IRCloggy #git 2012-02-05

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.

2012-02-05

→ dinduks___ joined00:00
← dinduks_ left00:00
→ dinduks_ joined00:00
← dinduks_ left00:01
→ dinduks_ joined00:01
← JackWinter left00:02
← twisted` left00:03
→ csherin joined00:03
→ kadoban joined00:03
→ twisted` joined00:03
← bobweinerjr_ left00:03
← dinduks left00:03
← imarcusthis left00:03
← freaksken left00:03
← amichair left00:04
→ dinduks joined00:04
← dinduks___ left00:04
→ dinduks__ joined00:05
→ JackWinter joined00:05
← dinduks_ left00:06
← tgunr_ left00:06
← vandal- left00:07
← dinduks__ left00:07
← MestreLion left00:07
→ __sorin__ joined00:07
→ osmosis joined00:08
→ dinduks__ joined00:08
← dinduks left00:08
→ dinduks joined00:09
← mantic0 left00:10
→ vandal- joined00:10
→ Skyx-mobile__ joined00:12
← dinduks__ left00:12
← nixmaniack left00:12
→ dinduks_ joined00:13
← TSK left00:14
→ yuriks_ joined00:14
→ dinduks___ joined00:15
← freezey left00:15
→ dinduks____ joined00:16
← dinduks left00:17
← yuriks left00:17
← dinduks_ left00:18
← mandric left00:18
← dinduks___ left00:20
← ajt left00:20
→ macrover joined00:21
→ kgadek joined00:22
przemoc86 → przemoc00:22
← Milossh left00:22
→ mandric joined00:23
← tewecske left00:23
← mandric left00:23
→ mandric joined00:24
← LekeFly left00:24
→ LekeFly joined00:25
→ Milossh joined00:25
← yuriks_ left00:25
→ yuriks_ joined00:25
yuriks_ → yuriks00:25
→ stringoO joined00:26
← albel727 left00:27
→ arohner joined00:27
→ dinduks joined00:28
→ dinduks_ joined00:29
arohner is there a better diff algorithm for clojure/lisp I can use? git merge always makes the worst mess when merging clojure code00:29
→ Chib joined00:29
← dinduks_ left00:30
→ dinduks_ joined00:31
← dinduks____ left00:31
→ dinduks__ joined00:32
← dinduks left00:32
← dinduks__ left00:33
→ dinduks joined00:33
← giallu left00:33
→ ericm joined00:34
TheMoonMaster_ → TheMoonMaster00:35
← dinduks_ left00:35
→ disappearedng joined00:36
← jacekmigacz left00:37
→ imarcusthis joined00:37
→ wiwo joined00:37
← wiwo left00:37
← dinduks left00:38
→ dinduks joined00:38
← kadoban left00:38
← __sorin__ left00:38
→ ericm_ joined00:38
← legnaleurc left00:38
← mandric left00:40
← TL left00:40
← ericm left00:40
→ albel727 joined00:41
← dinduks left00:41
→ dinduks joined00:41
← spyro left00:43
offby1 I am not aware of language-specific diffs00:43
← ericm_ left00:43
→ dinduks_ joined00:43
offby1 arohner: I wonder if your problems are caused by whitespace -- perhaps one side of the conflict is indented more than the other.00:43
In which case ... maybe getting "diff" to ignore whitespace would suffice00:44
← Bombe left00:44
→ caseymcg joined00:44
← imarcusthis left00:44
→ spyro joined00:45
← spyro left00:45
→ spyro joined00:45
→ frogonwheels joined00:46
← Skyx-mobile__ left00:46
← dinduks left00:47
← Demego left00:47
← Rahul_ left00:48
← niko left00:49
→ niko joined00:49
→ SundarNET joined00:50
← niko left00:50
→ niko joined00:50
← dinduks_ left00:51
← dSebastien left00:52
← csherin left00:53
→ Bombe joined00:53
→ dinduks__ joined00:53
← ph^ left00:54
← AlbireoX left00:54
← mikepack left00:55
← disappearedng left00:55
← magnus_ left00:56
→ dinduks_ joined00:57
← macrover left00:57
← lonewulf` left00:58
← Bombe left00:58
→ disappearedng joined00:58
← robotmay left00:58
← dinduks__ left00:58
→ dinduks____ joined00:59
← MariusKarthaus left00:59
→ magnus_ joined00:59
→ diamonds joined01:01
← diamonds left01:01
→ diamonds joined01:01
← dinduks_ left01:02
← dinduks____ left01:03
← iocor left01:03
← barr5790 left01:03
→ dinduks joined01:05
→ uu joined01:05
→ Katniss_ joined01:05
← cooldude left01:05
→ lonewulf` joined01:06
→ dinduks__ joined01:06
→ cooldude joined01:07
← Katniss left01:07
→ dinduks___ joined01:08
→ QaDeS_ joined01:09
← dinduks left01:10
→ Katniss joined01:10
← eijk left01:11
← dinduks__ left01:12
← QaDeS left01:12
← Katniss_ left01:12
→ dinduks joined01:13
← dinduks___ left01:15
← osmosis left01:16
→ dinduks_ joined01:16
→ panthrodeb joined01:17
→ mdel joined01:17
← zulax left01:17
← panthrodeb left01:17
← dinduks_ left01:18
→ zulax joined01:18
→ mezolio joined01:19
← ferai left01:20
← dinduks left01:20
→ themachine joined01:21
← themachine left01:21
→ themachine joined01:22
← themachine left01:22
← CannedCorn left01:23
→ jefferai joined01:23
← MissionCritical left01:25
← zulax left01:26
→ zulax joined01:29
→ Bombe joined01:31
→ sudoquai_ joined01:32
← xerora left01:32
→ linusoleander joined01:33
→ MestreLion joined01:33
← zulax left01:33
→ zulax joined01:33
← scelis left01:34
← sudoquai left01:34
→ gavin_huang joined01:35
→ kennethreitz joined01:37
← neurodrone left01:38
← Textmode left01:40
MestreLion guys, i have a question about workflow. Im new to git, and this is my most common scenario: 1 - I find a bug in a software, go to upstream url, clone the repo. 2 - work on the issue, with several WIP and tests commits. 3 - after done, i want to contribute by sending 1 single patch (usually I only work in a single feature/bug). 4 - if possible, keep my commit history for personal use only. 5 - leave my cloned repo in a state where it can pull upstream i01:40
f my patch ever gets accepted. What would be the indicated procedures for that?01:40
← ben225 left01:40
→ osmosis joined01:41
← arohner left01:41
← Targen left01:41
MestreLion (I asked a similar question a few days ago... but then i already had a repo with work done. now I have no constrains, can start afresh with whatever workflow you guys suggest me01:42
→ igotnolegs joined01:43
→ xerora joined01:44
← _obiJuan left01:46
← likewhoa- left01:47
milki if you leave your work in the branch01:50
→ luxurymode joined01:50
← edude03 left01:50
milki your master should be fine01:50
→ soulcake joined01:50
milki you might want to rebase that branch against the commit before your patch01:50
but otherwise, dont mess with the branch01:51
← zulax left01:51
→ zulax joined01:52
→ MissionCritical joined01:53
→ edude03 joined01:53
← twisted` left01:53
→ twisted` joined01:53
MestreLion so i should branch as soon as i clone?01:53
→ Textmode joined01:53
milki ya, leave all your work in a branch01:54
when you submit a patch, it would be a squash of the entire branch01:54
but since you want to keep the commit history01:54
MestreLion rebase against wich commt?01:54
milki when your patch is accepted and pushed to master01:55
you would rebase the branch against hte commit before that01:55
MestreLion milki: im not sure I i understood the timeline...01:56
milki hm01:56
so you clone the repo at A01:56
MestreLion i clone, branch, work on...01:56
milki you work on a branch01:56
A - B' - C' - D'01:56
you submit your patch as a squash of B' - C' - D'01:57
MestreLion A being master head, i assume01:57
milki at that point ya01:57
but your patch is accept after a couple commits to master01:57
A - B - C - your patch01:57
you would rebase your branch against C01:57
so the branch and your patch would be equilanet01:57
equivalent01:57
← twisted` left01:57
MestreLion ok... guess i got the general idea... thank you :)01:58
as for the details, how to generate a squashed patch?01:58
milki um...01:59
MestreLion say im in D' right now01:59
milki im not sure. but one way could be to do a squash merge01:59
and submit that commit as the patch01:59
← magnus_ left01:59
MestreLion if i run --generate-patch, I get a series of files, 1 for each commit01:59
canton7 just rebase -i A. Then you get the advantage of having rebased back on top of the original branch01:59
← bartman left02:00
→ magnus_ joined02:00
milki the branch originally comes from A, so i dont think it needs to be rebased02:00
the point of the rebase im suggesting is just to keep the commit history around for the patch02:01
MestreLion canton7: that will leave me with A - D (my rebased branch) and A - B' - C' - D' , correct?02:01
canton7 MestreLion, it'll throw away the branch B'-C'-D', but you can save it by creating a new branch before you rebase02:01
← CrazyGir_ left02:01
→ CrazyGir_ joined02:01
CrazyGir_ → CrazyGir02:02
canton7 milki, if the branch your feature branch was based off has moved forward in the meantime, quite a few projects want you to rebase back on top of the original branch02:02
milki yes, that woudl be before submitting the patch02:02
but it wouldnt be A at that point then02:02
and the patch might be accepted after more commits are added02:03
canton7 I thought we were talking about before submitting the patch?02:03
milki i think we want both02:04
since MestreLion wants to keep his history around as well02:04
MestreLion im trying to understand both approaches.. since im new to git, i can't even understand what would be the difference02:04
← Textmode left02:04
canton7 goes back to the top of the scrollback02:04
→ themachine joined02:04
MestreLion yes milki , that would be nice.. i want to keep the history, even if its in a separete, isolated branch02:05
milki so when you do step 2, always branch02:05
and work on the branch02:05
to do 3, first fetch the original branch, and rebase your branch against the original02:05
MestreLion ok02:06
milki then you need to generate that squashed patch somehow02:06
→ bartman joined02:06
milki which im not sure about myself02:06
but i think you can do it with a squash merge from the original branch02:06
canton7 one way: git checkout -b branch_to_be_submitted; git merge --squash original_branch; git commit -am "My final patch"02:06
→ Vampire0_ joined02:07
canton7 another way: git checkout -b branch_to_be_submitted; git rebase -i original_branch; squash as necessary02:07
← matthiaskrgr left02:07
canton7 (the latter does the rebasing step as part of it, so you don't have to do it separately)02:07
milki so basically, you would have a clone of your feature branch, but you squashed the commits02:08
→ matthiaskrgr joined02:08
milki and that is what you submit as a patch02:08
canton7 (and by original_branch, I mean the upstream's "mainline" branch, which you created your feature branch from)02:08
MestreLion milki: and that clone would be a 3rd branch? or merged into master?02:08
milki a 3rd one02:08
MestreLion i got that canton702:08
milki you dont want to mess with master02:09
← kgadek left02:09
canton7 If master is tracking upstream's master, don't ever merge anything into it yourself (except as a "test merge" to check that your branch merges cleanly)02:09
If you start fiddling with master and adding your own stuff, then upstream adds more stuff, including one of the patches you submitted to upstream, then things get quite messy quite quickly02:09
MestreLion ok... i got master branch and work branch... then i clone work to work_final , fetch master and rebase it against master.02:10
is that?02:10
canton7 "clond" is the wrong word, but yes, that's right02:10
*clone02:10
MestreLion so no git merge then?02:11
← Vampire0 left02:11
milki after the rebase against master, you squash all the commits02:11
canton7 You can git merge --squash, which isn't a "real" merge, and doesn't affect the branch you're "merging" into02:11
MestreLion oh, ok canton702:11
in terms of git commands, let me see if i know: git clone, checkout -b work, [several commits], checkout -b work_final, fetch (or pull?) origin... then the rebase (how exactly?)02:14
← SundarNET left02:14
← diamonds left02:14
← caseymcg left02:14
canton7 fetch from origin02:14
MestreLion would it be the "git merge --squash master" ?02:14
milki you can either fetch origin and rebase -i origin/master?02:14
canton7 pull merges into the *currenty-checked-out branch*, and you've got work checked out02:15
milki or checkout master; pull origin; checkout work_final; rebase -i master?02:15
canton7 yeah, both work02:15
milki dont worry about git merge anymore02:15
you will be squashing your commits when you do rebase -i02:16
→ Bearcat joined02:16
canton7 When the editor opens from rebase -i, change the "pick" in front of the relevant commits to "squash"02:16
or "fixup", if you want to discard the commit messages :P02:16
← zivester left02:17
← brian_g left02:17
← zulax left02:17
MestreLion canton7 / milki i liked the idea of a final commit message for submitting... i don't want it to be any of my commit's message02:17
→ zivester joined02:18
canton7 You'll be dropped into an editor to edit the commit message anyway (iirc), so it doesn't really matter02:18
if not, just commit --amend after you've rebased02:18
→ ipalaus joined02:18
Bearcat folks i'm using git 1.7.3.4-r1 on a 2 gentoo system in my lan. I have a directory that is a local git repo and i'm trying to make a git repo out of it on my server box so that i can pull from the server box to the other PCs. It looks like a simple process but every instruction set i've looked at has a sligtly different way of doing it and i'm confused. I'd love a bit of hand holding here.02:18
MestreLion sounds a nice workflow... thank you very much canton7 and milki !02:19
milki++02:19
→ devbug joined02:19
MestreLion canton7++02:19
canton7 cheers02:19
→ xiagox joined02:19
devbug Submodules. Should I watch out for anything?02:19
canton7 devbug, they tend not to auto-update when you switch branches. The heads are detached by default, so watch out02:19
← Lgb left02:20
canton7 Also, currently (I believe) it requires a little bit of effort to physically move a "supermodule" repo which has checked-out submodules02:20
Though they might have fixed that02:20
← disappearedng left02:20
← as3llus left02:20
Bearcat right now i have a directory called /var/git on the server machine that i want to to push this local repo to. according to book.git-scm.com i'm supposed to just do a 'git clone --bare [localdir] /var/git/reponame.git Does the file ownership matter? Should it be root?02:21
canton7 Bearcat, question 1: Choose your protocol. Your choices are http (smart or dumb, both fairly painful), ssh (lovely), or git (read-only, unless you like unauthenticated pushes)02:21
devbug canton7: Thanks, I'll post a beware sign.02:21
Bearcat canton7: thanks ssh would be preferred02:21
canton7 devbug, the latter will catch you out if a) you're trying to update the submodules from their upstream, or b) you're trying to work in them02:21
→ Chillance joined02:22
canton7 Bearcat, question 2: Do you want fine-grained access control, or is it just you using the repo(s) ?02:22
Bearcat canton7: it's just me. I'm just cloning a .vim directory02:22
MestreLion this clone --bare sounds much better than copying over the dir :D02:23
canton7 Bearcat, question 3: Do you want your git repos under a new, dedicated user, or under your normal user?02:23
→ Lgb joined02:23
← dougsland left02:23
Bearcat looks at the "Repo Access over SSH" section02:23
Bearcat canton7: which is safer?02:23
canton7 no difference, really. It's mostly aesthetics02:23
Bearcat canton7: *more secure02:23
canton7 You can give your git user a restricted shell, so I guess that option...02:24
Bearcat canton7: probably a dedicated user.02:24
devbug Shit.02:24
Bearcat canton7: what about just using nobody:nogroup?02:24
canton7 Bearcat, cool. Create your new user. In their home dir, "git init --bare my_repo_name.git". Make sure they can login over ssh02:24
devbug I just realized the number of users in here02:25
canton7 (you can set up key authentication, but for now let's roll with passwords)02:25
Bearcat canton7: cool. I'll set up a git user and make thier home /var/git02:25
canton7 Bearcat, then you can clone that repo over ssh with "git clone gituser@yourhost:my_repo_name.git"02:25
Bearcat canton7: thanks so much for the help. I assume i'll only need that user on the machine that holds the git server?02:26
jpalmer Bearcat: are you setting up your first ever git repo?02:26
canton7 Bearcat, if you're feeling secure, set the git user's shell to /usr/bin/git-shell (or wherever your distro has plonked it)02:26
Bearcat jpalmer: yup02:26
devbug Another question (it's something I'm meaning to get around to).02:26
git post hooks (or whatever they're called)02:26
jpalmer Bearcat: doing it manually is fine. and can help you understand HOW git works.. but many of us would suggest using something like gitolite to setup and maintain your repo.02:26
canton7 Bearcat, yup. The git user lives on the machine which hosts your repos, and you then ssh in as that user (or rather get git to) in order to interact with the repos02:27
devbug What would be the best way to setup a post hook that would let me build the commit02:27
canton7 yeah, gitolite's good if you need access control. It's his first repo and he didn't need any of that, so I avoided it02:27
starlon How do I make branch 'fixes' turn into 'master'?02:27
← D-Chymera left02:27
devbug starlon: merge02:27
canton7 starlon, define "turn into"02:28
← MissionCritical left02:28
starlon Well I want to remove master, and rename fixes 'master'02:28
jpalmer devbug: you'd probably want to use something like jenkins to build the product after a commit.02:28
starlon but that's likely the wrong way02:28
justfielding dang this channel is quite active tonight02:28
→ i8degrees joined02:28
devbug jpalmer: But let's say I want to write my own script02:29
canton7 starlon, git branch -d master; git branch -m master ? or maybe git checkout master; git reset --hard fixes; git branch -d fixes02:29
devbug How do I make the hook global?02:30
starlon that's likely deltree in git slang02:30
Bearcat thanks jpalmer. I thnk i'll try it manually first. I'm trying to *learn* git02:30
→ dotnull joined02:31
canton7 Bearcat, don't be fooled: gitolite builds on the stuff you're learning here, particularly key authentication if you get into that. If you can't do this, you'll fail with gitolite :P02:31
devbug Screw it, I'll just use Jenkins02:31
jpalmer devbug: http://progit.org/book/ch7-3.html look at the two parts about pre-recieve hooks, and updates02:32
Bearcat: learning manually is probably not a bad idea. gitolite does a lot of this for you, and adds auth. so, if you decide to use gitolite later.. you're learning a solid foundation now.02:33
→ t0rc joined02:33
devbug jpalmer: I'll look into it. Thanks.02:34
→ heppy joined02:34
→ Fandekasp joined02:35
Bearcat ahh02:35
← canton7 left02:36
→ munichlinux joined02:36
Bearcat installed gitolite02:37
← abetaha_ left02:39
→ as3llus joined02:39
← ipalaus left02:40
→ zulax joined02:40
→ D-Chymera joined02:42
→ caseymcg joined02:42
Bearcat jpalmer: ok, gitolite is installed02:44
← sudoquai_ left02:44
Bearcat jpalmer: and i've made the home dir /usr/lib/gitolite for the 'git' user02:44
→ Yuuhi` joined02:44
nevyn /usr/lib/gitolite eww02:45
/srv/gitolite please. let's be all fhsy02:45
← Yuuhi left02:45
← jakob left02:47
→ jakob joined02:47
EugeneKay uses /data/gitolite, but that's me02:48
→ maletor joined02:49
← churp left02:49
jpalmer uses /home/gitolite, cuz i have bacula setup to grab /home02:54
Bearcat looks for a 'how to set up a git server with geolite document'02:54
← mezolio left02:55
→ kukks joined02:55
→ ericm joined02:55
jpalmer Bearcat: the gitolite site has a tutorial, is that not working for you?02:55
MestreLion about fetch and pull: Suppose im in a new local branch, WORK, and i want to "update" my local MASTER branch so it syncs with upstream. Do I really have to checkout MASTER before the pull/fetch?02:56
← ericm left02:56
→ ericm joined02:56
Bearcat jpalmer: ah ha! Found the instructions.02:57
MestreLion i mean... is there a way to stay in WORK, unaffected, and just "please download and update whatever is new from the remote" ?02:57
jpalmer Bearcat: cool02:58
nevyn MestreLion: so you can do a git remote update while on any branch (which is the fetch)02:58
Bearcat jpalmer: this seems like a lot of abstraction..02:58
← magnus_ left02:59
nevyn MestreLion: however to do the pull or a merge you need to be on master02:59
→ MissionCritical joined02:59
jpalmer Bearcat: indeed. thats why you saw the advice you got a few minutes ago.02:59
MestreLion nevyn: and i need a pull to update all local branches to their new commits, correct?03:00
← ericm left03:00
→ magnus_ joined03:01
← LuvMySubi left03:01
nevyn MestreLion: yes.03:01
← Liq^m left03:01
← flavius left03:02
MestreLion this is something I failed to undersand when I read GitPro book: fetch downloads all new commits, but it does not update any local branches to their new heads, right? So, if a repo has 3 branches, and new commits were made to all of them, do I need to issue "pull" in each one?03:04
→ mandric joined03:04
nevyn yes03:05
MestreLion wow, really? theres no single command that can update all my local branches for their corresponding remote state?03:06
← heppy left03:07
→ imarcusthis joined03:07
→ ericm joined03:08
nevyn well..03:08
Bearcat MestreLion: do you know bash scripting?03:08
nevyn so the idea is that branches have local commits on them03:08
a decision needs to be made as to how that local commit get's applied to upstream's new state03:08
this is why local tracking branches that don't have work on them are stupid.03:09
you can always directly just checkout the remote's branch03:09
→ intripoon_ joined03:09
nevyn MestreLion: ^^03:10
MestreLion sorry, i was afk.. and yes i do Bearcat ... but I can't believe I would need to resort to that03:10
nevyn MestreLion: so there are multiple ways of integrating a new upstream with local changes03:11
MestreLion nevyn: I never do work on local tracking branches, I always create a new branch right after the clonie... but the clone automatically creates the local versions of remote's branches03:11
clone*03:11
← Destos left03:11
← ericm left03:12
nevyn so I use master as a "for upstream" branch03:12
→ SethRobertson joined03:12
← intripoon left03:12
→ oriba_ joined03:12
→ Goplat joined03:13
MestreLion the only local commits I usually have ar the ones in my new branch, which does not exist on remote03:13
nevyn well pull is probably a bad move on local work branches03:13
you may want rebase03:14
in fact you probably want rebase03:14
← oriba left03:15
MestreLion "this is why local tracking branches that don't have work on them are stupid." - I dont understand... clone automatically creates them, don't it?03:15
← imarcusthis left03:16
nevyn so making local branches of remote branches that you never commit to is silly03:16
or never merge local changes into03:16
it just becomes overhead to keep it up to date.03:17
MestreLion I did't, clone did03:17
nevyn no clone just made the branch not using it was a part of your workflow :)03:17
MestreLion hummm03:17
nevyn note I'm not saying you should work on master.03:18
MestreLion I know03:18
← Chappo left03:18
nevyn I usually rename master after the clone03:18
to make it something like "integration for upstream"03:18
→ Chappo joined03:18
MestreLion and I wonr work on master... I prefer to keep it "clean", and do all my work in a sepatare branch03:18
nevyn then git pull it first and then rebase featurebranches03:19
→ xith_kymous joined03:19
MestreLion but I have to switch to master before that, correct? can't I "update" the branches while staying on work?03:20
nevyn well yes.03:20
you can't merge into a branch that isn't checked out.03:20
MestreLion Say upstream has 3 branches... devel, testing, master.. I cloned repo, created "mywork"03:20
nevyn yep03:21
so you cloned the repo creating master which trackes origin/master (assuming head of the remote is master)03:21
MestreLion after clone command, my repo will have 7 branches... devel, origin/devel, master, origin/master, etc etc... and "mywork"03:22
nevyn it shouldn't03:22
right so origin/master origin/devel isn't a local branch03:22
MestreLion Oh, clone only creates 1 local branch?03:22
nevyn it's the remote branch03:22
← linusoleander left03:22
MestreLion clone does not create a devel local branch?03:23
nevyn if you git fetch git checkout origin/master you don't need to pull.03:23
brb03:23
MestreLion thats great... i don't even have to checkout for the rebase03:24
→ plusk joined03:24
→ BiggFREE joined03:24
→ MaximusColourum joined03:25
← themachine left03:25
nevyn so if you rebase to origin/master you're gold.03:26
→ Targen joined03:26
MestreLion looks like renaming local master instead of creating a new branch may simplify the workflow03:26
→ Bass10 joined03:27
MestreLion how would the renaming affect the "auto-tracking" records that clone creates?03:27
nevyn it doesn't03:28
← kukks left03:29
nevyn it would still track origin/master03:29
which means you could rebase to origin/master by just using git rebase03:30
MestreLion It would make "mywork" track origin/master?03:30
nevyn yep03:30
← bwidmer left03:30
→ bwidmer joined03:30
nevyn which may not be what you want.03:30
you may want to track development or testing03:30
← Eduard_Munteanu left03:31
→ neurodrone joined03:31
MestreLion would still be the same workflow, only for "testing" instead of "master"03:32
my concern is about future updates to origin/whatever-i-based-on to conflict with my renamed branch for fetches, pulls, etc03:33
nevyn don't use pulls03:33
a pull is a fetch + merge you know that right?03:33
← plusk left03:33
→ themachine joined03:34
MestreLion Sure, sure, of course! what kind of newbie you think I am?03:34
nevyn right03:34
MestreLion quickly fires up man git pull :P03:34
nevyn so I don't like merges03:34
particularly merges from upstream into work branches03:34
Bearcat doesn't know what's wrong with 'git pull03:34
Bearcat and thought that was stadard03:35
nevyn it is03:35
Bearcat is the kind of newbie that nevyn thinks he is ^>.>^03:35
nevyn and it's fine if you have a pure tracking branch with no local modifications03:35
MestreLion nevyn: I would not like that either, nevyn03:35
(I mean.. merging upstream in my work branch)03:36
nevyn so you want to rebase there.03:36
I almost never use pull03:37
I use git remote update ( a lot)03:37
and rebase.03:37
and checkout03:37
milki Bearcat: what is ^>.>^ o.O03:37
MestreLion hummm, remote update... that's gold03:38
It's the same as fetch for each of the configured remotes?03:39
← BiggFREE left03:39
nevyn yep03:39
it fetches for every configured remote03:39
← ceej left03:40
Bearcat cat ears + looking to the side. It's similar to *cough*03:40
← Fandekasp left03:40
MestreLion man, I could even configure my repo to track several forks of a given project, and then fetch them all at once?03:42
Sounds incredibly powerful03:42
milki learn to script03:42
cbreak git remote update03:43
MestreLion I do milki ... but it sounds like I don't even need to...03:43
cbreak fetches all remotes.03:43
no need to script03:43
← caseymcg left03:43
→ gxzaga joined03:43
← fisted_ left03:45
MestreLion oh, by the way... nevyn and Bearcat both reaize I was joking about that "newbie" stuff, right? Because, yes, I didn't know a pull was an internal fetch+merge (/me blushes)03:45
for me, it was like: fetch = download objects, pull = update location of named commits (heads from branches, tags, etc)03:47
I must re-read GitPro... It's incredible how much I've forgotten in 3 months03:47
(and, worse than forget... to mix the concepts in the wrong way)03:49
anyway, thanks nevyn03:49
nevyn++03:49
← oriba_ left03:49
gitinfo set mode: +v03:49
Bearcat i'm having to go back and read up on pgp keys. I have one, i've been using it but that doesn't mean i understand it all....03:49
gxzaga hello, I need to change the history of older commit messages. Since the history is non-linear, "git rebase -i" yields conflicts when I do "git rebase --continue". I am tempted to try a "git filter-branch --msg-filter". However, I have not used this before, any other ideas?03:50
→ fisted joined03:50
MestreLion Bearcat: ssh keys != pgp keys03:51
Bearcat er..yeah that's what i meant03:52
MestreLion which one? Git uses both for different purposes03:52
Bearcat i'm not sure..03:53
MestreLion Ok, remember I'm a total newbie...03:53
Bearcat i have a local git repo that i'm trying to now set up on my central server in my lan so i can pull it around. Eventually i want external access to it.03:53
← orafu left03:53
MestreLion I use ssh keys to connect to remote repos (for example, Github)03:54
Bearcat and gitolite was reccomended to me so i'm looking that up..03:54
MestreLion and pgp keys are used to sign commits03:54
It's probably refering to ssh keys03:54
Bearcat ahh, ok ssh keys03:54
that does make more sense03:54
MestreLion :)03:55
Let me guess, Mint/Ubuntu user too? :)03:55
→ amigojapan joined03:55
Bearcat wonders if it has to be a personal ssh certificate or if it can be the same one he uses for his server03:56
Bearcat MestreLion: gentoo03:56
nevyn do you have a ssh key?03:56
Bearcat i've never heard of an ssh key. Only certificates.03:57
nevyn certificates are ssl (completely different)03:57
MestreLion Bearcat: I said Mint/Ubuntu because I use Mint... and they have a great GUI for keys that handles both SSH and PGP (as well as Gnome-keyring), so it's really easy to confuse them since you manage them all in 1 place03:57
Bearcat MestreLion: ahh03:57
nevyn: i just started looking into security stuff and this is all a bit confusing.03:57
nevyn so..03:58
keys/certificates are essentially the same thing.03:58
← maletor left03:58
→ orafu joined03:58
nevyn they're digital uniqueness used for either autenticating an entity or protecting data.03:58
or both.03:59
Bearcat nevyn: well i have ssh set up so that i can log in remotely. When i do so it askes me to accept a 'fingerprint'.03:59
← magnus_ left03:59
nevyn right. that's the server side key03:59
Bearcat i doubt that's the key as i don't recall ever generating one.03:59
ahh ok03:59
MestreLion sits down ad carefully reads the nevyn 's lesson03:59
nevyn when you log in remotely. does it ask for a password?03:59
Bearcat yes03:59
nevyn so you don't have a ssh key04:00
a ssh key is used for authentication to replace the password04:00
gitolite relies on this heavily.04:01
→ jwmann joined04:01
→ magnus_ joined04:01
Bearcat i don't understand that, really. Wouldn't that mean i'd have to carry file-that-is-the-key around with me and if i loose that anyone can get into my stuff?04:02
→ HFSPLUS joined04:02
← gxzaga left04:02
nevyn yesish04:02
so the key should be encrypted with a passphrase.04:02
→ wuj joined04:03
nevyn loosing your key is a bad thing it's not something you want on the loose04:03
Bearcat is this still a private/public key pair topology?04:03
nevyn losing it is also bad.04:03
yes04:03
sitaram topology? I dont think that word means what you think it means...04:03
Bearcat and this is not the same thing as a gpg key?04:04
sitaram Bearcat: same *technology* but not same applictaion of it04:04
← toothr left04:04
sitaram application*04:04
Bearcat *nod nods*04:04
thanks for the help i'm learning a lot here04:04
HFSPLUS fuck niggers04:04
nevyn srsly?04:04
HFSPLUS they think they are cool04:04
them and there chicken04:04
sitaram I thought we had a website you need to go to get voice...04:05
Bearcat: http://sitaramc.github.com/gitolite/gl_ssh.html (how gitolite uses ssh) may be of use. Ignore the gitolite parts -- it explains the ssh part in good enough detail04:06
the "ssh basics" section at least04:06
Bearcat thanks sitaram. I'll shut up for a while and read that :")04:06
sitaram np04:06
you're welcome!04:06
HFSPLUS they think they are cool04:06
Bearcat and you too nevyn, thank you.04:06
← HFSPLUS left04:08
Bearcat much better.04:08
→ angelsl joined04:08
→ Katniss_ joined04:09
sitaram Bearcat: now that I read it again, if you "ignore the gitolite parts" as I first said, there isn't a lot <sheepish grin> but should still be useful04:09
Bearcat "If people can't communicate, the least they can do is to shut up" -- Tom Leher04:09
← MaximusColourum left04:10
nevyn Bearcat: so it's a stronger form of authentication than passwords.04:10
Bearcat sitaram: oh, i'll read that whole thing and then look up the gento specific ssh stuff04:10
nevyn in that keys are longer.04:10
and something you have rather than "just" something you know04:10
← Katniss left04:11
sitaram and in the "many users logging onto one user" context, you can selectively de-recognise one user while not affecting the others. (If you used passwords you'd have to change the password and tell everyone else again and they would have to <whatever>)04:12
here you just remove his pubkey from the authkeys file04:12
Bearcat ahh, cool04:12
→ simplechat joined04:13
← simplechat left04:13
→ simplechat joined04:13
jpalmer !ops04:14
gitinfo [!topsecret] If your situation is so secret that you can't tell us how to !reproduce it, there's really not much we can do. Perhaps you need to find or buy support you can trust? There are people with clearances even on #git but since clearance levels are classified and #git is not a secure channel, this may not help you. However, you are protected by the biggest secret of all: We just don't care.04:14
nevyn jpalmer: :(04:14
Bearcat ha ha ha ha04:14
jpalmer oh, didn't see his K-line quit.04:15
→ ericm joined04:15
Bearcat jpalmer: i was kinda hoping they'd choke on a chiken bone04:15
*chicken even04:16
Corey jpalmer: ?04:16
Yeah, you're a bit slow. :-p04:16
jpalmer ;)04:17
← xerora left04:17
sitaram hums "it's fun to eulogise // about people you despise // it's only for a week so have no feaaar // be graeful that it doesn't last all yeaaar" (since someone mentioned Tom Lehrer)04:18
Aleric Uh... I did commits and now I'm happy.. but then I find that I'm on "no branch".. I thought I was on 'master' :/. How can I tell git that where I'm now is master and to forget about the current master (which contains a partial patch that is already in my current 'no branch' but will cause collisions if I try to merge)?04:18
← MestreLion left04:18
nevyn Aleric: git checkout -b newstuff04:19
will make a new branch from where you are right now.04:19
this will let you check what's on master and balh04:19
← ericm left04:19
→ MestreLion joined04:19
Aleric I already know what is there... I just want to move the master tag04:20
→ f0i joined04:20
Aleric What about: git branch -d master; get checkout -b master04:20
→ Katniss joined04:20
MestreLion I know ssh is off-topic, but since we are at it... I don't understand this: the 1st time i log in a server using ssh my pub keys are not at server yet, so I authenticate using regular username/passwrd (assuming I do have one at server)04:21
nevyn MestreLion: that statement confuses multiple things.04:21
sitaram including "what is the question"?04:22
nevyn Aleric: so after making newstuff you do git reset master newstuff04:22
Aleric: then delete newstuff04:22
MestreLion nevermind nevyn , Aleric seems to have a legit Git question, I can read about ssh somewhere else04:22
← Katniss_ left04:22
nevyn Aleric: doing git branch -d master && git checkout -b master is terrifying and will remove the tracking link that's part of master.04:22
→ AlbireoX joined04:23
sitaram MestreLion: it's a lazy-ish Sunday morning. People have gone far more offtopic and ssh is not really offtopic since it's a valid transport protocol for git. Ask.04:23
→ kerrick joined04:23
nevyn s/morning/afternoon/ but yes04:23
Bearcat notes that where he is, it's still Saturday. He must be slower than jpalmer04:24
sitaram s/gone/in the past gone04:24
Aleric nevyn: That didn't work ...04:24
→ Rahul joined04:24
MestreLion ok, sitaram , thanks. But I guess before asking I need nevyn 's explanation about my confused statement04:24
nevyn what didn't work.04:24
Aleric I did:04:25
sitaram MestreLion: TBH I was confused too, as in "was there a question in there?"04:25
Aleric git checkout -b newstuff04:25
git reset master newstuff04:25
nevyn ok.04:25
Aleric git checkout master04:25
nevyn and.04:25
MestreLion there is, I just didnt get there yet... but... was the statement so far that wrong?04:25
Aleric and now I have not the code that I want/had... but the old code04:25
sitaram Aleric: "git reset master newstuff" does not compute, to me04:25
nevyn so if you checkout newstuff you get the newcode04:25
ChanServ set mode: +v04:25
nevyn actually I have that wrong04:26
git reset newstuff while having master checked out should work tho04:26
sitaram MestreLion: could be, if people assumed what your *question* was. Complete it...!04:26
→ modal2 joined04:26
← xith_kymous left04:27
MestreLion sitaram: Ok. So I log in, 1st time, using regular user/passwrd . Then I could transfer my pub key to the ~/.ssh auth file so I don't have to use my username/password all the time... sounds. BUT.. I would still need to enter my ssh passphrase. So, besides enhanced security, theres no real "no need for password" gain, isn't it?04:28
← wuj left04:29
nevyn so there's a client utility called an ssh-agent which loads your key into memory for ssh to use04:29
MestreLion ssh-agent only caches my passphrase for like 15 minutes (I guess), so I would still need to re-enter it almost every time I log in04:30
nevyn so for me I arrive at work. type my passphrase to load the key into the agent. then ssh all day...04:30
sitaram MestreLion: several differences between this and plain password...04:30
nevyn that's a bizzare configuration of ssh-agent04:30
certainly not standard.04:30
sitaram MestreLion: (1) this requires that file to be available, so an attacker would need *two* things -- your passphrase *and* the file04:31
MestreLion sitaram: but they are all security-related, correct? theres no "ease of use"04:31
nevyn MestreLion: it allows greater tracability on shared accounts04:31
sitaram MestreLion: wouldn't you rather I needed *two* things to hack your server account than just *one*?04:31
nevyn so for things like gitolite where every users ssh's to git@gitolitehost04:31
it allows identification of the individuals based on their differerent keys04:32
← Gitzilla left04:32
MestreLion nevyn: I never changed default settings of ssh-agent... well, maybe I just dont use Github that often... ddoes it cache for more than 1 day? or until logout?04:32
nevyn until logout by default I think04:33
MestreLion nevyn: that makes sense04:33
nevyn well.. it's a user process04:33
so you shouldn't run ssh-agent if you're using X you should be able to just use ssh-add04:34
← xiagox left04:34
MestreLion sitaram: which other file he needs? the pub key is already at the server... as for my private key, can't the hacker just use plain username/password, like I did in my 1st login?04:34
Bearcat it would be kinda nice if i could just use my php key for this, but looking at them both in vim they look to be not the same thing.04:35
^php^pgp geeze04:35
sitaram MestreLion: in most really secure setups the password access is disabled once pubkey is setup, or the password changed to something really hard to guess/use04:36
MestreLion ohhhh, NOW it makes sense04:36
nevyn sshd can be configured to disallow password authentication04:36
MestreLion so once my pub key is there, username/passwd acces is forbidden?04:36
sitaram MestreLion: you're not alone though. See http://lwn.net/Articles/369703/04:37
MestreLion: there're some situations where *password* is apparently better!04:37
→ pdtpatr1ck joined04:37
sitaram MestreLion: could be/should be forbidden, or the password changed to something really out there04:37
which cannot be mitm-ed or social engineered04:38
MestreLion sitaram: I already fired up a bunch of browser tabs for ssh tutorials since this nevyn started his "lesson" to Bearcat04:38
← zulax left04:38
MestreLion sitaram / nevyn : this "forbid password access" is a user or server setting?04:38
MestreLion googles for "mitm"04:39
MestreLion oh, Man in the Middle04:39
→ herdingcat joined04:40
MestreLion I mean, can I forbid password access to my account in a remote server myself, or do I require the server's sshd to be configured in that way?04:40
nevyn you need to configure sshd04:41
MestreLion Can I, i mean?04:41
← herdingcat left04:41
MestreLion or do I require any special privilege in the server?04:42
→ shiba_yu36 joined04:42
→ TL joined04:42
→ herdingcat joined04:42
MestreLion (man, my english is awlful today)04:42
→ guest4854 joined04:43
nevyn you require root on the server.04:43
→ toothr joined04:43
nevyn Aleric: did that make sense?04:43
15:23 < nevyn> git reset newstuff while having master checked out should work tho04:43
MestreLion 15:23 there? wow... it's not even 3am here :P04:44
← shiba_yu36 left04:44
MestreLion AM04:44
→ shiba_yu36 joined04:44
→ kryl99_ joined04:44
← kerrick left04:45
MestreLion so the only thing an ordinary user, accessing his univeristy account for example, is to change the apssword to someing exorbitant, as sitaram suggest, right? Other than that, hope that the university server is configured to forbid password access if ssh pub keys are present. Correct?04:46
nevyn you can change your password to a 80 character generated thing and store it in a gpg encrpted file or in a envelope locked in a safe/drawer04:47
→ Gitzilla joined04:47
nevyn likes securing things in the physical world more and more04:47
→ freezey joined04:49
nevyn so in an envelope locked in a real safe.04:49
MestreLion that URL sitaram pasted, about passwd vs ssh is gold... not the debate itself, but it already answered a lot of my questions about ssh in general04:50
gitinfo set mode: +v04:50
MestreLion wow, Ubuntu/Mint don't come with ssh server by default04:51
nevyn nope04:51
openssh-server04:51
MestreLion yup04:51
How is the default settings in most distros? Already set as "PasswordAuthentication no"04:52
?04:52
← techstein left04:56
nevyn nope04:57
default is passwordAuthentication yes04:57
← albel727 left04:57
→ techstein joined04:58
guest4854 Hello. I had a question which I can't seem to find much of anything on elsewhere. What causes these sort of errors: "Warn: [some-directory-here] doesn't contain commit [hash-here]" Also, how might one fix it? In my case its related to sub modules if that makes any difference.04:58
nevyn did you pull the submodule?04:59
← magnus_ left04:59
guest4854 yes04:59
→ magnus_ joined05:02
← lyddonb_ left05:04
→ yshh joined05:09
→ ceej joined05:12
← ceej left05:13
← guest4854 left05:16
→ Cam joined05:18
← niki left05:19
→ rchavik joined05:19
→ Filthy joined05:19
← soulcake left05:19
← zomothesweeper left05:19
→ soulcake joined05:21
← oc80z left05:23
→ oc80z joined05:23
← osmosis left05:24
→ zomothesweeper joined05:25
← zomothesweeper left05:25
→ zomothesweeper joined05:27
← mandric left05:29
← jakob left05:29
→ heppy joined05:32
→ albel727 joined05:35
→ aknewhope joined05:35
→ jakob joined05:36
← f0i left05:37
→ f0i joined05:40
← t0rc left05:42
→ the_cybersphinx joined05:46
← jakob left05:47
→ lnostdal_ joined05:47
→ abetaha joined05:48
← luxurymode left05:48
→ glennpratt joined05:49
← werdnativ left05:49
← cybersphinx left05:49
→ kensington joined05:50
→ jakob joined05:52
← justfielding left05:52
→ tongueroo joined05:52
tongueroo hi guys, is there a way to preload a git commit with a message in vim?05:53
i would like to type, "git commit" and when the vim editor opens , there to already be a message there05:53
← bgerber left05:54
Nugget you can do that via a prepare-commit-msg hook05:54
there should be sample code in .git/hooks/prepare-commit-msg.sample in any of your existing git repos05:54
→ xerora joined05:55
tongueroo Nugget: awesome05:55
thanks05:55
!05:55
→ bgerber joined05:55
kensington when building git, most binary names end with last eg. 'gitlast' 'git-helplast'. any ideas as to why that might be happening?05:56
← jakob left05:57
→ justfielding joined05:58
→ jakob joined05:59
→ kevwil joined05:59
← magnus_ left06:00
← macmartine left06:01
← justfielding left06:01
← kevwil left06:01
→ magnus_ joined06:01
milki how are you building it06:02
→ maleknet joined06:02
kensington ./configure && make06:03
← ormris left06:03
→ macrover joined06:03
← msponge left06:05
→ msponge joined06:07
→ macmartine joined06:08
← glennpratt left06:09
← bgerber left06:11
→ bgerber joined06:12
→ Skyx-mobile__ joined06:12
← alta189 left06:13
→ osmosis joined06:13
→ alta189 joined06:13
kensington oh, I figured it out, I had $X set, and that variable is appended to binary names for whatever reason06:14
→ arvind_khadri joined06:14
← jakob left06:14
← dotnull left06:14
→ jakob joined06:15
← macmartine left06:16
← zomothesweeper left06:17
← Bass10 left06:18
→ mezolio joined06:19
→ glennpratt joined06:20
← mezolio left06:21
milki lol06:22
← devbug left06:23
kensington I'm not sure which is stranger, that build behaviour or my global X=last :)06:24
→ ilius joined06:27
← aknewhope left06:29
→ akosikeno joined06:29
→ aknewhope_ joined06:29
← vandal- left06:30
← neurodrone left06:31
← whyking left06:31
→ mikepack joined06:34
← glennpratt left06:34
← soulcake left06:35
← lonewulf` left06:35
→ macmartine joined06:38
← arvind_khadri left06:39
← macmartine left06:42
→ Demego joined06:43
← tongueroo left06:44
← TL left06:44
← dreiss left06:44
← aknewhope_ left06:45
← yshh left06:45
→ aknewhope joined06:46
← Skyx-mobile__ left06:46
→ bcardarella joined06:47
← modal2 left06:47
← Demego left06:48
bcardarella So I mistakenly committed code to my local repo that I didn't intend to (wrong window open) is there a way I can uncomit the code and get back the changes I made as a dirty directory?06:48
← aknewhope left06:48
nevyn git reset --soft HEAD^06:48
→ Araxia_ joined06:49
← macrover left06:50
→ nixmaniack joined06:50
← Heimidal left06:50
← Chillance left06:53
← Araxia left06:53
Araxia_ → Araxia06:53
→ Chillance joined06:54
bcardarella oh, thank-you06:55
← Chillance left06:55
← jargon-_ left06:56
← edude03 left06:57
← Filthy left07:00
← magnus_ left07:02
→ magnus_ joined07:02
← herdingcat left07:03
→ dreiss joined07:05
→ csherin joined07:05
→ tilleps joined07:08
← maleknet left07:08
← alta189 left07:08
→ arvind_khadri joined07:09
→ alta189 joined07:09
gitinfo set mode: +v07:11
→ maleknet joined07:11
gitinfo set mode: +v07:12
tilleps If I git clone a repo (B) inside a git repo (A), would repo (B) be pulled/fetched if I do "git pull --recursive" ?07:12
→ Dragooon joined07:12
← jakob left07:13
→ lonewulf` joined07:14
→ jakob joined07:14
milki no07:14
→ imarcusthis joined07:17
tilleps what's the term I would need to search for to read up on how to accomplish this?07:17
← skylord5816 left07:17
← mikepack left07:17
selckin bash07:18
tilleps nevermind, found it: http://stackoverflow.com/questions/2140985/setup-a-git-external-for-remote-repo07:19
→ jacekmigacz joined07:19
→ CapnKernel joined07:20
selckin that still won't give you a single command to update07:21
tilleps care to indulge on what would?07:21
→ mikepack joined07:22
← mklappstuhl left07:22
→ werdnativ joined07:22
CapnKernel Hello. I have some files in a repo that are read-only. git pull is having trouble because they're read-only. Any suggestions?07:22
← QaDeS_ left07:22
→ tewecske joined07:23
selckin make them writable07:23
CapnKernel Yes git is allowed to change the perms in order to make the changes. I'm hoping there's some sort of option that says "faff around with perms to do what you need to"07:23
selckin git doesn't care about perms07:23
it doesn't track or change them in any way (except +x)07:24
CapnKernel Oh I see. It doesn't track the "w" bit?07:24
selckin no07:25
milki if you want them to be read-only, you probably need some git hooks to temporarily change them07:26
selckin or you know, not be dumb and having them readonly when they are tracked will need to be changed07:26
→ AmirBehzad joined07:27
← imarcusthis left07:27
← tm1000 left07:28
CapnKernel I have a website in git. Certain files in the website are read-only. I'd rather not have to have some sort of intermediate copy that gets shuffled back and forth.07:28
← uu left07:28
selckin git is not a deployment tool07:29
CapnKernel ok07:30
milki if you are using it as a deployment tool, i dont see any reason you couldnt write a post-deployment script to fix perms07:30
CapnKernel yes07:30
That's what I do07:30
selckin you can sorta make it into one, but won't be pretty07:30
CapnKernel I'd like to be able to have it two-way though07:30
→ tm1000 joined07:30
CapnKernel Anyway, I'll see what I can do07:31
milki ya, git wont do that for you07:31
selckin but it will still have the same effect as them being writeable07:31
CapnKernel Sure is useful to be able to develop a website on a laptop, then push it to testing and production websites.07:31
→ thiago joined07:31
milki i use post-receive hooks to do deployment07:31
CapnKernel I'll improve the way I do Unix permissions, in that they should all be writable for me, but not for the web server user.07:31
Then if I run git as me, it shouldn't be a problem :-)07:32
→ aknewhope joined07:33
← tilleps left07:38
→ Heimidal joined07:39
→ Fandekasp joined07:40
← joshteam left07:45
→ h0bbit joined07:45
→ Synaps joined07:47
gitinfo set mode: +v07:49
Synaps Guys, I need your help with understanding git. I have a repository, and I want to get the name of the author with most number of commits. Is there a way to do that?07:49
bob2 that's not really an understanding git question07:49
you'll need a script to output the author for every commit, so read the man page for git-log and use the required format07:50
then uniq -c or something07:50
possible git-stats (not part of git) has a thing already07:50
Synaps Thank you, I will try to do that07:51
→ starlays joined07:51
→ spirat joined07:52
teuf you can also use git shortlog output07:52
spirat Hello, I can't figure out how to clone this repository: http://repo.or.cz/w/qemu/stefanha.git, tried every url displayed...07:52
Synaps sorry, what is git shortlog output?07:53
teuf Synaps: just try "git shortlog"07:54
bob2 spirat, ?07:56
the git url works fine07:56
(it saves time if you explain what exactly you did and what exactly happened)07:56
teuf ah yeah working tow07:57
now07:57
first try, I got "fatal: The remote end hung up unexpectedly "07:57
so maybe there was a glitch on the server07:57
spirat yes07:57
but it seems strange, the maintainer sent me the url. I feel a little stupid now, I'll probably ask him to check things...07:58
→ BiggFREE joined07:58
← freezey left07:58
← lotsofcode left07:59
→ aspotashev joined07:59
→ imarcusthis joined08:00
bob2 maybe repo.or.vz is just overloaded08:00
← smw left08:00
← magnus_ left08:01
→ SegFaultAX joined08:02
← SegFaultAX left08:03
→ SegFaultAX joined08:03
→ magnus_ joined08:03
← Goplat left08:03
spirat with another url: warning: remote HEAD refers to nonexistent ref, unable to checkout.08:03
→ kp666 joined08:04
→ rurufufuss joined08:04
← CapnKernel left08:04
← starlays left08:06
→ starlays joined08:06
← orospakr left08:07
← abetaha left08:09
← imarcusthis left08:10
← amigojapan left08:10
spirat actually i was given this url: http://repo.or.cz/w/qemu/stefanha.git/shortlog/refs/heads/vlan-hub08:13
→ imarcusthis joined08:14
← Bearcat left08:14
← maleknet left08:14
→ maleknet joined08:14
→ patrick99e99 joined08:15
→ munichpython joined08:16
patrick99e99 how can I revert a commit?08:16
I just did git commit -a -m "blah" and I realized that I didn't want to do that yet...08:17
← munichlinux left08:17
munichpython → munichlinux08:17
patrick99e99 will git revert HEAD actually unstage everything and show me all my previous work when I do git status?08:17
or will I lose it?08:17
← aknewhope left08:18
spirat (concerning my issue, I mailed the maintainer, no need to investigate, thanks)08:18
← Lgb left08:20
← themachine left08:20
← marenz left08:21
→ mikepack_ joined08:21
→ koo5 joined08:21
← mikepack left08:22
→ Lgb joined08:23
fairuz_ patrick99e99: git reset --soft HEAD^ will help you to undo the last commit08:23
bob2 (as long as you haven't pushed it)08:23
patrick99e99 fairuz_: and how do I get my changed code?08:24
fairuz_: when I do git status now, it says my working directory is clean08:24
fairuz_ reset --soft will undo the commit, and the modified files will be still there08:25
← Cam left08:25
patrick99e99 hmm maybe I screwed everything up because I did git revert HEAD08:25
and it made a 2nd commit..08:25
so I did git reset --soft HEAD^ twice08:25
commit history shows my commits are gone..08:26
but so is my code...08:26
→ Vortex34 joined08:26
← munichlinux left08:28
← brokenshakles left08:29
patrick99e99 hmm it apparently was in another branch..08:31
so i am good..08:31
thanks08:31
→ brokenshakles joined08:32
→ leavittx joined08:34
→ mikepack joined08:34
← mikepack_ left08:34
→ jonas_g joined08:34
← thiago left08:35
← mikepack left08:36
→ mikepack joined08:37
→ jackdempsey joined08:40
→ psoo joined08:45
← eno left08:46
→ eno joined08:48
→ shruggar joined08:49
→ fr0stbyte joined08:49
← ericmuyser left08:49
fr0stbyte → ericmuyser08:49
→ themachine joined08:50
← gavin_huang left08:50
→ caseymcg joined08:52
Droggl_ → Droggl08:52
→ ehsan joined08:53
→ ben225 joined08:53
→ ph^ joined08:55
→ porco joined08:57
← themachine left08:59
← Stew-a left09:00
← jackdempsey left09:01
← magnus_ left09:02
→ Sigma joined09:02
← Sigma left09:02
→ Sigma joined09:02
→ thiago joined09:03
← eno left09:03
→ magnus_ joined09:04
→ eno joined09:05
→ jackdempsey joined09:05
← Katniss left09:07
→ ThomasLocke joined09:10
← Milossh left09:10
→ Katniss joined09:13
← jackdempsey left09:16
← zozeo left09:18
← Synaps left09:19
← eno left09:21
→ flijten joined09:22
← kryl99_ left09:22
← flijten left09:22
→ eno joined09:23
← caseymcg left09:24
← igotnolegs left09:25
→ whitman joined09:28
← csherin left09:30
← Fandekasp left09:33
→ zomothesweeper joined09:33
← mikepack left09:33
← stringoO left09:34
→ rendar joined09:35
→ d0k joined09:35
← Katniss left09:36
← eno left09:37
→ lotsofcode joined09:37
→ gusnan joined09:37
→ eno joined09:39
→ csherin joined09:39
→ Katniss joined09:41
→ Shusshu joined09:42
← BiggFREE left09:43
→ mabrand joined09:47
Rahul what is the prerequisites to get to know about git ? :)09:47
→ flavius joined09:47
Rahul gitinfo:what is the prerequisites to get to know about webkit ? :)09:47
← eno left09:48
selckin ~make sense09:48
← Rahul left09:49
→ eno joined09:50
← psoo left09:50
→ Dave^| joined09:50
← zomothesweeper left09:53
← ericmuyser left09:55
→ eijk joined09:55
→ cedeon joined09:55
← dorkmafia left09:57
→ arvind_k joined09:57
← arvind_khadri left09:58
← eno left09:58
→ robotmay joined09:59
→ eno joined10:00
→ _iron joined10:00
← Shusshu left10:00
→ kadoban joined10:01
→ Milossh joined10:02
← magnus_ left10:03
→ magnus_ joined10:04
← eno left10:06
← porco left10:06
→ FernandoBasso joined10:07
← ivan` left10:07
→ zarubin joined10:07
→ eno joined10:08
→ zomothesweeper joined10:09
← zomothesweeper left10:09
→ i42n joined10:10
→ ivan` joined10:11
← mabrand left10:11
→ tnt joined10:12
tnt Hi. I cloned a repo that was created by someone else with git-svn and I'm trying to update it (fetch new revision from the original SVN). But one hour later and I'm still nowhere. Any tip ?10:13
→ JohnBat26 joined10:13
← cedeon left10:14
← eno left10:15
→ eno joined10:17
← spirat left10:17
selckin not supported10:20
tar up his whole repo, or clone again from svn10:20
→ justfielding joined10:21
→ kryl99_ joined10:21
← eno left10:21
→ eno joined10:22
← koo5 left10:24
→ Shusshu joined10:24
← Vortex34 left10:24
→ koo5 joined10:25
← kryl99_ left10:26
← eno left10:27
→ airborn_ joined10:28
→ eno joined10:28
→ marenz joined10:28
← msponge left10:32
→ churp joined10:33
→ glcrazy joined10:33
→ kryl99_ joined10:33
← eno left10:34
→ giallu joined10:36
→ tnt_ joined10:36
→ eno joined10:36
← churp left10:36
← tnt left10:39
→ Liq^m joined10:43
→ mr_schlauch joined10:45
← eno left10:47
→ faber joined10:48
→ eno joined10:49
→ sansdro1 joined10:49
← sansdro1 left10:49
← aspotashev left10:50
→ d34df00d joined10:52
d34df00d Hi!10:53
← h0bbit left10:53
d1b_ → d1b10:53
d34df00d I have two branches, say, 'master' and 'vmime'. I want to revert master to vmime's state, but partially, only reverting changes to files in particular subdirectory.10:53
How that can be done in git?10:53
← shruggar left10:54
→ arvind__ joined10:55
→ canton7 joined10:55
wereHamster d34df00d: manually10:56
← arvind_k left10:57
← eno left10:57
→ eno joined10:59
← ShooterMG left11:00
d34df00d wereHamster: ah, that's sad.11:00
I guess I'd do with `git diff master vmime path/to/dir/` and applying the corresponding patch then.11:00
→ ShooterMG joined11:03
← eno left11:03
← magnus_ left11:04
→ eno joined11:04
→ magn0z_ joined11:04
→ Demego joined11:05
→ Synaps joined11:07
gitinfo set mode: +v11:07
← mansour_ left11:10
→ dharrigan joined11:10
← dharrigan left11:10
→ dharrigan joined11:10
→ mansour joined11:10
→ Lisimba2 joined11:10
← dharrigan left11:10
← eno left11:11
→ dharrigan joined11:11
← dharrigan left11:11
→ dharrigan joined11:11
mansour I commited my code and started some refactoring, without creating a branch .11:11
→ Fandekasp joined11:11
mansour what do I need to do, to more all these changes to a new branch ?11:11
← Lisimba left11:12
→ eno joined11:12
← eno left11:12
→ eno joined11:12
cbreak mansour: git checkout -b newbranch11:12
and then commit11:12
(Be aware that changes are NOT on any branch until you commit them)11:13
mansour perfect.11:13
thank you11:13
cbreak np :)11:13
→ barr5790 joined11:14
← Milossh left11:14
→ TSK joined11:15
→ Dmitr joined11:16
← Dmitr left11:16
→ Jackneill joined11:16
← Jackneill left11:16
→ Jackneill joined11:16
← eno left11:18
← tnt_ left11:20
→ eno joined11:20
→ joe-mac1 joined11:23
gitinfo set mode: +v11:23
joe-mac1 is it possible to always ignore a certain number of lines at the beginning of a file?11:24
→ Milossh joined11:24
cbreak with git? No11:24
you can take a look at smudge/clean filters though11:24
joe-mac1 the issue is I have two branches wh ere in the 'comment header' will always have one little difference, instead of 'production' it will say 'staging'11:24
cbreak joe-mac1: man gitattributes11:24
gitinfo joe-mac1: the gitattributes manpage is available at http://jk.gs/gitattributes.html11:24
joe-mac1 thank you cbreak11:25
cbreak joe-mac1: that sounds stupid :/11:25
writing scm meta data into a file?11:25
joe-mac1 cbreak: not really, use case if configuration files served out by puppet11:25
sort of11:25
I can fix it11:25
← glcrazy left11:25
→ legnaleurc joined11:25
joe-mac1 I have to convert some plain file resources to erb templates, and I think I can fix it11:25
← xoebus left11:25
joe-mac1 a couple find ./ with search and replaces...11:26
yea indeed I can just change 'staging' or 'production' to <%= environment %>11:26
that works11:26
← eno left11:26
joe-mac1 I am retarded11:26
I talk to myself til I figure these things out mostly11:26
→ xoebus joined11:27
→ LongBeach joined11:27
→ eno joined11:28
cbreak joe-mac1: http://c2.com/cgi/wiki?RubberDucking11:28
joe-mac1 cbreak: sweet, nice new term I learned today heh11:29
← legnaleurc left11:29
← Sigma left11:30
auke-_ → auke-11:31
← myme left11:31
→ legnaleurc joined11:31
← Fandekasp left11:32
→ norton- joined11:33
← no1home left11:34
Synaps Guys, how can I get a list of most frequently updated files? Is it possible to do it with git log?11:34
← eno left11:35
← legnaleurc left11:36
→ myme joined11:36
→ aspotashev joined11:36
cbreak Synaps: you can look at git log --name-only11:37
← shergill left11:37
koo5 → latchup11:37
← arvind__ left11:37
→ eno joined11:37
cbreak maybe with some other --pretty options11:37
→ arvind__ joined11:37
cbreak and write your own script to analyze it11:37
← arvind__ left11:37
→ iPeter joined11:37
→ Sigma joined11:37
← iPeter left11:38
→ Nathandim joined11:38
← leavittx left11:38
→ arvind_khadri joined11:38
← arvind_khadri left11:38
→ arvind_khadri joined11:38
→ LeMike joined11:40
← ivan` left11:40
→ iocor joined11:41
← eno left11:41
← albel727 left11:42
→ eno joined11:42
← eno left11:42
→ eno joined11:42
→ albel727 joined11:42
→ munichlinux joined11:43
→ mishok13 joined11:43
→ tnt joined11:45
← thiago left11:47
→ dougsland joined11:47
← dougsland left11:47
→ dougsland joined11:47
← aspotashev left11:47
← eno left11:48
joe-mac1 yea then put the data out to graphite and have pretty graphs!11:49
→ Akufen joined11:50
→ martinjlowm joined11:50
mva d34df00d: ping11:50
→ eno joined11:50
→ leavittx joined11:50
← rchavik left11:51
→ ivan` joined11:51
← ivan` left11:51
← LeMike left11:52
→ ivan` joined11:54
→ brian_g joined11:54
→ diamonds joined11:58
← diamonds left11:58
→ diamonds joined11:58
→ Dave^|| joined11:59
← eno left12:00
d34df00d mva: wut.12:02
← Dave^| left12:02
→ eno joined12:02
← Milossh left12:02
← eno left12:02
→ eno joined12:02
← magn0z_ left12:03
→ magnus_ joined12:06
→ Milossh joined12:12
→ Skyx-mobile__ joined12:12
← eno left12:12
← Shapeshifter left12:13
→ martinjlowm-lptp joined12:13
→ eno joined12:14
← TSK left12:15
← jacekmigacz left12:16
→ robottinosino joined12:20
→ Matrixiumn joined12:22
← cenuij left12:24
← osmosis left12:24
← Sigma left12:26
← eno left12:27
→ cenuij joined12:27
← cenuij left12:27
→ cenuij joined12:27
→ Element9 joined12:27
→ Sigma joined12:27
← Milossh left12:28
→ porco joined12:28
→ eno joined12:28
→ icwiener joined12:28
← riskynet left12:29
← tnt left12:29
← kensington left12:32
← matthiaskrgr left12:32
← selckin left12:33
gitinfo set mode: +v12:33
xoebus I've forgotten to sign off my last 3 commits with the "-s" flag. I haven't pushed yet. Is there a simple way of going back and adding these in?12:33
wereHamster xoebus: with rebase12:34
→ twisted` joined12:35
→ matthiaskrgr joined12:35
mva d34df00d: enter in gentoo.ru conference. Mellon asked you12:35
→ berserkr joined12:35
xoebus wereHamster: I had a quick go at that. `git rebase -i HEAD~3`. Then electing to rewrite the messages. Is there a simpler way than that?12:36
← Nathandim left12:36
wereHamster I don't think so12:36
xoebus okay, I'll do that. cheers12:36
← xerora left12:37
→ zomothesweeper joined12:38
← eno left12:41
d34df00d mva: I cannot for some reason :(12:41
mva: can he join some channel on cjr, like c_plus_plus?12:41
← robottinosino left12:41
→ eno joined12:42
← eno left12:42
→ eno joined12:42
→ jacekmigacz joined12:43
← twisted` left12:43
→ twisted` joined12:43
→ webclectic joined12:43
canton7 xoebus, you can elect to edit the commit, then use commit --amend -s12:44
← mishok13 left12:44
→ cccaldas joined12:44
→ whyking joined12:45
→ twisted`_ joined12:46
← Skyx-mobile__ left12:46
← ajpiano left12:46
→ ajpiano joined12:47
← twisted` left12:48
twisted`_ → twisted`12:48
← eno left12:48
→ psoo joined12:48
→ eno joined12:50
→ ipalaus joined12:50
→ merlin83_ joined12:54
→ mantic0 joined12:55
← kryl99_ left12:57
→ kryl99_ joined12:58
← zomothesweeper left12:58
← eno left13:01
→ eno joined13:03
← magnus_ left13:04
→ bitkiller joined13:05
→ herdingcat joined13:05
→ erichynds joined13:05
→ magnus_ joined13:06
← eno left13:07
→ eno joined13:08
→ pom joined13:11
← eno left13:14
→ eno joined13:16
← akosikeno left13:16
→ xissburg joined13:18
→ akosikeno joined13:19
xissburg I had "profile" in my gitignore, and now my project a folder Profile where I store some source code. I removed profile from the gitignore but I still can't add the files in the Profile folder :S13:19
→ johnkpaul joined13:21
→ storrgie joined13:22
← ehsan left13:23
← eno left13:25
→ eno joined13:29
← johnkpaul left13:30
→ thiago joined13:31
← Gitzilla left13:32
← kryl99_ left13:33
→ alex__ joined13:33
→ Sigma[Mtp] joined13:35
← Sigma left13:35
← fruight left13:36
← bitkiller left13:37
→ bitkiller joined13:38
← eno left13:39
← toothr left13:40
→ eno joined13:41
← pidus left13:41
_ikke_ xissburg: can't add?13:42
← alex__ left13:44
← airborn_ left13:44
→ alex__ joined13:44
← alex__ left13:45
→ Gitzilla joined13:45
← intripoon_ left13:48
← cccaldas left13:49
→ fruight joined13:49
cbreak in general, git itself is case sensitive13:51
← eno left13:51
← jonas_g left13:53
→ eno joined13:53
→ Airells joined13:55
Airells hi, i am looking for architecture diagram for GIT, like that one for svn: http://chestofbooks.com/computers/revision-control/subversion-svn/images/ch01dia1.png I need it for compare to svn in presentation.13:56
wereHamster [git] <-> [transport (any of git, http, local, ssh] <-> [git]13:58
← psoo left13:59
wereHamster there is no client library. Server (except git-damon) is provided by the ssh daemon, web server etc.13:59
→ Xizor joined13:59
← Aleric left13:59
→ cccaldas joined13:59
← ilius left13:59
wereHamster git itself is pretty monolithic.14:00
Airells thank you, Any graphic version? couldnt find14:00
wereHamster I'm not aware of any. Maybe in one of the books there is something14:00
!books14:00
!book14:00
gitinfo There are several good books available about git; 'Pro Git' is probably the best: http://progit.org/book/ but also look at !bottomup !cs !designers !vcbe and !parable14:00
→ Bass10 joined14:02
← storrgie left14:03
xissburg _ikke_: yea14:03
← eno left14:03
→ krake joined14:04
→ yshh joined14:04
← Bass10 left14:04
← magnus_ left14:04
→ eno joined14:05
← joe-mac1 left14:06
→ magnus_ joined14:06
→ MaximusColourum joined14:06
telemachus wereHamster: re books and links, what about http://think-like-a-git.net? Semi-random suggestion.14:11
tokkee_ → tokkee14:12
→ CR_SUAMI35TAHUNK joined14:12
→ mtkd joined14:13
← icwiener left14:13
→ zozo_ joined14:14
→ soulcake joined14:14
← eno left14:14
telemachus !vcbe14:14
gitinfo 'Version Control By Example' gives a good overview of the different VCSes available. The author will even mail you a dead-tree copy for free. http://ericsink.com/vcbe/14:14
← CR_SUAMI35TAHUNK left14:15
→ eno joined14:16
← Wooga left14:16
← bitkiller left14:18
← Chib left14:18
← Lgb left14:20
← eno left14:21
← porco left14:22
→ eno joined14:22
→ Lgb joined14:23
→ psoo joined14:24
← JohnBat26 left14:25
→ neurodrone joined14:28
Airells it is really hard to find something about technical aspect of GIT, database, flow, support for protocols etc14:28
← luke-jr left14:28
← eno left14:28
← psoo left14:28
canton7 Airells, !bottomup is good for learning about git's internals14:28
gitinfo Airells: 'Git from the bottom up' starts with explaining the building blocks of git and proceeds to tell you how they fit together. http://ftp.newartisans.com/pub/git.from.bottom.up.pdf14:28
canton7 Airells, http://stackoverflow.com/questions/5176225/are-gits-pack-files-deltas-rather-than-snapshots gives the gory details on git's delta compressing, although you probably don't want to go there14:29
→ Wooga joined14:29
→ eno joined14:30
canton7 Airells, protocols: smart + dumb http(s) (push and pull), ssh (push and pull), git protocol (unauthenticated push and pull, uses git-daemon)14:31
also can copy the git repo, git bundle to stuff a repo into a single file14:31
← erichynds left14:31
wereHamster Airells: http://code.google.com/p/support/wiki/DVCSAnalysis14:31
canton7 Airells, see !workflow for git workflows14:32
gitinfo Airells: Finding the right workflow for you is critical for the success of any SCM project. Git is very flexible with respect to workflow. See https://gist.github.com/1540906#workflow for a list of references about choosing branching and distributed workflows.14:32
→ Ksugzulut joined14:33
Airells thank you guys14:33
← eno left14:35
→ rdholder joined14:35
← Jackneill left14:36
→ eno joined14:36
← webclectic left14:36
← maleknet left14:38
→ Jackneill joined14:38
← Synaps left14:40
← eno left14:40
← zozo_ left14:42
→ eno joined14:44
← zarubin left14:45
← rdholder left14:45
→ jonas_g joined14:46
← Matrixiumn left14:46
→ rawtaz joined14:46
rawtaz hey guys and gals14:46
← FernandoBasso left14:46
→ Matrixiumn joined14:47
fairuz_ rawtaz: hey14:47
← Matrixiumn left14:47
→ Matrixiumn joined14:47
→ shergill joined14:51
rawtaz having fun with my first real git attempt with an existing project14:51
fired up Xcode 4 but it showed no files. but http://stackoverflow.com/questions/5996966/xcode-4-shows-no-files-after-loading-project worked14:51
→ |jonny| joined14:53
← arvind_khadri left14:55
← eno left14:57
→ eno joined14:58
← |jonny| left14:58
← gusnan left14:59
← Sweetshark left15:03
← frogonwheels left15:03
← rurufufuss left15:03
→ Sweetshark joined15:04
← Sweetshark left15:04
→ Sweetshark joined15:04
sitaram Airells: http://sitaramc.github.com/1-basic-usage/tias.html may help somewhat (although not in the way you might expect)15:06
← magnus_ left15:06
→ johnkpaul joined15:06
→ magnus_ joined15:07
← eno left15:07
← johnkpaul left15:08
→ eno joined15:10
→ MrWynand joined15:11
← lnostdal_ left15:13
← rubyruy left15:14
→ ehsan joined15:16
→ Run joined15:17
← Run left15:17
→ Run joined15:17
Run → Aleric15:17
← Favorit_ left15:17
→ Favorit joined15:18
← p3rror left15:19
→ rdholder joined15:20
← eno left15:21
→ orospakr joined15:22
→ eno joined15:23
← thiago left15:24
← Chappo left15:24
← lonewulf` left15:24
← rdholder left15:25
→ porco joined15:25
→ rdholder joined15:26
← rdholder left15:27
← c1de0x left15:29
→ zarubin joined15:30
→ bronson joined15:32
→ p3rror joined15:32
← gxk left15:33
→ bauruine joined15:34
→ g105b joined15:35
← eno left15:36
→ eno joined15:37
← eno left15:37
→ eno joined15:37
→ LeMike joined15:37
→ gxk joined15:37
→ SundarNET joined15:40
← munichlinux left15:42
→ Bass10 joined15:42
→ munichlinux joined15:42
← LeMike left15:42
← Gitzilla left15:45
→ wuj joined15:45
← martinjlowm-lptp left15:46
→ Cromulent joined15:48
← cyphase left15:49
← eno left15:49
← mantic0 left15:50
← SundarNET left15:50
→ weichkaes joined15:50
→ eno joined15:52
← eno left15:52
→ eno joined15:52
→ mantic0 joined15:53
← simplechat left15:53
← mantic0 left15:53
← shergill left15:54
→ shergill joined15:55
→ mantic0 joined15:55
→ vandal- joined15:57
← wuj left15:57
→ wuj joined15:58
→ edude03 joined15:58
→ Raging_Hog joined16:00
← angelsl left16:00
→ cyphase joined16:00
rawtaz oh god how not straight forward git can be when you havent used it much..16:01
→ legnaleurc joined16:01
rawtaz >_<16:01
→ Elenita_17_moren joined16:01
cbreak it's a command line app16:01
you need to know the words to talk with it :)16:01
rawtaz yeah im well aware of that16:01
yap16:01
→ lyddonb_ joined16:03
← eno left16:04
← Elenita_17_moren left16:05
→ maletor joined16:05
→ eno joined16:05
← magnus_ left16:05
→ pandeiro joined16:06
← eightyeight left16:06
→ eightyeight joined16:07
rawtaz here's the thing: for some reason Xcode (or i, in it, god knows) managed to delete two files in the project, and i did a `git reset --hard HEAD` to mitigate that. but now those two files are shown as untracked. the way way i look at it is that i asked git to completely revert/reset/whatever back to the initial checked out/cloned status, but if it did i dont think these files would be "untracked" right now16:07
→ magnus_ joined16:07
rawtaz probably misunderstanding some part here16:07
eightyeight → Guest4308616:07
← pandeiro left16:07
← legnaleurc left16:07
← shergill left16:07
← fruight left16:08
cbreak rawtaz: did you commit the files?16:08
rawtaz no, i have not done any commiting16:08
→ HG` joined16:08
cbreak well, then that's normal16:08
← Guest43086 left16:08
cbreak HEAD is the last commit16:08
rawtaz i gig git clone from github. then added the upstream repo16:08
cbreak if the files were not in the last commit16:08
→ pandeiro joined16:08
rawtaz but the files comes from the origin repo16:09
cbreak and you told git "make everything as it was at the last commit"16:09
then they are untracked16:09
→ eightyeight joined16:09
← AmirBehzad left16:09
← HG` left16:09
telemachus rawtaz: This may help you get a better sense of git reset http://progit.org/2011/07/11/reset.html16:09
→ HG` joined16:09
rawtaz i have not added these files, they are part of the source tree that is in the origin repo.16:09
telemachus: thanks, will read16:10
→ AmirBehzad joined16:10
rawtaz what i want to do here is effectively the same thing as just deleting this whole project folder, doing a new clone from github, adding the upstream repo again. then i would be at the point i want to tell git to "revert" to now16:10
→ Gitzilla joined16:11
cbreak telemachus: interesting16:11
← eno left16:11
cbreak rawtaz: then git reset --hard origin/master or something similar16:11
then git clean -fdx16:12
telemachus cbreak: Pretty pictures and charts - never hurts16:12
cbreak this will delete a whole lot16:12
→ cakehero joined16:12
rawtaz cbreak: produces http://paste.pocoo.org/show/BV5p2wkz8AQ6l9ZI8CwY/ if you care to see16:13
hm ok let me run clean as well16:13
← martinjlowm left16:13
cbreak not origin16:13
→ eno joined16:13
← diamonds left16:13
→ ajt joined16:13
cbreak reset does not touch untracked files16:13
← krake left16:13
rawtaz cbreak: i tried them all and none of them made the two files not untracked16:13
cbreak it just changes tracked files16:13
rawtaz ok16:13
cbreak clean handles untracked files16:13
rawtaz it seems clean deleted tham :D16:13
them*16:14
cbreak yes. that's the purpose16:14
rawtaz right16:14
← bcardarella left16:14
rawtaz i need to just throw this crap out and reclone then16:14
cbreak why?16:14
rawtaz well16:14
cbreak recloning is for morons.16:14
rawtaz well16:14
lol16:14
cbreak you already have the full repository16:14
downloading the same thing again won't give you anything16:14
rawtaz ill repeate myself: the two files that were now untracked, they are not something *i* added. they are part of the origin source tree. the should not be untracked16:15
cbreak no16:15
rawtaz i buy that i dont need to reclone, if i can gave git restore the files16:15
cbreak if you git reset --hard origin/master, then you are exactly where you are if you clone16:15
rawtaz ok16:15
trying :)16:15
cbreak on master branch of origin remote16:15
of course, if origin changed, you have to do a git remote update first16:16
or git fetch origin16:16
to get new changes16:16
but that's an incremental process16:16
rawtaz im sorry man but its not doing what you seem to be saying. or im doing something wrong.16:16
SethRobertson I know which one I'm betting on16:16
cbreak if you don't have those files, then that's because the remote didn't have those files.16:16
SethRobertson Why not !repro16:17
gitinfo Please paste (using https://gist.github.com/ or similar) a transcript of your terminal session. This will help immensely with troubleshooting.16:17
→ thiago joined16:17
rawtaz let me clarify. one of the untracked files were Resources/English.lproj/Credits.rtf. it is part of the origin sources. its from the repo. i have not added it16:17
← yshh left16:17
cbreak if it were part of the original repo, then you would have it.16:17
← lapistano left16:18
cbreak you don't have it? Then it wasn't part of the original repo. Easy.16:18
SethRobertson Are you on a case insensitive filesystem?16:18
cbreak it's OS X16:18
SethRobertson so yes16:18
cbreak it's an xcode project16:18
→ lapistano joined16:18
rawtaz i now did `git reset --hard origin/master`. the output is at http://paste.pocoo.org/show/lwmhi6RdBYrESZubkdUn/16:18
cbreak XCode projects are made for OS X16:18
→ amigojapan joined16:18
→ msponge joined16:18
cbreak rawtaz: see? It wasn't part of the remote repository either.16:18
rawtaz let me make sure that file is indeed coming from the repo on github.16:18
cbreak tell me the URL and I'll check :)16:19
rawtaz cbreak: look at https://github.com/rawtaz/CoRD/tree/master/Resources , its the Credits.rtf file16:19
→ urbanmonk_ joined16:19
← eno left16:19
rawtaz cbreak: to clarify further, i am NOT saying i am not the one misunderstanding things here, im just discussing this from my perspective. not questioning your suggestions16:19
cbreak https://github.com/rawtaz/CoRD/tree/master/Resources/English.lproj, no Credits16:20
canton7 rawtaz, there is no Resources/English.lproj/Credits.rtf on github16:20
rawtaz hmm16:20
SethRobertson Please run `git ls-tree -r origin/master | grep Credits.rtf`16:20
rawtaz lemme recap16:20
apparently i mixed up the file names16:20
← whyking left16:20
cbreak you probably told XCode that you wanted to localize the Credits file16:21
so it copied it into the english resource folder16:21
rawtaz ok yeah16:21
cbreak and then you reset16:21
which made the copy untracked16:21
→ eno joined16:21
rawtaz first thought was "ok i moved something" but as you say i probably asked it to add an english localization lol16:21
cbreak XCode has git integration16:21
rawtaz thanks so much :)16:21
cbreak if you make a new file with xcode, it will add the file itself16:22
rawtaz yeah. apparently it sucks though16:22
cbreak nah16:22
it's barely usable16:22
→ htoothrot joined16:22
rawtaz yeah16:22
cbreak if you want to see git integration that sucks, take a look at tortoise git16:22
or egit16:22
rawtaz haha16:22
brb16:22
← pandeiro left16:23
→ pandeiro joined16:25
→ munichpython joined16:25
← munichlinux left16:25
munichpython → munichlinux16:26
← sejo left16:26
→ sejo joined16:26
← latchup left16:26
rawtaz anyway i think ill go with CLI instead, assuming Xcode doesnt interfere. better to learn git for real16:26
i have no problems with that, just failed to realize git and i was talking about different files this time :)16:26
cbreak xcode doesn't interfere16:27
it just does convenience things like visualizing diffs/history16:27
and adding new files16:27
or offering the capability to commit (which I never tried)16:27
rawtaz mm :)16:27
cbreak also it uses the real git binaries, not some clone like eclipse16:27
rawtaz read http://oleb.net/blog/2011/05/on-git-support-in-xcode-4/ earlier today. didnt look that awesome16:28
thats nice though16:28
cbreak I think the best part of it is that it's not crap.16:28
← AmirBehzad left16:28
→ Djo joined16:28
cbreak better have only a few features that work instead of having something like tortoise that pretends git is svn16:29
← matthiaskrgr left16:29
rawtaz so you use it for the things Xcode does okay and then complement with cli when needed?16:29
← EricInBNE left16:29
→ vervic joined16:29
gitinfo set mode: +v16:30
Djo hi everybody16:30
rawtaz heya16:30
← fisted left16:31
Djo whats happen here ?16:31
rawtaz oh, the usual embarrasment.. :-)16:31
→ matthiaskrgr joined16:31
rawtaz cbreak helped me out with an issue16:32
cbreak rawtaz: usually I just ignore xcode's git integration16:32
→ kukks joined16:32
→ macmartine joined16:32
rawtaz ok16:32
cbreak the diff view isn't too bad16:32
and auto adding new files is not harmful either16:32
if you know it does it :)16:32
← eno left16:32
Djo rawtaz : yeap :/16:32
cbreak but most of the time I use git on the command line, or gitx for visualization16:32
← ajt left16:33
htoothrot → toothr16:33
Djo +1 cbreak16:33
urbanmonk_ → urbanmonk16:33
SethRobertson !thanks16:33
gitinfo Feeling thankful? Type "ExampleUser++", and ExampleUser will score karma points at http://carmivore.com (our preferred way to objectify self-worth). There's really no point to thanking me – instead, why not thank the person who made me help you?16:33
Djo why do u want to use git' xcode ?16:33
→ ceej joined16:34
→ eno joined16:34
cbreak yesterday I found "SourceTree" on the app store thing, couldn't try it out but maybe it's useful as visualizer16:34
→ fisted joined16:35
← gchristensen left16:36
← HG` left16:36
→ gchristensen joined16:37
← gchristensen left16:37
→ gchristensen joined16:37
→ lonewulf` joined16:37
→ legnaleurc joined16:39
Djo wahhou16:40
→ ehsan_ joined16:41
← ehsan left16:41
← msponge left16:41
← heppy left16:42
rawtaz question; i did a fetch from upstream and then merge upstream/master, and git status now says "Your branch is ahead of 'origin/master' by 1 commit." - shouldnt there be a bunch of changed files (the one commit in question should be https://github.com/rawtaz/CoRD/commit/b168cfeb0e4ff612b8d63b1f3ccae21ebeb101f1 ) ?16:44
hmm wait. no16:44
never mind16:44
← albel727 left16:44
rawtaz its merged in just fine here locally. but im not seeing it on github until i push or something16:44
canton7 yeah, merges are strictly in your local repo16:46
← eno left16:46
→ eno joined16:48
→ erichynds joined16:49
← herdingcat left16:49
← urbanmonk left16:50
→ pmenon joined16:50
pmenon i've added a file to git, it shows up as new file in "git status" but it doesn't appear in my "git diff"16:51
any ideas ?16:51
→ mishok13 joined16:51
→ Chillance joined16:51
Djo have git add - git commit ?16:51
m1sc pmenon: git diff --cached ?16:51
canton7 pmenon, have you 'git add'ed the file?16:51
Djo *have you16:51
cbreak git diff shows difference between working dir and index16:52
pmenon canton7: I have, m1sc what does the --cached do ?16:52
canton7 pmenon, the difference between your previous commit and your currently-staged files is given by diff --cached16:52
cbreak it makes it show the difference between index and HEAD16:52
canton7 pmenon, the difference between your previous commit and your modified, unstaged files it given by 'git diff'16:52
→ albel727 joined16:52
← albel727 left16:52
cbreak no16:53
git diff HEAD16:53
pmenon cbreak: so diff against HEAD shoes *any* changes, files added but not commited, files deleted but not committed etc?16:54
← g105b left16:54
→ albel727 joined16:54
← eno left16:54
← Tasser left16:54
pmenon why isn't that default diff behaviour /16:54
tizzo-afk → tizzo16:54
cbreak because it's inconvenient16:55
→ Aexel joined16:55
cbreak normally you only want to look at changes that you haven't already approved for the next commit16:55
→ maxydido joined16:55
cbreak or only at those you have16:55
→ pidus joined16:56
→ eno joined16:56
tizzo → tizzo-afk16:56
← pandeiro left16:57
→ JohnBat26 joined16:59
→ glennpratt joined17:00
← csherin left17:00
← eno left17:01
← Cromulent left17:02
→ eno joined17:02
← eno left17:02
→ eno joined17:02
← vervic left17:02
→ csherin joined17:03
← magnus_ left17:05
→ pandeiro joined17:06
← mishok13 left17:06
← D-Chymera left17:06
→ LeMike joined17:06
← p3rror left17:07
→ magnus_ joined17:08
← eno left17:08
← pandeiro left17:09
→ FernandoBasso joined17:09
→ hwit joined17:10
← Aexel left17:10
→ eno joined17:10
→ ajt joined17:10
← hwit left17:11
→ pandeiro joined17:11
→ D-Chymera joined17:12
← pidus left17:14
→ pidus_ joined17:14
← pidus_ left17:14
→ pidus_ joined17:14
← csherin left17:15
← bronson left17:15
→ gandy93 joined17:15
→ csherin joined17:16
→ h0bbit joined17:16
→ pernekova joined17:16
← oso96_2000 left17:18
← munichlinux left17:20
← eno left17:23
← amigojapan left17:23
→ eno joined17:25
← shiba_yu36 left17:30
← Djo left17:30
→ munichlinux joined17:34
→ kerrick joined17:35
→ eighty4 joined17:36
← porco left17:38
← lepine left17:39
← eno left17:39
← glennpratt left17:41
→ eno joined17:41
→ mishok13 joined17:43
→ bronson joined17:44
← eno left17:48
← tango_ left17:48
→ Milossh joined17:49
→ eno joined17:49
← norton- left17:50
→ sh4rm4 joined17:51
← lonewulf` left17:51
sh4rm4 one of my editors messed up the whitespace in one file in a bigger commit... can i somehow split the changes i made and redo the ones on that specific file, without the whitespace noise ?17:52
the stupid editor touched about 10K lines17:53
SethRobertson Is your repository all LF or all CRLF or something?17:53
sh4rm4 hmm. it's possible that file had CRLF17:53
usually its just LF17:54
← mishok13 left17:54
← eno left17:54
→ tango_ joined17:55
SethRobertson How many files did you touch for real in the commit? Did files that you did not make useful changes in get touched?17:55
→ eno joined17:55
SethRobertson convert17:55
Oh, and have you pushed?17:56
sh4rm4 no, i have changed 3 files, and the one with the noise is basically just 3 lines deleted17:56
no, i can still rebase17:56
SethRobertson Only lne file got poor EOL conversion?17:56
sh4rm4 indeed17:56
seems it was originally written by a windows dude17:57
i opened that one with a different editor though17:57
gnu nano...17:57
SethRobertson This should be fairly easy then. You have two choises, run a unix->windows EOL conversion program over that one file, or zap that file and reimplement the changes (via applying a patch, probably)17:57
sh4rm4 zap ?17:58
SethRobertson reset the file back to repository defaults17:58
→ khmarbaise joined17:58
→ icwiener joined17:58
← bgerber left17:59
sh4rm4 hmm, how would i do that ?17:59
SethRobertson I would first recommend trying to do the EOL conversion. Do you have the dos2unix or fromdos program installed?17:59
← munichlinux left17:59
SethRobertson We can easily try the other method if this fails17:59
sh4rm4 i got dos2unix, yeah17:59
← pidus_ left17:59
→ kgadek joined17:59
SethRobertson Well, I guess you need the other way around. Run unix2dos on that file.17:59
→ pidus joined17:59
SethRobertson Actually, before we do that...have you committed? Is it the most recent commit?18:00
it==the bad commit18:00
sh4rm4 yep, i could do a fixup18:00
SethRobertson OK, go ahead and run unix2dos on that file18:00
→ pidus_ joined18:01
← pidus_ left18:01
→ pidus_ joined18:01
← pidus left18:01
pidus_ → pidus18:01
SethRobertson Then run `git diff HEAD^ -- filename`18:01
← matthiaskrgr left18:01
→ bgerber joined18:04
← _ikke_ left18:04
→ matthiaskrgr joined18:04
SethRobertson ?18:05
sh4rm4 mhm, it seems it have touched any line as well18:05
so now commit it, and reabse/fixup ?18:05
*rebase18:05
SethRobertson No, stop18:05
`git diff -- filename` should report that it touched every line. `git diff HEAD^ -- filename` should report only your desired changes. Assuming your comment on what happened was accurate.18:06
Did that not happen?18:06
sh4rm4 no, git diff HEAD^ -- etc did show the whitespace noise18:07
← magnus_ left18:07
← thiago left18:07
SethRobertson Did `git diff HEAD -- filename` show whitespace noise?18:07
→ faceface joined18:08
faceface hello18:08
sh4rm4 no, its empty18:08
faceface how can I get git-status to give me the date the modification was made?18:08
I want to check if all the mods occured in one or two sesions18:08
SethRobertson sh4rm4: OK, perhaps you were mistaken. Try it the other way. Run `dos2unix` on the file.18:08
← LeMike left18:08
→ magnus_ joined18:09
sh4rm4 yep, that worked :)18:09
strange, nano converts LF to CRLF ? weird18:09
→ flyz_ joined18:09
← eno left18:09
SethRobertson faceface: `git status` cannot report modification time. You can post-process the output. Something like `ls -l $(git status --porcelain | colrm 1 3)` The colrm numbers might be wrong18:10
faceface sends git-status to file and 'fires up the emacs'18:10
faceface ty SethRobertson18:10
← flyz left18:10
flyz_ → flyz18:10
← ceej left18:10
faceface SethRobertson: can I pay $10 to get it coded ;-)18:11
→ TeckniX joined18:11
SethRobertson sh4rm4: OK, so `git diff HEAD^ -- filename` reported only your desired change, correct?18:11
sh4rm4 indeed18:11
→ eno joined18:11
→ tappi joined18:11
SethRobertson sh4rm4: OK, then run `git commit --amend` and you are done18:11
Well, after `git add filename`18:12
→ Skyx-mobile__ joined18:12
tappi is the difference between commit --amend and reset --soft ^HEAD and making another commit just that --amend replaces the existing commit while reset --soft ^HEAD + commit makes a new one?18:12
sh4rm4 SethRobertson, cool, it worked. thanks a lot.18:12
SethRobertson !thanks18:13
gitinfo Feeling thankful? Type "ExampleUser++", and ExampleUser will score karma points at http://carmivore.com (our preferred way to objectify self-worth). There's really no point to thanking me – instead, why not thank the person who made me help you?18:13
sh4rm4 SethRobertson++18:13
faceface ++18:14
SethRobertson tappi: ^HEAD is bad . Perhaps you mean HEAD^?18:14
faceface SethRobertson++18:14
tappi SethRobertson well yes18:14
faceface who's winning?18:14
tappi does ^HEAD even do anything?18:15
→ lonewulf` joined18:15
SethRobertson tappi: The difference is probably mostly in the fact that your old commit message is pre-populated. Both methods make a new commit.18:15
faceface: We don't know because the Carmivore people are too lame to report per-channel winner boards.18:16
tappi ah right18:16
→ shruggar joined18:17
→ _ikke_ joined18:17
tappi yeah, now that i checked help page says it straight up, "roughly equivalent for 'git reset --soft HEAD^' '... edit ...' 'git commit -c ORIG_HEAD'18:18
faceface heh18:18
column -t helped18:18
← gchristensen left18:19
→ nicxvan joined18:20
← tappi left18:20
← csherin left18:21
→ oriba joined18:23
→ mikepack joined18:25
→ csherin joined18:25
← eno left18:27
→ abetaha joined18:27
→ disappearedng joined18:28
→ eno joined18:28
Sigma[Mtp] → Sigma18:28
→ di-five joined18:30
← ehsan_ left18:30
→ ehsan joined18:30
→ zulax joined18:32
← lotsofcode left18:32
→ daviddoria joined18:33
← guns left18:33
daviddoria I have a repository that is 300MB, but there is really only 50MB worth of stuff in it. I ran 'git gc' to try to get rid of that old stuff, but it only brought it down to 260MB. Is there a way to keep the history of files that still exist, but get rid of old stuff that has been deleted?18:34
← werdnativ left18:35
→ DofLeFou joined18:36
mh_le for some reason git thinks I have made a commit one hour from now it says time + 1 ???18:37
← mgpcoe left18:37
→ ehsan_ joined18:38
← ehsan left18:38
→ gchristensen joined18:39
m1sc daviddoria: only with filter-branch (rewriting history)18:40
SethRobertson daviddoria: See !large_files18:40
gitinfo daviddoria: Git isn't yet great at large files (larger than RAM). See http://git-annex.branchable.com/ and https://github.com/schacon/git-media Find them (after gc) with: git verify-pack -v .git/objects/pack/pack-*.idx | grep blob | sort -k3nr | head | while read s x b x; do git rev-list --all --objects | grep $s | awk '{print "'"$b"'",$0;}'; done18:40
→ gusnan joined18:41
offby1 !delete18:41
gitinfo [!branch_delete] Deleting a branch is easy. `git branch -d branchname` (or -D). However, this will NOT delete the branch upstream, that requires `git push --delete origin branchname`. However, this will NOT delete remote tracking branches. EACH USER must `git remote prune origin`, but if any users have local branches, they must likewise run `git branch -d branchname` (or -D)18:41
offby1 hmm18:41
← eno left18:41
→ eno joined18:43
→ Guest27303 joined18:43
→ keller joined18:44
→ Cromulent joined18:46
← nelk left18:46
← Skyx-mobile__ left18:46
→ nelk joined18:46
rawtaz hm. Xcode shows a file missing even though the file is indeed in there18:47
in the filesystem18:47
← koo3 left18:47
→ latchup joined18:48
← Cromulent left18:49
rawtaz restart made it happier18:49
→ ericmuyser joined18:49
→ aknewhope joined18:49
→ jaalto joined18:49
← akiress left18:50
← latchup left18:50
→ lyddonb joined18:50
→ skylord5816 joined18:52
← skylord5816 left18:52
→ skylord5816 joined18:52
← skylord5816 left18:52
→ skylord5816 joined18:52
← Chillance left18:53
→ krake joined18:54
→ Chillance joined18:54
→ no1home joined18:54
→ QaDeS joined18:54
offby1 *shrug*18:56
software.18:56
← kerrick left18:57
→ kerrick joined18:57
→ plusk joined18:57
→ nick joined18:57
nick → Guest8041318:58
← Guest27303 left18:58
← lyddonb left18:59
avarab → avar19:00
← eno left19:00
← wald0 left19:01
→ eno joined19:01
→ lyddonb joined19:02
→ wald0 joined19:03
← techstein left19:03
← nelk left19:05
← AlbireoX left19:05
→ nelk joined19:05
→ TL joined19:07
← magnus_ left19:07
← eno left19:07
→ Alexander_ joined19:09
→ magnus_ joined19:09
→ eno joined19:09
← disappearedng left19:10
← ASIRINeiluj left19:10
Nickeeh hello guuyyss19:10
→ disappearedng joined19:11
Nickeeh I'm in the middle of a checkout -p and I'd like to edit a hunk. My hunk is "--++" (two lines --, two lines ++, consecutive)19:11
I'd like to keep the first line unaltered, and apply the second line.19:11
How should I change my hunk?19:11
→ cybersphinx joined19:11
→ ceej joined19:12
← ceej left19:12
← Liq^m left19:13
← the_cybersphinx left19:15
← Alexander_ left19:15
Mikachu Nickeeh: if i understand right, change the - on the line you want to keep to a single space, and remove the corresponding +line19:15
→ g105b joined19:16
← iocor left19:16
→ mindlogger joined19:16
mindlogger what does this msg mean? "Your branch is ahead of 'origin/admission' by 6 commits."19:17
→ aspotashev joined19:17
→ Alexander_ joined19:17
mindlogger I just merged it with master19:17
← Alexander_ left19:17
← daviddoria left19:18
→ Alexander_ joined19:18
Nickeeh Mikachu, that goes against what the editor says..19:20
← csherin left19:20
milki mindlogger: its ahead of the admission branch in the remote repository19:21
mindlogger ah right19:22
so I just need to push it19:22
milki yes19:22
Nickeeh: not sure myself, but the comments should tell you somewhere?19:23
← g105b left19:23
Nickeeh Hmpf, the patch doesn't apply/is corrupt.19:25
→ mandric joined19:25
← mantic0 left19:25
Nickeeh http://pastie.org/332309419:26
That's my problem.19:26
→ csherin joined19:27
← lyddonb left19:29
← jast left19:30
→ lotsofcode joined19:31
← cooldude left19:31
milki for add -p19:32
you change - to nothing19:32
and remove + lines19:32
→ jast joined19:32
→ cooldude joined19:33
← aknewhope left19:33
→ aknewhope joined19:34
Nickeeh milki, doing that gives me A' -B +B', right?19:34
just to make sure we're on the same page19:34
milki uh19:35
you want A B' to be the final outcome19:35
so it would be A -B +B'19:35
er19:35
a19:35
ya19:35
Nickeeh yeah.. hm, then it complains19:36
milki then i dunno19:36
~19:36
Nickeeh Unexpected line, etc. etc.19:36
milki usually gives up and manually fixes...19:36
Nickeeh I must learn19:36
to harness the POOWEERRRRRRRRRRRRRRRRRRRRR19:36
milki o19:36
→ lyddonb joined19:36
milki do you have line numbers at the top of the chunk19:36
you might need to change one of those numbers19:37
its start line number, final line number19:37
since you are only adding 1 line instead of 219:37
you mean need to decrement one of those19:37
well19:38
for patch diffs at least >.>19:38
← Dragooon left19:39
Nickeeh milki, I manually edited it.19:41
:(19:41
→ Goplat joined19:42
→ drindt joined19:43
drindt how to get a diff from between head and previous commit?19:44
Nickeeh drindt, git diff HEAD~1 HEAD19:45
or, the shorthand for that19:45
→ banseljaj joined19:45
Nickeeh drindt, git diff HEAD^119:45
← aknewhope left19:45
charon drindt: or simply 'git show' unless it's a merge commit19:45
banseljaj Hi. I mistakenly pulled a wrong branch. how do i undo it?19:46
charon Nickeeh: how is that a shorthand? ~1 and ^1 is the same, and omitting the HEAD means diff the worktree instead19:46
drindt thank you guys :) very friendly19:46
→ vervic joined19:46
Nickeeh charon, no, HEAD^1 is a range, HEAD~1 is a revision.. so that's the difference. I didn't know you could ommit HEAD. Thanks for that.19:46
banseljaj, git reflog19:47
charon Nickeeh: go read man gitrevisions, really19:47
gitinfo Nickeeh: the gitrevisions manpage is available at http://jk.gs/gitrevisions.html19:47
← lonewulf` left19:47
← FernandoBasso left19:47
charon Nickeeh: ^n means the n-th parent, and ~n means the n-th 1st parent ancestor. so as a special case, ^1 and ~1 (and ^ and ~, with implied n=1 in both cases) are the same19:48
→ lonewulf` joined19:48
banseljaj Nickeeh: I got the log, how do i tell it to rewind to a particular reviion? git reset?19:48
Nickeeh banseljaj, erm not.. sure.. reset or checkout. I think reset. --hard if you want to discard your worktree.19:49
But wait for someone else's reply. xD19:49
← eno left19:49
← pidus left19:49
Nickeeh charon, oh okay, I didn't kow that.19:49
banseljaj I want to discard my work tree. Discard anything that happened after the merge19:50
← pernekova left19:50
← drindt left19:50
→ eno joined19:51
Nickeeh then your probably want git reset --hard HEAD@{something}19:51
banseljaj Nickeeh: okay. Thanks a lot. :)19:51
← lyddonb left19:53
← pmenon left19:53
← Guest30232 left19:54
← Alexander_ left19:54
→ Invader joined19:55
← dreiss left19:55
← three18ti left19:56
← Milossh left19:56
→ Milossh joined19:57
→ AlbireoX joined19:58
banseljaj Nickeeh: It worked. Thanks a lot :)19:59
Nickeeh banseljaj, phew! Might've blown up your computer any other day. You're welcome.19:59
← fairuz_ left19:59
banseljaj Nickeeh: do you blow up computers regularly? :D If yes, i might have a job for you. :P20:00
← robotmay left20:01
→ fairuz_ joined20:01
← banseljaj left20:02
← wuj left20:03
milki lol20:03
i do20:03
but not on purpose20:03
and its always critical infrastructure20:04
→ dreiss joined20:04
→ DrHouse|TARDIS joined20:05
← pdtpatr1ck left20:10
→ pdtpatr1ck joined20:10
← csherin left20:11
← eno left20:13
← Jackneill left20:13
→ dorkmafia joined20:14
← plusk left20:14
← barr5790 left20:14
→ eno joined20:15
← pdtpatr1ck left20:16
→ fayimora joined20:16
← Lgb left20:20
← eno left20:21
mh_le when I cheked my git log the time said xx:xx + 1 even though the commit time was xx:xx - how can that be?20:22
← jacekmigacz left20:22
→ Lgb joined20:23
→ eno joined20:23
← mikepack left20:25
→ igotnolegs joined20:25
← magnus_ left20:26
→ caseymcg joined20:27
wereHamster mh_le: you mean the timestamp is off?20:27
by one minute?20:27
→ shergill joined20:28
→ smw joined20:30
← JohnBat26 left20:31
→ glennpratt joined20:32
→ magnus_ joined20:32
← faber left20:35
→ lyddonb joined20:37
← caseymcg left20:37
← Blaster left20:38
← h0bbit left20:39
← kyconquers left20:43
← hays left20:44
→ drizzd joined20:44
← echoSMIL1 left20:46
→ psoo joined20:46
← LongBeach left20:47
→ hays joined20:47
→ swombat joined20:47
← bronson left20:47
← leavittx left20:47
→ ehsan joined20:48
← ehsan_ left20:48
→ ceej joined20:49
→ echoSMILE joined20:50
→ timonia joined20:50
← aspotashev left20:50
← timonia left20:50
← gusnan left20:50
← nixmaniack left20:51
→ FernandoBasso joined20:51
→ nixmaniack joined20:52
→ JuanDaugherty joined20:54
← kennethreitz left20:54
JuanDaugherty has anyone used their regular PGP signature such as one posted on the MIT pgp server with github?20:55
mh_le wereHamster: no it said +1 litterally20:55
→ bronson joined20:55
patrick99e99 so I forked a project, made a branch, did some coding, pushed a reference to my github account... made a pull request.. And then this morning looked at my code and saw two small things I'd like to modify...20:57
canton7 mh_le, I don't know the exact context, but that sounds like a timezone to me20:57
mh_le wereHamster: I assume it means +1 hour20:57
patrick99e99 no one has responded to my pull request yet, so is it ok if i just do git push mine -f origin <my branch>20:58
or will the pull request get wonky?20:58
canton7 patrick99e99, so long as the pull request hasn't been merged in, github handles you force-pushing to them just fine (I've been told)20:58
mh_le canton7: yeah but as far as I can see I've set up my time zone correctly20:58
patrick99e99 canton7: cool thanks20:58
canton7 mh_le, and is your timezoen GMT+1 ?20:58
wereHamster mh_le: the time zone maybe?21:00
→ g105b joined21:00
wereHamster mh_le: what does `date` print?21:00
mh_le canton7: yeah, it said the commit was 19:26 + 1, but the correct time was 19:2621:00
canton7 mh_le, I know that. What timezone are you in? Answer wereHamster21:01
mh_le canton7: CET21:01
wereHamster: Sun Feb 5 22:00:55 CET 201221:01
wereHamster mh_le: that's +121:01
mh_le yeah but 19:26 + 1 is 20:26, no?21:01
canton7 there we go then. Your commit was made at 19:26, in the timezene GMT + 121:02
mh_le I see21:02
← jmho left21:02
← TL left21:04
← vervic left21:05
→ chip2n joined21:08
tizzo-afk → tizzo21:08
chip2n So, hmm, well, I'm kinda newbie at git usage, but is it really a good idea to use dropbox as a git server? I seem to get the "failed to read object" error sometimes when I pull/push :S21:09
bremner it's not a good idea afaict, because dropbox tries to be "clever".21:10
EugeneKay No.21:10
Use github or bitbucket.21:10
chip2n The project is sadly not open source, so I cant host it on github :/21:11
I got a server though, maybe I could set it up to host it?21:11
← JuanDaugherty left21:11
EugeneKay They have private repos and/or bitbucket has free ones21:11
canton7 bitbucket has free private repos21:11
EugeneKay Sure. If it's just you, set it up as a ssh:// remote repo21:11
If it's >1, !gitolite21:12
gitinfo Want to host as many git repos(and users!) as you like, on your own server, with fine-grained access control? You want gitolite: https://github.com/sitaramc/gitolite21:12
chip2n Alright, ill check out bitbucket! I found a guide about gitosis also, seemed cool21:12
EugeneKay Don't use !gitosis21:13
gitinfo gitosis is no longer maintained and supported by the author; we usually recommend gitolite instead which has much better documentation and more features: http://github.com/sitaramc/gitolite21:13
chip2n Ah, thanks for the info :D21:13
I'll be sure to check out both bitbucket and probably use it until I feel confident about gitolite21:14
← soulcake left21:14
← weichkaes left21:15
→ pdtpatr1ck joined21:15
→ weichkaes joined21:15
← nixmaniack left21:15
← mdel left21:15
← Katniss left21:19
← weichkaes left21:19
→ iocor joined21:19
← Dreamer3__ left21:20
← krake left21:20
→ krake joined21:21
← xissburg left21:21
← candybar left21:21
← krake left21:22
→ lekefly_ joined21:23
AlexMax So what happened with this?21:23
http://paste.pocoo.org/show/Drzk4QMUvky5dYZ6Zrkx/21:23
I tried to run this shell script in a git filter-branch21:23
It worked just great when I tested it on individual commits21:24
However, when I ran filter-branch, it had this really weird effect where the contents of ~/newlic.cpp.txt were pasted on top of the unmodified original file21:25
and then there were three lines that were filled with colons21:25
← LekeFly left21:25
← ajt left21:25
canton7 AlexMax, what's your filter-branch command?21:26
← zulax left21:26
AlexMax git filter-branch --tree-filter '/home/alexmax/workspace/skulltag/unlag.sh' -- …21:26
…--all21:26
ignore those elipses21:26
that's just fish21:27
← twisted` left21:27
← starlays left21:28
← di-five left21:29
→ acl_ joined21:29
→ Blaster joined21:30
→ twisted` joined21:30
Blaster → Guest7160121:30
← Guest71601 left21:31
AlexMax canton7: As you can see, it's nothing fancy21:31
tizzo → tizzo-afk21:32
canton7 AlexMax, I'm afraid I have no idea... Someone else might though21:32
→ blaster joined21:33
← twisted` left21:33
→ twisted` joined21:33
← whitman left21:33
bremner AlexMax: without looking, isn't tree-filter supposed to be for a subdirectory?21:33
blaster Hello. I am trying to remove a folder from the index so I typed git rm <folder name> but it says pathspec 'eec_design/' did not match any files. I am positive the folder is in the file tree.21:33
← Invader left21:33
← pdtpatr1ck left21:33
blaster The folder is also listed under 'Untracked files' when I type git status21:34
milki folders are not tracked by git21:34
← jonas_g left21:34
milki you want to rm everything within the folder21:34
canton7 (but if there's an untracked folder with contents, git will just list the folder as untracked)21:34
blaster I tried a recursive git rm <folder name> -R21:34
can you just remove the whole folder and it's contents from the index?21:35
canton7 blaster, the git rm's failing because "git rm" acts on things which git knows about. The thing you're trying to "git rm" is untracked, therefore git doesn't know about it21:35
if git claims that it's untracked, then it isn't in the index21:35
blaster oh ok21:35
so if it's listed under untracked then it's in the file tree?21:35
canton7 yup21:35
blaster ok thanks21:36
→ candybar joined21:36
← candybar left21:36
→ candybar joined21:36
→ italiano40 joined21:36
blaster if I have a folder titled .git in it can I recover the file tree and all commits some how?21:36
← matsebc left21:36
canton7 sure. All of the commits, etc, are safely inside it. If you want a checkout, then it depends whether it was originally inited as a bare repo, or a non-bare one21:37
→ bitkiller joined21:37
← mandric left21:38
blaster ok21:38
→ matsebc joined21:39
→ twisted`_ joined21:40
← zarubin left21:40
← fairuz_ left21:41
→ zarubin joined21:41
AlexMax bremner: Nope, it checks out the tree21:41
It's really odd21:41
← khmarbaise left21:41
AlexMax because the end result ends up being new_header ::: sv/unlagged.cpp ::: old_file21:41
instead of new_header old_rest_of_file21:41
http://paste.pocoo.org/show/546452/21:43
← zarubin left21:43
AlexMax What I thought was going to happen is that the top bit would end and then the rest of the file would follow21:43
← twisted` left21:43
twisted`_ → twisted`21:43
AlexMax but instead there's a bunch of colons, the name of the file, then the old file in its entirety21:43
blaster hey if I remove some files from the index and then commit that change, why would it remove the files from the file tree on my remote server after I pull there?21:43
AlexMax and it only happens with filter-branch, if I run it straight from the shell it works dandy21:43
→ fairuz_ joined21:43
← lyddonb left21:45
← hays left21:45
→ hays joined21:46
← i42n left21:46
mindlogger I just removed a branch how can I sync it with my remote repo?21:48
→ eijk_ joined21:48
← eijk left21:49
mindlogger I can't push that branch? how does the remote repo realize that this branch is no longer in the local repo?21:49
cbreak you can push the branch21:49
just push an empty one21:49
git push remotename :remotebranchname21:49
mindlogger error: src refspec thirdparties does not match any.21:50
← fisted left21:50
← pandeiro left21:50
→ fisted joined21:51
mindlogger cbreak: I get this error "error: src refspec thirdparties does not match any." when I do "git push origin thirdparties" ,21:51
cbreak yes21:51
that's normal21:51
do you have a local branch thirdparties?21:51
mindlogger not anymore21:51
I just removed it21:51
cbreak yeah, so it's normal21:51
you have to do what I told you above21:51
the : is not just there for decoration :)21:52
mindlogger cbreak: what do you mean by an empty branch?21:52
ah ok21:52
milki mindlogger: you forgot the :21:52
cbreak the syntax is git push remotename localbranchname:remotebranchname21:52
if you have a newer git, you can also just do git push --delete remotename remotebranchname21:52
but since you asked, you probably didn't find delete in your man page21:52
AlexMax I'm probably going to end up rewriting this in python21:53
cbreak so you don't have a git new enough for that21:53
milki oo, delete is new?21:53
haha21:53
fairuz_ Hi, is git rebase branchA HEAD is the same as git rebase branchA?21:53
cbreak milk: well. There was a time where it didn't exist21:53
→ uglyoldbob joined21:53
← uglyoldbob left21:54
← cccaldas left21:54
mindlogger cbreak: git push --delete origin thirdparties worked21:55
thanks21:55
cbreak so I guess you have a new enough git :)21:55
and just couldn't find "delete" in git push :)21:55
mindlogger yeah I was looking for delete in branch21:56
I didn't look for it in push :(21:56
→ lyddonb joined21:56
mindlogger is there a way to get how many line you have coded since firs commit until now?21:56
for all branches perhaps? or is it dumb and I just have to do it with a line counter wc perhaps21:57
← macmartine left21:58
EugeneKay mindlogger - man git-log, see --numstat22:00
gitinfo the git-log manpage is available at http://jk.gs/git-log.html22:00
→ ShadeHawk joined22:00
mindlogger EugeneKay: ty22:00
blaster how can you see what's different between your current file tree and an index you're about to pull?22:00
→ uglyoldbob joined22:00
ShadeHawk I don't envy gitster (Junio) trying to manage this herd of cats without offending anybody, losing important contributions, and lovering quality of patches...22:01
EugeneKay fetch it, then abuse diff to your heart's content22:01
milki EugeneKay: is that how you treat your diff!?22:01
cbreak blaster: just use git fetch and then it's local22:01
EugeneKay I named mine steve22:02
uglyoldbob im trying to figure out a daily build script for a specific branch of a git repository22:02
i want the script to build only when git tells me there were changes to the branch22:03
→ diogogmt joined22:03
milki on a bare repo, you could do a post-receive hook22:05
→ wroathe joined22:05
wroathe So lets say I created a remote branch called feature_branch and I want to pull the most recent changes in master into that branch after not working on it in a while22:06
How would I go about doing that?22:06
uglyoldbob milki: i thought about that, but i dont want it to apply to every user of the repository, just the build machine22:06
wroathe: "git checkout feature_branch" and "git pull" i think, depending on how you created the local branch22:07
→ cccaldas joined22:07
→ werdnativ joined22:07
← thomas_b left22:07
wroathe git pull would just pull from the feature_branch on the server wouldn't it?22:07
I'm extremely new to git :P22:08
milki uglyoldbob: explain the setup then22:08
→ macmartine joined22:08
tizzo-afk → tizzo22:09
→ thomas_b joined22:09
uglyoldbob i want to have a specific computer run a script every day that checks to see if a specific branch had any changes (commits or merges)22:09
it it did, then it will do a complete build and upload a certain file to a web server22:10
milki where are the repos22:12
← Ksugzulut left22:12
uglyoldbob they are on my git server (im running gitolite)22:13
milki and the specific computer is not the git server?22:13
→ Sharas joined22:13
uglyoldbob that is correct22:13
tizzo → tizzo-afk22:14
→ znq joined22:14
milki you could setup mirroring from specific compter with gitolite server, and have specific computer run post-receive hook22:14
← Dave^|| left22:15
uglyoldbob is there not a way to do it with a shell script run by a cron job?22:15
← znq left22:15
← knittl left22:16
← unreal left22:16
→ znq joined22:16
milki then its up to you to check for changes22:16
← Xizor left22:16
milki fetch objects, if objects found, then changes happened22:16
← QaDeS left22:17
← ShadeHawk left22:17
← lyddonb left22:18
→ unkmar joined22:19
uglyoldbob what about parsing git log after doing a pull?22:19
unkmar git'er done.22:19
how do I pull from I haven't setup as a dedicated repo?22:19
gitinfo set mode: +v22:20
milki unkmar: can you reword that?22:20
Sharas hi22:20
trying to run git-daemon on windows 7 with --detach22:20
milki uglyoldbob: sounds fine if you want to do it that way?22:21
Sharas it just doesnt start22:21
nothing listens on default port22:21
anybody knows what might be wrong?22:21
starts all right with console22:21
unkmar milki: I cloned a repo that I previously created. I can't reach the original repo at the moment. Yet, the clone is still available. How can I pull from it?22:21
→ centipedefarmer joined22:21
← wroathe left22:22
milki unkmar: you want to pull from the cloned repo instead of the original repo?22:22
→ frsk_ joined22:22
← frsk_ left22:23
milki you can setup a new remote with git remote add22:23
unkmar yes22:23
blaster if you have a file in the tree and then you add it to .gitignore so it stops tracking it then how do you remove that file from the index?22:25
← icwiener left22:25
→ icwiener joined22:25
blaster i have various server-specific config files and I just want them to stay as they are on each of the servers22:25
but when I do a git rm --cached <file> and then commit it's removing it from my other servers22:25
after I pull from those servers22:26
unkmar milki: I'm sorry, I don't quite follow.22:26
← cccaldas left22:26
← bauruine left22:26
milki unkmar: your original repo is a remote called origin22:27
you just need to add your cloned repo as a sepearate remote22:27
unkmar I have a clean system I want to clone the repo to. but my original isn't available.22:27
milki and then fetch from the new remote22:27
o22:27
← uglyoldbob left22:27
milki why dont you just clone the clone then?22:27
→ looopy joined22:27
looopy is it possible to branch off of a branch?22:27
← TeckniX left22:27
milki yes22:27
→ bauruine joined22:28
looopy say i make an experimental branch and i don't want that to be a master per say...but a'experimental master' experimental/new_feature22:28
how would i go about that? not sure what to search for o_O22:28
milki git branch newbranchname22:28
?22:28
← ThomasLocke left22:29
looopy o_O;22:29
chaoflow looopy: while on the branch you want to branch from: git branch experimental/new_feature22:29
looopy hah...no sheeet -_-22:29
EugeneKay This won't work if you already have a branch named 'experimental'22:30
looopy now what if i wanted to merge one particular branch from that experimental branch to master just git merge blah/blah_2422:30
EugeneKay Because 'experimental' can't simultaneously be a file(which is what the branch pointer is) and a directory22:30
While on master, yes.22:30
→ Cam joined22:30
← joshsmith left22:30
EugeneKay Or you can do it the other direction, 'git merge master' from blah/blah_2422:30
unkmar looopy: while in the brance you want to make a branch of.22:31
→ fruight joined22:31
looopy gotcha22:31
EugeneKay These will behave mostly the same, but the parent commit of your merge commit wll be reversed22:31
→ airborn_ joined22:31
EugeneKay Merge strategies involving --theirs or --ours will also be different, ofc22:31
← bitkiller left22:31
← echoSMILE left22:31
looopy ok22:31
unkmar milki: I would but I guess I'm an just too ignorant on HOW.22:32
nm22:33
→ Chappo joined22:33
unkmar ught22:33
milki you said you had access to the clone22:33
looopy alright that worked...i'll be sure to return in 5 hours with a merge conflict =)22:33
milki lol22:33
unkmar I finally got it.22:34
That makes my life so much easier.22:36
← faceface left22:37
→ johnkpaul joined22:37
tizzo-afk → tizzo22:38
unkmar milki: ty, ty, ty, ty, ty. Many many thank you's. I wish I had taken time to learn that 2 weeks ago.22:38
milki you will want to replace the remote with the original repo location when its available though22:39
→ frogonwheels joined22:40
unkmar That was almost too easy.22:40
milki: nope. Running from a LIVE CD. but in the near future. My original repo is to be moved. Dang network there isn't stable enough. :(22:40
← kerrick left22:41
→ EricInBNE joined22:41
blaster if you have a file in .gitignore should it still show up under 'Changes not staged for commit'?22:41
← canton7 left22:41
unkmar blaster: it shouldn't22:41
milki i believe .gitignore is only for untracked files?22:41
Tabstar → Tabmow22:42
milki ya22:42
says so in the description22:42
blaster how do you untrack a file but not cause it to show up in git status?22:42
milki git rm first then add it to ignore?22:42
blaster its already in ignore, and when I do git rm it removes the file... but I need to keep the file because it's some server specific config22:43
I tried git rm --cached and it removes the file from my other repos when I pull22:43
← znq left22:43
blaster they have the same file on the other repos with their server specific config22:43
→ joshteam joined22:44
frogonwheels blaster: ~pull22:44
!pull22:44
gitinfo pull=fetch+merge (or with flags/config also fetch+rebase). It is thus *not* the opposite of push in any sense. A good article that explains the difference between fetch and pull: http://longair.net/blog/2009/04/16/git-fetch-and-merge/22:44
blaster shouldnt pull be the opposite of push?22:44
frogonwheels blaster: it's not really.22:45
← skylord5816 left22:45
← wald0 left22:45
frogonwheels blaster: or it is kind of, but under very narrow circumstanses.22:45
blaster: actually, no, I'll go back to gitinfo on this one. it's not really the opposite.22:46
← Airells left22:46
→ koo4 joined22:46
frogonwheels blaster: imho pull is useful when you don't need to push :) otherwise use fetch and merge / rebase, until you understand what you're doing :)22:47
→ wald0 joined22:47
← JayBazuzi left22:48
← cakehero left22:48
AlexMax Here is my question from an hour ago in stackoverflow form...22:49
http://stackoverflow.com/questions/9153824/shell-script-works-fine-by-itself-but-produces-unexpected-results-when-run-thro22:49
→ koo5 joined22:49
← erichynds left22:52
SethRobertson AlexMax: If you use tail -n +34 instead of more +34, does it work?22:53
← g105b left22:53
AlexMax SethRobertson: Trying now22:54
← fairuz_ left22:56
← Shusshu left22:56
→ isxek joined22:59
← acl_ left22:59
→ Chappo_ joined23:01
← Chappo left23:01
Chappo_ → Chappo23:01
airborn_ → airborn23:01
← blaster left23:02
→ cakehero joined23:02
← johnkpaul left23:02
→ echoSMILE joined23:03
← mastro left23:03
← Sharas left23:05
→ codewrangler joined23:08
← icwiener left23:09
→ Davey joined23:09
← Davey left23:09
→ Davey joined23:09
→ Dreamer3__ joined23:11
→ bigkm joined23:13
← isxek left23:15
→ thiago joined23:17
← thiago left23:17
→ thiago joined23:17
→ johnkpaul joined23:19
← Heimidal left23:20
← dharrigan left23:21
← johnkpaul left23:21
← d0k left23:22
→ wuj joined23:23
→ wroathe joined23:23
← rendar left23:24
← akosikeno left23:24
→ mar77i joined23:25
→ vinod joined23:25
mar77i hey, is it possible for me to get every commit diff since a certain commit?23:25
← Chappo left23:25
thiago git log -p thatcommit..23:25
SethRobertson git log certaincommit...23:25
thiago including the two dots23:26
← ph^ left23:26
thiago also: git format-patch thatcommit..23:26
← ianmcorvidae left23:26
mar77i oh awesome23:27
← vinod left23:27
→ bitkiller joined23:28
→ ianmcorvidae joined23:30
← FernandoBasso left23:30
mar77i format-patch spit out multiple files. I'm fine with it all in a single file and went with git log -p commit.. >some_file.log23:30
thiago git format-patch --stdout23:30
← cakehero left23:31
→ Pessimism joined23:31
→ deeebug joined23:31
→ dinduks joined23:32
← glennpratt left23:32
→ acl_ joined23:33
← diogogmt left23:33
mar77i I'll note that.23:33
→ diogogmt joined23:33
→ __machine joined23:36
← shruggar left23:36
dinduks Hi, I just did my first merge with git, and there's something I don't understand, so if can anyone explain: I merged branch "auth" with master, but some new files stayed in auth where I created them, they don't appear in master, as if I've done no merge. Is that normal? Thanks.23:37
__machine im using git-svn to get familiar before i switch our repos... i have my Subversion remote and git-svn branch on it... and i have an origin/master remote branch at bitbucket... i was always pushing to subversion first, then pushing to bitbucket to keep it updated...23:37
now i have about 10 or so commits in Subversion/git-svn AND origin/master, but they have different commit hash and timestamps... not sure how this happened, or how to fix it?23:38
→ dharrigan joined23:38
← dharrigan left23:38
→ dharrigan joined23:38
thiago push the Subversion commits to GitHub23:39
← psoo left23:39
__machine they are already on the origin/master (bitbucket)... but they seem to have different commit hash and timestamp?23:40
← vandal- left23:41
SethRobertson dinduks: That is not normal. We can work through understanding what happened.23:41
__machine when i try push again i get... "failed to push some refs" ... "To prevent you from losing history, non-fast-forward updates were rejected" ... "Merge the remote changes (e.g. 'git pull') before pushing again." ...23:41
SethRobertson faq non-ff23:42
gitinfo Your push would lose changes on the remote; more details available at http://jk.gs/gitfaq#non-ff23:42
→ Skyx-mobile__ joined23:42
__machine yes... but why does the remote have commits that i dont have? they are the same identical changes, but with a different timestamp and commit hash... not sure how that happened... i always pushed to git-svn first after making changes... then just pushed to origin to keep the bitbucket mirror updated...23:43
← deeebug left23:43
tizzo → tizzo-afk23:44
SethRobertson You either rewrote public history (rebased or amended a commit after you pushed) or someone (including you) made a change in a different clone of the upstream repo and uploaded it23:44
Sounds like you probably rebased or amended the commit23:44
__machine i dont think i amended... but git-svn always does git svn rebase when it pulls down latest changes?23:45
→ Chib joined23:45
SethRobertson Are you mixing a svn upstream with a non-svn upstream?23:46
← diogogmt left23:46
→ diogogmt joined23:46
→ glennpratt joined23:46
__machine i started with git svn clone ... then i added an "origin" remote at bitbucket as well... when working i pull from Subversion/git-svn, which rebases my local commits on top of latest svn code... then i push to Subversion/git-svn ... then i push to origin/master to keep the bitbucket remote up to date23:47
→ psoo joined23:47
SethRobertson That, specifically the rebase nature of git-svn, would do it. You need to push -f to bitbucket.23:48
← Dreamer3__ left23:48
→ pl0sh joined23:48
__machine the Subversion/git-svn commits are 12 minutes *after* the origin/master commits, according to the timestamps... the origin/master commits have the svn commit metadata on them...23:48
dinduks Thanks SethRobertson, I don't know from where to start thought. :/ How can I fix that?23:48
→ barr5790 joined23:48
SethRobertson First make sure there are no outstanding changes. `git status`23:49
__machine i see... still not sure what i did wrong to cause this though? i did plenty of other commits this way before now and did not get this problem... i must have changed the workflow slightly this time, but im not sure how...23:49
→ TL joined23:50
SethRobertson __machine: Any time you have a local commit which is rebased, it changes and gets a new SHA. If you have pushed that commit somewhere else, then they will need a force push. Either you previously hadn't uploaded to bitbucket yet or you had uploaded to svn23:50
→ robotmay joined23:51
← giallu left23:51
← drizzd left23:51
__machine SethRobertson: could it be... i pushed to svn... then rebased from origin... so my commits which were already pushed to svn but not on origin, were rebased and then pushed... so what i should do is... local commit, git svn rebase... then pull from origin (NOT rebase)... and then push to origin?23:53
SethRobertson If you are not pushing your changes to svn, any svn change will cause your local changes to be rebased and thus bitbucket will need a force push23:54
→ Dreamer3 joined23:55
→ al4nc4ds joined23:58
__machine thanks... i think i fixed it... did a reset to the last git-svn commit, and force pushed that to bitbucket23:58
SethRobertson That doesn't sound like a good idea, but if you are happy23:59
__machine why is it not a good idea?23:59
SethRobertson Because you just lost your local commits!!23:59
← krisfremen left23:59

Logs Search ←Prev date Next date→ Channels Documentation