IRCloggy #git 2011-01-24

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.

2011-01-24

→ nazgul102 joined00:01
← andrewvos left00:02
← spaceonline left00:02
← Modius left00:02
← nazgul101 left00:02
→ pdelgallego joined00:03
← chrischambers left00:03
→ pdelgallego_ joined00:04
← nevyn^1 left00:04
→ markoso joined00:04
← paintchimps left00:05
→ juangiordana joined00:05
← tizbac left00:06
← workmad3 left00:07
→ ROBOTARMY joined00:07
← John___ left00:08
→ chrischambers joined00:10
← __name__ left00:11
← Cro_Crx left00:12
← project2501a left00:13
→ chrisdarroch_ joined00:14
← p3rror left00:15
→ Lalaland1125 joined00:17
← Agos left00:17
← chrischambers left00:19
← airborn left00:21
kanzure do i want a post-receive or post-update hook? i want to render pdfs from each commit00:21
patrikf IMHO, neither00:22
→ abstrakt joined00:22
thiago_home probably post-update, so that they don't block the push acceptance00:22
← oracal left00:23
kanzure patrikf: why neither?00:23
thiago_home: thank you, that's a good point00:23
→ John___ joined00:23
patrikf kanzure: do you really need that PDF up-to-date at all times?00:23
banisterfiend why is it wrong to push to a non-bare repo?00:24
Unknown[NF] → Unknown[OFF]00:24
kanzure i lied, i'm actually rendering images of hardware schematics, and yes it needs to be for each revision00:24
→ nevyn^1 joined00:24
thiago_home banisterfiend: because it does things you don't expect it to00:24
banisterfiend: it modifies a branch, but doesn't update the checkout00:24
banisterfiend thiago_home: like what my friend00:24
thiago_home banisterfiend: you need to read faq nonbare00:24
← tewecske1 left00:24
thiago_home huh...00:24
banisterfiend: faq non-bare00:24
jast banisterfiend: Pushing to non-bare repositories is discouraged: please see https://git.wiki.kernel.org/index.php/GitFaq#non-bare [automatic message]00:24
banisterfiend thiago_home: oh ok, so the working tree is the same as it was before the push00:24
→ Cro_Crx joined00:24
← felixge left00:24
← Bigsista left00:25
→ surikator joined00:26
→ luxurymode joined00:29
← gilaniali left00:29
← thiago_home left00:30
← nevyn^1 left00:30
← Tuller left00:30
← Lalaland1125 left00:31
→ John_____ joined00:31
← d0k left00:32
→ dreiss joined00:32
← Cro_Crx left00:33
→ p3rror joined00:33
← John___ left00:33
← SirFunk_ left00:34
banisterfiend can i define an alias that is local to a repo?00:34
bremner yes00:34
banisterfiend how bremmy?00:34
bremner with git config ?00:34
banisterfiend can you give a simple example bremmx0r ?00:34
← Yuuhi left00:35
bremner see man git-config00:35
jast the 'git-config' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-config.html [automatic message]00:35
← John_____ left00:35
→ Cstep4 joined00:35
banisterfiend thanks brems00:35
bremner search for "alias", takes you right to it00:35
do you massacre everyone's nick?00:36
banisterfiend bremner: also, out of interest, what merge tool do you use?00:36
bremner I use kdiff3 and emerge00:36
banisterfiend bremner: im just being affectionate00:36
← Cstep4 left00:36
banisterfiend ah ok00:36
talexb Hmm .. gitk is broken now I've updated git. :(00:36
banisterfiend bremner: is emacs git support pretty good?00:36
bremner it is if you install magit00:36
banisterfiend great00:37
thanks00:37
bremner welcome, banny-wanny00:37
banisterfiend ;)00:38
bremner ;)00:38
→ Cro_Crx joined00:38
← p3rror left00:38
→ stringoO joined00:38
← dpino left00:42
← dfr|mac left00:43
→ Slix` joined00:44
← warthog9 left00:45
→ SethRobertson joined00:48
yaotti → yaotti_aw00:48
ChanServ set mode: +v00:48
→ dfr|mac joined00:49
→ p3rror joined00:50
chrisdarroch_ so I'm using git-svn, and the repo has a branches folder. I've cloned it using `git svn clone https://path/to/svn/repo/' -T trunk -b branches`00:50
my problem is that there's a folder inside branches/ called branches/private/, which has several folders inside for individual user's branches00:50
I'd like to simply ignore that entire folder forever, but it seems rather difficult to do00:50
If I add an ignore-path="branches/private", git-svn still runs over the folder and makes a commit for it, but doesn't add content to the commit00:51
anybody have ideas on a better way of making git-svn completely ignore that folder's existence?00:52
patrikf is it an issue, though?00:52
I mean, you can just ignore that branch...00:52
→ verto joined00:52
chrisdarroch_ it's not a gigantic problem, though because of the way git-svn has interpreted the branches/private folder...00:53
occasionally I'll get about 20 commits to that folder when I fetch —all00:53
similarly, when it shows up in log history it's annoying to wade through00:53
I usually just `git branch -d -r private` after each fetch00:54
bit of a pain, is all00:54
← nairb774 left00:54
← rgr left00:55
patrikf okay, I see your point (don't have any advice though)00:55
→ nairb774 joined00:56
chrisdarroch_ lord help me if anything from that private branch gets merged in to the product, though :P00:56
← dfr|mac left00:56
→ utu joined00:57
← cenuij- left00:57
talexb Stupid question warning .. where can I get the latest gitk? All I can find is stuff like http://freshmeat.net/projects/gitk which is eight years old.00:58
bremner talexb: it ships with git now00:58
→ Bigsista joined01:00
Bigsista lo01:00
← pheaver left01:00
→ pheaver joined01:01
talexb bremner .. Hmm. I built git from source this afternoon -- but apparently didn't get gitk as well?01:01
patrikf that's not even git's gitk, lol. already wondered how it could be eight years old...01:01
Bigsista if i need qiuck teamwork. can i work on remotes/master and send my commits directly? or do i need to commit to local/master and then push to remotes/master?01:01
→ flaguy48 joined01:02
bremner talexb: dunno about that. Maybe it is configuration option. It needs tk01:02
patrikf Bigsista: the remotes are only local mirrors. never work on them.01:02
Bigsista i'm used to svn. so don't blame me for asking =/01:02
patrikf s/remotes/remote-tracking branches under refs/remotes/, to be exact01:03
patrikf is not blaming anyone01:03
patrikf if you want svn-like behaviour, you can pull+push after each commit.01:03
Bigsista patrikf: if i switch to a remote tracking branch in eclipse e.g. - where do my commits go then? are they stil only local until i push them?01:04
ah ok01:04
← stringoO left01:04
Bigsista we need to sort that out in our team, if we all change to git... and i am the one who's gonna tell them how =)01:05
patrikf you can commit to remote-tracking branches, but that will irritate some of git's assumptions.01:05
← jfelchner left01:05
patrikf so don't do it, the only reason you would checkout remote-tracking branches is to merely look at the contents01:05
← priidu left01:05
Bigsista patrikf: so... this would NOT be a good way to work then01:06
i see01:06
→ RahulT joined01:06
talexb bremner .. I'm getting 'Error in startup script: expected version number but got "1.7.4-rc2"' ..01:06
← PointMan_ left01:06
← shruggar left01:07
patrikf Bigsista: also, git's power lies in the command line, so starting by working from eclipse might not be the best idea01:07
← hagabaka left01:07
Bigsista i'm a shell nerd... but the others are not01:07
patrikf just saying01:08
Bigsista patrikf: most docs i read gave the same impression to me. it gets really powerful when you start using the shell01:09
← panike left01:10
→ gebi joined01:12
talexb Hmm, http://comments.gmane.org/gmane.comp.version-control.git/164487 .. doesn't like the RC bit at the end.01:14
← jonanin left01:15
→ yhager joined01:16
talexb bremner -- Yep, that fix did it. If I knew more shell I'd submit a patch.01:18
bremner talexb: it looks like there is already a patch in that thread01:18
← kipras left01:19
talexb Ah.01:20
Anyway, just wanted to pass that tidbit along.01:21
→ iSikh joined01:21
Unknown[OFF] → Unknown[NF]01:21
→ hagebake joined01:22
→ bryantluk joined01:22
← miloshadzic left01:23
← necromancer left01:24
hagebake → hagabaka01:24
← nb left01:25
Unknown[NF] → Unknown[OFF]01:25
→ rgr joined01:26
← sonnym left01:27
Unknown[OFF] → Unknown[NF]01:27
→ sonnym joined01:27
← iravid left01:27
← MikhX left01:27
→ iravid joined01:28
→ panike joined01:28
← panike left01:28
← chrisdarroch_ left01:29
→ chrisdarroch_ joined01:29
← Cro_Crx left01:31
→ warthog9 joined01:32
→ necromancer joined01:33
← bitkiller left01:34
→ MikhX joined01:34
Unknown[NF] → Unknown[OFF]01:34
← alberto56 left01:35
magcius Ok01:35
← pheaver left01:36
magcius If I have branch1-B-C-D-E, and want to get commits C and D onto branch2 to look like branch2-C-D, but not B or E, is there an easy way to do this?01:36
→ Cro_Crx joined01:36
magcius I could rebase -i and make branch1 become branch1-C-D-B-E, and then reset branch2 to become D, but I don't want to do that.01:36
← ROBOTARMY left01:37
talexb magicus Use the cherrypick feature.01:37
→ nb joined01:38
magcius talexb, won't that duplicate C and D?01:38
talexb, and if I want to merge branch1 back into branch2, will it not become confused?01:38
Unknown[OFF] → Unknown[NF]01:39
→ surikator_ joined01:39
← surikator left01:39
→ gilaniali joined01:39
talexb magicus You want to make changes C and D to branch2, right?01:40
The cherrypick feature will do that. And no, it won't get confused -- it'll see it's already done that change.01:40
→ laj joined01:41
← chessguy left01:43
→ ProperNoun joined01:43
→ auntie joined01:43
→ dr_win joined01:44
dwmw2_gone → dwmw2_BNE01:44
← ProperNoun left01:45
→ fivetwentysix joined01:45
→ gretch joined01:47
← auntie left01:48
necromancer does anyone here know why you can't follow organizations on github?01:48
bremner maybe ask on #github?01:49
← xiong left01:52
kanzure my tree id for a commit was 'A'; then when i pushed three commits over, for each of those commits, the tree id was 'B'; the only change in all 4 commits was a change to a certain file01:53
why would the tree id change?01:53
or not change in the case of the three that i pushed all at once?01:53
→ jarodluebbert joined01:54
← surikator_ left01:57
→ surikator joined01:57
→ nazgul101 joined02:00
→ Remixman joined02:01
← nazgul102 left02:01
← dreiss left02:02
→ klunky joined02:02
← surikator left02:02
necromancer oh shit there's a #github? lol02:02
← scelis left02:05
← klunky left02:05
→ klunky joined02:06
→ jzacsh joined02:07
→ adnap joined02:09
← klunky left02:09
← yhager left02:09
← Cro_Crx left02:10
→ chessguy joined02:11
← sraue left02:11
Unknown[NF] → Unknown[OFF]02:13
→ QinGW joined02:14
Unknown[OFF] → Unknown[NF]02:14
← Bigsista left02:14
→ dl9pf_ joined02:14
← dl9pf left02:15
→ mikepack joined02:16
← iSikh left02:16
← rjrodger left02:18
→ HuntingBears joined02:18
dwmw2_BNE → dwmw2_gone02:18
→ dfr|mac joined02:19
← sako left02:20
→ alan_k joined02:22
→ sraue joined02:23
→ pheaver joined02:26
→ silanus_ joined02:27
→ Vampire0_ joined02:28
← pdelgallego left02:28
← silanus left02:29
necromancer here we go! about to init repos for all of my logic pro projects02:29
← pdelgallego_ left02:29
← rgr left02:31
← necromancer left02:31
← Vampire0 left02:32
← utu left02:32
adnap hello, i've just started learning to use git. i have a question: what should i do if i need to use a makefile for my project that nobody else will need to use? should i keep this out of any commits, causing it to constantly show up in my status, add it to my ignored file, or add it in a commit under a different name?02:33
kanzure what's the best way to get a list of files updated in a certain commit?02:33
→ bburhans_ joined02:33
← bburhans left02:33
kanzure adnap: why not add it to your ignore file?02:33
adnap kanzure: git diff, or git diff --cached02:33
kanzure hmm i guess there's git show --pretty="format:" --name-only02:33
adnap kanzure: that was one of my thoughts. do you think that's the best?02:33
kanzure adnap: sure02:34
← engla left02:35
→ ROBOTARMY joined02:35
adnap also, is it possible to rename a file in my repo, but never make that change known?02:36
the makefile in my repo that i don't want to use is named GNUmakefile, which is the first thing make uses. i'd prefer that make use my makefile by default, so i'd like to rename GNUmakefile, but not have that change reflected when i do any commits.02:37
kanzure yes, you could move it on your own then put it into your ignore file; however, when you check out the file again, it will come back to life02:37
ChanServ set mode: +v02:37
→ bcardarella joined02:39
← HuntingBears left02:39
adnap okay, ty02:41
← Linc left02:42
bburhans_ → bburhans02:43
banisterfiend adnap; hello02:46
← webPragmatist left02:46
→ yhager joined02:46
→ ysiad joined02:47
ysiad what's the standard process for conflict resolution?02:48
like if I edit the file by hand02:48
and I want to mark it "resolved"02:48
how do I do that02:48
chrisdarroch_ just git add the file after you've resolved the conflicts02:48
adnap banisterfiend: hello. what's up?02:51
→ Bass10 joined02:55
← yhager left02:57
→ yhager joined03:00
← SimonP86 left03:00
← Mab879 left03:02
← ROBOTARMY left03:03
← jarodluebbert left03:03
→ ROBOTARMY joined03:04
jzacsh hello, i'm having a bit of trouble making gitolite and gitweb work together. it seems to be a common cause for new forum threads, but none seem that clear. I'm not sure what is wrong, but I've followed instructions both of these places: http://codeslife.com/?p=99 http://groups.google.com/group/gitolite/browse_thread/thread/851ded374e5ae4c903:05
does anyone know a definitive guide that explains how to do it?03:05
→ shentino joined03:06
← shentino left03:06
→ shentino joined03:06
→ Cro_Crx joined03:08
bremner just followed the gitolite docs03:09
bremner jzacsh: can you be more specific about what is going wrong? permissions are the usual problem03:09
→ Mab879 joined03:09
jzacsh bremner: there was no where that you had to specify what dir/ to serve from ? (through apache)03:09
← storrgie left03:10
bremner jzacsh: I guess that depends on your gitweb config. It's true I do some symlinking from the gitolite dir to where gitweb looks. But pointing gitweb at the right place should work too.03:10
jzacsh bremner: i guess i haven't read enough (though this is not the first go around)... I don't really know what's wrong.03:11
bremner: what gitweb conf are you editing? just the /etc/gitweb.conf?03:11
← gilaniali left03:11
bremner yeah03:12
← Cro_Crx left03:13
→ necromancer joined03:14
jzacsh bremner: is that conf based on `git-daemon(1)` ?03:14
bremner nope. just some ad-hoc perl03:14
jzacsh bremner: i'm thinking that the only thing missing is a specification as to _where_ to serve gitweb from. i feel like I've seen that in a config somewhere, but don't recall. do you have that specified in gitweb.conf?03:16
bremner I guess you want $projectroot03:17
← ysiad left03:18
← necromancer left03:18
→ necromancer joined03:18
jzacsh bremner: sorry, i didn't mean where the repo is -- i ment where git gitweb.cgi script is.03:18
bremner oh, that is configured in apache I guess.03:18
yaotti_aw → yaotti03:19
← alan_k left03:19
← kulp left03:19
← zombor left03:19
jzacsh bremner: okay, so than I guess I'm not missing anything too obvious. you can see http://code.jzacsh.com/ -- but it just gives you a dir/ listing and shows you the cgi script sitting there.03:20
→ gilaniali joined03:21
jzacsh my gitweb.conf in /etc/ is http://dpaste.de/2Mcd/raw/ --- both ~git/repositories and ~git/projects.list have a group of "code" and are group readable and executable.03:21
← necromancer left03:22
← towski left03:23
jzacsh these is the ~git/.gitolite.rc relevant lines: http://dpaste.de/R3O7/raw/03:23
← abstrakt left03:24
→ stringoO joined03:24
→ kulp joined03:24
jzacsh bremner: do you have anything in /etc/gitweb.conf that it looks like i definitely need in mine?03:24
bremner I think you have a basic CGI/apache problem, nothing fancy03:25
jzacsh bremner: oh! awesome03:25
bremner: you mean... me not having a mod_cgi turned on or osmething..? (looking)03:25
bremner I installed the gitweb debian package, but I have something like this http://paste.debian.net/105492/03:26
banisterfiend bremner: have you seen03:26
→ intripoon_ joined03:26
banisterfiend bremner: 'the forsyte saga' ?03:26
bremner nope03:26
jzacsh bremner: thanks a bunch :) this is a good lead for me :)03:27
banisterfiend bremner: http://en.wikipedia.org/wiki/The_Forsyte_Saga03:27
semes like your kind of thing somehow03:27
← Amorphous left03:28
bremner I'm not _that_ old03:28
← Remixman left03:28
→ necromancer joined03:29
← intripoon left03:29
← bryantluk left03:31
→ hyperair joined03:32
← warthog9 left03:33
← jxpx777 left03:34
← QinGW left03:34
→ sagsousuke joined03:34
banisterfiend bremner: im watching the series on tv now, the father just confessed a rape to his daughter03:35
bremner: she seemed sympathetic03:35
→ sako joined03:39
→ towski joined03:40
→ Amorphous joined03:43
→ jmcantrell joined03:44
→ shyam_iyer joined03:46
→ QinGW joined03:47
→ timj_ joined03:52
← pyther left03:54
Unknown[NF] → Unknown[OFF]03:54
→ webPragmatist joined03:55
← webPragmatist left03:55
→ webPragmatist joined03:55
← timj left03:55
← Elfe left03:56
→ Elfe_ joined03:56
← webPragmatist left03:56
← sagsousuke left03:58
← luxurymode left03:59
Unknown[OFF] → Unknown[NF]04:02
← jvv left04:02
→ envi joined04:04
→ t0rc joined04:05
← juangiordana left04:07
← vitor-br left04:07
→ sagsousuke joined04:07
→ jonshea joined04:08
← glennhartmann left04:08
← chessguy left04:09
→ NewtonianB joined04:10
→ webPragmatist joined04:11
← webPragmatist left04:11
→ webPragmatist joined04:11
bremner sitaram: I'm thinking about generating a bunch of keydir/* files in a cron job, and I'm loath to check them into git. Is this crazy talk? Can I just manually run the right hook?04:13
→ xiangfu joined04:13
→ theoros_ joined04:13
→ warthog9 joined04:14
← Gryllida left04:15
→ Gryllida joined04:15
← linusoleander left04:15
← theoros left04:15
sitaram bremner: not at all; see src/gl-setup-authkeys (documentation in top of file as comments)04:16
→ SirFunk_ joined04:16
bremner sitaram: great, thanks.04:16
← justin-george left04:16
→ OOPMan joined04:17
necromancer → necromancer`bzzy04:18
sitaram bremner: what I forgot to say in that docs is that then you probably want to set GL_SETUP_NO_AUTHKEYS in the main rc file so that a push to the admin repo does not end up overwriting the authkeys file with the (few, if any) keys in the normal keydir04:19
(doc patch welcome after you test it, heh!)04:19
→ urbanmonk joined04:19
← urbanmonk left04:20
← necromancer`bzzy left04:20
bremner sitaram: ok, I'll let you know how it works out. probably not right away, I just wanted to catch you before I went to sleep.04:20
← theoros_ left04:21
→ clarkfischer joined04:21
sitaram huh?04:22
I'm at work -- it's 9:52 am :-)04:22
offby1 it's a biiiig world ... ♩♬♫♪04:22
← webPragmatist left04:22
bremner right. I'm the sleepy one ;)04:23
→ theoros joined04:23
← theoros left04:23
→ theoros joined04:23
talexb Close to midnight here, but I got a lot accomplished today. And git flow rocks!04:24
→ urbanmonk joined04:26
→ BiggFREE joined04:26
← bpc415 left04:36
← talexb left04:36
← kukks left04:36
← theoros left04:37
sitaram bremner: I'm the sleep one, it seems. I *just* realised you wrote "before I went to sleep". God alone knows how I read it as "before you went to sleep" <sheepish grin>04:40
sleepy04:40
.oO(ok time for a cuppa... this is horrible even for me)04:41
banisterfiend sitaram: what's git flow?04:42
→ theoros joined04:42
SethRobertson Google can help with that04:42
It is a method of managing branches and workflow04:42
banisterfiend SethRobertson: can you tell me about it, please SethRobertson.04:42
→ deryl joined04:43
sitaram banisterfiend: or talexb, since he was the one who said it rocks :)04:43
SethRobertson http://nvie.com/posts/a-successful-git-branching-model/04:43
deryl Think I'm totally missing something here. I no longer need my 'testing' branch. I have main (HEAD), testing, and alpha. I am trying to completely and utterly delete the testing branch, keeping just main (HEAD) and alpha. i though this was accomplished with git branch -r -D origin/testing ; git push but that fails utterly to update the remote repo04:44
→ henry_ept joined04:44
banisterfiend SethRobertson: do you use git flow? the program, not just the model?04:44
SethRobertson git push :testing04:44
deryl: Then git remote prune or something like that. See man git-remote04:44
jast deryl: the 'git-remote' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-remote.html [automatic message]04:44
deryl SethRobertson: thanks04:44
SethRobertson bf: No, I don't believe in the model. I just know some people like it and use it04:45
← towski left04:45
banisterfiend SethRobertson: what's your model SethRobertson04:47
sitaram banisterfiend: why do you repeat the name banisterfiend? :-)04:49
SethRobertson Very simple. master for future work plus release branches for the run up to release and post release patching. People generally use their local repo as a temporary feature branch as necessary.04:49
banisterfiend sitaram: it's respectful04:51
...sitaram ;)04:51
SethRobertson: oh ok. One other question --- when people say 'git flow' do they typically refer to that model or do the program that facilitates that model?04:52
or to*04:52
← verto left04:52
← mikepack left04:52
→ utu joined04:52
offby1 hey, ho, go with the flow04:53
banisterfiend sounds like osmeting snoop dogg would say offby104:53
SethRobertson I'm sure the answer is both.04:54
← xiangfu left04:54
← shyam_iyer left04:55
→ napster_123 joined04:56
← t0rc left04:58
← tchan left04:59
← sako left04:59
→ tchan joined05:00
← tchan left05:01
← gilaniali left05:01
← deryl left05:02
→ tchan joined05:02
→ towski joined05:03
← theoros left05:06
← beatak left05:10
→ escape joined05:11
→ Mage3 joined05:11
← Mage left05:12
→ xiangfu joined05:12
→ gozoner joined05:13
← ISF left05:16
← gebi left05:17
← shentino left05:17
← markoso left05:17
→ luxurymode joined05:17
→ sako joined05:19
← fivetwentysix left05:19
← NewtonianB left05:20
← up_the_irons left05:20
← nairb774 left05:22
→ emptyflask joined05:27
← emptyflask left05:27
→ emptyflask joined05:27
← TheMoonMaster left05:29
→ theoros joined05:29
← urkud left05:29
→ Utkarsh_ joined05:34
→ nairb774 joined05:35
← Utkarsh left05:36
→ offby1` joined05:37
← BiggFREE left05:38
← banisterfiend left05:38
← offby1 left05:39
← towski left05:39
→ towski joined05:39
→ devanjan joined05:39
→ theoros_ joined05:40
← theoros left05:43
← starcoder left05:43
→ starcoder joined05:43
→ LongBeach joined05:44
→ nohk joined05:44
← nohk left05:44
→ urkud joined05:44
offby1` → offby105:45
← towski left05:45
→ towski joined05:46
→ mct joined05:48
→ theoros joined05:49
← theoros left05:49
→ theoros joined05:49
← theoros_ left05:51
→ theoros_ joined05:52
Chillaway → Chillax05:52
→ coderdad joined05:53
← theoros left05:54
← duckinator left05:55
→ theoros joined05:55
← theoros left05:55
→ theoros joined05:55
← theoros_ left05:56
← madewokherd left05:57
→ BiggFREE joined05:59
← jfkw left06:03
→ _pingu joined06:03
→ freimatz joined06:06
→ doener joined06:07
← RahulT left06:09
← jmil left06:11
← NickyP left06:12
← towski left06:13
→ joel_falcou joined06:13
→ towski joined06:14
→ ThomasLocke joined06:14
← towski left06:18
→ maleknet joined06:19
← Yuffster left06:19
← giallu left06:20
→ guardian joined06:21
← joel_falcou left06:22
← coderdad left06:23
← escape left06:24
← Pewpewarrows left06:25
→ towski joined06:26
← Octalot left06:26
→ up_the_irons joined06:27
← sako left06:28
→ khmarbaise joined06:29
← _pingu left06:30
← warthog9 left06:30
← benishor_ left06:31
→ pedrobelo joined06:31
← pedrobel_ left06:31
← adamm left06:31
← starcoder left06:34
→ starcoder joined06:34
← hagabaka left06:34
→ banisterfiend joined06:36
→ banisterfiend`` joined06:39
Unknown[NF] → Unknown[OFF]06:40
Unknown[OFF] → Unknown[NF]06:40
Unknown[NF] → Unknown[OFF]06:41
← banisterfiend left06:41
→ hagebake joined06:41
Unknown[OFF] → Unknown[NF]06:42
hagebake → hagabaka06:42
Unknown[NF] → Unknown[OFF]06:43
Unknown[OFF] → Unknown[NF]06:43
← orafu left06:44
→ orafu joined06:46
← dfr|mac left06:46
→ sako joined06:47
→ ph^ joined06:48
→ nohk joined06:48
← Bass10 left06:48
← nohk left06:49
→ pratz joined06:49
→ nevyn^1 joined06:49
← tlonim left06:51
← starcoder left06:51
← gozoner left06:53
← guardian left06:53
← ph^ left06:56
→ starcoder joined06:57
→ krosenvold joined06:58
→ TrineLocke joined06:58
← abetaha left07:03
← yhager left07:04
→ yhager joined07:04
← utu left07:07
← Error404NotFound left07:07
→ joel_falcou joined07:10
← pratz left07:14
← pheaver left07:14
→ pratz joined07:14
→ Error404NotFound joined07:15
← nevyn^1 left07:15
← vdv left07:19
→ uniquex joined07:19
← uniquex left07:20
→ chairabanta joined07:20
← LouisJB left07:21
→ shentino joined07:23
← shentino left07:23
→ shentino joined07:23
← OOPMan left07:24
→ henrikbjorn joined07:24
→ pheaver joined07:25
→ j0ran joined07:26
→ pht joined07:28
pht hi07:28
→ mastro joined07:28
pht how to show 'git log' containing only commits done by a particular person?07:28
← stringoO left07:29
doener --author or --committer, depending on which one you want to filter07:29
→ warthog9 joined07:31
→ giallu joined07:32
→ StevenBlack joined07:34
← emptyflask left07:34
→ stringoO joined07:35
almaisan-away → al-maisan07:35
← pheaver left07:37
→ tewecske joined07:38
← ROBOTARMY left07:39
→ DrNick2 joined07:39
→ Pieter joined07:39
→ opalka joined07:40
← DrNick left07:40
→ saschpe joined07:41
pht doener, cool.07:42
→ Zhonghua joined07:42
→ guardian joined07:43
← bcardarella left07:43
← StevenBlack left07:43
← yagami_i left07:46
→ ph^ joined07:47
← madduck left07:47
→ radarek joined07:48
← ph^ left07:49
← Textmode left07:49
→ ph^ joined07:50
→ yagami_i joined07:51
→ thiago_home joined07:52
→ priidu joined07:54
→ davidc_ joined07:55
← hagabaka left07:55
→ ludde joined07:55
→ jhoepken joined07:55
→ madduck joined07:57
Chillax → Chillaway07:57
← sako left07:58
→ ningia joined07:59
→ hagabaka joined08:00
← hagabaka left08:00
→ hagabaka joined08:00
→ benishor_ joined08:00
→ mhale joined08:02
→ fivetwentysix joined08:04
← angelsl left08:04
← fivetwentysix left08:04
← sagsousuke left08:05
ChanServ set mode: +v08:09
mhale hi all08:09
i have a question about the --signoff parameter in the command line git08:09
DrNick2 then maybe you should ask that question08:10
→ teweWork joined08:10
mhale i want to set up a git workflow such that person A works on a feature in a branch and commits it, then person B does a code review and signs off on that feature08:10
DrNick2 → DrNick08:10
mhale this seems easy to do if the people are working on different machines08:11
is there some way to do it if they are working at the same keyboard and mouse?08:11
→ j0ran1 joined08:11
← j0ran left08:11
mhale ie. an extra --signed-off-by [email@hidden.address] parameter that im not aware of?08:11
← hagabaka left08:12
→ fivetwentysix joined08:12
← fivetwentysix left08:13
← banisterfiend`` left08:14
mhale it appears i cant do what i want but i thought it was worth asking here08:14
→ kreemer joined08:15
→ borwar joined08:17
al-maisan → almaisan-away08:17
→ hagabaka joined08:18
← hagabaka left08:18
→ hagabaka joined08:18
← chairabanta left08:18
→ drl joined08:19
→ workmad3 joined08:19
→ j416 joined08:19
→ banisterfiend joined08:20
→ vdv joined08:20
← sivy left08:21
→ chrischambers joined08:24
← laj left08:24
→ j0ran joined08:31
← j0ran1 left08:32
→ sgillet joined08:38
→ galderz joined08:38
malfy_ → malfy08:39
→ dharrigan joined08:39
← Mage3 left08:40
→ crashanddie joined08:44
← crashanddie left08:44
→ crashanddie joined08:44
→ nevyn^1 joined08:45
→ darkredandyellow joined08:46
← doener left08:46
→ doener joined08:46
→ ziro`_ joined08:48
← vdv left08:49
→ psoo joined08:49
← tyteen4a03 left08:50
silex → Silex08:50
← workmad3 left08:50
← hyperair left08:50
→ dc5ala joined08:51
→ tyteen4a03 joined08:51
→ serard joined08:51
← urkud left08:52
← sandbags left08:54
→ tatsuya joined08:54
→ sandbags joined08:54
← sandbags left08:54
→ sandbags joined08:54
← theoros left08:55
← nevyn^1 left08:56
→ ntoll joined08:57
← mhale left08:57
← sandbags left08:57
→ goshawk joined08:58
→ matjas joined08:58
← Zhonghua left08:59
→ Zhonghua joined09:00
← dean__ left09:01
→ thiago joined09:01
→ LouisJB joined09:03
← Sigma left09:05
← gbjk left09:07
→ torvalamo joined09:09
→ Titosemi joined09:10
← banisterfiend left09:11
→ OOPMan joined09:11
→ banisterfiend joined09:11
→ Sigma joined09:11
→ nicoulaj joined09:12
→ StrongOrder joined09:14
← ericlee left09:14
→ ericlee joined09:14
→ dpino joined09:16
→ PigeonFriend joined09:16
← Slix` left09:16
→ rjrodger joined09:16
→ snearch joined09:17
→ vdv joined09:17
→ Cro_Crx joined09:18
← Rapp left09:27
← napster_123 left09:28
← nicoulaj left09:28
→ nicoulaj joined09:30
→ yotaff joined09:33
→ shruggar joined09:33
almaisan-away → al-maisan09:34
shruggar does anyone know the status of the "remote helper" interface, and whether various remote-helpers are planned to be merged into the main tree? (for example, the remote helper for mercurial)09:34
← banisterfiend left09:35
← bauruine left09:35
→ banisterfiend joined09:35
→ tchan1 joined09:40
← tchan left09:41
← scp1 left09:42
→ agateau_ joined09:42
→ hyperair joined09:42
← agateau left09:45
← Buglouse left09:45
→ bluenovember joined09:45
→ angelsl joined09:45
→ miloshadzic joined09:47
→ bentob0x joined09:48
→ airborn joined09:48
→ pfrenssen joined09:50
→ Rapp joined09:51
→ skoop joined09:51
→ PointMan joined09:52
← stringoO left09:54
← hyperair left09:57
→ hyperair joined09:58
→ ereslibre joined09:59
← ereslibre left09:59
→ ereslibre joined09:59
→ YouSirOne joined09:59
Unknown[NF] → Unknown[OFF]09:59
YouSirOne hi i got error: Entry 'gitosis.conf' not uptodate. Cannot merge. when trying to run git pull from gitois-admin, what to do?10:00
gitosis-admin*10:00
cbreak_work do a git diff10:00
YouSirOne i got some differences10:01
cbreak_work remove them10:01
tango_ or stash them if you want to keep them for later10:01
cbreak_work by committing, stashing, just checking out the original...10:01
← angelsl left10:02
YouSirOne i just want to overwrite my own changes10:02
how do i do that?10:02
→ erez joined10:02
→ p3rg2z joined10:03
← erez left10:03
→ erez joined10:03
cbreak_work git checkout HEAD -- filename10:03
YouSirOne ok thanks10:04
it worked :)10:04
→ workmad3 joined10:05
→ d0k joined10:06
→ lanthan_afh joined10:07
→ kirthi joined10:07
← clarkfischer left10:09
← QinGW left10:09
← sraue left10:11
← dharrigan left10:12
→ tizbac joined10:12
→ makkes joined10:14
→ Michael joined10:15
Michael → Guest4617310:16
→ _ollie joined10:16
← hyperair left10:16
← MikhX left10:16
← _ollie left10:17
→ drizzd_ joined10:18
→ rolfb joined10:20
← drizzd left10:22
→ surikator joined10:22
→ sraue joined10:23
← StrongOrder left10:23
→ candybar joined10:26
← candybar left10:26
→ candybar joined10:26
→ chairabanta joined10:27
← gretch left10:29
← ntoll left10:30
→ bottiger joined10:31
← candybar left10:38
← kirthi left10:39
→ kirthi joined10:39
→ candybar joined10:39
← tyteen4a03 left10:41
→ bauruine joined10:42
→ nevyn^1 joined10:43
al-maisan → almaisan-away10:45
→ angelsl joined10:46
→ tyteen4a03 joined10:46
← kreemer left10:47
→ Titosemi1 joined10:48
← Drakevr left10:50
← Titosemi left10:50
→ MikhX joined10:50
→ ysiad joined10:51
← j416 left10:52
→ ctp joined10:53
→ fr0sty-away_ joined10:53
→ pheaver joined10:53
← fr0sty-away left10:54
← galderz left10:55
← tyteen4a03 left10:55
← ysiad left10:56
yaotti → yaotti_aw10:56
← j0ran left10:57
→ j0ran joined10:57
→ tyteen4a03 joined10:57
→ TomyLobo joined10:58
← HopsNBarley left10:59
→ HopsNBarley joined11:00
← HopsNBarley left11:00
→ HopsNBarley joined11:00
← tyteen4a03 left11:02
→ Drakevr joined11:03
← pasky left11:04
← mastro left11:04
← miloshadzic left11:05
→ tyteen4a03 joined11:06
← ereslibre left11:09
← tyteen4a03 left11:10
→ unon joined11:10
→ xiong joined11:11
→ mahound` joined11:13
← nadim_ left11:13
← mahound` left11:13
→ mahound joined11:14
mahound hello11:14
→ tyteen4a03 joined11:14
mahound does anybody know of a conflict solving tool that displays where each file comes from, i mean, the commit11:14
?11:14
TomyLobo git blame11:16
cbreak_work git blame tells you where each line in the file comes from11:16
TomyLobo git blame <filename> [-L <from>,<to>]11:17
cbreak_work make your terminal wider11:17
or you won't see anything :)11:17
mahound no, i mean, when conflicts exist in a merge/rebase operation, git mergetool opens a diff tool. however the filenames are shown as .LOCAL .REMOTE... is there a way to display commits instead?11:17
← luxurymode left11:17
mahound i always get terribly confused11:17
TomyLobo or output to a text file and open it11:17
→ clarkfischer joined11:18
mahound brb, lunch11:19
← steph021 left11:19
← goshawk left11:21
→ arnavk joined11:21
← sneakret_ left11:21
→ sneakret joined11:21
→ steph021 joined11:22
→ napster_123 joined11:22
← nanotube left11:22
→ ereslibre joined11:22
→ picasso_ joined11:22
← tyteen4a03 left11:23
→ troyt joined11:23
← sftp left11:23
→ nanotube joined11:24
→ sftp joined11:24
→ mastro joined11:25
→ avinashhm joined11:26
→ kampasky joined11:27
← kampasky left11:27
← workmad3 left11:27
→ kampasky joined11:27
→ tyteen4a03 joined11:27
→ vitor-br joined11:27
→ MetaCosm joined11:29
← MetaCosm left11:29
→ MetaCosm joined11:29
← MetaCosm_ left11:30
→ spaceonline joined11:30
← RobertMeta left11:30
→ GarethAdams|Work joined11:32
→ miloshadzic joined11:32
← zorzar_ left11:33
surikator hi guys... is there anyone here using git in ocamlforge? i can't find the git url for developers to commit to a particular project. thanks!11:34
→ jankoprowski joined11:35
→ samtuke joined11:36
← pheaver left11:37
→ pyther joined11:37
→ Cromulent joined11:38
samtuke How do I forget local commits and change my local copy, and commit history, to match the remote (non-bare) head? I've tried git reset --hard, git checkout ., but my local repo still states that it is two commits ahead of origin/master. I want to forget these two commits altogether!11:38
jankoprowski Hi11:39
→ artefon joined11:39
jankoprowski What is dedicated fastest method to serve Git repositories over https11:39
cbreak_work it's faster to do it via ssh11:40
← markelik_ left11:41
← torvalamo left11:41
→ Athox joined11:41
Athox → torvalamo11:42
← nevyn^1 left11:43
jankoprowski cbreak_work: Yeap - I know that. Unfortunately in my company this should work over https11:44
cbreak_work: becase of delivering ssh public keys issue11:44
→ krfs joined11:44
→ tvw joined11:45
← j0ran left11:45
→ j0ran joined11:46
divVerent how is https better there?11:46
delivering https client certificates is better? ;)11:46
jankoprowski We have our internal SSL certificate center11:47
shruggar samtuke, try: git remote update && git reset --hard <the/remote/ref, eg: origin/master>11:47
→ zorzar_ joined11:49
samtuke shruggar: Yes that worked, I just found it independently in a forum post. I used git reset --hard <remote_commit> and local commits were forgotten. Thanks!11:49
jankoprowski git-http-backend over cgit?11:49
or in backword?11:49
shruggar samtuke: if this is a security issue, note that the commits are not actually "forgotten" yet11:50
they're just not reachable from the current HEAD11:51
samtuke shruggar: hmm, when will they be forgotten? presumably they are somewhere on my local machine?11:51
divVerent jankoprowski: and now the two evil questions :P11:51
→ bitkiller joined11:52
divVerent - how do you deploy the SSL client cert so that git finds it?11:52
← BiggFREE left11:52
shruggar samtuke, at the moment you can see them with git log HEAD@{1} (the position of HEAD 1 action ago)11:52
divVerent (don't know how, but it sure is possible)11:52
← devanjan left11:52
divVerent - how do you do the same, with the extra requirement that git shouldn't ask EVERY SINGLE TIME for the client cert passphrase?11:52
→ MetaCosm_ joined11:52
divVerent just saying, git with https works (especially together with gitolite)11:53
but is a bit annoying11:53
samtuke shruggar: interesting, and important to bear in mind. Thanks for the tip11:53
shruggar samtuke, they're all exactly where they were before, and are referenced in the reflog. You can kill them with "git gc --prune", but don't do that unless it's really a security problem having those commits around11:53
← arnavk left11:53
divVerent so much, I bet many people will end up saying "screw the certs, I'll use user/pass"11:53
thiago jankoprowski: ssh keys aren't delivered. They are generated and uploaded to the server.11:53
→ marcob joined11:53
divVerent thiago: same thing11:53
thiago jankoprowski: I doubt that Git will use client-side SSL certificates11:53
divVerent git will11:53
if you manage them to put them so libcurl finds them ;)11:53
some dot-directory11:54
probably strace is the easiest way to find out where it is supposed to be :P11:54
thiago yep11:54
divVerent still... in corporate environments, you often DO deliver SSH private keys11:54
like you do for SSL11:54
→ luxurymode joined11:54
jankoprowski Simplicity11:55
divVerent with a "CA" server providing encrypted archives with SSH private keys, and authorized_keys files11:55
jankoprowski I will use username and password11:55
← AlexP left11:55
divVerent and doing so has one simple security reason...11:55
jankoprowski because in SSH scenartion someone must upload ssh public key to gitolite11:55
divVerent no big diff :P11:55
typically, CA servers or other "ultra secure" instances do not allow upload of ANY data11:55
but download only11:56
shruggar anyone know of a good online guide which would explain all of these concepts? I'll be needing to set up some internal SSL stuff eventually11:56
divVerent so the typical way is to generate on that server pub and priv key11:56
→ cesario joined11:56
divVerent jankoprowski: you can automate that easily, though11:56
but sure11:56
do as you please :P gitolite supports it, so why not11:56
jankoprowski But my manager told me I they don't give me time for this11:56
so just answer for my question11:57
divVerent just be prepared that setting up https nicely is a bit difficult11:57
jankoprowski instead trying to be smart11:57
thiago the key is always generated on the local machine11:57
both for SSL and SSH11:57
divVerent thiago: only for private setups :P11:57
thiago for all11:57
divVerent corporate setups often disallow that11:57
thiago the public part of the key is uploaded to the server11:57
divVerent thiago: also for public CAs11:57
thiago: again, in theory11:57
thiago the private part never leaves the local machine11:57
divVerent I worked at a bank11:57
there the ONLY way was to have the CA server generate both11:57
thiago web browsers have functionality to generate keys like that11:57
thiago would NEVER trust a key generated by someone else11:58
jankoprowski join #git-pl11:58
divVerent thiago: I agree11:58
_I_ didn't trust it eihter11:58
the bank did ;)11:58
thiago my bank had similar features11:58
divVerent just saying, this is how it usually works in big corporates11:58
thiago my browser generated the key11:58
← Cromulent left11:58
thiago it uploaded the cert with the public key to the bank11:58
divVerent browser generates server certificate?11:58
that is new :P11:58
thiago client certificate11:58
divVerent yes, client certs, sure11:58
for that there is functionality11:58
thiago server certs is from the server11:59
divVerent again, NORMALLY11:59
in my bank setup, the standard openssl way was simply NOT SUPPORTED11:59
only way was downloading an archive with both private and public part11:59
→ workmad3 joined11:59
divVerent and I even see why they did that11:59
thiago well, that's no different than password mechanisms then11:59
clarkfischer What's the best way to delete the latest commit?11:59
divVerent any file uploaded from client to the CA server might compromise the CA12:00
that file may e.g. contain overflow exploits against the CA12:00
← freimatz left12:00
← LouisJB left12:00
divVerent so they are very paranoid about from where to where data flows12:00
only freeform field to enter was the CN of the cert12:00
so only one thing where they can possibly screw up and compromise the CA12:00
if you think of typical code quality in big corporate environments12:01
→ freimatz joined12:01
divVerent you understand why they are paranoid about direction of data flow12:01
because, it is the only security they have :P12:01
→ AlexP joined12:01
divVerent jankoprowski: anyway, want my apache config snippet for git+https?12:02
clarkfischer git reset --soft HEAD~112:02
divVerent jankoprowski: http://paste.debian.net/105517/ - works for http and https12:04
→ samjam joined12:04
divVerent and so you know it: main danger of this is that the web server SHOULD be quite dedicated when doing this12:04
as any php script has full access to the git repo12:04
is mitigated by /home/xonotic/http-push/gitolite-gl-auth-command/ being a shell script calling a suidperl script checking the username against the users in the digest_pw file12:05
so the attacker who exploited a php script is limited to the union of powers password-authed users have, and cannot e.g. change "master"12:05
→ arnavk joined12:07
← giallu left12:08
Caelum how do you update the msysgit full installer?12:11
jankoprowski divVerent: Thanks12:11
divVerent: Now I'm thinking is cgit can be faster12:11
divVerent maybe, but you still need the git http backend server :P12:12
jankoprowski ok12:13
→ corni joined12:13
← samtuke left12:16
← ericlee left12:17
YouSirOne hi i got this message when trying a "git pull". Your configuration specifies to merge the ref 'master' from the remote, but no such ref was fetched... what is wrong? how can i fix it?12:19
→ project2501a joined12:19
→ DiGitalX joined12:20
← errordeveloper left12:21
jankoprowski thanks12:22
→ pdelgallego joined12:23
→ pdelgallego_ joined12:23
← pyther left12:24
→ simonj joined12:25
jankoprowski But I still don't know which tool (gitweb or cgit) should I use.12:26
Any advices?12:26
Caelum jankoprowski: you could also try Gitalist12:28
→ nona joined12:28
jankoprowski Caelum: I'm checking right away12:31
gitalist is beutiful12:31
I love it12:31
I love them12:31
yaotti_aw → yaotti12:32
jankoprowski But example.gitalist.com looks is broken12:33
LD12:33
:D12:33
YouSirOne hi i got this message when trying a "git pull". Your configuration specifies to merge the ref 'master' from the remote, but no such ref was fetched... what is wrong? how can i fix it? anyone got an idea?12:34
this is a newly created repo, can i just remove it and start over again with the same repo name?12:34
→ icwiener joined12:36
→ HG` joined12:37
→ ExtraSpice joined12:37
← project2501a left12:38
almaisan-away → al-maisan12:39
→ boombatower joined12:40
→ Cromulent joined12:44
→ koltroll joined12:45
← tyteen4a03 left12:54
→ tyteen4a03 joined12:54
← artefon left12:54
→ gozoner joined12:56
→ goshawk joined12:56
← MikhX left12:56
→ gusnan joined12:56
mahound back12:57
← simonj left12:57
mahound so, does anyone have any tips on merging/rebasing, so that one doesn't get lost?12:58
cbreak_work define "get lost"12:58
selckin it shifts in time when you pull a big lever12:59
→ roger_padactor joined13:00
mahound selckin: haha :D13:02
cbreak_work: when i am rebasing and there are conflicts, git mergetool normally opens a diff window, but the file versions are only referred to as LOCAL and REMOTE13:03
i often get confused :/13:03
→ smuf joined13:04
cbreak_work just use the command line13:04
or a text editor13:04
mahound cbreak_work: what do you mean?13:04
selckin you'll just have to look it up and remember/learn, same with ours & theirs, its confusing as hell at the start and always will be13:04
← gozoner left13:04
cbreak_work mahound: why do you use mergetool?13:04
use a text editor instead13:04
that's what I mean13:04
mahound cbreak_work: i see13:05
→ t0rc joined13:06
← EricInBNE left13:06
mahound http://stackoverflow.com/questions/3051461/git-rebase-keeping-track-of-local-and-remote13:06
→ alberto56 joined13:07
→ galderz joined13:09
← drl left13:09
→ coderdad joined13:09
→ ch077179 joined13:09
← clarkfischer left13:10
→ NickyP joined13:11
← d0k left13:12
→ nevyn^1 joined13:14
← kylehayes left13:15
← OOPMan left13:15
← rjrodger left13:15
← arnavk left13:15
← SirFunk_ left13:16
yaotti → yaotti_aw13:20
→ juanjoc joined13:21
← borwar left13:23
← soidexe left13:25
→ soidexe joined13:25
← _rane left13:26
→ hyperair joined13:27
→ _pingu joined13:28
→ arnavk joined13:29
→ StevePro joined13:30
→ gfto joined13:30
← jankoprowski left13:33
→ galderz1 joined13:33
← galderz left13:33
← YouSirOne left13:34
→ rjrodger joined13:34
→ jaql joined13:34
← Error404NotFound left13:34
← ningia left13:38
← galderz1 left13:38
← NickyP left13:39
→ shiba_yu36 joined13:42
→ LiamH joined13:43
→ tstclair joined13:44
→ galderz joined13:44
← pgib left13:46
→ jankoprowski joined13:47
yaotti_aw → yaotti13:47
→ bcardarella joined13:50
← goshawk left13:51
← surikator left13:52
← ben_alman left13:53
← tizbac left13:53
← _pingu left13:54
← StevePro left13:56
← banisterfiend left13:57
→ LouisJB joined13:58
→ surikator joined13:59
→ engla joined14:00
← engla left14:01
→ engla joined14:01
→ banisterfiend joined14:02
→ banisterfiend`` joined14:04
← sejo left14:04
jettero for one of my remotes, I like to config push = refs/heads/*, but I wish there were some way to avoid pushing branches that start with _ ... is there a way to glob with exceptions?14:05
→ errordeveloper joined14:05
→ TheMoonMaster joined14:06
→ Pewpewarrows joined14:06
← banisterfiend left14:07
→ shentino_ joined14:07
tango_ jettero: you could try refs/heads/[a-z]* assuming no heads start with something other than [a-z_]14:07
← bauruine left14:07
← shentino left14:07
← krfs left14:08
→ kukks joined14:08
→ Mannequin joined14:09
← banisterfiend`` left14:10
→ banisterfiend joined14:10
jettero tango_: that's a great idea14:11
tango_: er... but it's not actually globs at all apparently14:11
→ urkud joined14:11
jettero fatal: Invalid refspec 'refs/heads/[a-zA-Z0-9]*'14:11
→ NickyP joined14:11
tango_ oh14:11
then I'm afraid you're out of luck14:12
jettero it seems so14:12
good idea though14:12
← skoop left14:12
← LouisJB left14:13
← tchan1 left14:13
← priidu left14:13
→ EricInBNE joined14:14
→ tchan joined14:14
→ zombor joined14:14
← pratz left14:15
← kirthi left14:15
→ LouisJB joined14:15
← TheMoonMaster left14:16
→ jxpx777 joined14:18
→ psundaram joined14:19
← j0ran left14:19
→ j0ran joined14:20
→ booginga joined14:20
→ montylounge joined14:21
→ bauruine joined14:21
← p3rg2z left14:21
← makkes left14:22
← Swimming_bird left14:22
← p3rror left14:22
→ Migaaresno joined14:24
→ mishok13 joined14:24
← gdmfsob left14:25
→ priidu joined14:26
→ nohk joined14:26
← EricInBNE left14:26
→ sejo joined14:26
← sejo left14:26
← nohk left14:26
→ ghankstef joined14:27
→ icwiener_ joined14:27
→ sejo joined14:27
Migaaresno Urrghh. can i undo my last local commit?14:28
divVerent git reset --hard HEAD^14:28
wait14:28
do you want to undo just the commit command14:28
or also the changes of it?14:28
Migaaresno Yes, i wanted to try out git commit -a -m "What does this do?"14:29
And it added all .pyc files.14:29
divVerent git reset HEAD^14:29
then14:29
this leaves your working copy as is14:29
and just undoes commit -a14:29
← icwiener left14:30
→ SirFunk_ joined14:30
→ kylehayes joined14:30
jettero maybe add *.prc to your ignore file ... then -a wouldn't have added them14:30
Migaaresno Thanks14:30
→ j0ran1 joined14:30
← j0ran left14:31
← psundaram left14:31
← kylehayes left14:33
← napster_123 left14:33
← ziro`_ left14:36
ghankstef ok am fairly new to git after working with svn for some time and trying to setup remote branches to push to and pull from. Here is what I have done so far: http://pastie.org/149274114:36
→ ziro`_ joined14:36
← ziro`_ left14:37
ghankstef what do I need to do to create remote branches to push and pull from -- say dev and production branches?14:37
Silex just push your local branches to origin14:37
git push origin foo14:37
tango_ ghankstef: a remote branch is created by pushing into it. git push remotename localbranch:remotebranch14:37
ghankstef I see thank you. Trying now14:38
Silex tango_: right, hum would simply foo instead of foo:foo also work?14:38
the real question is does "foo" expand to "foo:foo" ?14:38
tango_ I don't remember if that works when the remote branch doesn't exist yet14:39
ghankstef so create a local dev branch, check it out and pushh ot to origin?14:39
tango_ no, foo exands to foo:<config set as the remote branch for branch foo>14:39
→ p3rror joined14:39
Silex tango_: thanks14:39
tango_ so on creation it probably won't work14:39
unless you --set-upstream14:39
Silex tango_: yeah14:39
ghankstef hey I think that worked thanks!14:40
← elb0w left14:40
← sejo left14:40
ghankstef sweet!14:41
→ goshawk joined14:41
→ Ryback_ joined14:41
ghankstef yippee! on the way14:41
← maleknet left14:41
← radarek left14:42
← urkud left14:42
→ panike joined14:42
→ icwiener joined14:43
→ juangiordana joined14:44
← HopsNBarley left14:45
→ HopsNBarley joined14:45
← HopsNBarley left14:45
→ HopsNBarley joined14:45
← icwiener_ left14:46
al-maisan → almaisan-away14:48
← blee left14:48
→ Neo joined14:48
→ roxlu joined14:49
roxlu hi14:49
→ patabugen joined14:50
→ jrmuizel joined14:51
patabugen Hey, I'm having some troubles with contacting remotes over SSH and am not sure what else to check. It only works when I connect to GitHub, any other I get "fatal: The remote end hung up unexpectedly". I don't recall doing anything special for GitHub but could have and forgot. Any ideas what I should be checking?14:52
← shiba_yu36 left14:52
← alberto56 left14:53
→ OOPMan joined14:53
← khmarbaise left14:55
→ obeattie_ joined14:56
jettero you should be checking your firewall...14:56
wereHamster patabugen: does normal ssh work?14:56
jettero or perhaps your .profile or .bashrc are printing something git doesn't like.14:57
I've seen that before, but it's rare14:57
→ icwiener_ joined14:57
patabugen I can't find any problems on the firewall, regular SSH works, I'm on Windows 714:57
jettero I should say, I've seen it screw up openssh scp, but not git14:57
obeattie_ Hi, I'm having a bit of a funny problem with a git server of mine. When I try to clone over SSH I get a "fatal: protocol error: bad line length 8205", anyone know what causes this?14:57
← bentob0x left14:57
patabugen I can use my GitHub accounts over SSH too. I'm wondering if somewhere I've told something how to SSH to GitHub and it's trying that for my other server now. But I can't think what14:58
→ sivy joined14:58
→ gozoner joined14:58
bluj i have a dir with a whole bunch of .git repos within.. can anyone give me a hint of how i can 'find' each of these dir's and then git log --pretty=oneline -1 of each to show the top commit id?14:59
obeattie_ I see by looking at the source it's coming from a check in pkt-line.c but honestly I have no idea what it's doing14:59
bluj using find -exec with cd doesntw ork as i would have thought14:59
← Mannequin left14:59
jettero patabugen: you might try temporarily moving your $HOME/.ssh/ folder to another name to see if it's a setting or something14:59
→ TheMoonMaster joined14:59
← icwiener left15:00
patabugen jettero: A good idea :) But the same problem :(15:00
hyperair hmm why can't i push --tags and --all together?15:01
→ Mannequin joined15:01
jettero patabugen: when you ssh to the machine(s) that's not working... does it print a bunch of things? echo hi! echo loaded your profile?15:01
hyperair it feels so counterproductive15:01
yrlnry I love git.15:02
patabugen jettero: When I do a regular SSH there is a MOTD which comes up. Could that cause trouble? It comes up after the password though, and this is never asking me for the password.15:02
jettero patabugen: no, probably not. The motd normally doesn't cause trouble... but try this ssh host ls -115:02
does it print things besides files?15:02
yrlnry Is there any reason why this 'gitsh' is worth anything more than a curt dismissal? http://www.codedanger.com/caglar/?p=86815:03
To me it seems utterly worthless, but maybe there is something I am missing.15:03
patabugen Nope, just filed15:03
Chillaway → Chillax15:03
jettero k, no idea then.15:03
patabugen It did ask me to confirm the fingerprint though15:03
roxlu When I've got a repos already And I want to create 2 branches with totally different files, would it be possible to do: git init, git add remote origin [existing repos], git branch "mynewbranch", git add . , git commit -am "Update", git push origin mynewbranch ?15:03
jettero patabugen: oh, well, that could be the problem there15:03
panike bluj: find . -name '.git' -print | while read dirname; do (cd ${dirname%/*};git log --oneline --all --decorate --no-walk); done15:03
patabugen jettero: Thanks for trying :)15:03
It's not better =[15:03
bluj panike, cheers15:03
→ j0shua joined15:04
j0shua how do i find out where a branch was branched from ?15:04
Caelum how do I checkout a file from a previous commit?15:04
→ pgib joined15:05
← sgillet left15:05
panike j0shua: Assuming it branched from master: git merge-base HEAD master15:05
→ abetaha joined15:05
← unon left15:05
→ kylehayes joined15:06
Caelum aha, git checkout sha file # works :)15:06
→ ISF joined15:07
← curtana left15:07
j0shua panike: i don't know where it was brnached from, that is what i am trying to find out15:07
← angelsl left15:08
→ trochala joined15:09
← obeattie_ left15:09
→ obeattie__ joined15:09
→ dfr|mac joined15:10
→ mabrand joined15:11
← trochala left15:11
→ trochala joined15:11
→ dean[w] joined15:11
panike j0shua: git for-each-ref --format='%(refname)' | sed 's/refs\/heads\///'s/refs\/remotes\///;s/refs\/tags\///;' | while read ref; do git merge-base $ref HEAD; done will give you a list of branch points. Then you have to search through them manually to find the one you are interested in.15:12
→ ziro`_ joined15:12
← smuf left15:12
patrikf roxlu: yeah. you can also do git symbolic-ref refs/heads/<totallynewbranch>15:12
dean[w] I just messed up a commit. I commited to a branch I normally pull from, then tried to push it back and it came back with lots of errors. How can I undo a git push15:12
→ scelis joined15:12
dean[w] and also the commit15:12
← henrikbjorn left15:12
← obeattie__ left15:12
dean[w] current status is:15:12
# On branch devel15:13
# Your branch is ahead of 'origin/devel' by 1 commit.15:13
#15:13
→ icwiener joined15:13
patrikf roxlu: sorry, git symbolic-ref HEAD refs/heads/<totallynewbranch>15:13
← opalka left15:13
← hyperair left15:13
← blongden left15:13
→ shyam_iyer joined15:14
patabugen jettero: I fixed it! By setting up passwordless authentication in PuTTy/pAgent15:14
jettero oh, right15:15
it woudln't know how to ask you for a password from git I wouldn't think15:15
← nicoulaj left15:15
serard → Whysky15:16
← LouisJB left15:16
← icwiener_ left15:16
patabugen It worked from home that way, just not from my work machine. Could be because I'm on Linux at home.15:16
jettero sure sure, whatever mode it uses for grabbing your password doesn't work in windows maybe15:17
panike j0shua: git log HEAD --not $(git for-each-ref --format='%(refname)' | sed 's/refs\/heads\///'s/refs\/remotes\///;s/refs\/tags\///;') will help you out I think. You just use the parent of the last commit from the log.15:17
→ cannonball joined15:17
← cannonball left15:17
→ cannonball joined15:17
Whysky → serard15:17
← pgib left15:18
→ Anubisss joined15:18
← kukks left15:18
→ LouisJB joined15:19
← ymasory left15:20
← freimatz left15:20
← Tommy[D] left15:21
→ nicoulaj joined15:21
← NickyP left15:21
← jrmuizel left15:22
← dr_win left15:22
→ amondracorp joined15:22
← ph^ left15:23
→ wedgeV joined15:23
→ jvv joined15:23
wedgeV hi, i want to undo a commit, a friend did "git checkout HEAD^" now its detached from any head, what should i do next?15:24
← sonnym left15:24
yrlnry Is the commit you want to undo the most recent one on its branch?15:26
If so, you can discard it completely with "git reset --hard branch^".15:26
← xiong left15:26
→ mezgani_ joined15:27
cbreak_work wedgeV: first, get back to a branch15:27
just do git checkout branchname15:27
yrlnry Or you can check it out, change it around to make it the way you wish it had ben and then use "git commit --amend" to discard the bad version of the commit and make a replacement for it.15:27
← LouisJB left15:27
wedgeV yrlnry: i want to undo it completely it was prematurely merged15:27
→ Tommy[D] joined15:28
← p3rror left15:28
wedgeV yrlnry: do i need to be on the branch when doing "git reset --hard branch^" ?15:28
→ spearce joined15:28
← Cro_Crx left15:28
yrlnry Yes.15:28
The goal here it to make the ref point to the commit before the one that it currently points to.15:29
wedgeV ok, thanks yrlnry and cbreak_work15:29
cbreak_work git reset is a command to change the current branch after all :)15:29
← corni left15:29
yaotti → yaotti_aw15:30
yaotti_aw → yaotti15:30
← mezgani_ left15:31
wedgeV (just a theoretical question: when i would want to re-add the commit, could i just do "git reset --hard sha-of-previous-head"?)15:31
yaotti → yaotti_aw15:32
← PointMan left15:32
→ tlonim joined15:32
← gfto left15:33
tango_ wedgeV: that doesn't readd it, it scratches everything and brings your status to that hash15:33
→ Cro_Crx joined15:33
tango_ wedgeV: maybe you want to cherry-pick it?15:33
wedgeV tango_: i don't really need it, i was just wondering15:34
← jonshea left15:34
cbreak_work wedgeV: reset is not a command to remove commits15:34
it's a command to change where your current branch points to15:34
(among other things)15:34
you can make it point to a previous hash of course15:35
→ MattDiPasquale joined15:35
cbreak_work but --hard makes it nuke uncommitted changes too, which you won't get back.15:35
← bauruine left15:35
← icwiener left15:36
→ icwiener joined15:36
← ericdc left15:37
→ jds joined15:37
→ ericdc joined15:37
← Cro_Crx left15:38
→ alvaro_o joined15:40
← ericdc left15:40
→ ericdc joined15:40
← ericdc left15:41
← t0rc left15:41
→ ericdc joined15:42
→ Yuuhi joined15:42
→ mezgani_ joined15:43
← bottiger left15:43
← xiangfu left15:43
← jankoprowski left15:43
→ _jesse__ joined15:45
→ thana joined15:45
thana hey15:45
is there a way to tell git the new name of a file when i rename it without git mv?15:46
← _jesse_ left15:46
thana so git add foobar; mv foobar barfoo; git movedMagically foobar barfoo; or so ;-)15:46
broonie thana: git rm foobar ; git add barfoo.15:46
Which is all that git mv does anywa.y15:46
thana but then i loose the history...15:46
broonie git log --follow15:47
← ericdc left15:48
Unknown[OFF] → Unknown[NF]15:48
← Cromulent left15:49
thana broonie: uh nice. how does git keep track of the renaming?15:50
broonie thana: It doesn't, it infers a rename from having new and old blobs which are similar in a commit.15:51
→ abstrakt joined15:51
→ ankushbansal joined15:51
broonie ie, there's no record of a move but the tools that do output can infer one and pretty print to show any moves they detect.15:51
abstrakt hi, so I have a unique situation (which actually may not be so unique) but seems like a difficult one to deal with and I've had a hard time to this date figuring out how to make this fly with svn (which I've used up to this point until I just recently switched over mostly to git and git-svn)... so... my situation is: ...15:52
thana broonie: mmh ok. and is there a way to confirm git in its inference?15:52
abstrakt I have a web app, for this web app, plugins/extensions/modules can be written15:52
I would like to maintain essentially two different repos inside the same source tree15:53
thana so that i do not always have to add the --follow switch …15:53
abstrakt one repo for the web app and one for only my module files15:53
→ icwiener_ joined15:53
← dfr|mac left15:53
abstrakt however, they will both share some of the same folder structures, they are intertwined in that way - otherwise, all the module's actual files are separate and won't be included in the actual app repo and vice versa15:53
modules never need to alter core code15:53
there are actually lots of web apps (most of them, I would say, in fact) that are like this15:54
heck even lots of web application frameworks are basically like this15:54
so, question is, how can I have both of these two things versioned at the same time15:54
reason being, for example, say the 3rd party web app creator (e.g. wordpress or magento or whoever) makes an update15:55
I want to be able to git pull wordpress master for example and update the main/core app files15:55
← shentino_ left15:55
abstrakt and then say that a team member makes some updates to the module, I want to also then be able to git pull mymodule master to pull updates to the module15:55
← dc5ala left15:55
← icwiener left15:56
abstrakt because if I just keep one monolithic repo, I don't want the added files for my module to make it in to the main line development source tree for the base application when I do a git commit/git push15:56
also, if I do a monolithic repo, I'll have one monolithic repo for the main web app stack, and then a separate, duplicate copy of that repo essentially, with my mods included15:57
so updates to the main line development tree will be, essentially inaccessible (or at least I don't know how I would access them)15:57
→ jrmuizel_ joined15:57
← galderz left15:58
→ galderz1 joined15:58
→ shentino_ joined15:58
← benishor_ left15:59
→ winterweaver joined16:00
→ csanz joined16:01
→ ben_alman joined16:01
→ jfkw joined16:01
→ jfelchner joined16:01
→ Swimming_bird joined16:02
→ unon joined16:02
← cbreak_work left16:02
→ kukks joined16:02
← mezgani_ left16:03
→ TheMoonM_ joined16:03
→ beatak joined16:03
→ xiong joined16:03
→ smolyn joined16:03
→ ph^ joined16:04
Weust`afk → Weust`16:04
→ ROBOTARMY joined16:04
← TheMoonMaster left16:04
→ jstemmer joined16:05
→ mrsri joined16:06
← aah left16:06
← jhoepken left16:07
→ jfelchne_ joined16:08
samjam I've just suddenly got: fatal: object c53a1da3a49b8e220747b415fd40ee8d221c0a19 is corrupted16:08
I've another repo where that commit is OK. What's the spell to fix and fetch. git fetch of the branch with the corrupted item didn't fix it16:08
← jfelchner left16:08
→ transplant joined16:09
→ SimonP86 joined16:09
← icwiener_ left16:10
→ icwiener_ joined16:10
← TomyLobo left16:10
→ cptG joined16:11
→ sonnym joined16:11
→ jonshea joined16:12
← envi left16:12
← ROBOTARMY left16:13
→ smuf joined16:13
← funfun left16:14
← priidu left16:14
→ Yuffster joined16:14
← selckin left16:15
→ selckin joined16:15
← tty1 left16:15
← tchan left16:15
→ funfun joined16:16
→ tty1 joined16:16
→ kenichi joined16:17
→ tchan joined16:17
→ soimort joined16:17
→ stliu joined16:17
→ sako joined16:19
→ mezgani_ joined16:19
→ NickyP joined16:20
← yhager left16:21
samjam I can use git-pack-object to grab a good copy from the other repo, but git unpack-objects -n < packer-345d3d05cb3cebc9f656de46ffdcb5401e6029ce.pack doesn't replace the corrupted file16:21
→ Cromulent joined16:21
← soimort left16:21
→ bnovc joined16:22
→ soimort joined16:22
bnovc is there a simple way to find what revision removed a line without manually checking out various versions and looking for it?16:22
bremner bnovc: man git-blame16:23
jast bnovc: the 'git-blame' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-blame.html [automatic message]16:23
scelis bnovc: git log -S<string>16:23
bnovc that works for added lines but removed16:23
→ icwiener joined16:23
bnovc ok i'll try that thanks16:23
← sako left16:23
← icwiener left16:23
→ icwiener joined16:23
← csanz left16:23
← mrsri left16:24
samjam I've tried moving to /tmp the corrupt file and then fetching from the remote repo a branch beginning at that commit, but it doesn't do it16:26
← icwiener_ left16:26
→ priidu joined16:27
← rjrodger left16:27
→ pheaver joined16:27
panike samjam: How much work would it be to clone the remote and then push your changes on top of the new clone until you get to the corruption?16:29
→ zivester_ joined16:30
samjam the branch that has the corruption has not been developed since so I could just push the other branches and then fetch the backup. But I thought now is a good time to understand git more.16:30
panike I never had git corrupt my files. It might be a disk problem, so you really would not be understanding git better.16:31
samjam Well my power was yanked at the end of last week so I think that may be the cause.16:31
← ankushbansal left16:32
panike Indeed16:32
samjam As for udnerstanding git, I'm at the low level plumbing and want to just force it to re-accept an object even if it thinks it has it16:32
→ ankushbansal joined16:32
→ yhager joined16:32
panike git is designed to explicitly stop that from happening, either intentionally or unintentionally16:32
samjam fancy that.16:33
transplant if it thinks it has it and fsck is OK, the object should be there and in good state16:33
samjam Oh, the fsck is not ok16:33
so re-injecting the corrupt object is my idea of how to fix it, as man git-fsck offers no ideas16:34
jast the 'git-fsck' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-fsck.html [automatic message]16:34
→ _psychic_ joined16:35
samjam thanks jast with clenched teeth16:35
→ madewokherd joined16:35
panike To inject an object into the database, use git hash-object -w16:36
← mezgani_ left16:36
→ mezgani_ joined16:37
samjam Is there the opposite command I can run on the good repo?16:37
On order to get the thing to invoke git hash-object on, and know the type?16:38
panike Do you mean git cat-file?16:38
samjam could be16:38
panike git cat-file can identify the type with -t option16:38
→ Dave^| joined16:38
← Dave^| left16:39
panike Gotta run sorry16:39
← panike left16:39
→ Dave^| joined16:39
samjam thats fine, thanks kindly16:39
← jfelchne_ left16:39
samjam I think I have it16:39
← _psychic_ left16:40
→ EM03 joined16:40
← icwiener left16:40
→ icwiener joined16:41
picasso_ → picasso16:41
EM03 sorta new to git, used it for a few years now for simple stuff but, say i only want some directories to track my changes, how should I go about this?16:41
samjam great panike, that did it just fine!16:42
→ icwiener_ joined16:43
teuf EM03: I don't understand what you want to do16:43
Utoxin EM03: Uh... same way you'd track any changes?16:44
← abetaha left16:44
EM03 I'm not really sure what I need to do either16:44
← ajpiano left16:44
Utoxin It's going to be quite difficult to help you then.16:44
EM03 I use Drupal and want to have all my changes. Sometimes though I might install 10 modules and dont want to have to authorize each file / what not16:44
← mabrand left16:45
→ ajpiano joined16:45
← MattDiPasquale left16:45
Utoxin EM03: git add . ; git commit -a16:45
EM03 i use git x and when i make changes i go through and comment each one etc16:46
← koltroll left16:46
EM03 will i have to do that using that method for files etc ?16:46
→ artefon joined16:46
← m1sterf left16:47
Utoxin 'git add .' adds all files (under the current directory) to commit, including currently untracked. 'git commit -a' adds all currently tracked files' changes to the commit, and asks you for the message.16:47
← steph021 left16:47
← soimort left16:47
← icwiener left16:47
EM03 so i dont have to message each thing individually?16:47
Utoxin EM03: You've never /had/ to do that. You chose to.16:47
samjam I guess "git add ." won't add the .git folder16:48
Utoxin samjam: Correct. Or any files ignored by any method.16:48
EM03 ok16:48
this should work well then16:48
→ galderz joined16:50
← cesario left16:50
← galderz1 left16:50
→ MattDiPasquale joined16:50
charon hagabaka: you'd change the script slightly so that it does the same logic on 'git diff C^ C' instead, and then amends it to have a subject of 'fixup! ...'. or some such.16:51
hagabaka: i mean, i would have made a contrib patch or something if it was actually a well-thought-through feature. i just posted it as "this hack worked for me"16:51
→ pgib joined16:52
← thana left16:53
→ dreiss joined16:53
hagabaka oh16:53
→ panike joined16:54
← pheaver left16:54
hagabaka I guess the cause of my problem was that I was mising different sets of files with commits, e.g. A changes 1,2, B changes 2,3 C changes 1,2, so it was hard to reorder them16:54
← patabugen left16:55
← panike left16:55
hagabaka but I ended up just creating new commits based on the final versions of the files16:55
→ abetaha joined16:55
← yhager left16:56
→ yhager joined16:56
← towski left16:56
← ormaaj left16:57
← transplant left16:57
← OOPMan left16:57
→ _rane joined16:57
← stamina left16:57
→ SimonP86_ joined16:57
→ ormaaj joined16:58
← pgib left16:58
← SimonP86 left16:59
← squentin left16:59
→ Dave^|| joined17:00
→ jankoprowski joined17:00
jankoprowski Hi17:00
How force TortoiseGit to remember password?17:00
→ Schmallon joined17:00
← Dave^| left17:01
selckin use ssh keys and putty (pageant)17:01
← teweWork left17:01
← winterweaver left17:02
→ tjholowaychuk joined17:02
jankoprowski selckin: Unfortunatly my repository is not available via SSH17:07
→ napster_123 joined17:08
→ aah joined17:08
→ sako joined17:08
← smuf left17:08
← psoo left17:09
→ sejo joined17:09
→ d0k joined17:10
→ rjrodger joined17:10
← mastro left17:10
← rettub left17:11
→ justin-george joined17:11
jankoprowski I found issue raised in last year :|17:11
← dreiss left17:11
jankoprowski I can't imagine something so basic wasn't implemented yet17:11
→ __name__ joined17:11
← __name__ left17:11
→ __name__ joined17:11
→ rettub joined17:11
→ pheaver joined17:12
← pheaver left17:12
→ silence joined17:12
silence → Guest6007417:13
ben_alman if i have a short sha, is there any way to get the full sha from it with git?17:13
→ pgib joined17:13
→ ningia joined17:13
ben_alman i know that git show shortsha will show me the full sha if it's a commit17:13
patrikf rev-parse17:13
← ajpiano left17:13
ben_alman awesome!17:14
→ jhoepken joined17:14
→ ajpiano joined17:14
→ arkonova joined17:14
← saschpe left17:15
→ surikator_ joined17:17
← Guest60074 left17:17
← surikator left17:18
← surikator_ left17:18
→ albel727 joined17:18
← kelvie_ left17:20
← juangiordana left17:20
→ kelvie_ joined17:20
→ dr_win joined17:22
→ _psychic_ joined17:23
← workmad3 left17:24
← nevyn^1 left17:24
→ ROBOTARMY joined17:24
→ juangiordana joined17:24
← jankoprowski left17:25
← Mannequin left17:25
→ Mannequin joined17:26
→ moog_ joined17:28
eno__ → eno17:28
← ningia left17:28
← eno left17:28
→ eno joined17:28
→ henrikbjorn joined17:30
← juangiordana left17:30
→ hyperair joined17:30
← sivy left17:30
← _psychic_ left17:31
← nicoulaj left17:34
← Samual left17:34
→ bauruine joined17:34
→ benishor joined17:34
← eletuchy_ left17:35
→ apipkin joined17:36
← darkredandyellow left17:36
→ PointMan joined17:36
→ juangiordana joined17:37
← apipkin left17:37
← mezgani_ left17:38
→ urkud joined17:38
→ Samual joined17:38
→ ki__ joined17:39
← SirFunk_ left17:39
→ nadim_ joined17:39
→ oriba joined17:40
→ jmil joined17:40
→ sivy joined17:40
→ WebDragon joined17:41
→ _psychic_ joined17:41
← ROBOTARMY left17:42
→ radarek joined17:42
← PointMan left17:44
← napster_123 left17:44
→ ceej joined17:44
→ simosx joined17:44
← simosx left17:44
→ noahcampbell joined17:45
→ obeattie_ joined17:46
→ _pingu joined17:46
← coderdad left17:47
→ ROBOTARMY joined17:48
← joel_falcou left17:48
→ rorist joined17:49
← priidu left17:49
→ napster_123 joined17:53
→ ChUrP joined17:53
← shruggar left17:53
→ Xorlev joined17:54
→ towski joined17:55
→ Bass10 joined17:55
← Bass10 left17:56
→ Bass10 joined17:56
→ MikhX joined17:57
→ pheaver joined17:57
← pheaver left17:57
→ mezgani_ joined17:57
→ pheaver joined17:58
← MikhX left17:58
← _pingu left17:58
← alvaro_o left17:58
→ alvaro_o joined17:58
← towski left17:59
← moog_ left18:00
← rolfb left18:01
← serard left18:01
← banisterfiend left18:02
→ ningia joined18:05
→ towski joined18:06
→ banisterfiend joined18:07
→ sraue_ joined18:07
→ fowlduck joined18:08
← phantomcircuit_ left18:08
← sraue_ left18:08
→ phantomcircuit joined18:08
← ereslibre left18:09
→ eletuchy joined18:09
→ Mannequin1 joined18:09
← Mannequin left18:10
← guardian left18:10
← sraue left18:11
← _psychic_ left18:11
← tchan left18:11
→ sraue joined18:12
← banisterfiend left18:12
consumerism why does git diff tell me the mode of a file is 755 when permissions are rwxrwx--- ?18:13
→ _psychic_ joined18:15
→ krfs joined18:15
← TheMoonM_ left18:16
← ziro`_ left18:16
→ MikhX joined18:18
← radarek left18:19
yrlnry I think git doesn't actually track file permissions, except for whether the execute bit is set.18:22
→ SparFux joined18:22
→ mytoh joined18:23
→ rgr joined18:24
← ChUrP left18:25
SparFux Hi all. Why do I get "*(no branch)" when I do "git checkout <branchname>" ?18:25
← arnavk left18:26
hyperair what does "git log --describe -1" say?18:26
cbreak SparFux: what did you tell git was your branch name?18:27
SparFux hyperair: not a sane git command.18:27
hyperair SparFux: what?!18:27
SparFux cbreak: remotes/origin/master as in "git branch -a"18:27
cbreak that's not a branch18:27
that's a remote tracking branch18:27
hyperair SparFux: did you run it with or without quotes?18:27
cbreak you can't check those out18:27
→ moog_ joined18:27
SparFux hyperair: I think log doesnot know a --describe arg?18:28
hyperair SparFux: yes it does.18:28
cbreak you have to create a local branch for them with git checkout -t origin/master18:28
SparFux cbreak: ah! thx.18:28
hyperair SparFux: try it before rejecting it outright, but what cbreak said is right18:28
cbreak hyperair: I think SparFux is using an ancient git18:28
SparFux cbreak: I am? :-P18:28
git version 1.7.2.318:28
hyperair pfft18:29
that's newer than mine18:29
cbreak not that ancient18:29
hyperair SparFux: and yes, there's a --describe arg. please try it before rejecting it outright18:29
it just makes you look stupid18:29
SparFux fatal: unrecognized argument: --describe18:29
I couldn't find it in the man page. Anyway.18:30
→ coderdad joined18:30
cbreak that's because it's not a valid flag for git log18:30
SparFux cbreak: indeed. :-)18:30
→ tal joined18:30
hyperair what18:30
oh shit18:30
--decorate18:30
hyperair has been using git logd for too long =_="18:30
← amondracorp left18:31
SparFux http://pastebin.com/p7CEE264 is my branches. The local branches are without remote in the beginning, if I am right. I can delete them then.18:31
hyperair: ok :-P18:31
cbreak SparFux: why delete them?18:31
looks like you have three local branches18:32
and a few more remote tracking branches18:32
← gozoner left18:32
→ jarodluebbert joined18:32
cbreak you already have a local master branch18:32
← Cromulent left18:32
cbreak just do a git checkout master to switch to it18:32
hyperair cbreak: recent gits have checkout on demand18:32
SparFux cbreak: my problem was the local master was not updated for some reason.18:32
cbreak in general, the only branches you can check out are in git branch18:32
hyperair cbreak: i.e. git checkout newbranch automatically tracks origin/newbranch18:32
cbreak hyperair: I am aware of that18:32
→ Zephyrus joined18:32
cbreak and of it's many shortcommings18:33
hyperair what shortcomings?18:33
→ irc2samus joined18:33
cbreak SparFux: you have to update it yoruself18:33
hyperair i think it's wonderful18:33
cbreak hyperair: see above :)18:33
(it doesn't work if you use a fully qualified name)18:33
hyperair cbreak: that's because some wisecrack decided to checkout remotes/origin/master instead18:33
that's a pebkac18:33
er pbkac18:34
← HG` left18:34
hyperair not a git shortcoming18:34
irc2samus hi guys, I've exported GIT_AUTHOR_NAME and GIT_AUTHOR_EMAIL but when I commit they're not recognized, any ideas of what can be the problem? thanks18:34
cbreak see? it doesn't work :)18:34
SparFux cbreak: I had it the way that I only had to issue "git pull" to update to the latest git.18:34
→ HG` joined18:34
cbreak SparFux: you have to pull from somewhere18:34
is your upstream configured?18:34
hyperair SparFux: so basically master is out of date and you want to bring it up to date?18:34
SparFux: how about git checkout master; git pull origin master18:35
← ningia left18:35
patrikf irc2samus: how do you check? are you sure you're not actually looking at the committer info?18:35
← Guest46173 left18:35
SparFux hyperair: hm... still the log has latest entry of 8th dec 2010.18:36
irc2samus patrikf: git commit tells me so18:36
hyperair git log --pretty="%an <%ae>"18:36
irc2samus: commit first, *then* check18:36
irc2samus patrikf: and it recommends to use --amend --author=...18:36
→ tchan joined18:36
hyperair O_o18:36
irc2samus hyperair: I am commiting18:36
hyperair SparFux: can you do... git log master..origin/master18:37
← shyam_iyer left18:37
SparFux thinks he has to read the git book again, too long ago :-\18:37
SparFux hyperair: this is empty.18:37
hyperair SparFux: therefore the branch is up to date.18:37
SparFux hyperair: but git log give me last entry of 8th dec and the http://cgit.freedesktop.org/nouveau/linux-2.6/ says, master has changed 17hours ago. :-\18:38
hyperair SparFux: can you do git remote show origin18:38
SparFux: i suspect that your "origin" remote repository is not the nouveau git repository18:39
→ asmega joined18:39
irc2samus this is the console output: https://gist.github.com/79368518:39
→ Ownage joined18:40
Ownage can I tell git to ignore ssl warnings18:40
hyperair irc2samus: the Author is correct. the Committer is wrong.18:40
irc2samus: and git is telling you to set your user.name and user.email so that your Committer will be correct.18:41
→ ChUrP joined18:41
→ Cromulent joined18:42
SparFux hyperair: but it is: http://pastebin.com/UM0Nu9uc18:42
irc2samus hyperair: by setting that is the env still recognized?18:42
← rjrodger left18:42
hyperair irc2samus: yes.18:42
irc2samus ok thank you :)18:42
hyperair irc2samus: if you don't set your author name and author email, it will default to user.name and user.email in the config. same goes for committer name and email18:43
← Mpenz left18:43
→ radarek joined18:43
→ Evious joined18:43
Silex → silex18:43
← jcdny left18:44
hyperair SparFux: that's weird. can you do git log -1 origin/master?18:44
← unon left18:45
hyperair SparFux: by the way, it looks like you've done something funny with the remote branches18:45
SparFux yes: http://pastebin.com/b9kUb3sG Well, the former paste says, the local master would be out of date.18:45
hyperair no wait18:45
→ patmaddox joined18:45
hyperair do git remote prune18:45
SparFux hyperair: which name to prune?18:45
hyperair er nevermind18:45
that can come later18:45
for now...18:45
git fetch origin18:45
what output do you get from that?18:45
SparFux I tried that already, but I need a name to prune.18:45
hyperair SparFux: git fetch origin18:46
now.18:46
SparFux hyperair: fetch origian pauses 1 sec and then stops without any output.18:46
hyperair weird18:46
SparFux hyperair: no output.18:46
hm...18:46
hyperair SparFux: git fetch -v origin18:46
← dpino left18:47
SparFux http://pastebin.com/9U3Q7FrC18:47
strange !18:47
← tatsuya left18:47
SparFux it says up to date.18:47
hyperair urgh18:47
← jonshea left18:48
SparFux perhaps the website is wrong about the update 17hours ago.18:48
hyperair no18:48
i just fetched from that repository18:48
it shows me the correct commit18:48
SparFux oh :-P18:48
hyperair try cloning again18:48
SparFux but with --depth=0 ok?18:48
→ elmob joined18:48
hyperair er no18:49
i don't know18:49
i don't do shallow clones18:49
if you're worried about the speed...18:49
SparFux hm...18:49
→ ciapecki joined18:49
hyperair do something like git clone --reference=/path/to/original/repository/18:49
that'll use the existing objects from that repository18:49
SparFux I am worried about putting network load on the dev servers.18:49
hyperair use the --reference flag18:49
← avinashhm left18:49
SparFux ah, ok.18:49
hyperair and either way, i'm sure they have plenty of bandwidth to spare18:50
→ Mpenz joined18:50
hyperair i've never had an issue with freedesktop's anongit18:50
← ch077179 left18:50
hyperair SparFux: alternatively..18:50
back to the original repository..18:50
do something like18:50
git remote rm origin18:50
git remote add origin git://anongit.freedesktop.org/nouveau/linux-2.618:51
SparFux will it just update my existing dir? or create a new one by using objects from reference?18:51
→ mw joined18:52
← kpreid left18:52
hyperair create a new one by using objects from reference18:52
on the other hand, if you choose to just delete the remote called origin, and add it again18:52
and fetch18:52
it should only download the new refs18:52
→ pheaver_ joined18:53
SparFux hm... I did the latter.18:53
← irc2samus left18:53
hyperair which one?18:53
SparFux still only 8th dec.18:53
hyperair okay, can you do me a favour?18:54
SparFux the git remote rm origian && git remote add origin ...18:54
hyperair git log --decorate origin/master18:54
i suspect you've gone and screwed your branches over grandly18:54
SparFux warning: refname 'origin/master' is ambiguous. !!!18:54
hyperair e.g. if you create a local branch refs/heads/origin/master18:54
yeah, see18:54
→ guardian joined18:54
hyperair that's what happened18:54
git br -D refs/heads/origin/master18:55
do that18:55
and then git log origin/master18:55
i suspect what you did was git branch origin/master18:55
or git checkout -b origin/master18:55
SparFux error: branch 'refs/heads/origin/master' not found.18:55
hyperair wtf18:55
find .git/refs18:56
as in, type that in the terminal18:56
and see what output you have18:56
→ shyam_iyer joined18:56
← Evious left18:56
SparFux http://pastebin.com/jHfRewSZ18:56
← pheaver left18:57
pheaver_ → pheaver18:57
SparFux have to get afk 20 min. bb later. sry.18:57
hyperair i need to sleep.18:57
you'll have to ask someone else when you return.18:57
→ Evious joined18:57
← yotaff left18:58
← krfs left18:58
hyperair oh hey18:58
it's git branch -D origin/master18:58
not refs/heads/origin/master18:58
← Schmallon left18:59
→ TheMoonMaster joined18:59
← moog_ left18:59
← spaceonline left18:59
← gusnan left18:59
→ j416 joined19:00
← bauruine left19:01
← mezgani_ left19:01
← _psychic_ left19:01
← snearch left19:02
→ bauruine joined19:02
← jrmuizel_ left19:02
→ VAB3LY5 joined19:03
→ excess^ joined19:03
→ Textmode joined19:03
→ napster_123_ joined19:03
← napster_123 left19:05
napster_123_ → napster_12319:05
→ kpreid joined19:05
← xiong left19:05
→ peshalto joined19:05
← urkud left19:06
→ kipras joined19:07
→ ch077179 joined19:07
→ TheFox21 joined19:08
→ VAB3LZY joined19:08
← ch077179 left19:08
→ ch077179 joined19:09
← wedgeV left19:10
→ spaceonline joined19:10
← ptl left19:10
← PigeonFriend left19:11
← kipras left19:12
→ Buglouse joined19:12
← Buglouse left19:13
→ dreiss joined19:13
→ bartek joined19:14
→ PigeonFriend joined19:15
→ dunpeal joined19:15
SparFux hyperair: yeah, said to just delete the local branches :-P19:15
→ theoros joined19:16
dunpeal Hi. I have head foo that branched off master at some point. How do I get the commit just before the branch point, i.e. the last common ancestor of master and foo?19:16
→ enherit joined19:18
j416 dunpeal: have a look at git-mergebase19:18
← Zhonghua left19:19
← coderdad left19:19
← ISF left19:19
← engla left19:20
→ kipras joined19:20
bartek Hi there. Just converted our team to git today and one question: The git tree is riddled with various "Merge branch 'master' of <origin url>" commits. We have 6 people on our team and it's basically when I pull in changes from someone else. This seems silly and I think we're doing something wrong. Would --no-ff be the solution to make these merges actually show useful information?19:20
benishor git pull --rebase origin master19:20
or whatever your branch is19:21
bremner you could, uh, type useful commit messages.19:21
bartek bremner: We do?19:21
bremner oh, I see. yeah, maybe no-ff would help, not sure.19:21
benishor bartek: have you read about rebasing?19:21
→ krfs joined19:21
bartek I think something is not right with our flow, but I want to verify. The tree just has a lot of it, but we're playing around a lot right now19:21
← Mab879 left19:21
bartek benishor: I understand rebasing in the form of taking a set of commits and re-organizing and/or deleting them (git rebase -i for example), but not within a pull19:22
← obeattie_ left19:22
← luxurymode left19:22
→ engla joined19:22
← rgr left19:22
benishor bartek: try it out19:22
→ Mab879 joined19:22
← EM03 left19:22
bartek Alright :)19:22
benishor basically it makes a merge with remote and then puts your changes on top19:22
so the log won't be filled with ridiculous merges19:23
bartek Ah very nice19:23
← Cromulent left19:23
bremner bartek: generally many merge commits means the merges are non trivial. You may find rebasing the topic branch against the "master" branch before you merge helps.19:23
cannonball benishor: I've not tried the rebase during the pull. I always checkout master, pull, checkout master_topic_branch and then rebase against master. You're saying that I can just stay in my topic branch and git pull --rebase master and it will do all those steps for me?19:23
farktronix bartek: you may want to use a more structured branching/merging model. my team has been using this one for a couple months now and it works great (as long as everyone follows the rules) http://nvie.com/posts/a-successful-git-branching-model/19:24
← bcardarella left19:24
bartek Now if I'm doing git pull --rebase, but the designers are doing just a git pull -- There won't be any issues when they eventually push their changes will there?19:24
→ ericdc joined19:24
_jesse__ → _jesse_19:24
benishor cannonball: that's how I play around ... do work, commit, git pull --rebase origin master, git push origin master19:24
bartek farktronix: We have looked into that. Our strategy right now, since many of our team is new to git is simply to let our model evolve and get everyone used to git first. I really want to get into using that model, though19:24
← tjholowaychuk left19:25
bartek But thanks for the link, it's a nice model to follow19:25
→ tjholowaychuk joined19:25
→ joel_falcou joined19:25
→ pyther joined19:26
← galderz left19:27
← engla left19:27
→ banjiewen joined19:27
← kipras left19:27
→ engla joined19:28
→ _psychic_ joined19:28
← engla left19:29
→ engla joined19:29
→ kipras joined19:29
← jxpx777 left19:29
← kipras left19:29
→ richcollins joined19:30
→ bmahe joined19:31
→ Chillance joined19:31
← pfrenssen left19:34
← Nickeeh left19:34
→ oracal joined19:34
→ alberto56 joined19:35
→ Nickeeh joined19:35
→ stamina joined19:37
← bluenovember left19:37
← patmaddox left19:38
→ kipras joined19:38
→ galderz joined19:38
← galderz left19:38
→ galderz joined19:38
→ Utkarsh joined19:38
→ emptyflask joined19:39
← icwiener_ left19:39
→ coderdad joined19:40
← vitor-br left19:41
← Utkarsh_ left19:41
← kipras left19:41
→ Utkarsh_ joined19:42
→ andrewvos joined19:43
→ banisterfiend joined19:44
← ankushbansal left19:44
← Utkarsh left19:45
→ vitor-br joined19:45
← tgunr left19:46
← artefon left19:46
→ Modius joined19:46
→ kipras joined19:47
← kipras left19:47
→ kipras joined19:49
→ Utkarsh__ joined19:49
← kipras left19:49
← PerlJam left19:50
→ Eridius joined19:51
→ PerlJam joined19:51
← kpreid left19:51
→ khmarbaise joined19:51
← napster_123 left19:51
← ph^ left19:52
← Utkarsh_ left19:52
→ Utkarsh_ joined19:53
→ Buglouse joined19:54
← Utkarsh__ left19:55
→ normanrichards joined19:56
→ Cromulent joined19:56
→ kipras joined19:57
← matjas left19:58
← kipras left19:58
→ workmad3 joined19:59
→ kpreid joined19:59
→ panike joined20:00
→ _psychic__ joined20:01
→ nfiniteset joined20:01
→ rjrodger joined20:01
→ bentob0x joined20:02
→ matjas joined20:02
Chillax → Chillaway20:02
← _psychic_ left20:03
_psychic__ → _psychic_20:03
→ kipras joined20:03
← Cromulent left20:04
→ NickyP_ joined20:04
jast samjam: you're welcome ;)20:05
← MikhX left20:05
→ laj joined20:06
SparFux My git checkout -f leads to this: http://pastebin.com/VEb0mnav20:06
→ cesario joined20:07
→ tizbac joined20:08
→ Agos joined20:08
→ Cipher-0 joined20:08
← Tommy[D] left20:09
← _jesse_ left20:09
← NickyP_ left20:09
← tvw left20:10
← pheaver left20:10
→ Tommy[D] joined20:10
→ pheaver joined20:10
← workmad3 left20:11
→ _jesse_ joined20:11
← krfs left20:12
→ MikhX joined20:12
← laj left20:12
← shentino_ left20:13
→ shentino_ joined20:14
← DiGitalX left20:14
← shentino_ left20:14
→ Shentino joined20:15
← Shentino left20:15
→ Shentino joined20:15
← Shentino left20:15
→ shentino joined20:15
→ Cromulent joined20:16
→ laj joined20:16
WebDragon during a topic branch off of develop, I created a third branch to hold some speculative changes to a particular file that's otherwise remained unchanged, and the forgot about it for a while. the parent branch was merged into develop, and later on in develop I added a minimal skeleton to that file. now I want to rebase that third branch so that it starts from current develop head, but the file conflicts slightly with the skeleton. How20:16
ever, I want to keep everything in the branch I'm rebasing preferentially overwriting anything that had happened to it in the develop branch20:16
is there a way to tell (git rebase develop) that I want to keep everything in the current branch over anything that it would possibly conflict with in the branch I'm rebasing this onto20:17
→ ereslibre joined20:18
→ elmo2k3 joined20:18
← elmob left20:19
→ ptl joined20:20
→ chairabanta_ joined20:20
← doener left20:20
← chairabanta_ left20:20
← Zephyrus left20:22
← goshawk left20:23
← jas4711 left20:23
← panike left20:23
→ krfs joined20:24
→ panike joined20:25
→ delinquentme joined20:25
panike git rebase -s ours develop will do the trick I think20:25
→ urkud joined20:26
→ lteo joined20:28
lteo i came across an ancient server where the git version installed is 1.4. will the latest git inter-operate with it?20:28
→ jcdny joined20:28
Eridius yes20:29
lteo Eridius: just to clarify, is that in response to my question? :)20:30
jast yes20:30
Eridius yes20:30
:p20:30
lteo lol thanks :D20:30
→ EricInBNE joined20:31
← ChUrP left20:32
WebDragon oho, there it is20:34
WebDragon goes to TIAS20:34
→ patmaddox joined20:34
← marcob left20:35
WebDragon raises eyebrows20:36
WebDragon ok something evidently didn't do what I expected20:36
panike: http://fpaste.org/SLrA/ this happened, however the rebased file did NOT keep the changes I expected it to -- i.e. the rebased branch has lost all the changes it originally contained, and now just contains the 'skeleton' changes that develop had before I moved this topic branch up the line20:38
→ ChUrP joined20:38
WebDragon am a little confused about 'already applied'20:38
→ p3rror joined20:38
WebDragon using 1.7.2.3 here20:39
(evidently I need to update soon)20:39
← jpieper left20:39
SparFux I guess my problem was, because I was using some "detached HEAD" named "master" which obviously is a very wicked thing to have.20:39
WebDragon how do I undo this and get myself back to where I was previously before attempting this rebase?20:39
panike I think you can do git reset --hard HEAD@{1} to get back to your previous HEAD.20:40
→ ysiad joined20:40
← j416 left20:40
bartek I see some blog posts saying to rebase master into your branch and some others saying don't. I have a long running branch that is fairly behind, is there anything *wrong* with rebasing master into it?20:41
← richcollins left20:42
→ gozoner joined20:42
→ SIGBUS joined20:42
panike Already applied means that the same chunk was on both branches20:43
Utoxin bartek: Have you shared it with /anyone/?20:43
bartek Utoxin: The branch? Yes, it's between two developers right now20:43
ayust rebasing is probably bad then20:43
Utoxin bartek: Then a rebase will cause problems, most likely.20:43
ayust rewriting public history tends to cause problems20:43
bartek How about a merge of master into the branch?20:44
ayust merging is fine20:44
Utoxin That's the correct way, yes.20:44
Doesn't re-write history.20:44
ayust and git will handle the eventual merge back to master properly20:44
WebDragon will that move (develop)->(productindex) back to (integration [branch off develop])->(productindex[branch off integration])...<-(develop)20:44
bartek Ugh, so blog posts with differing methods20:44
← pgib left20:44
bartek Thanks guys20:44
Utoxin bartek: Rebase is fine if you haven't shared the branch.20:44
← p3rror left20:45
ayust in general "shared history = don't rebase"20:45
bartek Hmm20:45
Good to know20:45
Utoxin I still prefer merges over rebase even for private branches, but that's a personal preference thing.20:46
panike WebDragon: Yes, that is what I am saying20:46
→ koruptid joined20:46
WebDragon okay20:46
WebDragon crosses fingers :)20:46
← tal left20:46
Utoxin I try to avoid anything that re-writes history.20:46
← koruptid left20:46
→ tal joined20:46
← albel727 left20:47
→ squentin joined20:47
WebDragon whew20:48
Cipher-0 You'd make a good Timelord, Utoxin.20:48
bartek So a workflow of: git checkout -t -b newbranch; *hack hack for a week*; git merge master; *hach hack*; git checkout master; git merge newbranch; git branch -d newbranch -- This is my understanding of how it should be done in a high-level, one-line kind of way. Anything wrong with it?20:48
WebDragon ok, that got the changes back20:48
bartek Only merging master if I really should20:48
Utoxin Cipher-0: ... How'd you know? Blast. Need to work on my disguise more.20:48
panike Maybe try git rebase -s theirs and see if that produces the right tree. If needed you can do the git reset command to get back20:49
← dunpeal left20:49
→ ISF joined20:49
← Simkin left20:49
ayust bartek: that workflow should be fine20:49
bartek Great ..20:49
Utoxin bartek: Looks good to me.20:49
bartek: That's effectively what we do here.20:49
panike Sorry, meant Maybe try git rebase -s theirs develop and see if that produces the right tree. If needed you can do the git reset command to get back20:50
bartek awesome :)20:50
WebDragon panike: ok20:50
← hagabaka left20:50
← batrick left20:51
→ Simkin joined20:51
→ batrick joined20:51
WebDragon panike: what about git rebase -s recursive ours develop ?20:51
I can't quite tell how that differs from git rebase -s ours20:51
panike Go ahead and try it. That is the only way to learn20:51
git reset is your friend when messing around20:52
WebDragon yeah, but I hate easter-egging blackbox style on my code20:52
:D20:52
← kpreid left20:52
→ p3rror joined20:53
SparFux hyperair: the solution would be for me to git checkout -f, git clean -f, git checkout -b dummy, git branch -d master, git checkout -b master, git branch -d dummy.20:53
WebDragon hm the syntax of that is incorrect20:54
the manpage says " The recursive strategy can take the following options: "20:54
panike I think git will treat "ours" as a branch rather than a merge strategy if you do "git rebase -s recursive ours develop"20:54
WebDragon one of which is 'ours' but it doesn't give an example of how to write the command .. -s recursive=ours also does not work20:54
WebDragon headscratches a bit20:55
WebDragon not sure how to pass the 'ours' option to recursive under these conditions20:55
maybe a later version cleaned up the manpage some.. let me check online20:55
panike ours, recursive, and theirs are different merge strategies, so it does not make much sense to combine them20:55
← SparFux left20:56
ayust " The recursive strategy can take the following options:20:56
ours20:56
"20:56
WebDragon http://fpaste.org/Rwz3/20:56
this is what my manpage looks like20:56
← vitor-br left20:56
← theoros left20:56
→ timepilot joined20:56
→ matjas_ joined20:56
ayust WebDragon: -X20:56
for merge strategy options20:57
-s recursive -X ours20:57
WebDragon votes to have that section of the manpage spell that out20:57
ayust yeah.. it's spelt out in the git-merge manpage20:58
and there is a -X listing in the git-rebase manpage20:58
WebDragon gave me a usage message http://fpaste.org/0kBH/20:58
ayust hm20:58
w20:58
what git version?20:58
← cm100 left20:58
WebDragon 1.7.2.320:59
→ benzen joined20:59
← matjas left20:59
← benzen left20:59
ayust it's *possible* that it was only added in 1.7.3?20:59
← ISF left21:00
→ benzen joined21:00
→ ISF joined21:00
WebDragon oh well. I can always resolve the merge conflict manually with mergetool -- I was just hoping there was a way to fast-track keeping the changes in the topic branch21:00
← Vampire0_ left21:00
matjas_ → matjas21:00
→ theoros joined21:00
← theoros left21:00
→ theoros joined21:00
WebDragon I should probably upgrade git anyway, but I have a distro-release upgrade to do as well, so I need to do that first anyway21:01
ChanServ set mode: +v21:01
← _psychic_ left21:01
benzen hi21:01
i've got a question about git-deamon21:01
← TrineLocke left21:01
WebDragon once I get the distro-release upgraded then I can see what version of git comes with the current distrover and go from there21:01
← LongBeach left21:02
panike Best of luck21:02
benzen I would like to know is there a "good way " to set up a git-deamon which will start with the server and have a custom port ?21:03
→ ereslibre_laptop joined21:03
Utoxin benzen: xinetd, at a wild guess.21:04
← cesario left21:04
→ simplechat joined21:04
WebDragon thanks panike and ayust21:05
← ereslibre left21:05
← ysiad left21:05
→ kpreid joined21:05
WebDragon do any modern distro releases even install xinetd by default anymore?21:06
→ haploid joined21:06
→ _psychic_ joined21:06
→ raton joined21:06
benzen Utoxin: ok i'll search about that, do you know another way. I mean without install anything else ?21:06
← ThomasLocke left21:07
WebDragon benzen: http://progit.org/book/ch4-9.html21:07
← quentusrex_ left21:08
→ Schmallon joined21:08
Chillaway → Chillax21:08
→ matjas_ joined21:09
← joel_falcou left21:09
← banisterfiend left21:09
→ quentusrex_ joined21:10
← bitkiller left21:10
→ joel_falcou joined21:10
→ ysiad joined21:10
benzen WebDragon: Thank you very much, i look like it's what i was looking for.21:11
← p3rror left21:11
haploid Is there any reason merging a branch back into master wouldn't create files in master that are in the branch ?21:11
→ mmw joined21:11
← matjas left21:11
benzen Utoxin: thank for your help21:11
← miloshadzic left21:11
Utoxin benzen: No problem.21:11
thiago_home haploid: the reason is that your assumption is wrong21:12
haploid: are you sure that the files are in the other branch? git log branchname -- filename21:12
← __name__ left21:13
→ p3rror joined21:13
← ptl left21:13
← NickyP left21:13
→ ptl joined21:13
haploid thiago_home: yeah, they're there and committed21:13
thiago_home ok21:14
now execute the merge. Did it succeed?21:14
← ROBOTARMY left21:14
→ ROBOTARMY joined21:15
← Neo left21:15
haploid thiago_home: yes, no conflicts. files still aren't there21:15
thiago_home git whatchanged -- filename21:16
anything output?21:16
was it deleted?21:16
→ miloshadzic joined21:17
haploid no output at all. no error, no response21:17
or rather, blank line21:17
thiago_home try again with the branch name. Do you see output?21:17
→ Vampire0 joined21:18
← TheFox21 left21:18
→ TeckniX joined21:18
→ nbjoerg joined21:19
nbjoerg fast-import question. can a marker be used for both a blob and a commit?21:19
→ Neo joined21:19
haploid actually, it worked just fine with a freshly-pulled repository21:19
nbjoerg (and do the right thing depending on the context)21:19
thiago_home nbjoerg: the last overwrites the older mark.21:19
nbjoerg thiago_home: so it won't remember both the content and the hash for the commit?21:20
← elmo2k3 left21:20
haploid into another directory, that is. maybe this particular repo is just corrupted21:20
← jhoepken left21:20
→ banisterfiend joined21:20
← ptl left21:22
thiago_home nbjoerg: a mark is exactly one SHA-121:22
nbjoerg: it can be the SHA-1 for a blob or for a commit or for a tag21:22
← khmarbaise left21:23
→ ice799 joined21:23
nbjoerg ok. I'm asking to make sure before I start depending on the behavior21:24
→ LouisJB joined21:24
← Titosemi1 left21:26
← nbjoerg left21:26
→ jas4711 joined21:26
→ bitkiller joined21:27
← quentusrex_ left21:28
→ ptl joined21:28
← ChUrP left21:30
→ m104 joined21:31
→ xelister joined21:32
xelister if there is a project on git hub, how easiest to fork it?21:32
← haploid left21:32
thiago_home I think if you click on the fork link, it forks21:33
← ericholscher left21:33
→ pantsman joined21:34
← urkud left21:34
← raton left21:35
→ urkud joined21:35
→ ChUrP joined21:37
→ priidu joined21:37
← PigeonFriend left21:37
← priidu left21:37
→ priidu joined21:38
→ ShooterMG joined21:38
← p3rror left21:39
← ptl left21:40
← bartek left21:40
→ hagabaka joined21:41
← hagabaka left21:41
→ hagabaka joined21:41
→ arnavk joined21:41
→ skoop joined21:41
→ richcollins joined21:42
picasso trying to set up a post-receive hook to update my website21:42
the site itself is a cloned repo21:42
so my post-receive hook is: cd /path/to/site && git pull21:43
← fowlduck left21:43
→ pervy_sage joined21:43
picasso it works fine when i run it myself, but i keep getting this error when it's run when doing a push:21:43
pervy_sage Is it possible to just stash away *some* changes to your work with git stash?21:43
picasso remote: fatal: Not a git repository: '.'21:43
pervy_sage Like certain files?21:43
thiago_home picasso: is /path/to/site a Git repo?21:44
← HG` left21:44
picasso thiago_home: yeah21:44
when i run post-receive myself, it updates just as you'd expect.. just when it's run from git, it somehow isn't working21:44
thiago_home picasso: inside a hook, GIT_DIR is set21:45
you should unset GIT_DIR and GIT_WORK_TREE21:45
← tchan left21:45
← emptyflask left21:45
← Cipher-0 left21:46
→ haploid joined21:46
picasso beautiful, thank you!21:46
→ tchan joined21:46
patrikf pervy_sage: --keep-index21:46
← nona left21:49
pervy_sage patrikf: When it says, "Changes alreadya dded to the index are left in tact?" What does that mean?21:50
If I do git add foo then I stash, foo is not included in the stash?21:50
Let's say I modify foo and bar. I want to stash bar only, I add foo first, then stash and foo will still remain with my modifications?21:51
Eridius `git stash` by default stashes both the staged and unstaged changes21:51
`git stash -k` will leave the staged changes alone21:51
← jarodluebbert left21:52
pervy_sage I see.21:52
→ tgunr joined21:52
← logankoester left21:52
→ fedesilva joined21:52
← zombor left21:52
← ludde left21:52
→ p3rror joined21:52
→ ptl joined21:52
← pyther left21:54
← shorg left21:54
pervy_sage Is it possible to add empty directories, I forget.21:54
Or do I have to put a file in there :-/21:54
← alvaro_o left21:55
← ChUrP left21:55
→ ChUrP joined21:55
→ tfoote joined21:55
consumerism can git add be smarter about interpreting filenames? i have a bunch of long paths in my repository and i'd like to be able to just say git add file.ext rather than git add /some/huge/long/annoying/path/to/file.ext. what can i do?21:56
← jaql left21:56
← matjas_ left21:56
← cptG left21:58
← p3rror left21:58
← TheMoonMaster left21:58
→ gusnan joined21:58
→ Mosterd joined21:59
← kpreid left21:59
patrikf can't you just cd there? how do you specify the path to other programs?!21:59
but you can of course use find21:59
→ ChUrP_ joined22:00
← ChUrP_ left22:00
← ChUrP left22:00
patrikf or add -i22:00
← _psychic_ left22:01
← robinbowes left22:01
→ Athox joined22:01
← wyvern` left22:02
← jcdny left22:02
← eletuchy left22:02
← torvalamo left22:03
→ benishor_ joined22:03
← jstemmer left22:04
→ vitor-br joined22:04
→ _psychic_ joined22:04
← Cromulent left22:04
→ matjas joined22:04
arkonova Does nesting git repositories a bad thing?22:05
Context: Managing Apache's sites-available with a git repo while having /etc/ maintained with etckeeper.22:05
→ toni1 joined22:05
← benishor left22:06
← matjas left22:06
← mmw left22:07
← crashanddie left22:08
← oracal left22:08
← banisterfiend left22:09
← montylounge left22:09
Mosterd I got a hard to understand error.22:09
http://pastebin.com/tX7a3FgB22:09
→ psoo joined22:09
← toni1 left22:10
← Migaaresno left22:10
→ brokencycle joined22:10
Mosterd And when I try it again, it loops forever.22:10
← Athox left22:10
Mosterd s/loops/hangs (It just freezes completely)22:11
← benzen left22:11
← kylehayes left22:12
pervy_sage Eridius: ha, worked like a charm22:13
→ pheaver_ joined22:13
→ p3rror joined22:13
→ markelikalderon joined22:13
ChanServ set mode: +v22:14
← asmega left22:14
brokencycle Hi!22:14
→ artefon joined22:15
→ robinbowes joined22:15
→ torvalamo joined22:15
→ raboof joined22:15
raboof is it possible to have a git repo on one machine, and gitweb (or some other web interface) on another? I'd like to give users the possiblity of downloading the lastest master snapshot as a tarball without installing git themselves.22:16
← pantsman left22:16
brokencycle I have a local and a remote repo, with local = remote - 1. Now I'd like to delete the additional commit in remote, but didn't yet figure out how.22:16
← pheaver left22:17
pheaver_ → pheaver22:17
brokencycle raboof: I do this with redmine + plugin, but redmine needs to run on the same machine where you have the repo. Maybe you can use NFS or something?22:17
← stliu left22:17
← panike left22:18
← Dave^|| left22:18
→ mikepack joined22:19
→ banisterfiend joined22:19
← zivester_ left22:21
→ workmad3 joined22:21
→ jcdny joined22:22
→ friskd joined22:24
← ereslibre_laptop left22:25
→ makkes joined22:26
← psoo left22:27
← makkes left22:27
← starcoder left22:27
← pheaver left22:27
← LiamH left22:28
← Layke left22:29
→ pheaver joined22:30
→ PigeonFriend joined22:30
← mw left22:30
→ starcoder joined22:32
← Mannequin1 left22:32
Davey So... I'm trying to replay changes to a repo... is it possible to "commit" (local repo) under someone elses name?22:33
so that when I push it to github it is attributed to them22:33
← pdelgallego_ left22:33
← pdelgallego left22:33
ayust yes22:34
--committer or --author22:34
as appropriate22:34
← Ryback_ left22:34
ayust actually only --author for "git commit"22:35
git log can take either :)22:35
← ExtraSpice left22:35
ayust you can also use -c <sha> to copy the info from another commit22:35
Davey ayust: I'm coming from SVN22:36
→ Mannequin joined22:36
ayust okay, then you'll want the --author="Foo Bar [email@hidden.address] syntax22:36
→ project2501a joined22:36
Davey trying to get a perfect git history — luckily it was all one branch (trunk) up till the last 90-something commits22:36
j22:36
ust22:36
gonna patch by hand and commit as I go :)22:36
ayust: will it say it was committed by me then? I'd rather it be identical to as if they had done it22:37
ayust it will say Author: Them; Commiter: You22:37
Davey if I set both to same thing, it'll be as if they had done it?22:37
or can I just do committer?22:37
charon Davey: look up the variables in man git-commit-tree and perhaps also look into git-filter-branch which can do a batch rewrite of author info22:38
jast Davey: the 'git-commit-tree' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-commit-tree.html [automatic message]22:38
ayust I don't believe you can set committer via command line options; you might be able to do it via environment variables tho22:38
← rjrodger left22:38
Davey oh, and what about date/time? I can set that in svn after-the-fact, can I do that in git?22:38
Davey thinks his answer might be on that page22:38
ayust GIT_COMMITTER_NAME="Foo Bar" GIT_COMMITTER_EMAIL="fbar@domain.com" git commit ...22:38
← sonnym left22:38
→ freq9 joined22:39
→ shmore joined22:39
← radarek left22:39
ayust you can set the date via the GIT_COMMITTER_DATE var22:39
charon Davey: you can't set anything after the fact; as one more eloquent gitter put it, you cannot change, you can only rewrite and forget22:39
(forget the pre-rewrite commits, that is)22:39
Davey there are GIT_AUTHOR_DATE and GIT_COMMITTER_DATE, I wonder if those apply for `git commit` too22:40
charon they do22:40
Davey well, problem solved :)22:40
of course, svn dates are different formatted, :)22:41
ayust shell scripts are you friend22:41
→ Bisu[Shield] joined22:42
→ patuc joined22:42
← patuc left22:42
← guardian left22:43
charon Davey: if you do an import from svn, why aren't you letting git-svn handle the gory details?22:43
→ patuc joined22:43
Davey because of some boneheaded moves in SVN that it doesn't like22:43
→ pdelgallego_ joined22:44
→ pdelgallego joined22:44
→ eletuchy joined22:45
← Buglouse left22:45
← dr_win left22:46
← SIGBUS left22:46
← pheaver left22:46
patuc Is it the expected behaviour that git doesn't create submodules in the filesystem if I have a .gitmodules that's already populated with submodules in an otherwise empty repo and run git submodule update --init?22:46
git submodule add works as expected, but I expected the above to do the same.22:48
← banisterfiend left22:49
← p3rror left22:49
Chillax → Chillaway22:50
→ angelsl joined22:50
← abstrakt left22:50
Weust` → Weust`afk22:50
scelis Davey: I ran into a similar situation. I worked around it by doing multiple clones in git-svn and then rebasing the multiple git repos into a single git repo22:51
→ lol_h joined22:51
scelis Davey: it worked surprisingly well22:51
← ch077179 left22:51
← raboof left22:51
← bentob0x left22:51
← angelsl left22:51
→ angelsl joined22:52
→ fr0sty-away joined22:52
← ciapecki left22:53
→ gilaniali joined22:53
← gilaniali left22:53
← artefon left22:53
→ clarkfischer joined22:54
← fr0sty-away_ left22:54
scelis Davey: (at least this is for when the root of the project within the svn repository moves, which yeah, git-svn hates)22:55
yaotti_aw → yaotti22:55
← Yuffster left22:56
→ kpreid joined22:56
← arkonova left22:57
← ki__ left22:58
← kpreid left22:59
← clarkfischer left22:59
← simplechat left22:59
← Agos left23:00
→ fedesilva_ joined23:00
→ autojack joined23:01
← fedesilva_ left23:01
→ Mannequin1 joined23:02
← Mannequin left23:02
autojack I need to create a git post-receive hook that knows what branch was just updated. i.e. when someone does a git push for branch "foo" I need to be able to pull "foo" into a variable in my hook script to do stuff with hit. anyone know how I can do this?23:02
→ pheaver joined23:02
← pheaver left23:03
← jmil left23:03
→ pheaver joined23:03
← Mannequin1 left23:04
→ pfrenssen joined23:05
← fedesilva left23:05
→ radarek joined23:05
← mikepack left23:05
→ Yuffster joined23:05
ayust autojack: the post-receive hook receives input on stdin which lists each ref being updated.23:06
grep for refs/heads and then parse out the rest of the ref name, and you'll have the branch name.23:06
autojack ayust: cool, that sounds good. thanks!23:07
→ zombor joined23:07
ayust if your hook only needs to run once per push, even if that push updates multiple refs, you may wish to use the post-update hook instead23:07
nevermind, i was thinking of update vs. pre-receive. ignore that last line23:08
← skoop left23:09
→ linusoleander joined23:09
← patuc left23:09
→ patuc joined23:09
→ Pupeno joined23:09
→ KipMacy joined23:09
← excess^ left23:10
← Utkarsh_ left23:10
← m104 left23:11
→ sonnym joined23:11
← zombor left23:11
→ M1ghtyDOC joined23:11
← beatak left23:12
→ beatak joined23:12
← spaceonline left23:12
← beatak left23:12
← rorist left23:12
→ beatak joined23:13
patuc can I create submodules in an otherwise empty repo by creating/dropping in a .gitmodules and then running git submodule update --init (or some other command?) rather than adding them one at a time with git submodule add?23:14
→ carutsu__ joined23:15
← coldboot left23:15
→ demet8 joined23:15
Eridius patuc: that should work23:16
← carutsu__ left23:16
→ carutsu joined23:16
← ben_alman left23:17
carutsu how can set the commit date to today instead of when the commit happened when merging a branch?23:19
→ theoros_ joined23:19
← shyam_iyer left23:19
← cappslocke left23:19
← andrewvos left23:20
patuc eridius: I expected it to, but it doesn't seem to fetch any of the submodules. That command just returns instantly with no output. Any ideas?23:20
→ andrewvos joined23:21
demet8 I have git set up on my computer. I am having trouble pushing files to git hub. When I enter cd ~/.ssh to check the key on my mac I am able to go in the ssh directory but the only item listed there is known_hosts. It should be id_rsa id_rsa.pub known_hosts. I thought I set up my key but maybe something went wrong. What could be the issue?23:21
← theoros left23:22
← TeckniX left23:22
→ coldboot joined23:22
→ rjrodger joined23:22
← tizbac left23:23
patuc demet8: did you use ssh-keygen to generate the keys?23:23
→ zombor joined23:23
theoros_ → theoros23:24
demet8 I thought I did so this is not making much sense to me.23:24
← project2501a left23:24
→ Utkarsh joined23:25
demet8 is known hosts a key file?23:25
← d0k left23:26
f00li5h what an odd question23:26
thiago_home demet8: no23:27
demet8: your key file is missing. Put it there.23:27
← galderz left23:27
← chrischambers left23:27
demet8 @thiago_home ok thanks. I will generate a key. I thought I had but obviously not. Thanks again.23:27
patuc demet8: Follow the instructions here: http://help.github.com/mac-key-setup/23:27
← andrewvos left23:28
patuc demet8: you can skip to "Generating a Key"23:28
demet8 @patuc that's exactly where I am headed. Thanks23:28
→ andrewvos joined23:29
patuc demet8: no worries. Also, there's #github on this server.23:29
→ freq9_ joined23:29
demet8 @patuc I didn't know that thanks again.23:29
← kc8pxy left23:30
→ Pupeno- joined23:30
patuc demet8: you're welcome.23:30
← patuc left23:30
Unknown[NF] → Unknown[OFF]23:30
← zombor left23:30
← pfrenssen left23:31
← normanrichards left23:31
← Pupeno left23:31
Pupeno- → Pupeno23:31
← lol_h left23:32
← freq9 left23:32
freq9_ → freq923:32
← coderdad left23:32
→ milleja46 joined23:33
milleja46 hi23:34
← andrewvos left23:34
→ andrewvos joined23:35
milleja46 if i'm getting "git not installed" while trying to clone a git repository onto my computer though i've got tortoisegit installed what would i do to get rid of the error?23:35
← EricInBNE left23:36
→ towski_ joined23:37
thiago_home install git23:37
← eletuchy left23:37
← Utkarsh left23:37
thiago_home tortoisegit isn't git, I guess23:37
Chillaway → Chillax23:37
cbreak no23:38
it's just a GUI that uses git23:38
→ eletuchy joined23:38
milleja46 well it's supposed to be a conversion of tortoisesvn to git, huh weird...i guess i should check the page where i got it from....23:39
cbreak http://code.google.com/p/msysgit/downloads/list?can=323:39
get the first in the list23:39
you need msysgit, and ifyou want the sucky tortoise UI, also tortoise git23:39
← towski left23:39
← vitor-br left23:41
← Mosterd left23:41
← booginga left23:41
milleja46 but i already downloaded that it wouldn't boot the git stuff to clone a git unless i installed it which i went through and installed that and then told it where the mysysgit was and it appeared to load and everything until i told it where to clone23:41
cbreak hmm23:43
→ project2501a joined23:43
cbreak well, windows is weird23:43
try using the git bash shell23:43
and clone with command line23:44
→ ki__ joined23:44
← andrewvos left23:45
→ normanrichards joined23:45
milleja46 would i just run the git clone git://synfig.git.sourceforge.net/gitroot/synfig/synfig command to get this source for that program??23:46
← chairabanta left23:46
→ xiong joined23:46
← timepilot left23:47
cbreak milleja46: yeah23:47
milleja46 actually how would i tell it specifically where to output the files of that clone?23:47
cbreak by default current work dir23:49
← haploid left23:49
cbreak but the second argument tells it exactly23:49
Chillax → Chillaway23:49
← banjiewen left23:50
→ banjiewen joined23:50
← MattDiPasquale left23:50
milleja46 well when i tell it to get those files it says that the directory is not an empty directory, and i don't know what to do to fix it23:50
→ koruptid joined23:51
cbreak well, give it a path to a directory that does not exist23:51
← koruptid left23:51
milleja46 which i would do how?23:51
cbreak use your keyboard to type it's name23:52
← priidu left23:53
← starcoder left23:53
← radarek left23:54
← theoros left23:54
milleja46 i know but where in that git command would i specify it?23:54
cbreak at the end23:54
← Yuffster left23:54
cbreak git clone ssh://url/to/repo.git local/path/to/repo23:54
→ radarek joined23:55
→ Buglouse joined23:55
← carutsu left23:55
→ theoros joined23:56
→ andrewvos joined23:56
← project2501a left23:56
← andrewvos left23:57
→ andrewvos joined23:57
← cannonball left23:57
→ ben225 joined23:58
← freq9 left23:58
milleja46 um i did that just adding C:\Users\<my direc>\Documents\Synfig source but it gave the too many arguments error23:59
→ starcoder joined23:59
cbreak you have to use a normal phat23:59
not that windows stuff23:59

Logs Search ←Prev date Next date→ Channels Documentation