
→ n1x joined08:06
← Streward left08:07
→ Guest32821 joined08:07
← AWinterman left08:08
← Tuna-Fish left08:09
Josh1986 http://pastebin.com/gnS9nbwv08:09
→ nixmaniack joined08:09
← ace_me left08:10
→ galaris joined08:10
→ chirayuk joined08:12
→ jaseemabid joined08:12
→ Tuna-Fish joined08:12
Josh1986 thiago: I pasted with some debug info08:12
→ threesome joined08:12
→ johnkevi_ joined08:13
← Watabou left08:13
← johnkevinmbasco left08:13
→ chuckharmston joined08:14
→ h0rrorvacui joined08:14
← shoshin left08:15
← dhruvasagar left08:15
thiago Josh1986: your server didn't authenticate you08:15
Josh1986: that's not a Git problem08:15
Josh1986: please fix your SSH server first08:15
→ dhruvasagar joined08:15
Josh1986 thiago: mmmm I dont understand, where is exactly my problem?08:16
← jaseemabid left08:16
← sam113101 left08:16
thiago Josh1986: in your SSH daemon08:17
Josh1986: it's not authenticating properly. Fix it.08:17
→ jaseemabid joined08:17
→ shoshin joined08:18
Josh1986 thiago: I`m using windows...08:18
← chuckharmston left08:18
← graphitemaster left08:19
→ graphitemaster joined08:19
thiago Josh1986: that has nothing to do with it08:19
Josh1986: the problem is in the server08:19
Josh1986: fix the server08:19
canton7 if I'm reading the log correctly, you don't have a private key on your local machine?08:20
thiago he's authenticating by password08:20
lb1a anybody can explain this strange gitolite behaviour to me ? http://fpaste.org/54723/84762828/ or shall i ask in #gitolite?08:21
canton7 cool, if that's what he intends to do (i.e. he's not using gitolite)08:21
→ Ozch joined08:21
canton7 lb1a, don't use the repositories/ prefix08:21
Josh1986 canton7: I have! I was generated and saved public key on server machine08:21
→ mklappstuhl joined08:21
canton7 lb1a, line 6 doesn't have that prefix ;)08:22
thiago Josh1986: your SSH server offered to negotiate via keys. You didn't send them.08:22
canton7 Josh1986, well you haven't saved the private key, with the right name and permissions, in ~/.ssh08:22
thiago Josh1986: lines 8 to 10 show that your SSH client found no keys08:22
Josh1986: do you want to authenticate via key mechanism?08:22
← teran left08:23
lb1a canton7, hm i failed. shame on me. thanks08:23
→ shruggar joined08:24
← mtk left08:25
← jaseemabid left08:25
lb1a damn gitweb is trying to fuck me :/08:26
→ mooperd joined08:26
→ jaseemabid joined08:26
Josh1986 canton7: I just run ssh-keygen -t rsa -C "my@email.com"08:26
→ teran joined08:26
Josh1986 thiago: Yes, working with gitlab needs key08:27
canton7 Josh1986, right, and I've told you that there's more you need to do08:28
there are a thousand and one tuts out there on getting this working; I'm afraid I need to go to work08:28
← Guest94419 left08:28
← maslowbeer_ left08:28
→ raar joined08:30
→ SimonNa joined08:30
raar → Guest5250608:30
→ mfeltner joined08:30
lb1a canton7, dont be afraid of going to work ;)08:30
Josh1986 Just one question where exactly I should run ssh-keygen on the client or on server machine?08:30
→ Balliad joined08:30
← abrkn left08:30
← nixmaniack left08:30
→ halu_ joined08:31
← teran left08:32
← jwmann left08:32
← Balliad left08:33
→ Balliad joined08:33
← halu left08:33
→ AWinterman joined08:34
← mfeltner left08:35
← dhruvasagar left08:35
→ dhruvasagar joined08:35
← h0rrorvacui left08:35
→ sam113101 joined08:36
← Devastator left08:36
→ Devastator_ joined08:36
→ mtk joined08:37
→ sm0ke joined08:38
→ dreiss joined08:39
→ aChatir joined08:41
← aChatir left08:41
→ joel_falcou joined08:42
→ maslowbeer_ joined08:42
→ johnkevinmbasco joined08:43
← guest242 left08:43
← dhruvasagar left08:43
← johnkevi_ left08:44
→ dhruvasagar joined08:44
→ arvind_khadri joined08:44
← Khisanth left08:44
Josh1986 question... I can't use windows command line? I always should use git-bash?08:45
→ Asenar joined08:46
← Georg3 left08:47
→ TomyWork joined08:47
thiago Josh1986: you can08:47
Josh1986 I can clone [email@hidden.address] via git-bash, but via cmd - not08:48
← AWinterman left08:48
← jaseemabid left08:51
→ jaseemabid joined08:52
← JeremyWei left08:54
→ bitrixdo joined08:55
← dhruvasagar left08:56
→ dhruvasagar joined08:56
bitrixdo i need help ((=08:57
there are some of you have free time?08:57
← mklappstuhl left08:58
→ vtunka joined08:58
→ Khisanth joined08:58
→ Anticom joined08:58
→ quetz joined08:59
← yetet left08:59
bitrixdo is there anyone free?)09:00
→ guest242 joined09:00
→ trusktr joined09:02
← Rob___ left09:02
lb1a i really want to use gitolite for my project hosting once and for all. but now i need to integrate a webviewer too. i kind of hate gitweb and i wanted to use this viewer: https://github.com/jonashaag/klaus. it runs as a wsgi app. now... how to handle the permissions correctly?09:03
← Search12 left09:03
← shruggar left09:03
← dhruvasagar left09:03
← guest242 left09:04
→ guest2421 joined09:04
→ dhruvasagar joined09:04
→ ixe joined09:04
lb1a little paste for this :/ http://fpaste.org/54727/47655231/09:05
→ Rob___ joined09:06
lb1a of course www-data != git soo... should i setup gitolite in a different way? should i use custom acls? or what would you do?09:06
→ Norm_ joined09:09
← mooperd left09:09
moritz I don't think you need custom ACLs09:10
← ixe left09:10
→ mooperd joined09:10
→ jasonroyle joined09:11
← graphitemaster left09:11
→ graphitemaster joined09:11
moritz maybe make www-data the default group of use git, and change the group of the existing files09:11
or maybe configure your apache with suexec to execute it in the context of the git user or group09:12
→ chuckharmston joined09:12
lb1a moritz, hm suexec seems to be a good option.09:13
← chuckharmston left09:13
← n1x left09:13
lb1a but since the repos have 700 permissions, changing the group would not function out of the box , am i right?09:14
→ chuckharmston joined09:14
moritz right09:14
lb1a so suexec would be better i think09:14
→ AWinterman joined09:15
→ shruggar joined09:16
← oznt left09:16
→ teran joined09:17
← chuckharmston left09:18
→ Jackneill joined09:19
← thegladiator_ left09:20
→ nesQuick joined09:20
← Layke left09:20
kalleth_ → kalleth09:21
← jaseemabid left09:21
→ jaseemabid joined09:22
→ Kayra joined09:22
juvenal_away → juvenal09:23
→ snql joined09:23
→ h0rrorvacui joined09:23
shoshin → shoeshine09:24
shoeshine → BlueOrg09:24
← rchavik left09:24
← maloik_ left09:24
→ Psy-Q_ joined09:27
← Psy-Q_ left09:27
→ Sonderblade joined09:28
← Sonderblade left09:29
← AWinterman left09:29
← terminalRepeat left09:31
← lite_ left09:31
→ mfeltner joined09:31
← h0rrorvacui left09:31
← dreiss left09:33
← boombatower left09:34
← mfeltner left09:35
→ Layke joined09:36
← guest2421 left09:37
← diegoviola left09:37
→ lite_ joined09:39
← trusktr left09:39
← graphitemaster left09:39
→ graphitemaster joined09:39
← mooperd left09:40
→ quetz^^ joined09:40
← harish left09:40
→ mooperd joined09:42
← quetz left09:43
← ktosiek_ left09:44
← johnkevinmbasco left09:44
← jaseemabid left09:44
← thiago left09:44
→ marhaban joined09:45
→ guest242 joined09:45
← hyperair left09:45
→ Ange_blond joined09:46
← weiyang left09:46
← dyng left09:47
← cbreak-work left09:47
← marhaban left09:47
← rurufufuss left09:48
→ Plouj- joined09:49
→ mikecmpbll joined09:49
→ ktosiek joined09:50
→ heroux joined09:51
→ cbreak-work joined09:52
← shruggar left09:52
→ freimatz joined09:52
← Plouj left09:52
← boblee left09:52
→ beautifulmind joined09:53
→ boblee joined09:53
← Kayra left09:56
→ AWinterman joined09:56
_rgn git show parameter for shorter output with only files and their +-?09:56
→ Kayra joined09:57
← M00sL0gB0t left09:57
→ rafeca joined09:57
→ M00sL0gB0t joined09:57
→ LeMike joined09:57
→ irqq joined09:58
→ ni__ joined09:59
→ jaeckel joined09:59
_rgn ah, --stat10:00
moritz _rgn: --stat10:00
moritz too slow10:00
← ni_ left10:00
→ pixelgremlins joined10:00
lb1a moritz, by 2 seconds ;)10:00
← Davey left10:01
→ rurufufuss joined10:01
← beautifulmind left10:02
freimatz thiago: git should be enough for Windows? is there a server included?10:02
cbreak-work that was a question?10:02
→ raonyguimaraes joined10:03
← guest242 left10:03
→ thegladiator_ joined10:05
Odd_Blok1 → Odd_Bloke10:06
freimatz from me: yes ;-)10:06
← echo-area left10:07
→ eijk joined10:07
cbreak-work You only need a server if you want to serve something.10:08
normally, git is not served directly, it is used via ssh as transport, so a simple ssh server will do.10:08
but on windows... no idea.10:08
→ spaceonline joined10:08
→ mfeltner joined10:09
→ JeremyWei joined10:10
← AWinterman left10:10
→ hs366 joined10:11
→ chuckharmston joined10:12
→ ewnd9 joined10:12
→ dyng joined10:13
← mfeltner left10:13
← mooperd left10:13
→ koda joined10:14
→ LimeBlast joined10:15
← hs366 left10:16
← chuckharmston left10:17
→ beautifulmind joined10:18
→ canoon joined10:18
→ twoface88 joined10:18
← chirayuk left10:19
→ Advocation joined10:19
← timthelion left10:20
← heroux left10:20
→ matanya joined10:20
juvenal → juvenal_away10:21
← bitrixdo left10:22
→ bitrixdo joined10:22
→ heroux joined10:22
→ harish joined10:24
→ mooperd joined10:25
→ kewah joined10:25
juvenal_away → juvenal10:25
← hoactzin left10:26
→ hs366 joined10:28
→ Misan joined10:29
← Kayra left10:33
→ jaseemabid joined10:34
← hs366 left10:34
→ guest242 joined10:34
← guest242 left10:34
→ guest242 joined10:35
← skorgon left10:35
→ xobb joined10:35
→ jeffreyio joined10:36
→ deuti joined10:36
← npnp left10:37
→ AWinterman joined10:37
← k_sze[work] left10:37
→ npnp joined10:37
→ mlechner joined10:38
← LimeBlast left10:38
→ Search12 joined10:39
← rob_w left10:40
← pixelgremlins left10:40
→ iamjarvo joined10:40
juvenal → juvenal_away10:41
→ LimeBlast joined10:42
← luc_ left10:42
← iamjarvo left10:42
→ corygibb_ joined10:43
→ SHyx0rmZ joined10:43
← mooperd left10:43
→ Sonderblade joined10:44
← corygibbons left10:45
→ burzum joined10:45
← wrouesnel1 left10:45
← h_kon left10:46
← dyng left10:48
juvenal_away → juvenal10:48
→ shruggar joined10:49
→ DanielHollands joined10:49
← LimeBlast left10:49
→ tma996 joined10:50
→ skorgon joined10:50
→ penguin42 joined10:51
→ mooperd joined10:51
← AWinterman left10:52
← GeertV left10:53
→ GeertV joined10:53
← guest242 left10:55
→ guest242 joined10:55
← rjc left10:55
→ rob_w joined10:55
→ xpe joined10:56
freimatz thx, we intended to have a server wit an repository used as origin for developers normally to push and pull and as base for jenkins10:56
← GeertV left10:57
→ h_kon joined10:58
← heroux left10:59
→ mdiers joined10:59
cbreak-work linux or mac os works fine for that.10:59
← guest242 left10:59
cbreak-work (with gitolite if you want to manage permissions)10:59
→ heroux joined10:59
cbreak-work maybe set up a small VM or so?11:00
→ guest242 joined11:00
← DanielHollands left11:00
← jnavila left11:00
→ jnavila joined11:00
freimatz we will take a look - but now i have to restart PC :-(11:01
Windows ;-)11:01
← Advocation left11:01
← Ivo left11:02
← freimatz left11:02
→ LimeBlast joined11:02
→ Advocation joined11:02
BlueOrg → worries11:03
worries → BlueOrg11:04
juvenal → juvenal_away11:04
→ guest2421 joined11:04
→ Davey joined11:04
← guest242 left11:04
→ jnavila_ joined11:05
→ dyng joined11:05
juvenal_away → juvenal11:05
← daito left11:05
→ wrouesnel joined11:06
→ rhombus joined11:08
rhombus → Guest3026011:08
← jnavila left11:09
→ Vinnie1991 joined11:09
→ mfeltner joined11:10
← Vinnie1991 left11:10
Guest30260 → rhombus11:10
← guest2421 left11:10
→ chuckharmston joined11:12
→ eaj joined11:12
→ jburcsik_ joined11:12
← Davey left11:13
← rhombus left11:13
→ rhombus joined11:13
← mfeltner left11:14
← jburcsik left11:14
jburcsik_ → jburcsik11:14
← LimeBlast left11:16
BlueOrg → shoshin11:16
← chuckharmston left11:17
← h_kon left11:17
→ AWinterman joined11:18
→ geekbri joined11:18
← geekbri left11:18
→ davividal joined11:19
→ Ivo joined11:20
← nesQuick left11:21
← _5kg left11:21
→ _5kg joined11:21
→ salamanderrake joined11:21
→ quetz joined11:21
← quetz^^ left11:22
→ timthelion joined11:22
← donmichelangelo left11:22
← jeffreyio left11:22
→ donmichelangelo joined11:22
← MattMaker left11:24
→ MattMaker joined11:25
← m0ltar left11:26
→ m0ltar joined11:28
← rouf left11:30
← m0ltar left11:31
→ the_cybersphinx joined11:31
→ j0ran joined11:32
← AWinterman left11:32
→ rouf joined11:33
→ axfcampos joined11:33
→ pfFredd joined11:34
← canoon left11:34
→ Sebastian_ joined11:35
← Sebastian_ left11:36
→ RaceCondition joined11:36
→ j0ran1 joined11:36
← j0ran left11:36
shoshin → BlueOrg11:37
→ spheros joined11:38
← User458764 left11:39
→ Kabaka joined11:39
← sqlnoob left11:41
← RaceCondition left11:41
← harobed left11:41
→ Gavrisimo joined11:43
→ tapout joined11:43
→ Kayra joined11:44
→ tvw joined11:44
← j0ran1 left11:45
→ bschindler joined11:46
→ oriba joined11:46
← Josh1986 left11:47
→ RaceCondition joined11:47
→ khushildep joined11:47
← Kayra left11:48
→ tadasZ joined11:49
← mikecmpbll left11:49
← pfFredd left11:49
← mooperd left11:50
← mattcen left11:50
→ girfas joined11:51
girfas Hi, Do you guys know any place to practice basic git features?11:52
→ soooga_ joined11:52
→ mooperd joined11:52
→ mattcen joined11:52
lb1a girfas, !trygit11:53
gitinfo girfas: A quick, interactive, in-browser introduction to git is available at http://try.github.io11:53
→ mikecmpbll joined11:53
girfas thanks I will check it out11:54
→ soooga joined11:54
→ User458764 joined11:55
← soooga_ left11:55
→ asQuirreL joined11:55
→ pfFredd joined11:56
← harish left11:57
→ AWinterman joined11:58
MattMaker wat11:59
lb1a MattMaker, !wat11:59
gitinfo MattMaker: Wat: https://www.destroyallsoftware.com/talks/wat11:59
MattMaker why is github's pull request generator comparing to my master repo even when I point it at a branch11:59
→ johnkevinmbasco joined11:59
MattMaker I like that talk11:59
← davividal left12:00
MattMaker ah. it's a bug12:00
I have to force-reload the page12:00
→ rjc joined12:01
← Kabaka left12:01
→ Criztian joined12:02
MattMaker doesn't seem to happen every time, either. *shrug*12:03
← Advocation left12:03
lb1a MattMaker, sounds like browser caching, but if you actually think it's githubs fault -> #github12:04
MattMaker thanks. I guess I'll leave it unless/until I can 100% reproduce it12:04
→ Gamecubic joined12:05
← sm0ke left12:05
← paulweb515 left12:06
→ Kayra joined12:06
→ quetz^^ joined12:06
→ mklappstuhl joined12:07
← quetz^^ left12:08
→ quetz^^ joined12:08
→ abelsiqueira joined12:09
← cdleonard left12:09
← quetz left12:10
→ mfeltner joined12:11
← jaseemabid left12:11
← mklappstuhl left12:12
→ chuckharmston joined12:12
← twoface88 left12:12
← AWinterman left12:13
← BlueOrg left12:13
→ j0ran1 joined12:14
← Z6 left12:14
← soooga left12:14
→ jaseemabid joined12:15
← mfeltner left12:15
← tadasZ left12:15
→ Kabaka joined12:15
→ Snetty joined12:16
← chuckharmston left12:17
→ yacks joined12:17
→ shoshin joined12:18
→ fightback joined12:19
← kewah left12:19
→ tadasZ joined12:20
← Rix left12:21
← dyng left12:22
← lite_ left12:24
→ lite_ joined12:25
← LeMike left12:26
← kcj left12:27
→ echo-area joined12:27
→ russoneto joined12:28
→ snag joined12:30
→ kewah joined12:30
← Entomo left12:31
→ freimatz joined12:33
Snetty on a remote with multiple urls, is it possible to suppress any errors so that a push completes all the urls?12:33
← shruggar left12:33
Silex Damnit, git is too smart. http://ideone.com/3ubVi112:34
I made a commit have its parent as one of its children, but git log refuses to enter an infinite loop!12:34
→ D4N3 joined12:36
← tadasZ left12:36
→ harobed_ joined12:36
← lite_ left12:36
← alexbligh left12:36
← D4N3 left12:37
→ D4N3 joined12:37
← asQuirreL left12:38
→ kirillt joined12:38
kirillt Hi, guys.12:38
gitinfo kirillt: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.12:38
→ dsdeiz_ joined12:39
← j0ran1 left12:39
→ lite_ joined12:39
→ AWinterman joined12:39
← dsdeiz left12:40
kirillt What can be wrong with my git config? `git config author.name' and `git config --global author.name' give me proper value A when author of commit becomes wrong B.12:40
How can it be?12:40
I checked config right before commit.12:40
→ LimeBlast joined12:40
→ Pullphinger joined12:42
← johnkevinmbasco left12:43
← ewnd9 left12:44
→ twoface88 joined12:44
→ johnkevinmbasco joined12:45
← dhruvasagar left12:45
← Kayra left12:46
→ sebhoss joined12:46
→ rnsanchez joined12:46
→ JPeterson joined12:46
← [1]JPeterson left12:48
← mooperd left12:49
→ j0ran1 joined12:50
← andytoshi left12:50
← shoshin left12:51
→ Rix joined12:51
Sonderblade kirillt: just guessing, character encoding issue?12:51
→ abrkn joined12:52
← j0ran1 left12:52
→ stephCoue joined12:53
→ Roots47 joined12:53
← AWinterman left12:54
→ Squarism joined12:54
→ spheros1 joined12:55
← bgerber left12:56
rethus → rethus_ZZzzzZZZ12:56
→ mooperd joined12:56
kirillt Sonderblade, no, "bad" name is my name in native language and "good" name is name in english12:57
Sonderblade, not random stuff12:57
Sonderblade, even can't find where I could set my native version of name12:57
← npnp left12:58
→ h_kon joined12:58
← ciampix left12:58
kirillt SSH public key seems to not contain name (and I don't think that it should).12:58
← spheros left12:58
→ franksta-imac joined12:59
→ asQuirreL joined12:59
← abrkn left12:59
→ donnib joined13:00
donnib hi, i am trying to convert a svn branch to git local branch with following command git branch -f -t mybranch refs/remotes/mybranch13:01
→ johnkevi_ joined13:01
lb1a kirillt, did you check it using "git config --list" ?13:01
← bitrixdo left13:01
donnib but it fails with : fatal: Cannot setup tracking information; starting point 'refs/remotes/mybranch' is not a branch.13:01
← snag left13:01
← johnkevinmbasco left13:01
donnib what am i doing wrong ? i am using git version 1.8.4.msysgit.013:02
gitinfo set mode: +v13:03
donnib hi, i am trying to convert a svn branch to git local branch with following command git branch -f -t mybranch refs/remotes/mybranch13:03
→ bgerber joined13:03
donnib but it fails with : fatal: Cannot setup tracking information; starting point 'refs/remotes/mybranch' is not a branch.13:03
what am i doing wrong ? i am using git version 1.8.4.msysgit.013:03
lb1a donnib, we heard you, the first time you pasted....13:04
donnib lb1a: really sorry about that13:04
i thought it did not go thru, i got the spam protection message13:04
→ harish joined13:04
donnib so i thought i had to paste again13:04
rethus_ZZzzzZZZ → rethus13:04
lb1a donnib, you'd get a notifictaion from your IRC client that your messaged failed.13:04
← maleknet left13:05
← dsdeiz_ left13:05
donnib yeah i was wondering why i didn't do that but there was a reason ;)13:05
lb1a and that anti-spam system is currently disabled due to an outage of the responsibly bot13:05
→ VictorCL joined13:05
→ tmcmahon joined13:05
← ni__ left13:06
← SimonNa left13:06
← keller left13:06
→ kcj joined13:07
→ dfaught joined13:07
→ j0ran joined13:08
← tmcmahon left13:08
dmamills_zz → dmamills13:08
Guest49362 → kamal_13:09
← beautifulmind left13:09
→ cannonball joined13:10
penguin42 are there any git tools that 'know' about function boundaries? i.e. I want to find the history of changes to a C function (in this case I know that somewhere between two versions it lost a couple of parameters - I want to know what happened to them)13:10
→ danecando joined13:10
→ jonatasnona joined13:10
← stingsay` left13:11
grawity the -L option of `git log` may help, or -G to just search for when a specific word first appeared13:11
→ mfeltner joined13:12
→ renatow joined13:12
→ Kayra joined13:12
→ chuckharmston joined13:12
→ lavong joined13:12
penguin42 grawity: Thanks! the -L looks good, I'd been using -S quite a bit13:13
→ PeterBooker joined13:14
← Fandekasp left13:14
grawity -S and -G are somewhat different; -S matches when the /count/ of the occurences changes, and -G matches when the word was in the commitdiff at all (so even if it was moved within the same line)13:14
→ johnkevinmbasco joined13:15
← mfeltner left13:16
penguin42 yeh, I think I should start using -G generally rather than -S13:16
donnib no ideas ?13:16
← chuckharmston left13:16
← mtk left13:17
← johnkevi_ left13:17
← Kayra left13:18
lb1a donnib, what is it you dont understand?13:18
donnib, you cant track a ref.13:18
but you asked git branch for it13:18
← rafeca left13:19
donnib i am trying to convert the branch to a local branch from an svn branch13:19
that's when i get this13:19
sp i don't know how to create that branch then13:19
lb1a donnib, did you import the svn repo into a git repo?13:20
donnib yes13:20
and i just did a git svn fetch13:20
before trying to do the above13:20
→ AWinterman joined13:21
donnib i have a bunch of branches in \svn\refs\remotes13:21
← dmamills left13:21
donnib but nothing in refs\remotes13:21
→ dzan joined13:21
lb1a donnib, do you plan to interact with svn further or was it a one-time import?13:21
→ mtk joined13:21
donnib i plan to sync the svn -> every day13:22
dzan hi, if I want to run astyle over my whole codebase to force the code style guidelines but keep the author information is there a way to do that?13:22
lb1a donnib, so use git svn branch and not git branch.13:22
donnib doing git svn fetch13:22
so will that create a local git branch ?13:22
that i then can push to the git server13:23
lb1a donnib, and your error is complaining about "cannot track a ref" so just drop the "-t" (--track) parameter from your command13:23
→ MattStratton joined13:23
donnib fyi, i don't commit into the git, right now i am in the playmode to migrate to git13:23
so by removing the -t i won't be tracking it anymore meaning ?13:24
→ mklappstuhl joined13:24
lb1a donnib, http://git-scm.com/book/en/Git-Branching-Remote-Branches#Tracking-Branches13:25
donnib ok so that means i won't be tracking the svn branch anymore, to get those changes into git i need to remove the git branch and recreate it ?13:26
Sonderblade i got branchA which is base for branchB which is base for branchC, how do i then move branchC from its base on branchC to branchA (master)?13:26
kirillt lb1a, your advice helped. git config --list shoed that I haven't configured "user.name"; I don't know where from git takes name for it but I manually set value for it and commits are ok now13:26
donnib but why is it that it can't track it ?13:26
→ sohail joined13:27
← sohail left13:27
→ sohail joined13:27
lb1a kirillt, \o/13:27
donnib, because it can only track a remote branch13:27
you are pointing it to a ref. that's not the same13:28
donnib but it is a remove branch but a svn branch or am i missing somthing ?13:28
← mklappstuhl left13:28
kirillt btw, what's difference between "user" and "author"?13:30
→ War2 joined13:30
← quetz^^ left13:30
→ quetz joined13:31
→ rafeca joined13:31
← sohail left13:33
→ scampbell joined13:33
_ikke_ kirillt: you mean comitter and author13:34
?13:34
← mikecmpbll left13:35
← AWinterman left13:35
← _Qman left13:36
← Raging_Hog left13:36
→ Plummer joined13:38
→ Advocation joined13:38
→ paulweb515 joined13:38
← MattStratton left13:39
← danecando left13:39
← fightback left13:40
→ ilhami joined13:40
ilhami Hey13:40
How can I push my project to google code from the terminal?13:40
→ mcruz joined13:40
→ alexbligh joined13:41
→ GeertV joined13:42
→ freeed joined13:42
← mcruz left13:43
→ quetz^^ joined13:43
_ikke_ ilhami: Have you cloned from google code?13:44
ilhami yes13:44
I did13:44
→ ewnd9 joined13:44
moritz make your changes, git add, git commit, git push13:45
kirillt _ikke_, I thought that commiter and author are same13:46
→ wet88 joined13:46
kirillt _ikke_, I am about "user.name" and "author.name" in config13:46
→ decotii joined13:46
← j0ran left13:47
← quetz left13:47
← takamichi left13:47
_ikke_ committer and author aren't necessarily the same13:47
kirillt: have you looked at man git config?13:48
gitinfo kirillt: the git-config manpage is available at http://jk.gs/git-config.html13:48
thegorn kirillt: in git committer and author can be different, although it doesn't show in regular git log13:48
→ Snibble joined13:48
→ dsdeiz joined13:48
thegorn kirillt: if you think about things like the linux kernel where patches are accepted off a mailing list generally speaking.. git keeps track of who wrote the patch (person who emailed it) but also keeps track of who actually made the commit and added it as part of linux kernel, they are both useful pieces of information on projects like that13:49
← parcs left13:50
ilhami _ikke_: what do I do after ?13:50
← mooperd left13:50
→ fightback joined13:50
← fightback left13:50
→ parcs joined13:50
→ Sendoushi joined13:51
→ j0ran joined13:52
→ breakingthings joined13:53
→ mj12albert joined13:53
← Rix left13:53
← Sendoushi left13:53
← mj12albert left13:53
→ Sendoushi joined13:54
MattMaker if I'm submitting a pull request, and in addition to the actual content, it contains an unrelated commit and a reversal of that unrelated commit, is that going to make it any more difficult for the recipient to merge?13:54
→ sgambino joined13:55
← juril left13:56
grawity MattMaker: no, it's not, but the recipient is very likely to ask you to clean up the pull request13:56
_ikke_ ilhami: after what13:56
ilhami after I cloned it?13:56
grawity MattMaker: just to reduce clutter in the stuff they're about to merge13:56
← Advocation left13:57
grawity MattMaker: it's common to clean up feature branches before merging them, using `rebase -i`13:57
→ faranjessani joined13:57
← Grum left13:58
MattMaker grawity: it is indeed a feature branch... I've never used rebase yet though13:58
→ spal joined13:58
ilhami _ikke_: ?13:58
kirillt _ikke_, thegorn, understood, thanks13:59
→ Grum joined13:59
→ Rix joined13:59
→ mooperd joined13:59
grawity MattMaker: see !sausage_making, !perfect13:59
gitinfo MattMaker: Some developers like to "hide the sausage making", transforming their commits before presenting them to the outside world. See http://sethrobertson.github.com/GitBestPractices/#sausage and !perfect13:59
MattMaker: [!postproduction] So, you want to make your commit history look pretty before pushing? http://sethrobertson.github.com/GitPostProduction talks you through how to use 'rebase -i' to do this.13:59
→ dyng joined13:59
→ TheBrainWork joined13:59
← svector left13:59
MattMaker "your commits sprung full-formed in utter perfection into your git repository" haha thanks, reading14:00
wait, did the git bot just try to invoke itself from its own answer?14:01
→ svector joined14:01
→ hellopat joined14:01
spal Is there a git option to do this: cd /foo; git clone <url>? In mercurial, this can be done as: hg --cwd /foo <url>. Anyway to do this in git?14:01
grawity no, it was invoked twice14:01
spal: git clone <url> /foo14:01
→ AWinterman joined14:01
→ griotspeak joined14:02
grawity spal: also, if I remember correctly, the next git release will have -C that works the same way as --cwd in hg, or -C in make.14:02
not that there's anything wrong with `cd` >_>14:02
spal grawity, that would set /foo as the repository directory. whereas I want to set /foo/reponame as the repository directory. so I probably want: git clone <url> /foo/reponame14:02
grawity spal: yeah, that14:02
spal thanks14:02
→ pixelgremlins joined14:04
← harobed_ left14:05
← bgerber left14:05
→ ni__ joined14:05
MattMaker auugh, I can't read this right now. I'm just going to re-create the feature branch. it's only 4 lines changed14:05
← Pyrogerg left14:05
MattMaker and i use SourceTree :-/14:05
→ guns joined14:05
→ Advocation joined14:06
grawity which does not prevent you in any way from using other Git tools (including command-line git)... although I guess it takes longer to learn them than otherwise.14:07
MattMaker I have to balance between the pure learnings and the billable time....14:08
→ FortySix2 joined14:08
grawity I suppose14:08
thegorn and to be fair, sourcetree is the best git gui i've seen by far, in at least when i go to a coworker's desk they haven't renamed every freaking git feature14:09
→ takamichi joined14:09
MattMaker thegorn: a thing I'd really like, is for every dialog box and checkbox to have a tooltip that explains what that particular UI feature is asking git to do underneath14:10
← tvw left14:10
→ bgerber joined14:10
thegorn oh i hadn't seen that, i'm just used to going over and seeing someone using tortoisegit, they ask me to help with something trivial and i get pissed and leave after i realize that any sort of actual git knowledge is useless for helping them14:11
→ Kayra joined14:11
cbreak-work thegorn: I help them. "just type git blahblah flags labels paths and so on"14:12
→ chuckharmston joined14:12
thegorn but by far i still prefer command-line git, but on windows the command line git is sorta painful sometimes, and some people i work with just type so freaking slow that maybe a gui is the right thing14:12
→ mfeltner joined14:12
thegorn cbreak-work: yea, that's probably a better way to go about it14:13
← Kayra left14:13
→ juril joined14:13
← jburcsik left14:13
→ Kayra joined14:14
← insejn left14:14
MattMaker thegorn: for me, I do a ton of staging hunks and individual lines, because I multitask and the changes don't belong together... and seeing all the hunks visually and in color and clickable all on the same screen, rather than being asked yes/no on each one by itself, is an entirely different experience to me14:14
thegorn MattMaker: git gui can do that and is built in, i actually do use that gui a lot for stuff like that14:15
→ mklappstuhl joined14:15
← AWinterman left14:16
MattMaker tortoisegit, ugh, agree. Windows Explorer Shell extensions are a great idea in theory, but in practice, at least in the past, they are a way to lock up your whole OS's UI14:16
thegorn well, in any case, if you are taking the time to do that, you will enjoy rebase -i, it helps a lot with the "oops, i forgot to add a line to that split out commit from earlier"14:16
← snql left14:16
← delasteve left14:17
← mfeltner left14:17
→ snql joined14:17
← j0ran left14:17
← chuckharmston left14:17
thegorn because you can squash a later commit into an earlier commit, so before i clean things up i'll have a lot of commits like "FIXME: add to the usb addition patch" or something like that14:17
← snql left14:18
← ilhami left14:18
→ snql joined14:18
MattMaker I can see how I might get into that, yeah14:18
← Kayra left14:18
← Chwing left14:19
← graphitemaster left14:19
→ graphitemaster joined14:19
MattMaker idly, I wonder if there aren't a bunch of things you can ask git to do that literally make no sense, but it will do them anyway14:19
I know I've gotten a few repos into inconsistent states where everything was an error14:20
← blahdodo left14:20
→ mizu_no_oto joined14:20
← _5kg left14:20
→ Kayra joined14:20
MattMaker doesn't happen anymore, but only because the things I do now are less random and more intentional14:20
→ _5kg joined14:20
grawity there *are* some sanity checks in the 'porcelain' commands, but Git is not one of those programs that would say "thou shall not touch the repository contents directly", so yes, it's possible to corrupt repositories in various ways14:20
thegorn yea, i dunno, i'm actually very comfortable with git in the sense that if i have ever made a commit with objects, i feel safe that i can always get them back14:21
grawity that too14:21
→ j0ran joined14:21
MattMaker I have wandered off into some brambles before14:21
thegorn but you can get into some weird states if you don't know what's going on, detached heads, etc.14:21
MattMaker usually because I misunderstood what an action would mean14:21
thegorn or pushing to checked out branches can get a little confusing, heh14:21
MattMaker thegorn: I feel that way, but only because I keep an extra working copy around that I can flee to if I screw mine up14:22
← sebhoss left14:22
MattMaker I use a lot of external, tree-style visual merge/diff tools14:22
looking at entire projects from above14:22
→ insejn joined14:23
MattMaker so it's handy to check out two branches locally and compare them with that14:23
← vipzrx left14:23
MattMaker github made a feature like this recently, but it's still just a flat list14:23
thegorn yea, i get that, there's a handy tool on linux called git new-workdir that lets you create a new git repo clone but it is linked to the other repo14:23
← Snetty left14:24
thegorn i.e. if you change a branch in one it changes in the other (no need to push, etc.)14:24
→ Droolio joined14:24
MattMaker I get you, that's neat14:24
← hyperboreean left14:24
→ Craig910 joined14:25
→ chuckharmston joined14:25
← Kayra left14:25
MattMaker I'm likely to use that all the time if I can get it on OS X14:25
← Kabaka left14:26
→ vipzrx joined14:26
→ Snetty joined14:26
MattMaker because I have this constant dual-working-copy workflow14:26
← vipzrx left14:26
thegorn MattMaker: oh i thought you were on windows14:26
MattMaker: you can definitely do it on os x, it is just a shell script in the contrib/ directory in the git source code14:26
put it in your path, and then git new-workdir should work fine14:27
it uses a lot of symlinks in .git/ which is why i doubt it'll work well on windows14:27
← netiul left14:28
thegorn MattMaker: https://github.com/git/git/blob/master/contrib/workdir/git-new-workdir14:28
MattMaker whoa, hello there, "git gui". it's doing all the things SourceTree does, but with slightly less usability and discoverability14:28
'Visualize All Branch History'14:29
thegorn MattMaker: yea, it isn't a 100% replacement for sourcetree, but i do like how you can stage lines with it14:29
MattMaker *nod*14:29
→ vipzrx joined14:29
thegorn and it is builtin, which is a plus if you are on another machine14:29
← vipzrx left14:29
MattMaker definitely not ovious where to grab the split panes, in these windows :(14:30
→ kriks joined14:30
MattMaker *obvious14:30
→ harobed joined14:30
← kcj left14:30
← spudz_ left14:30
MattMaker not bad, though.14:31
← Devastator_ left14:31
→ Devastator_ joined14:31
MattMaker thanks, I'll check out get-new-workdir. no pun intended.14:31
← mizu_no_oto left14:32
← Criztian left14:32
→ macabre joined14:33
→ slopjong joined14:33
← macabre left14:33
→ CheckDavid joined14:33
← mooperd left14:33
thegorn MattMaker: yea, tcl/tk doesn't make the most beautiful gui's14:33
but they are cross platform at least14:34
→ mooperd joined14:34
→ spudz joined14:34
→ Kayra joined14:35
[emmajane] → emmajane14:36
→ dhruvasagar joined14:36
← spheros1 left14:36
→ Kabaka joined14:36
→ nzjrs joined14:36
gitinfo set mode: +v14:37
nzjrs Hey all. I'm halfway through resolving a complicated merge and want to do something else. How can I pause/stash the merge and resume resolving it later?14:38
← jason237 left14:38
← Craig910 left14:39
→ joshbetz joined14:39
→ jason237 joined14:39
← nnyk_ left14:40
grawity the simplest way that comes to mind right now is to just `git clone` the repository again, and do the changes there14:40
moritz !backup and work in the backup14:40
gitinfo Worried about your data while trying out stuff in your repo? To back up commit history on all branches/tags: `git clone --mirror`. To backup everything, including work tree and staging area: `tar cf repo-backup.tar repodir`. Or do your experiment in a throwaway clone instead. See also http://sethrobertson.github.com/GitBestPractices/#backups14:40
→ Dougie187 joined14:42
→ nnyk_ joined14:42
→ AWinterman joined14:43
← j0ran left14:43
← oriba left14:43
→ huli_ joined14:43
→ tty-1 joined14:44
← sagittarian left14:44
← Kayra left14:44
→ mfeltner joined14:45
nzjrs OK, hmm. I thought that sort of use-case might be supported in a simple way. I will just clone it again14:45
→ DolphinDream joined14:46
→ natechan joined14:46
← hholtmann left14:46
← h_kon left14:46
cbreak-work nzjrs: you can clone locally14:47
DolphinDream howdy14:47
cbreak-work nzjrs: or use the new workdir contrib script from git.git14:47
← mfeltner left14:47
→ m0ltar joined14:47
nzjrs cbreak-work: ok, I'll check that out14:47
thanks all14:48
→ vipzrx joined14:48
DolphinDream why is it that when i rename files with mv oldName newName the git status shows file deleted and file added.. but when i do the git mv newName oldName the git status shows file as renamed ?14:48
← vipzrx left14:48
← quetz^^ left14:48
→ quetz joined14:48
→ jwmann joined14:49
→ slopjong_ joined14:49
moritz DolphinDream: because in the first place, you didn't tell get to stage the deletion of the file14:49
← asQuirreL left14:49
DolphinDream oh :) so once i stage it status is going to reflect a rename ?14:49
arand DolphinDream: becase git mv stages the deletion and addition, and git runs rename detection against the staging area14:49
moritz so there's no renaming, just a copy14:49
DolphinDream oh.. i see. wonderful14:49
moritz git doesn't track renames, it just infers from the changes if a rename happened14:50
→ pandeiro joined14:50
← twoface88 left14:50
← khushildep left14:50
hakril_ → hakril14:50
← Search12 left14:51
← juril left14:51
← slopjong left14:51
← Spyro left14:51
→ Spyro joined14:51
← ezekielnoob left14:52
← jaseemabid left14:52
→ treaki joined14:52
← donnib left14:52
← isleshocky77 left14:52
→ phile joined14:52
→ AaronMT joined14:53
→ bamj0rb joined14:53
gitinfo set mode: +v14:53
← Gavrisimo left14:53
← Snetty left14:53
→ ojacobson_ joined14:54
→ Snetty joined14:54
← ojacobson_ left14:54
→ Gavrisimo joined14:54
← freestyl3r left14:55
← Guest86971 left14:56
← bgerber left14:56
← nzjrs left14:56
→ tvw joined14:56
← dyng left14:56
→ tadasZ joined14:57
← AWinterman left14:57
→ sohail joined14:57
← ojacobson left14:58
← Gavrisimo left14:58
← mdiers left14:58
→ jburcsik joined14:59
→ StarFire2 joined14:59
→ el3ctron joined14:59
← donmichelangelo left15:00
→ thiago joined15:00
← soee left15:00
→ donmichelangelo joined15:01
→ joelrobichaud joined15:01
→ dyng joined15:01
→ PaulCapestany joined15:01
→ bgerber joined15:01
← tadasZ left15:02
→ vipzrx joined15:02
→ RogerWS joined15:02
← Roots47 left15:03
← Ange_blond left15:03
→ robertnix joined15:04
→ pmdarrow joined15:04
← rethus left15:04
← joelrobichaud left15:05
→ tadasZ joined15:06
→ mizu_no_oto joined15:07
→ mikecmpbll joined15:07
→ macabre joined15:08
← kbart left15:09
← rouf left15:09
← Advocation left15:09
→ johnnyfuchs joined15:09
← huli_ left15:10
→ rouf joined15:10
← rob_w left15:11
→ hholtmann joined15:11
← hholtmann left15:11
→ hholtmann joined15:12
→ marcosRz joined15:14
Xiti` → Xiti15:14
→ SynthesizerPatel joined15:14
← robertnix left15:15
marcosRz Hi guys. I have a single repo with n projects. I was wondering if there is a way to create and separate this repo in n repos in such a way that the changelog is not lost for each project.15:15
→ Kayra joined15:15
→ robertnix joined15:15
moritz marcosRz: man git-filterbranch15:15
marcosRz but the projects are not in different branch, they are all in master15:16
moritz sorry, man git-filter-branch15:16
gitinfo the git-filter-branch manpage is available at http://jk.gs/git-filter-branch.html15:16
grawity doesn't matter15:16
moritz marcosRz: yes, I got that.15:16
grawity create a separate clone for each project15:16
→ juril joined15:16
grawity or separate branches, same thing15:16
moritz marcosRz: the tool filters a branch, for example by directory15:16
grawity then filter each clone or branch to just contain files from a specific subdirectory15:16
moritz there are lots of tutorials on splitting git repos on the internet15:16
google them if you don't like the man page :-)15:16
→ Roots47 joined15:16
← Kayra left15:16
marcosRz great, thanks :D15:17
grawity specifically, --subdirectory-filter15:17
← freimatz left15:17
← SHyx0rmZ left15:17
→ Kayra joined15:17
← mooperd left15:18
← robertnix left15:18
→ mfeltner joined15:18
← arvind_khadri left15:18
→ macabre^ joined15:18
← matanya left15:19
← timthelion left15:20
→ robertnix joined15:20
← dsdeiz left15:20
← macabre left15:21
← robbyoconnor left15:21
→ joelrobichaud joined15:21
→ SHyx0rmZ joined15:21
← salamanderrake left15:21
→ whaley joined15:21
← Kayra left15:22
→ matteo joined15:23
→ dharmaturtle joined15:23
→ mmitchell joined15:23
→ AWinterman joined15:24
← lite_ left15:24
→ petey_ joined15:24
→ lite_ joined15:25
→ acharlieh joined15:26
← whaley left15:26
← papegaaij left15:27
→ ojacobson joined15:28
→ whaley joined15:28
→ salamanderrake joined15:29
← wet88 left15:30
← lite_ left15:30
← scampbell left15:30
← girfas left15:30
→ john2x joined15:30
→ dort joined15:31
← ni__ left15:31
← dhruvasagar left15:33
← nfroidure left15:33
→ petey__ joined15:34
→ danman joined15:34
→ bwreilly joined15:34
→ herdingcat joined15:35
← mikecmpbll left15:35
← john2x left15:35
← corygibb_ left15:35
← arand left15:36
← petey_ left15:36
→ k4rtik joined15:37
→ timthelion joined15:38
dharmaturtle Hi, I'm having trouble removing a sensitive file from Windows - I'm getting a permisison error, I think. http://pastebin.com/8jzTDQCa is my attempts to follow https://help.github.com/articles/remove-sensitive-data15:38
← AWinterman left15:38
→ CKyle joined15:39
← bwreilly left15:39
← nnyk_ left15:40
milki dharmaturtle: fix the permission problem then? its all local to your computer15:41
→ rouf_ joined15:41
bamj0rb Ewww, spaces.15:41
moritz spaces and ' in the path15:41
what could possibly go wrong? :-)15:41
dharmaturtle yeah that was a pretty bad idea x_x15:41
→ nnyk_ joined15:42
← jburcsik left15:42
dharmaturtle Well, I can delete the files from explorer, so I'm not sure why I'm getting a permission error. Maybe run it as Admin?15:42
how should I specify where the file is, with an absolute path like the first or a relative like the second?15:42
← Norm_ left15:43
← lb1a left15:43
cbreak-work dharmaturtle: windows has its issues with open files.15:43
→ Davey joined15:44
← rouf left15:44
→ lavaman joined15:45
→ kevwil joined15:45
← macabre^ left15:45
→ macabre joined15:46
dharmaturtle :( well, I closed everything and I'm still getting that error with "/C#/plsremoveme.cs" as the path15:47
← macabre left15:47
→ infynyxx joined15:47
→ macabre joined15:47
bamj0rb dharmaturtle: You may want to remove that leading /... I'm not sure if Git would consider that the root of the repo or root of the file system.15:47
→ Gavrisimo joined15:48
bamj0rb Which OS? Certainly working outside of your "home" is a bit non-standard (unless you are an admin or it's just a data partition).15:48
Also, Windows is just PITA.15:48
→ shinnya joined15:49
cbreak-work C#? hmm :/15:49
← StarFire2 left15:49
dharmaturtle yeah, the more I program, the more I become a linux fan15:50
→ FRANGIUL joined15:50
→ Advocation joined15:50
← War2 left15:50
← FRANGIUL left15:50
bamj0rb Protip: it is a bit awkward to name a directory after a particular data type. :) Just my /pedantic.15:50
← dyng left15:51
→ FRANGIUL joined15:51
bamj0rb Apparently you can change that directory (.git-rewrite) with an option. So you could try moving that somewhere else, like maybe $TMP (though that may be a security consideration).15:51
← tadasZ left15:52
→ Streward joined15:52
← FRANGIUL left15:52
MattMaker dharmaturtle: it could also be an antivirus program locking the file to "scan" it15:52
dharmaturtle Removing the leading "/" seems to work. I'm not getting an error message now, but it doesn't seem to be updated when I visit the repo on github. Is it cached or something? http://pastebin.com/4g4CZtX015:53
← timthelion left15:54
bamj0rb You will have to delete the repo on GitHub and re-push (Danger Will Robinson! This will affect all other users of it).15:55
dharmaturtle @_@ wait really15:55
I have to nuke the entire repo?15:55
→ keller joined15:55
→ King_Hual joined15:56
bamj0rb If you want to eliminate all traces, yes. Even after rewriting the history, the objects will still exist in the remote. They will still be accessible by SHA1.15:56
→ Xetius joined15:56
← Sendoushi left15:56
→ guide joined15:56
← CKyle left15:56
Xetius Hi15:56
dharmaturtle I'm not worried about the hash, it's not *that* sensitive15:56
bamj0rb If you had access then you potentially garbage collect them away, but I don't think you have that option with GitHub. So the easiest way (maybe the only way) is to delete the GitHub repo and create a new one.15:56
Xetius Can I use rebase if I have pushed to my main repository?15:57
gitinfo set mode: +v15:57
bamj0rb dharmaturtle: You'll need to push --force in order to change the ref in the GitHub repo (e.g., origin/master for you). This will still affect users (if anybody already has this repo cloned then you can't force them to clean their repo).15:57
gitinfo set mode: +v15:58
→ AWinterman joined15:58
ojacobson so it turns out that passing an entire git repo's object graph (not just commit graph, all the tags & trees & blobs too) to dot can take a looooong time to render15:58
so far, all weekend15:58
bamj0rb Xetius: Depends. As a rule, you shouldn't. You can, but it could cause problems. Any clones will need to also rebase. So that means any other users of your repo would need to.15:58
ojacobson: xD15:58
ojacobson 88k objects15:59
→ CKyle joined15:59
ojacobson well15:59
88k lines of .dot file15:59
-Nshape=point for obvious reasons15:59
dharmaturtle bamj0rb: Oh *duh*. Thanks very much x_x pretty new and stupid to all this15:59
Xetius so the only way for everyone else to pick up changes would be to do the rebase themselves as well?15:59
ojacobson https://gist.github.com/ojacobson/98d624f54036907360fc if you're bored16:00
→ lite_ joined16:00
← marcosRz left16:00
Xetius or, if the fetch/merge would they pick up the rebase?16:00
bamj0rb dharmaturtle: No worries. It takes some practice to wrap your head around it.16:00
→ arietis joined16:01
bamj0rb Xetius: They will get the new commits from the rebase, but they won't lose the original commits. They'll end up merging the two together most likely.16:01
→ ZujkisNx joined16:01
Xetius bamj0rb: thanks16:01
and arse16:01
bamj0rb Xetius: The other solution is to rebase, push -f, and then tell everybody else that they need to reclone.16:01
← macabre left16:01
→ bwreilly joined16:02
→ macabre joined16:02
Xetius ok, thanks. That just broke everything I spent today writing :)16:02
bamj0rb Xetius: They can take care to preserve any unpushed changes of their own that depend on the rebased commits, but it's an advanced maneuver so newbies will struggle with it and it probably won't be worth trying it.16:02
← macabre left16:03
← ikrabbe left16:03
→ nwertman joined16:04
→ hyperboreean joined16:04
→ tadasZ joined16:05
← chuckharmston left16:05
→ dhruvasagar joined16:07
→ jburcsik joined16:07
← el3ctron left16:07
← mlechner left16:07
→ rc111 joined16:08
→ timthelion joined16:08
→ kukks joined16:10
← natechan left16:10
← Snibble left16:10
→ natechan joined16:10
← tty-1 left16:10
→ asQuirreL joined16:11
→ FrodoTheHobbit1 joined16:12
← juril left16:12
→ obiwahn joined16:12
obiwahn hey there16:13
which hook is triggered wehn i do a git pull?16:13
← petey__ left16:13
← cbreak-work left16:13
← shadok left16:14
→ petey_ joined16:14
→ shadok joined16:14
← shadok left16:14
→ shadok joined16:14
→ shruggar joined16:14
milki hm, probably post-merge?16:15
← Sonderblade left16:15
← FrodoTheHobbit left16:15
→ Cromulent joined16:15
milki lol, its even explicit in the manpage16:15
Dougie187 When using a submodule that was created with the -b flag, whenever the submodules commit changes you have to commit to the repo using the submodule so that a clone gets that commit?16:16
← hike left16:16
← tadasZ left16:16
← tma996 left16:16
→ WaffleCat joined16:16
milki the later is always true, i dont know what you mean by -b flag16:16
grawity milki: git submodule add -b <branch>, of course16:17
← arietis left16:17
Dougie187 `git submodule add -b <branch_name> <url> <path>` sets the submodule up to track a branch16:17
← Xetius left16:17
milki o, i havent learned the behaviour of branch tracking yet16:18
→ Kayra joined16:18
← ThomasLocke left16:18
→ macabre joined16:18
→ sanguisdex joined16:18
Dougie187 It seems like even if you set it up like that, it still needs a commit in the "super project" to force a clone to track the latest commit on the submodule.16:18
→ SeptDePique joined16:19
→ joelm joined16:19
obiwahn is post-commit called when post-merge is used?16:20
ojacobson obiwahn: !xy16:20
gitinfo obiwahn: Woah, slow down for a bit. Are you sure that you need to jump to that particular hoop to achieve your goal? We suspect you don't, so why don't you back up a bit and tell us about the overall objective...16:20
milki and read man githooks16:21
gitinfo the githooks manpage is available at http://jk.gs/githooks.html16:21
→ kenperkins joined16:21
← infynyxx left16:22
obiwahn ok we want to write the hash of the current checkout to a file whenever the repo is changed and wonder which hooks are needed. i suggested post-commit and post-checkout.16:22
← Kayra left16:22
← GeertV left16:22
→ psyl0n joined16:22
→ sebhoss joined16:23
→ GeertV joined16:23
sanguisdex I have some new changes on aa feature branch that I would like to merge in behind the changes in head (so that the head changes are not over written by the changes here, should I use rebase?16:23
← Gavrisimo left16:23
→ tadasZ joined16:24
→ Robarelli joined16:24
→ infynyxx joined16:24
→ nfroidure joined16:24
← psyl0n left16:24
→ psyl0n joined16:24
→ chuckharmston joined16:24
→ leeN joined16:25
→ arietis joined16:26
→ LuxuryMode joined16:27
rm_work|away → rm_work16:27
← GeertV left16:27
← sebhoss left16:28
→ hwrd|work joined16:28
← jnavila_ left16:28
ojacobson obiwahn: run `git describe` in your build/packaging/deployment script.16:28
Git's not good at modifying the work tree except as described by versioned contents16:28
(or git rev-parse HEAD, if you want the bare hash, but `git describe` is generally more useful)16:29
← rath left16:29
← Misan left16:29
bamj0rb sanguisdex: Rebasing might be an option, though I guess that depends on what "head" you're referring to, and whether other people have it.16:30
→ m0viefreak joined16:32
← sohail left16:32
bamj0rb sanguisdex: As a rule, your feature branch should only do One Thing(tm), and if you're merging that thing then it should be good. If you don't want to merge everything that it did then you should probably rebase the feature branch.16:32
← Ozch left16:32
→ sohail joined16:33
← sohail left16:33
→ sohail joined16:33
← asQuirreL left16:33
→ Ozch joined16:33
qsusOFFcs → qsuscs16:33
→ zhost joined16:34
→ asQuirreL joined16:34
← joel_falcou left16:35
→ sebhoss joined16:36
bamj0rb sanguisdex: And you can always add new commits to undo those things before merging... It depends on what you want history to reflet.16:37
reflect*16:37
sanguisdex thanks bamj0rb you have given me plenty to think of16:38
→ Snibble joined16:38
→ therealkoopa joined16:39
← hwrd|work left16:39
← halu_ left16:42
→ ngladitz joined16:42
← ngladitz left16:42
→ ngladitz joined16:42
← lite_ left16:42
← etcetera left16:42
← johnkevinmbasco left16:43
→ lite_ joined16:44
← Balliad left16:45
→ mdiers joined16:46
← w0ng left16:47
→ SocalNick joined16:48
← JeremyWei left16:49
→ shoshin joined16:50
← AWinterman left16:50
→ halu joined16:51
← m0ltar left16:51
→ AWinterman joined16:51
← shruggar left16:52
→ scampbell joined16:52
← Davey left16:52
→ h0rrorvacui joined16:53
← user258467 left16:53
← User458764 left16:53
← asQuirreL left16:54
← griotspeak left16:54
→ mooperd joined16:54
← halu left16:55
→ jnl joined16:57
← robertnix left16:57
← SeptDePique left16:58
← Snetty left16:58
jnl is there a script-friendly way of doing 'git branch --contains'? (ie without the *-marker, detatched head etc)16:59
← ehmatthes left16:59
← bwreilly left17:00
← paulweb515 left17:00
← dhruvasagar left17:00
→ davividal joined17:01
→ Skrypter joined17:01
← Anticom left17:01
← calcifea left17:02
→ halu joined17:03
← infynyxx left17:03
← rhombus left17:04
→ kermit joined17:04
→ russ_ joined17:05
→ ciampix joined17:05
→ aberrant joined17:06
← wrouesnel left17:06
→ infynyxx joined17:06
← mizu_no_oto left17:06
ojacobson jnl: you can assemble one; I'd use `git branch --contains HASH | cut -c 3-'17:08
← halu left17:09
← nnyk_ left17:09
← ewnd9 left17:10
← xpe left17:10
← sebhoss left17:10
→ ehmatthes joined17:11
→ anderse joined17:11
← psoo left17:11
← sohail left17:11
jnl yeah, and then "grep -v '(detatched'", and then i thought maybe there is a better way :) but ill make do17:12
→ dhruvasagar joined17:12
→ imachuchu joined17:12
← CKyle left17:12
→ hwrd|work joined17:12
grawity maybe hack something with 'show-ref' and 'rev-list' or some other commands17:12
→ halu joined17:12
→ devbug joined17:12
grawity I don't know what's closest to --contains :(17:12
→ SeptDePique joined17:13
← irqq left17:14
→ mizu_no_oto joined17:14
→ rethus joined17:14
ojacobson jnl: I can make something up along the lines grawity suggests if you really want me to17:14
← rurufufuss left17:15
→ cff_ joined17:15
→ m0ltar joined17:15
jnl ojacobson: thanks! but no worries :) --contains will do17:15
→ sohail joined17:16
→ Shusshu joined17:16
← Shusshu left17:16
← Katafalkas left17:16
→ Kayra joined17:16
ojacobson https://gist.github.com/ojacobson/019203856fef10ac9fb8 (untested) is probably a tolerable starting point17:17
→ etcetera joined17:17
→ maestrojed joined17:17
← SeptDePique left17:17
jnl cheers17:17
← mishok13 left17:17
ojacobson turns out --shell is not what you want there17:18
reasons why left as an exercise ;)17:18
also if you only want branch names, not ref names, you'll need to tweak the echo17:18
→ royho joined17:18
← Kayra left17:18
jnl yeah i could add a pattern to for-each-ref17:20
→ matanya joined17:20
jnl oh, sorry, misread17:20
← mmitchell left17:20
← koda left17:20
jnl no full names are only a plus, its going into other git-commands anyway :)17:21
→ mmitchell joined17:21
→ bwreilly joined17:21
← pfFredd left17:22
→ CKyle joined17:23
← teran left17:23
← joshbetz left17:24
→ teran joined17:24
← h0rrorvacui left17:24
→ joshbetz joined17:24
← matanya left17:25
DolphinDream somewhat (not) of a git question:.. how is git able to suggest options via tab completion ? i guess this is more of a unix command /tabcompletion thing.. but anyhow.. how can you make any application return options on tab completion . ideas?17:25
← thiago left17:26
← snql left17:26
→ twoface88 joined17:27
ojacobson DolphinDream: it's shell-dependent17:27
read `man bash`, `man zsh`, or whichever man page is appropriate for your shell17:27
→ lolfrenz joined17:27
grawity or better, read the completion definitions themselves17:27
if the shell uses those (bash and zsh do)17:28
→ matanya joined17:28
← tvw left17:28
→ tvw joined17:28
← teran left17:29
→ dvide joined17:29
← kukks left17:29
← dhruvasagar left17:29
← Roots47 left17:29
← Cassiel left17:30
← donmichelangelo left17:30
→ donmichelangelo joined17:31
← shoshin left17:31
← matanya left17:32
→ erichynds joined17:33
→ matanya joined17:33
← ikoo left17:33
← rethus left17:34
← Asenar left17:34
← PaulCapestany left17:34
← AWinterman left17:34
→ PaulCapestany joined17:36
← whaley left17:36
→ GeertV joined17:36
→ m8 joined17:39
← Guest32821 left17:39
→ billy_ran_away joined17:41
→ boombatower joined17:41
→ Sonderblade joined17:42
← Sonderblade left17:42
→ CorySimmons joined17:42
→ War2 joined17:42
← hahuang65 left17:42
CorySimmons I tried to rebase -i to squash some commits and I ended up with another branch somehow? http://i.imgur.com/Vdr2cU3.png17:42
How can I fix this so there aren't 2 branches anymore?17:43
→ Sonderblade joined17:43
CorySimmons And what happened anyway?17:43
grawity basically, when you rewrite history, you aren't modifying the old commits (that's pretty much impossible) but rather creating new, rewritten ones17:44
← SocalNick left17:44
grawity after the rewrite, git points the branch to the new latest commit, and leaves the old ones collecting dust17:44
(dangling commits get garbage-collected after a while)17:44
but the old ones are still reachable by their commit-hash and the like17:45
← mklappstuhl left17:45
grawity so if you somehow give your GUI both the new and old histories, it shows them just like that17:45
← matanya left17:45
→ paulweb515 joined17:45
→ matanya joined17:46
grawity although I have no idea what command you ran to open that graph17:46
CorySimmons gitx17:46
grawity ah17:46
maybe you did that while having checked out the 'old latest' commit, or maybe it found that in some other way, I dunno17:46
→ SocalNick joined17:46
← billy_ran_away left17:47
← shinnya left17:47
→ kukks joined17:47
← kukks left17:47
→ kukks joined17:47
CorySimmons Is there any way to combine them from this point?17:49
It's weird, it usually doesn't do this but sometimes it does17:49
→ nnyk_ joined17:49
→ teran joined17:49
→ hyperair joined17:49
→ Asenar joined17:49
→ shoshin joined17:50
imachuchu CorySimmons: git reset changes which commit your ref points to, and a branch is just a pointer to a commit anyway, so checkout the branch you want to move and reset --hard it to where you want it to point to. Of course if you haven't seen it: !rewrite17:51
gitinfo CorySimmons: [!rewriting_public_history] Rewriting public history is usually bad. Everyone who has pulled the old history have to do work (and you'll have to tell them to), so it's infinitely better to just move on. If you must, you can use `git push -f` to force (and the remote may reject that, anyway). See http://goo.gl/waqum17:51
← undecim left17:51
→ mikecmpbll joined17:51
→ nafg joined17:51
← tadasZ left17:52
→ spheros joined17:53
→ stodan joined17:53
← matanya left17:53
→ juanmabc joined17:53
grawity CorySimmons: you don't want to combine them, just think about it17:53
BlaDe^ I have a git submodule which i added to .gitmodules however it doesn't show up in git submodule status17:53
← vtunka left17:53
BlaDe^ what else do I need to do?17:53
grawity BlaDe^: git submodule init17:54
BlaDe^ tried that, didn't do anything17:55
← takamichi left17:55
→ matanya joined17:55
→ _iron joined17:56
← arjun_ left17:56
← psyl0n left17:56
→ mishok13 joined17:56
← tvw left17:56
← infynyxx left17:57
→ tadasZ joined17:57
← matanya left17:59
→ abbycabs joined17:59
→ matanya joined18:01
→ AWinterman joined18:02
→ matanya_ joined18:03
Dougie187 BlaDe^: `git submodule update --init`?18:03
BlaDe^ Dougie187: nope :( should I run that in my git repo or the submodule? neither seemed to work18:04
Dougie187 If you're repo18:04
→ SpNg joined18:04
Dougie187 Why don't you just `git submodule add` rather than manually editing .gitmodules?18:04
s/If/in18:04
BlaDe^ it didn't do anything18:04
CorySimmons grawity imachuchu: Thanks guys. Got it right with git push -f. I understand the cons associated with it, but it's a local thing anyway so no one is working from it (safe to force push)18:04
grawity aha18:04
→ shwaiil joined18:05
grawity so maybe gitx was just showing origin/master oddly unlabeled18:05
shwaiil hi18:05
Dougie187 BlaDe^: using `git submodule add` didn't do anything?18:05
← matanya left18:06
← Dreamer3 left18:06
← petey_ left18:06
→ marhaban joined18:06
← juanmabc left18:06
→ chirayuk joined18:07
← rafeca left18:07
CorySimmons grawity: Yeah that too, had to restart gitx and everything looked fine after the force push18:07
So weird18:07
BlaDe^ Dougie187: the problem is I added it myself before so I get: 'library/merchant-sdk-php' already exists in the index18:07
Dougie187 BlaDe^: why not remove .gitmodules, and add it using the infrastructure.18:07
?18:07
→ ntio joined18:07
Dougie187 Though, you have to remove the directory too.18:07
→ beastd joined18:08
→ thiago joined18:08
← twoface88 left18:08
Dougie187 BlaDe^: basically, clean the index. And use `git submodule add`18:08
→ dreiss joined18:08
→ faBimon joined18:09
← CorySimmons left18:10
→ a-l-e joined18:11
BlaDe^ Dougie187: i just removed the .git folder and commited it to my repo18:11
submodules are awlays a slight headache18:11
← marhaban left18:11
Dougie187 They aren't that bad.18:11
Though, if you don't like submodules, and you want to keep the "subproject" separate. You should look into subtree18:11
← eaj left18:11
← VictorCL left18:12
→ regreddit joined18:13
→ snql joined18:14
← lolfrenz left18:14
→ eaj joined18:14
← pmdarrow left18:14
← kewah left18:15
→ juanmabc joined18:16
← quetz left18:16
→ Kayra joined18:16
emmajane → emmajane|food18:16
← stodan left18:17
→ takamichi joined18:18
← takamichi left18:18
← kukks left18:19
← eaj left18:19
← echo-area left18:20
→ infynyxx joined18:20
→ tureba_ joined18:20
← Kayra left18:21
← cff_ left18:23
← tureba left18:24
→ kesha joined18:24
→ tty-1 joined18:24
→ ahmedelgabri joined18:26
← mooperd left18:26
→ Gabri joined18:26
← dreiss left18:26
→ hahuang65 joined18:26
← Advocation left18:28
← ahmedelgabri left18:28
← galaris left18:29
→ iamjarvo joined18:29
matanya_ → matanya18:29
→ griotspeak joined18:29
→ mooperd joined18:30
← iamjarvo left18:30
→ iamjarvo joined18:31
→ Bass10 joined18:31
→ Norm_ joined18:31
→ SocalNick_ joined18:31
← Exio4 left18:31
← mooperd left18:32
→ mooperd joined18:32
→ Exio4 joined18:32
← Gabri left18:32
→ ahmedelgabri joined18:33
← Exio4 left18:33
→ psyl0n joined18:33
← SocalNick left18:34
SocalNick_ → SocalNick18:34
← scampbell left18:35
→ takamichi joined18:35
← matanya left18:35
→ angasulino joined18:36
← kermit left18:36
→ foucist joined18:36
← ahmedelgabri left18:37
→ matanya joined18:37
→ Dreamer3 joined18:37
→ ahmedelgabri joined18:37
angasulino I'm looking for information on git subtree (the apparent replacement to submodules?) but I'm not finding much more than the man page, any good links that explain how to use it and how it works internally?18:38
← Cromulent left18:38
← rc111 left18:38
foucist when i do git rebase -i HEAD~5 and then change some of the picks to squash, and :wq , i end up not in any branch and it just says "Could not apply 786bdad... merge organization model" what do i do?18:38
Dougie187 angasulino: http://blogs.atlassian.com/2013/05/alternatives-to-git-submodule-git-subtree/18:38
← ahmedelgabri left18:38
→ pmdarrow joined18:39
Dougie187 foucist: First step. Read the rest of the error message. It stops after a patch application that fails. It expects you to clean it up (you can see what files need to be merged with `git status`). Second step, clean it up, and `git add` the cleaned files. Then final step is `git rebase --continue`18:40
← natechan left18:41
foucist Dougie187: yeah except there wasn't any files to merge, nothing shows up in git st, and git rebase --continue didn't do anything18:41
Dougie187 foucist: I think that's about as much help as I can be. But if you want to get back to where you were, you can do `git rebase --abort`18:42
→ shoshin_ joined18:42
← shoshin_ left18:42
→ sq-one joined18:43
← griotspeak left18:44
→ oojacoboo joined18:44
→ sarkis joined18:46
← jonatasnona left18:47
← dort left18:47
→ jonatasnona joined18:48
→ griotspeak joined18:48
← thelorax123 left18:49
→ mklappstuhl joined18:49
→ ewokchewy joined18:50
← SocalNick left18:50
← matanya left18:50
→ thelorax123 joined18:51
→ richwestcoast joined18:52
→ SocalNick joined18:52
→ whaley joined18:52
→ alexim joined18:52
→ whunt_ joined18:52
← freeed left18:52
whunt_ → whunt18:52
→ matanya joined18:52
→ freeed joined18:53
← jonatasnona left18:54
← shoshin left18:54
→ jonatasnona joined18:55
angasulino Dougie187, thank you!18:56
← timthelion left18:56
angasulino Dougie187, I had already read that article, though :)18:56
← ewokchewy left18:57
→ irqq joined18:57
→ LBV_User joined18:57
angasulino hates horizontal scrollbars when there is plenty of space on both sides...18:57
← alexim left18:57
→ petey joined18:58
→ timthelion joined18:59
← War2 left18:59
← TomyWork left18:59
← oojacoboo left19:00
→ oojacoboo joined19:01
← lavong left19:01
→ endiruna joined19:03
← kevwil left19:04
← eletuchy left19:04
← timthelion left19:05
← ZujkisNx left19:06
→ rouf joined19:07
← k4rtik left19:07
→ rafaqueque joined19:07
← r00tWearsRedHats left19:07
← hoschi left19:07
← rouf_ left19:09
← mikecmpbll left19:10
→ alexim joined19:10
→ Seus joined19:11
→ tang^ joined19:11
← alexim left19:11
← davividal left19:11
→ kewah joined19:11
→ ZujkisNx joined19:12
← mizu_no_oto left19:13
→ mizu_no_oto joined19:14
→ thegladiator__ joined19:14
→ kukks joined19:14
← kukks left19:14
→ kukks joined19:14
→ Roots47 joined19:15
← slopjong_ left19:15
→ TGiFallen joined19:17
← juanmabc left19:17
← vipzrx left19:17
← takamichi left19:17
← matanya left19:18
→ matanya joined19:18
→ aard joined19:19
← minty left19:19
← nwertman left19:19
→ pr0ton joined19:20
→ dreiss joined19:20
→ sfz joined19:20
← pr0ton left19:20
→ nwertman joined19:21
← aartist left19:22
→ mintyFresh joined19:22
mintyFresh → Guest6882719:22
→ davividal joined19:23
→ MacWinner joined19:23
→ guzzlefry joined19:24
guzzlefry hello19:24
I have existing code that's not tracked in Git, I'd like to add it to a new branch of an existing repo. Is this sane/possible?19:25
← spheros left19:25
_ikke_ Is it related to the existing repo?19:25
guzzlefry yep19:25
jrn guzzlefry: you mean a new branch of the existing remote repo?19:25
→ hs366 joined19:25
→ spheros joined19:25
← spheros left19:25
← dropdrive left19:25
guzzlefry jrn: That's more accurate, yes.19:26
jrn guzzlefry: the maintainer of git does something like this with his "todo" branch19:26
_ikke_ guzzlefry: Do you want an 'emty' branch/19:26
jrn guzzlefry: within his clone of git.git, he makes a separate repo named Meta19:26
guzzlefry: like this: "cd git; git init Meta"19:26
guzzlefry: he sets it up to push to the same remote repo: "(cd Meta && git remote add origin ssh://.../git.git)"19:27
← ZujkisNx left19:27
jrn guzzlefry: then he makes commits and pushes them like usual from there19:27
canton7 guzzlefry, so you want to start tracking it within the existing repo?19:27
guzzlefry canton7: if possible.19:27
canton7 if so, you dont' want to create a new git repo inside your existing one19:27
just create a new branch as normal, paste in your files, add commit19:28
guzzlefry I might just create a new one if this is going to be a ton of work.19:28
ah19:28
jrn guzzlefry: it's possible (and likely) I misunderstood what you were looking for19:28
guzzlefry: do you mean that you just want a new branch with more files in it?19:28
guzzlefry: i.e., it should have the same files as "master" plus a few more?19:28
guzzlefry I have another developer who did some changes to existing code, basically a rewrite.19:28
canton7 I imagine that someone took a snapshot from a git repo, worked on it, and he wants to re-integrate that snapshot19:28
guzzlefry But he wasn't using Git. :/19:29
→ ZujkisNx joined19:29
jrn guzzlefry: ah, that's way simpler than what I was talking about :)19:29
guzzlefry oh :)19:29
canton7 yeah, new branch (branched off the point where the snapshot was made), paste in the files19:29
emmajane|food → emmajane19:29
jrn guzzlefry: I'd do "git checkout -b topic <branched off point where the snapshot was made"19:29
guzzlefry: then "git rm -fr ."19:29
guzzlefry: then extract the snapshot there and "git add ."19:29
→ rcu joined19:30
jrn guzzlefry: then "git diff --cached" to see if it looks sane19:30
guzzlefry: commit, and you're done19:30
← arietis left19:30
→ kevwil joined19:30
← m0ltar left19:31
→ eletuchy joined19:31
→ alvaro_o joined19:31
← endiruna left19:32
← oojacoboo left19:32
→ endiruna joined19:33
→ timthelion joined19:33
→ v_y joined19:34
v_y when i run git config, does it apply to the repository i'm working in or globally?19:34
→ pantsman joined19:34
← pantsman left19:34
→ pantsman joined19:34
v_y i want to set "git config core.filemode false" on a MS-FAT usb stick fs19:34
← iamjarvo left19:35
→ iamjarvo joined19:35
→ eruMelkor joined19:36
→ Devastator joined19:36
Tectu_ → Tectu19:36
← Devastator_ left19:36
← sarkis left19:37
← mklappstuhl left19:38
→ shruggar joined19:38
DolphinDream how do i undo the last push ? (assuming noone fetched latest push)19:38
← shruggar left19:38
← endiruna left19:39
← kevwil left19:39
→ endiruna joined19:39
jrn DolphinDream: when you pushed it should have shown the old and new locations of the pushed branch, like this: fcafab3..87ca8cfg19:41
DolphinDream: you can rewind the ref with a command like "git push origin +fcafab3:master"19:41
← v_y left19:42
← endiruna left19:42
→ endiruna joined19:43
→ iamjarvo_ joined19:43
← lite_ left19:43
← eruMelkor left19:43
→ mfulz joined19:45
endiruna how can i merge a file from a different commit with the current one. not the whole thing19:45
← iamjarvo_ left19:46
→ TomyLobo joined19:46
→ iamjarvo_ joined19:46
← iamjarvo left19:46
→ vipzrx joined19:47
← vipzrx left19:47
← tty-1 left19:47
→ tty-1_ joined19:47
← matanya left19:48
→ mklappstuhl joined19:48
← phile left19:49
→ phile joined19:49
foucist endiruna: git show <SHA of diff commit>:path/to/file/from/git/root > file19:50
git add file19:50
→ matanya joined19:50
→ btree joined19:50
_ikke_ That won't merge the file19:50
It will overwrite it with the one from the other branch19:50
foucist git amend19:50
← kukks left19:50
foucist oh19:50
guzzlefry This is more of a headache than I thought.19:51
← kesha left19:51
endiruna i can of course put it somewhere and than merge the two. just thought that there is some standard way in git19:51
→ dort joined19:52
bamj0rb A bad way might be to: git diff -- path > patch && patch -p1 < patch. You're probably just doing it wrong though. :)19:52
_ikke_ endiruna: It's not a very common usecase, because you normally merge history, not single files19:52
bamj0rb Git doesn't track "files". It considers the entire tree.19:52
With proper history, you would be able to merge just fine. If you need to merge just a single file then you're probably doing history wrong. :)19:53
← DonRichie left19:53
obiwahn mh my git started recently to show escape codes instead of colors19:53
← russoneto left19:53
obiwahn all other applications behave normal19:53
→ eruMelkor joined19:53
_ikke_ obiwahn: export | grep -i less19:53
← mooperd left19:53
obiwahn declare -x LESSCLOSE="/usr/bin/lessfile %s %s"19:54
declare -x LESSOPEN="/usr/bin/lessfile %s"19:54
← eruMelkor left19:54
obiwahn type less19:54
less is aliased to `less -R'19:54
→ m0ltar joined19:55
→ eruMelkor joined19:55
_ikke_ and pager?19:55
DolphinDream jrn, what i did is revert the last change .. git reset --hard lastCommitSHA .. then git push -f19:55
foucist endiruna: so yeah do what i suggested. . git show <SHA of diff commit>:path/to/file/from/git/root > file and then add that file to your current commit .. seems like the simplest approach?19:55
_ikke_ obiwahn: echo $PAGER19:55
→ mooperd joined19:56
obiwahn empty19:56
← Skrypter left19:56
→ Skrypter joined19:56
obiwahn happens on debian sid19:56
_ikke_ try unset less19:57
and then again19:57
bamj0rb foucist: That isn't merging. That is overwriting. You can just git checkout <SHA1> -- <path> for that.19:57
jnl endiruna: how about merge with --no-commit, and zap everything except that file?19:57
_ikke_ bamj0rb: same!19:57
← tty-1_ left19:57
_ikke_ bamj0rb: That will overwrite also19:57
bamj0rb _ikke_: That's what I said. ;)19:57
_ikke_ ah, I misunderstood you then19:57
bamj0rb Perhaps the shell no longer interprets colors (or ANSI colors)?19:58
← eruMelkor left19:58
obiwahn git config --global core.pager 'less' fixed it19:58
← abbycabs left19:59
CareBear\ the shell never interprets colors19:59
the terminal emulator does19:59
jnl endiruna: or maybe --squash, depending on how you want your history19:59
bamj0rb CareBear\: Fair enough. :)19:59
→ Snibble1 joined19:59
foucist _ikke_: bamj0rb: overwriting what?19:59
assume the other commit is in a different branch19:59
← Snibble1 left19:59
_ikke_ foucist: The file in your worktree19:59
foucist lets say the file doesn't exist in the worktree20:00
← Snibble left20:00
← matteo left20:00
_ikke_ Then git checkout is the easiest20:00
← spaceonline left20:00
_ikke_ git checkout <sha> -- <file>, as bamj0rb said20:00
→ Mars joined20:00
Mars → Mars`20:00
obiwahn mh20:00
foucist ah20:00
← Skrypter left20:00
obiwahn i would like to fix it for all20:00
user20:00
→ Skrypter joined20:00
← matanya left20:01
← sgambino left20:01
← Devastator left20:02
→ Devastator joined20:02
→ macabre^ joined20:02
→ matanya joined20:03
← Droolio left20:03
← ZujkisNx left20:03
→ ZujkisNx joined20:04
→ rc111 joined20:05
← parcs left20:05
← macabre left20:05
→ sgambino joined20:05
← endiruna left20:05
→ jnavila joined20:06
→ vladgh joined20:06
→ kevwil joined20:06
← mklappstuhl left20:06
→ tvw joined20:07
← Grum left20:08
← matanya left20:09
→ Grum joined20:09
← macabre^ left20:11
← Ozch left20:13
→ tombee joined20:13
→ crater2150 joined20:13
← tombee left20:13
→ tombee joined20:13
← erichynds left20:14
← hs366 left20:14
→ mmitchel_ joined20:14
→ avantgeek_ joined20:15
← hwrd|work left20:15
← avantgeek left20:15
guzzlefry Anyone accessing gitolite via Windows?20:15
← m0ltar left20:16
→ Ozch joined20:16
→ VictorCL joined20:16
→ erichynds joined20:16
← avantgeek_ left20:17
charon !anyone20:17
gitinfo Usually, it does not help to ask for someone specific to help you. Without knowing your specific problem, nobody knows if they can be of assistance. Please ask your questions and wait until somebody speaks up.20:17
guzzlefry ah, good call20:18
← mmitchell left20:18
← a-l-e left20:18
→ avantgeek_ joined20:18
cbreak guzzlefry: it works.20:18
gitolite is transparent as far as clients are concerned20:19
guzzlefry I have a dev working in Windows, and I having a bit of a time getting him setup to use gitolite as a remote.20:19
cbreak: yeah, it's getting ssh/git setup on Windows that seems to be the issue.20:19
← KillYourTV left20:19
cbreak installing git should be easy with that msysgit installer thing20:19
just select openssh and then treat it like a normal computer20:20
→ KillYourTV joined20:21
obiwahn the color issue really drives me mad20:22
→ aknewhope joined20:22
← aknewhope left20:22
→ lawltoad joined20:23
← eletuchy left20:23
→ vipzrx joined20:23
← richwestcoast left20:23
→ aknewhope joined20:23
→ richwestcoast joined20:24
← vladgh left20:24
← aknewhope left20:25
→ gkze joined20:25
→ macabre joined20:25
→ |Freeman| joined20:26
← Mars` left20:26
← kewah left20:26
→ vladgh joined20:27
→ silviof1 joined20:29
← btree left20:29
← joki left20:29
← silviof left20:30
← nnyk_ left20:30
→ zeograd joined20:31
→ ni_ joined20:33
→ revyd joined20:34
→ joki joined20:34
← mfeltner left20:34
← ZujkisNx left20:34
→ hwrd|work joined20:35
→ ajm203 joined20:35
← petey left20:35
→ petey joined20:35
← dort left20:35
→ nickpresta joined20:36
→ cff_ joined20:37
→ mfeltner joined20:38
→ Snibble joined20:38
← stephCoue left20:39
→ jefferai joined20:39
→ runa_ joined20:39
← Skrypter left20:39
→ Skrypter joined20:39
→ eletuchy joined20:39
→ schemanic joined20:39
← Skrypter left20:40
runa_ I'm not sure how, I but I managed to add an empty directory to my repo. I suspect I did *something* related to submodules. How can I do to remove it?20:40
→ Skrypter joined20:40
← petey left20:40
← Skrypter left20:41
→ petey joined20:41
→ Skrypter joined20:41
→ calcifea joined20:42
imachuchu runa_: do you mean remove it from this point on or remove it from all history?20:42
guzzlefry runa_: Check to see if it's a submodule. Run `git submodule` in the root directory.20:42
← abelsiqueira left20:43
→ kukks joined20:43
← kukks left20:43
→ kukks joined20:43
← davividal left20:43
← Skrypter left20:45
← revyd left20:46
← Squarism left20:46
→ axl_ joined20:47
← vipzrx left20:47
← frogonwheels left20:48
← gkze left20:49
mooperd If I have an old version. How do I reload my local clone20:49
or reclone20:49
or20:49
here____ → he_re20:50
← rcu left20:50
jrn mooperd: what do you have an old version of?20:50
← leeN left20:51
→ wet88 joined20:51
→ ilhami joined20:51
→ Skrypter joined20:51
ilhami Hey how do I push my whole project to google code server?20:51
from the terminal20:51
mooperd jrn, I have a local clone that is a really old version of some scripts and stuff.20:51
I want to grab the most recent versiion20:51
lov ilhami: ./toTheCloud.sh20:52
jrn mooperd: take a look at man git-pull20:52
gitinfo mooperd: the git-pull manpage is available at http://jk.gs/git-pull.html20:52
jrn mooperd: !pull = fetch + merge20:52
gitinfo mooperd: pull=fetch+merge (or with flags/config also fetch+rebase). It is thus *not* the opposite of push in any sense. A good article that explains the difference between fetch and pull: http://longair.net/blog/2009/04/16/git-fetch-and-merge/20:52
ilhami lov: wait.20:52
→ nnyk_ joined20:52
← pmdarrow left20:52
lov too late, it's already in the cloud20:53
it's gone20:53
whoosh20:53
← pandeiro left20:53
ilhami lov: I need to add all files to index first. how do I do that?20:53
mooperd jrn, thanks20:53
← shwaiil left20:53
jrn sure20:53
lov ilhami: not to be rude, but have you read any git tutorials yet?20:53
→ groglogic joined20:53
ilhami lov: I am looking for one20:53
lov ilhami: you might want to look at the first chapter or two of the !book20:53
gitinfo ilhami: There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable20:53
ilhami I will read vogellas. I don't like that book you are referring you...20:54
to*20:54
lov ok, up to you.20:54
→ aChatir joined20:54
← SocalNick left20:54
bremner vofellas?20:54
jrn ilhami: also see man gittutorial20:54
gitinfo ilhami: the gittutorial manpage is available at http://jk.gs/gittutorial.html20:54
→ revyd joined20:55
→ abbycabs joined20:55
ilhami you dont know Vogella bremner ?20:55
bremner nope.20:55
→ Mars` joined20:55
← bgerber left20:55
ilhami He has made a lot of tutorials. Search his name20:56
bremner sounds like a !blog20:56
gitinfo Blog posts, while helpful and informative, are quite often outdated, give bad advice, or are just plain wrong. Please don't rely solely upon them, or treat them as authoritative.20:56
→ SocalNick joined20:56
obiwahn * 0 /usr/bin/lv 80 auto mode20:56
wtf?20:56
lawltoad Hi. So if I have commit A that is the parent of commits B1 and B2 each a different branch. While on B1 if I check out File F from B2, the underlying file object still points to the object in B2 right? so if I remove B2, the file's individual history has a commitless object in it?20:57
bremner ilhami: it looks a bit java centric for my taste. but whatever makes you happy I guess.20:57
→ dropdrive joined20:57
ilhami bremner: yeah I like Java :D20:57
→ bgerber joined20:58
lov ilhami: just a bit of warning, vogella makes a lot of tutorials, but they're not necessarily GOOD tutorials.20:58
→ vipzrx joined20:58
lov he might suggest you do something that's out and out wrong, for example.20:58
→ Georg3 joined20:58
ilhami e.g.?20:59
do you have any examples?20:59
← longbeach left20:59
imachuchu lawltoad: "underlying file object" are you saying what git is storing?21:00
← mfeltner left21:00
← hwrd|work left21:00
← sohail left21:00
← tureba_ left21:00
→ tureba joined21:00
ilhami how do I add my project files to the folder I cloned. It's empty.21:00
→ yetet joined21:01
← MacGyver left21:01
← bamj0rb left21:01
← _iron left21:02
bremner ilhami: man gittutorial answers that question21:02
← Nickeeh left21:02
gitinfo ilhami: the gittutorial manpage is available at http://jk.gs/gittutorial.html21:02
ilhami I am looking there now.21:02
← roflin_ left21:02
→ hwrd|work joined21:02
ilhami I am trying to follow it.21:03
imachuchu lawltoad: checkout just modifies your working copy, nothing about your history will change unless you commit the changes (or more complex git commands)21:03
← skorgon left21:03
→ mfeltner joined21:03
ilhami I moved to the git folder. the problem is I don't know how to move my project. Do I need to export is as an archive file ?21:03
→ spaceonline joined21:04
← aChatir left21:04
→ MacGyver joined21:04
← MacGyver left21:04
→ MacGyver joined21:04
→ skorgon joined21:05
→ Nickeeh joined21:05
← LBV_User left21:06
→ pr0ton joined21:08
← iamjarvo_ left21:08
← pr0ton left21:09
→ iamjarvo joined21:09
← spaceonline left21:09
← acharlieh left21:09
→ roflin joined21:09
→ pr0ton joined21:09
← pr0ton left21:10
ilhami can anyone help me ? I really need to push this21:10
→ pr0ton joined21:11
← anderse left21:11
lov ilhami: the problem is that no one wants to guide you, step by step, through your own problem that is easily solvable if you follow intro tutorials, because then we will literally be writing a tutorial, just for you.21:11
except that it will only be applicable to ilhami21:11
← SocalNick left21:11
ilhami lov: I know what you mean. If you can just help me now I promise I will read about it.21:12
lov we're happy to answer specific questions if you're confused about something, but we're not going to Do Your Work For You21:12
ilhami It doesn't have to be step to step.21:12
→ tpollard joined21:12
ilhami I just need to know how to copy my project to the hidden git folder. THat's all. The rest I will handle myself.21:12
lov Are you having a particular problem (which is more granular than "hey guys how do I use git?")21:12
you uh21:12
you don't actually want to do that.21:12
that's not how git works.21:12
look.21:12
git add .21:13
git commit -m "My first commit!"21:13
git push21:13
ilhami yes I did that.21:13
DolphinDream what do u guys have to say about his idea: to check if any user pushes a change that breaks the code (e.g does not compile) a push to remote will trigger the pull on a build machine that pulls the latest changes, compiles the code and if there are any errors emails the last committer that they broke the code21:13
lov assuming that you have a repo set up21:13
ilhami I did that lov but that didn't work. Let me give you the "error"21:13
grawity DolphinDream: sounds like Travis or Jenkins21:13
ilhami lov: nothing to commit (create/copy files and use "git add" to track)21:13
lov DolphinDream: It sounds like you reinvented Jenkins!21:13
bekks DolphinDream: There are build servers like Jenkins doing exactly that.21:13
lov ilhami: is there any content in your repo?21:13
grawity DolphinDream: in fact, I've already been emailed by Travis when I broke a project once21:13
← mizu_no_oto left21:14
ilhami no thats what Im talking about. My repo is empty.21:14
→ rcu joined21:14
← petey left21:14
← vipzrx left21:14
lov ok, so, put content into there (NOT the .git directory, its parent directory)21:14
ilhami I'm using Eclipse and have my workspace in another folder.21:14
lov >_<21:14
so, see21:14
ilhami Wait. I think I have it now.21:14
lov this is why I don't want to help.21:14
ilhami Wait21:14
→ petey joined21:14
ilhami don't hate me but I think I have solved this.21:14
DolphinDream grawity, ?:) you mean i'm not the first one to think of this ?? :p21:14
what's Travis or Jenkins ?21:15
lov DolphinDream: if only there was a company that provided a way to search the internet for things.21:15
Perhaps you could invent that too?21:15
grawity DolphinDream: what's Google?21:15
DolphinDream lov, :) ok ok21:15
you guys are better than google21:15
ilhami lov: so my .git folder should be in the same as the project folder, right?21:15
bremner what is this, #jeopardy?21:15
ToxicFrog DolphinDream: the search keys you want here are "jenkins", "build server" and "continuous integration"21:16
DolphinDream google can give me 10 posts.. and 5 of them are crap.. you guys give me the BEST reference that i can trust ;)21:16
→ Kayra joined21:16
ilhami Use Bing instead. Much better. :D21:16
→ lite_ joined21:16
DolphinDream thx ToxicFrog21:16
lov ilhami: yes. your structure should be something like /content/ , /content/.git/21:17
ilhami wait lov21:17
→ DrShoggoth joined21:17
← fionnan left21:17
ilhami how do you make space in terminal?21:18
e.g. in a file name?21:18
lov ....21:18
come on bro :(21:18
ojacobson ilhami: quote the filename or escape the space21:18
and then read `man bash`21:18
(`man zsh`, whatever)21:18
ilhami ok21:18
lov either "like this" or like\ this21:18
ilhami sure21:19
osse if the file already exists you can tab-complete it and the shell will deal with it for you21:19
ojacobson lov: fwiw I habitually use 'single quotes' for filenames, because bash will "helpfully" expand vars and history markers inside "double quotes" and it's never what I want21:19
ilhami Nice osse . :) it did21:19
lov ojacobson: that's a fair reason.21:19
← tadasZ left21:20
ilhami and move folder is like this, right? mv21:20
its just mv?21:20
lov git doesn't understand the concept of folders.21:20
osse ojacobson: it has to expand at some point :P21:20
lov but sure, mv is mv from bash21:20
ilhami ok. good21:20
ojacobson osse: oh, sure, and when I want that, I use double quotes :)21:20
do not get into fights with me about bash tokenization, I will win every time /bravado21:20
← mmitchel_ left21:20
osse ojacobson: oh, I got the impression you didn't want expansion inside those either21:21
← lite_ left21:21
← Kayra left21:21
ojacobson no, I just don't want expansion most of the time, at all21:21
When I want it, I know how to get it21:21
→ acharlieh joined21:21
DolphinDream i assume that jenkins works well with gitolite21:21
Dougie187 In what sense?21:22
DolphinDream in the sense that when i push to remote server managed by gitolite, the jenkins system can be triggered to have builds made on remote build machines.. (I'm still reading about jenkins)21:23
ilhami guys I will solve my problem, don't worry and don't hate.21:23
lov DolphinDream: Jenkins has the ability to periodically poll source control21:23
ilhami Haters gonna hate.21:23
:D21:23
lov you can probably set up a trigger or something21:23
but, polling is usually good enough21:23
DolphinDream lov, what if i want the pull to happen immediately after a push ?21:24
is it? ok21:24
← iamjarvo left21:24
lov what some teams do is combine Gerrit with Jenkins, so that Jenkins will ensure that the proposed patch builds21:24
ilhami lov, look I am in my project folder now21:24
lov DolphinDream: sure?21:24
Dougie187 DolphinDream: It will happen once for each commit. gitolite isn't anything to jenkins21:24
ilhami I will clone from google source now. OK?21:24
GOogle code*21:24
→ iamjarvo joined21:24
Dougie187 Well, I forget if it actually tests each commit, but it does in buildbot. so I assume it does on jenkins too21:24
lov ilhami: you probably want to use add remote. Really, I'd VERY VERY VERY STRONGLY RECOMMEND that you read a few tutorials so that you HAVE ANY CLUE WHATSOEVER WHAT YOU ARE DOING.21:24
I also VERY VERY VERY strongly recommend that you make a copy of all of your source code, in case you screw it up.21:25
ilhami Hahahaha21:25
Yeah21:25
DolphinDream Dougie187, did not say gitolite means anything to jenkins.. not knowing yet how jenkins works i just wanted to make sure there are no issues here21:25
lov Because I'm getting the feeling that you are going to do exactly that21:25
← etcetera left21:25
lov because you're not bothering to learn what you're doing21:25
at all21:25
ilhami lov: I have used git before.. Just with Egit. Thats something else. Now I want to learn it with terminal21:25
← Jackneill left21:26
← leprechau left21:26
→ leprechau joined21:26
→ Fenikkusu joined21:26
osse DolphinDream: you could investigate if Jenkins has some sort of REST/URI/POST thing. Maybe you can do 'curl --data job=yesplease http//your-jenkins-instance/something/something'21:26
← GeertV left21:26
→ abelsiqueira joined21:26
→ nutron joined21:27
osse in a hook on the server I mean21:27
→ GeertV joined21:27
→ etcetera joined21:27
→ Kayra joined21:27
ilhami lov: code uploaded to Dropbox just in cae. Hehehehe21:27
case*21:27
DolphinDream osse, thx21:28
ajm203 DolphinDream: what are you using for source repo?21:28
DolphinDream git21:29
← iamjarvo left21:29
Fenikkusu I've a software (Magento) that is notorious for an upgrade consisting of starting with a fresh copy of the code. What is the best way to accomplish this on a git repo and not lose the past history? IE: Create New Branch, Remove All Files, Add New Files, Etc.21:29
DolphinDream what do u mean ajm203 ?21:29
→ SocalNick joined21:29
← foucist left21:29
ojacobson yes, you can trigger jenkins builds remotely21:29
writing a hook to do it on git-push is trivial21:30
the jenkins git plugin documents how it works21:30
(there's a URL that you can GET to cause jenkins to re-poll all the jobs for a given repository)21:30
→ iamjarvo joined21:30
ajm203 DolphinDream: gitolite can operate as a remote, so you can push to staging/production and it'll verify the build before deploying21:30
osse ojacobson: i see there is only a pre-push hook, so I suppose one needs to delay the job somehow.21:31
← GeertV left21:31
ilhami lov, you know what?21:31
ojacobson osse: you install the hook on the recieving repo (post-receive or post-update).21:31
eg. in gitolite.21:32
ilhami lov, say "what?"21:32
← RaceCondition left21:32
ojacobson That way the hook doesn't fire until the data is definitely already available for jenkins21:32
← rcu left21:33
→ valheru7 joined21:33
osse ojacobson: then I follow. That's what I would do. I just misunderstood you whrn you wrote "do it on git-push"21:33
ilhami ojacobson: IT WORKED!!!! YES21:33
→ bamj0rb joined21:33
valheru7 so I'm still getting used to git. with svn, you could use svn:remote properties to assign a remote repository to a certain folder in your repo. How do you go about doing this in git?21:33
gitinfo set mode: +v21:33
← timthelion left21:34
ilhami SVN is so old school. :D Use git21:34
lov valheru7: look into !submodules21:34
gitinfo valheru7: [!submodule] git-submodule is ideal to add subsidiary git repositories to a git superproject when you do not control the subprojects or more specifically wish to fix the subproject at a specific revision even as the subproject changes upstream. See http://book.git-scm.com/5_submodules.html21:34
ojacobson osse: ah, yeah, I see how I misspoke :)21:34
valheru7 ilhami, thats what I'm working on ;)21:34
← lawltoad left21:34
valheru7 thanks lov21:34
ilhami thanks lov21:35
bamj0rb Of course, while you're looking into submodules, consider whether what you're doing is really what you want to do anyway. Subversion might have taught you some backwards ways of doing things. >:)21:35
← faranjessani left21:35
→ hopkin joined21:35
← lamont left21:36
valheru7 bamj0rb, haha good point. Basically I want to be pulling in an open source project to my project and be able to update any bugfixes from upstream without having to re download a zipfile and just replace files everytime21:36
→ asQuirreL joined21:36
← erichynds left21:37
ilhami how do I go out of a log? :D21:37
it keeps saying end21:37
LO21:37
LOL21:37
→ RaceCondition joined21:37
→ endiruna joined21:37
ilhami ok found out.21:37
← kukks left21:38
lov valheru7: no problem. Just be aware that submodules have limitations, and aren't exactly like svn:remote21:38
→ amcsi joined21:38
← bschindler left21:38
→ petey_ joined21:39
→ grafi joined21:39
grafi hello peoples!21:39
gitinfo grafi: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.21:39
valheru7 lov ok thanks21:39
grafi so i've just found out i have ssh access to my crappy webspace21:40
now im trying to set it up for the "git workflow"21:40
bit it doesnt seem to work21:40
*but21:40
lov grafi: OK, so what's going wrong?21:40
→ tadasZ joined21:41
grafi not sure, i did git init --bare project.git inside the project directory21:41
i set up the remote21:41
_ikke_ grafi: Why did you do that?21:41
grafi because the internet told me21:41
:D21:41
lov an excellent reason21:41
grafi am I confused?21:41
:P21:41
→ vipzrx joined21:41
→ lamont joined21:41
lov if you were not, you would not be here :D21:42
← vipzrx left21:42
_ikke_ So if the internet told you to do sudo rm -rf /*, you'd do that also?21:42
m0viefreak a repo in a repo -- gitception!21:42
← petey left21:42
→ spaceonline joined21:42
grafi accidentally i know what rm -rf /* does, so yes of course :D21:42
→ kukks joined21:42
← kukks left21:42
→ kukks joined21:42
grafi but dont i have to have some kind of repo on the wepspace side of things?21:42
_ikke_ grafi: nope21:43
ilhami lov: I pushed my changes to the remote. It worked! :)21:43
_ikke_ grafi: You can just have a local repo all on it's own21:43
lov ilhami: congratulations, well done.21:43
grafi so how do I push to the webspace then?21:43
_ikke_ grafi: webspace?21:43
← LuxuryMode left21:43
grafi the server, if you will21:43
ilhami I can see the files under source in Google Code.21:44
→ vipzrx joined21:44
ilhami I think I did it correctly.21:44
→ seanz joined21:44
→ FernandoBasso joined21:45
← petey_ left21:45
ilhami But the thing with password I haven't figured out yet. How to store the password.21:45
→ petey joined21:45
grafi _ikke_ so what do I do to be able to push from my local machine to the server?21:45
_ikke_ grafi: Why do you want to push to the server?21:46
grafi so I dont have to upload the files via ftp?21:46
← ilhami left21:46
m0viefreak rsync21:46
→ Someguy123 joined21:46
grafi on a windows machine?21:47
:x21:47
→ kcj joined21:47
_ikke_ grafi: Using git for deployment is not really advisable21:47
← endiruna left21:48
grafi and here I thought that was what everybody was talking about :x21:48
so how do I do it?21:48
_ikke_ What kind of server is it/21:48
?21:48
← thiago left21:48
→ cn28h joined21:48
grafi well I have restricted access to it, im surprised im able to use ssh21:49
→ HicksD_ joined21:49
grafi what should I look for to answer your question to your satisfaction?21:49
_ikke_ And are you able to install git on it?21:49
→ mribeirodantas joined21:49
cn28h what's the significance of including push = refs/tags/* in a remote config? does this mean that tags will automatically be pushed, or simply to allow them to be pushed?21:49
grafi git is apparently installed21:49
← HicksD left21:49
← mribeirodantas left21:50
cn28h and by "automatically" I mean without having to explicitly use --tags or call the tag out by name21:50
_ikke_ grafi: look at !deploy21:50
gitinfo grafi: Git is not a deployment tool, but you can build one around it(in simple environments) or use it as an object store(for complex ones). Here are some options/ideas to get you started: http://gitolite.com/the-list-and-irc/deploy.html21:50
→ endiruna joined21:50
grafi I just discovered that, I was under the impression that it was the typical webspace/you get ftp access type of deal21:50
_ikke_ shared hosting21:50
← tadasZ left21:50
grafi yeah well you can have a vserver and that would be shared hosting too, no?21:50
thanks for the link21:51
reading now :p21:51
_ikke_ vserver?21:51
→ galaris joined21:51
grafi virtual?21:51
seanz Greetings. If I initialize a git repository as shared, will the FETCH_HEAD also retain group write permissions?21:51
← SocalNick left21:51
→ thiago joined21:51
_ikke_ a vps is not shared hosting21:51
→ RogerWS_ joined21:51
← guzzlefry left21:51
← hholtmann left21:52
_ikke_ seanz: You seem to have problems with it?21:52
← RogerWS left21:52
grafi well you share the actual hardware with other people though21:53
or at least you probably are21:53
ojacobson cn28h: the normal behaviour of a remote created with `git remote add foo URL` is that `git push foo some-branch` will push some-branch to foo, and push any tags pointing to commits that are in some-branch's history to foo21:53
_ikke_ yeah, but it's not really called shared hosting21:53
ojacobson with that ref in place, then `git push foo` will push all tags, possibly in addition to whatever other behaviour is configured21:54
seanz _ikke_: I've got a git repo on a server that the release engineers update, but whichever was the last engineer to do a fetch ends up owning FETCH_HEAD, causing the next person to need to use sudo.21:54
_ikke_ grafi: Unless you are on a real cheap vps provider, you do have your own memory21:54
seanz I'm looking for a fix for that.21:54
It looked like core.sharedRepository set to group...might be a solution.21:54
_ikke_ seanz: Sounds like a non-bare repo?21:54
seanz _ikke_: Yes, non-bare.21:55
← freeed left21:55
← axl_ left21:56
← spaceonline left21:56
← Ozch left21:56
← ni_ left21:57
← galaris left21:58
grafi _ikke_ real cheap is always what im going for ;)21:59
← Robarelli left21:59
← DrShoggoth left22:00
milki seanz: since you are talking about release engineers and stuff, you should probably invest in a !gitolite setup22:00
gitinfo seanz: Gitolite is a tool to host git repos on a server. It features fine-grained access control, custom hooks, and can be installed without root. Download: https://github.com/sitaramc/gitolite Docs: http://gitolite.com/gitolite/22:00
seanz milki: No way. We just switched from gitolite to GitHub Enterprise.22:00
This is for a deployment.22:00
milki seanz: then you should read !deploy22:01
gitinfo seanz: Git is not a deployment tool, but you can build one around it(in simple environments) or use it as an object store(for complex ones). Here are some options/ideas to get you started: http://gitolite.com/the-list-and-irc/deploy.html22:01
→ psoo joined22:01
seanz milki: Thanks!22:01
← TheBrainWork left22:02
← pr0ton left22:02
cn28h ojacobson, aha, interesting22:03
→ pr0ton joined22:03
← Snibble left22:03
→ sword joined22:03
← pixelgremlins left22:04
_ikke_ seanz: Or use a proper deploy system22:04
→ pixelgremlins joined22:04
seanz _ikke_: As in packages built for the target OS?22:04
ojacobson or, hell, rsync22:05
milki hell rsync22:05
seanz Oh gosh. I don't want to go that far. That's where we came from. That can be messy.22:05
ojacobson there are almost always some steps between "I have a pristine, source-controlled tree" and "the service is running/packaged/available"22:05
milki thats what scripts are for22:05
seanz Either git or packages. Something that lets me work in "units", as it were.22:05
ojacobson whether it's compiling sources, or setting config values, or restarting apache22:05
← sq-one left22:05
← cannonball left22:06
→ iamjarvo_ joined22:06
← schemanic left22:06
_ikke_ seanz: As in a application that is specialized in deploying code on servers22:06
← mdiers left22:07
← irqq left22:08
ojacobson you can build OS packages (debs, RPMs, whatever) from your git tree, sure22:08
→ jkremser joined22:08
← jkremser left22:08
→ jkremser joined22:08
ojacobson or use capistrano/fabric/ansible/etc to handle rollout, or22:08
→ SocalNick joined22:08
juvenal → juvenal_away22:09
← iamjarvo left22:09
→ Fandekasp joined22:10
← iamjarvo_ left22:10
→ hashguy joined22:11
→ sgu joined22:12
← abbycabs left22:12
← zapb left22:13
→ GODDOG joined22:13
angasulino I'm reading the article http://blogs.atlassian.com/2013/05/alternatives-to-git-submodule-git-subtree/ about git subtree, but I'm not clear on how it works (plus the examples all require horizontal scrolling... wtf?), especially committing and pulling in the subrepo, and I'm assuming having a subrepo in one branch but not in another is fine? It's not like with submodules?22:14
sgu Hi all, after an failed merge, I have unmerged path. When I try to use `git checkout --theirs`, I got an error "fatal:--ours/--theirs is incompatible with switching branched". What does that mean? Whan can I do if I want the changes from the another branch?22:14
→ galaris joined22:15
← mishok13 left22:16
→ knuth joined22:17
← kaendfinger left22:17
← jkremser left22:17
ojacobson sgu: 'git checkout --theirs -- path' or 'git checkout --ours -- path'22:17
do _not_ include branch names in the checkout command for this22:17
← DolphinDream left22:18
ojacobson also, it's rarely what you want, as both options overwrite the changes from one branch or the other22:18
juvenal_away → juvenal22:18
← ValicekB left22:18
→ SoftwareMechanic joined22:19
← rnsanchez left22:19
← pr0ton left22:19
← SoftwareMechanic left22:20
← Fenikkusu left22:21
sgu ojacobson: Does it have the same effect of `git merge --squash -s recursive -X theirs B`?22:21
← psoo left22:21
Guest68827 → minty22:21
→ refried_ joined22:22
ojacobson No.22:22
← beastd left22:23
sgu ojacobson: What are the differences?22:23
ojacobson They don't do anything like the same thing.22:24
→ kaendfinger joined22:24
ojacobson `git merge --squash ...options...` creates a new commit with a single parent, whose diff is the aggregate of the diffs on B22:24
milki oo, so concise22:25
→ ValicekB joined22:25
→ uknownalgorithms joined22:25
ojacobson `git merge B` creates a new commit with two parents (HEAD and B); if it runs into a conflict, it exits so you can resolve it. `git checkout --theirs -- PATH` marks PATH as resolved, using the contents of PATH exactly as they are in B, regardless of what happened on HEAD since B diverged.22:25
It's a travesty of UI development that `git merge` even has a --squash option -- in any sane UI it would be its own command, because it's strictly not a merge.22:26
→ macabre^ joined22:26
ojacobson (For some useful definitions of "merge", anyways)22:26
It does happen to use a lot of the merge machinery, which is probably why it's an option on merge (git has a bad case of implementer-oriented UIs)22:26
← kirin` left22:27
← jnavila left22:28
→ kirin` joined22:28
→ pr0ton joined22:28
← hellopat left22:28
uknownalgorithms Git was telling me i was ahead of origin master by 1 commit but wouldn't let me push the changes. I did a pull --rebase which broke the code im guessing when it merged automatically. I did a reset head --hard to my earlier commit with all the correct code but now my local master has diverged from origin master with 1 and 3 commits respectively. How can i fix this, ideally i have the correct version of the files right now and i22:28
← hopkin left22:29
milki thats wierd, your rest --hard should have put you 1 ahead of origin/master22:29
← macabre left22:30
← donmichelangelo left22:30
milki also, pull shouldnt have done anything at all22:30
o.O22:30
→ donmichelangelo joined22:31
ojacobson milki: it was "one ahead" because refs/remotes/origin/master was stale22:31
git pull runs git-fetch, which updated it22:31
← ValicekB left22:31
ojacobson he's in the same state before and after, other than no longer having stale remote-tracking branches :)22:31
milki oo22:31
uknownalgorithms: do you want to ignore those 3 new commits in your remote?22:32
→ mklappstuhl joined22:32
→ mdiers joined22:32
milki o wait22:32
uknownalgorithms That would be fine as i have the latest and correct version of the code in my working directory22:32
milki theres a dnalging i22:33
sgu ojacobson: Does `git merge --squash B` look at the common ancestor of B and HEAD?22:33
ojacobson sgu: yes.22:33
milki uknownalgorithms: you _could_ just push -f, but think of !rewriting22:33
gitinfo uknownalgorithms: [!rewriting_public_history] Rewriting public history is usually bad. Everyone who has pulled the old history have to do work (and you'll have to tell them to), so it's infinitely better to just move on. If you must, you can use `git push -f` to force (and the remote may reject that, anyway). See http://goo.gl/waqum22:33
← TGiFallen left22:33
uknownalgorithms initially i was trying to git push --force but the remote was rejecting it22:34
im not sure rewriting is the way to go as theres other devs working on it22:34
← infynyxx left22:34
→ axl_ joined22:35
→ dkpi joined22:35
uknownalgorithms if i try to git push -f now it gives me error:denying non-fast forward refs/heads/master. you should pull first22:36
← mklappstuhl left22:36
→ srayner joined22:38
← nutron left22:39
→ halu_ joined22:39
sgu ojacobson: so `git merge --squash -s recursive -X theirs B` and (git merge --squash B;git checkout --theirs -- all-unmerged-path) are the same? They both create a single parent commit and the diff are the aggregate of diff of B?22:39
ojacobson uknownalgorithms: okay so your server is configured to not let you destroy history22:39
→ infynyxx joined22:39
ojacobson now you have to decide how you want to combine the remote changes with your local changes22:39
welcome to source control22:39
← ngladitz left22:39
ojacobson 'git merge origin/master' is one option, 'git rebase origin/master' is another22:39
milki uknownalgorithms: !fixup, i suggest doing a revert of those 3 commits, then redoing your changes22:39
gitinfo uknownalgorithms: So you lost or broke something or need to otherwise find, fix, or delete commits? Look at http://sethrobertson.github.com/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history!22:39
→ ValicekB joined22:40
→ TGiFallen joined22:40
← pixelgremlins left22:40
→ takamichi joined22:40
← vladgh left22:41
← lavaman left22:41
← halu left22:41
← macabre^ left22:42
← danman left22:42
→ ZujkisNx joined22:42
→ tinti joined22:43
→ mribeirodantas joined22:43
← Zarthus left22:44
→ jiffe99 joined22:44
← jiffe99 left22:45
→ Zarthus joined22:45
← breakingthings left22:45
→ macabre joined22:46
→ corygibbons joined22:46
→ snuffeluffegus joined22:46
→ nutron joined22:46
← macabre left22:47
← ValicekB left22:47
← skorgon left22:47
→ macabre joined22:47
← threesome left22:47
← dropdrive left22:48
← cff_ left22:48
→ DanielHollands joined22:49
← marthinal left22:50
← renatow left22:51
← takamichi left22:51
← LimeBlast left22:52
← endiruna left22:53
← joelrobichaud left22:53
← uknownalgorithms left22:53
← zarubin left22:53
→ macabre^ joined22:53
← Pullphinger left22:54
→ renatow joined22:54
→ ValicekB joined22:55
← whaley left22:55
← nickpresta left22:56
← m8 left22:56
← CKyle left22:56
sgu Hi all: is there an easy way to undo "git checkout --theirs -- file" and put file back to conflict state(unmerge-path)?22:56
→ midnightmagic joined22:56
← macabre left22:56
ojacobson git checkout -m -- PATH22:57
→ ni_ joined22:57
ojacobson as it says in man git-merge22:57
gitinfo the git-merge manpage is available at http://jk.gs/git-merge.html22:57
← YoungFrog left22:57
← devbug left22:58
→ undecim joined22:58
→ quackgyver_ joined22:58
← Roots47 left22:59
→ pandeiro joined22:59
← Dougie187 left23:01
← TomyLobo left23:01
← vipzrx left23:01
← petey left23:02
→ h0rrorvacui joined23:02
← GODDOG left23:02
← ntio left23:02
→ kermit joined23:02
→ petey joined23:02
→ ntio joined23:03
← bwreilly left23:03
→ bwreilly joined23:04
qsuscs → qsusOFFcs23:04
← hwrd|work left23:04
→ SimonNa joined23:05
← axl_ left23:05
← pilif left23:06
→ nickpresta joined23:06
← acharlieh left23:06
← mitz_ left23:06
← macabre^ left23:07
← aard left23:07
sgu Hi all: `git status` shows unmerge path, however when I try to run `git checkout --theirs -- path`, nothings happened and the path is still in the unmerge path. What might be the problem?23:07
← srayner left23:08
→ endiruna joined23:08
→ mitz_ joined23:08
sgu never mind. I am being stupid, I need to run `git add` to staged the changes23:10
← rc111 left23:10
← jburcsik left23:10
← tinti left23:11
→ acharlieh joined23:11
← mooperd left23:12
← endiruna left23:12
→ mooperd joined23:12
pure Is it at all possible to version my git hooks?23:12
ojacobson Sure: store them in the repo, and put a symlink in .git/hooks/ (or make .git/hooks a symlink).23:12
← FernandoBasso left23:12
→ endiruna joined23:12
ojacobson There's no way to automatically configure that for new clones, because sending people code to run isn't really a great idea in the general case :)23:13
(it can work for simple cases where everyone trusts all collaborators)23:13
pure Alrighty. :)23:13
← endiruna left23:14
→ endiruna joined23:14
← pantsman left23:15
← nutron left23:16
→ canoon joined23:17
← tombee left23:17
← halu_ left23:17
← pr0ton left23:17
← Sonderblade left23:18
→ hwrd|work joined23:19
→ pr0ton joined23:19
→ marhaban joined23:19
← hyperair left23:20
← ajm203 left23:21
← knuth left23:21
← endiruna left23:21
→ endiruna joined23:22
← infynyxx left23:23
← marhaban left23:24
← cn28h left23:24
jayne has anyone found a workaround for the git-svn assertion failure when attempting to commit a rename?23:25
→ petey_ joined23:27
← hwrd|work left23:29
→ primitive joined23:29
→ joelrobichaud joined23:30
← petey left23:30
← mooperd left23:31
← zeograd left23:32
← pr0ton left23:32
← petey_ left23:32
← kadoban_ left23:33
→ pr0ton joined23:33
→ mooperd joined23:34
← mribeirodantas left23:35
← mdiers left23:35
→ vipzrx joined23:36
→ kadoban joined23:36
← vipzrx left23:36
→ frogonwheels joined23:36
juvenal → juvenal_away23:36
← nwertman left23:37
→ dropdrive joined23:37
← SHyx0rmZ left23:37
← snql left23:37
← mooperd left23:37
← kevwil left23:38
← Kicer86 left23:38
→ nwertman joined23:38
← wet88 left23:39
→ mooperd joined23:39
← mfeltner left23:39
← frogonwheels left23:40
→ frogonwheels joined23:40
← mfulz left23:40
← dropdrive left23:41
juvenal_away → juvenal23:41
→ jimmyeatworld joined23:42
→ zjdriver joined23:42
→ mdiers joined23:42
← Kayra left23:46
→ lavaman joined23:46
← groglogic left23:47
← ValicekB left23:48
← sgambino left23:49
→ schemanic joined23:49
← kriks left23:50
→ mfeltner joined23:51
→ nutron joined23:53
→ dropdrive joined23:53
← insejn left23:53
← asQuirreL left23:53
→ ValicekB joined23:55
← King_Hual left23:55
← bgerber left23:55
← harobed left23:56
→ Kicer86 joined23:57
← nwertman left23:58
→ insejn joined23:58
← amcsi left23:58
← ojacobson left23:58
→ King_Hual joined23:59
→ wirrbel joined23:59
← wirrbel left23:59
→ nwertman joined23:59

Logs Search ←Prev date Next date→ Channels Documentation