Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2016-08-31

| Channels | #git index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:00 Raed|Laptop joined #git
00:01 clemf joined #git
00:04 vbgunz !backup
00:04 gitinfo Worried about your data while trying stuff out in your repo? The repository in its entirety lives inside the .git directory in the root of your work tree so to backup everything `cp -a path/to/workdir path/to/backup` or equivalent will suffice as long as the repo is not modified during backup. See also http://sethrobertson.github.com/GitBestPractices/#backups
00:05 DolpheenDream joined #git
00:08 sleepee joined #git
00:08 netj joined #git
00:12 alansaul joined #git
00:23 danimal2 joined #git
00:24 danimal2 joined #git
00:28 sangy joined #git
00:29 mingrammer joined #git
00:32 cust_ joined #git
00:34 sdothum joined #git
00:35 azwieg103 joined #git
00:35 vassagus joined #git
00:39 clemf joined #git
00:46 SwiftMatt joined #git
00:49 mischief- git brunch --contains sandwich
00:53 warthog9 joined #git
00:54 Gsham joined #git
00:55 xiongmao86 joined #git
00:56 M-shine joined #git
00:57 ljc joined #git
00:57 xiongmao86 Hello, I wonder why 'git log --stat -1' doesn't show the file stat?
00:57 xiongmao86 but 'git show --stat -1' does
00:58 Orion3k joined #git
01:00 MaddHatter I get the same output for both
01:00 MaddHatter do you have an alias or something?
01:00 i7c Here they both do. Note that the -1 is nonesense for stat command.
01:00 i7c show command*
01:01 xiongmao86 MaddHatter: are there any configs that I can check against?
01:01 chaos-zhang joined #git
01:01 chigang_ joined #git
01:01 xiongmao86 i7c: Thanks.
01:02 i7c You can check ~/.gitconfig, ~/.config/git/config
01:03 i7c Or you could run git config --list
01:03 xiongmao86 [filter "lfs"]
01:03 xiongmao86 clean = git-lfs clean %f
01:03 xiongmao86 smudge = git-lfs smudge %f
01:03 xiongmao86 required = true
01:04 vrishab joined #git
01:04 xiongmao86 i7c: other than the user part, this is it in ~/.gitconfig
01:05 i7c There’s also a project-local gitconfig in the .git directory of each project. I'd just grep git config --list tho
01:05 cyphase joined #git
01:06 xiongmao86 ======
01:06 xiongmao86 credential.helper=osxkeychain
01:06 xiongmao86 user.name=xiongmao86
01:06 xiongmao86 user.email=superkeen@sina.cn
01:06 xiongmao86 filter.lfs.clean=git-lfs clean %f
01:06 xiongmao86 filter.lfs.smudge=git-lfs smudge %f
01:06 xiongmao86 filter.lfs.required=true
01:06 xiongmao86 ======
01:06 xiongmao86 i7c: this is config list
01:06 i7c Uhm please use a paste service if you wanna paste more than 2 lines and just provide the URL here, in the future.
01:07 i7c And yeah, obviously no alias there.
01:07 i7c What git version are you running?
01:07 xiongmao86 git version 2.9.2
01:08 ahr3n joined #git
01:08 xiongmao86 i7c: mac os x platform
01:08 i7c Okay, then I have no idea why you don't see the file stat. Sorry.
01:09 xiongmao86 i7c: Thanks all the same.
01:11 alansaul joined #git
01:12 docnvk joined #git
01:14 anonymuse joined #git
01:17 Dougie187 joined #git
01:18 dmto joined #git
01:20 romerocesar joined #git
01:22 pikajude joined #git
01:22 pikajude joined #git
01:24 mingrammer joined #git
01:24 safe joined #git
01:29 nwmcsween joined #git
01:29 M-shine joined #git
01:29 durham_ joined #git
01:30 daynaskully joined #git
01:30 ivar joined #git
01:31 magicturtle joined #git
01:32 PaulCapestany joined #git
01:33 jhass joined #git
01:33 danslo joined #git
01:36 vassagus joined #git
01:36 evanwang joined #git
01:38 l4v2 joined #git
01:39 warthog9 joined #git
01:39 cqi joined #git
01:40 Kulrak joined #git
01:40 Kulrak joined #git
01:41 daynaskully joined #git
01:46 pks joined #git
01:46 sanketdg joined #git
01:46 Emperor_Earth__ joined #git
01:48 ilbot3 joined #git
01:48 Topic for #git is now Welcome to #git, the place for git help and doorbells | Public logs at http://goo.gl/BuUi5o | Current stable version: 2.9.3 | First visit? Read: http://jk.gs/git | Getting "cannot send to channel"? /msg gitinfo .voice | Knock knock. Who's there? Git. Git-who? Sorry, 'who' is not a git command - did you mean 'show'?
01:51 HoierM joined #git
01:52 fakenerd joined #git
01:53 harish joined #git
01:53 Vampire0_ joined #git
01:53 d^sh joined #git
02:00 Anupkumar joined #git
02:01 rominronin joined #git
02:03 jaguarmagenta joined #git
02:03 Aces_Charles joined #git
02:04 BenderRodriguez joined #git
02:06 ytl joined #git
02:06 dsantiago joined #git
02:09 chachasmooth joined #git
02:11 mycyber joined #git
02:12 Gsham joined #git
02:19 BSAlb joined #git
02:20 BSAlb joined #git
02:20 BSAlb joined #git
02:21 BSAlb joined #git
02:21 BSAlb joined #git
02:24 hexagoxel joined #git
02:25 Darren_ joined #git
02:34 danslo joined #git
02:36 fstd_ joined #git
02:40 daivyk joined #git
02:44 shgysk8zer0 joined #git
02:44 jaguarma_ joined #git
02:45 alansaul joined #git
02:45 snowkidind joined #git
02:45 lagothrix joined #git
02:52 jaguarmagenta joined #git
02:52 chrisshattuck joined #git
02:53 ahr3n_ joined #git
02:53 Dougie187 left #git
02:54 msafi joined #git
02:54 msafi Hey folks, I have this situation in my git repo https://i.imgur.com/OAoi63s.png
02:55 msafi Is it possible to squash the 2 gold commits and have them appears as part of master?
02:55 msafi And then squash the pink ones the same way?
02:58 mozzarella joined #git
03:00 msafi All while maintaining the authors of the commits...?
03:01 hahuang65 joined #git
03:04 aavrug joined #git
03:05 thiago msafi: not easily
03:05 thiago it's possible, but given that you're branching and merging left and right, that makes the operation more difficult
03:06 na__ joined #git
03:07 hobodave joined #git
03:11 mizu_no_oto joined #git
03:11 alansaul joined #git
03:12 mingrammer joined #git
03:14 dhollinger joined #git
03:15 anonymuse joined #git
03:16 chaos-zhang_ joined #git
03:20 raijin joined #git
03:20 sangy joined #git
03:21 watabou joined #git
03:22 msafi joined #git
03:22 msafi Sorry, got disconnected
03:22 msafi thiago: it's all by mistake
03:22 msafi That's why I wanna fix it
03:27 mizu_no_oto joined #git
03:27 mycyber joined #git
03:28 raijin joined #git
03:30 Andrew_K joined #git
03:30 AnarchyAo joined #git
03:32 b1tchcakes joined #git
03:33 msafi Okay, an easier question then, how do I just get rid of that "no message" commit?
03:37 vassagus joined #git
03:40 BinGOs joined #git
03:43 Kuukunen msafi: probably the easiest way to do the thing would be interactive rebase
03:45 Kuukunen I think it should retain the authors
03:45 e14 joined #git
03:47 d4rklit3 joined #git
03:48 miklcct joined #git
03:54 thiago msafi: a rebase will linearise history
03:55 Kuukunen yea well, I assumed that would be fine
03:55 msafi Yeah, the problem is the commit that merges master into the branch
03:55 msafi Actually the goal is to linearize master
03:55 Kuukunen then rebase
03:55 msafi Like how it was https://i.imgur.com/fEm3ZLo.png
03:56 Kuukunen msafi: but of course the first question to ask is... is this is a repo you have pushed for other people to read?
03:56 romerocesar joined #git
03:56 msafi Kuukunen: the problem is the second commit from the top
03:56 msafi Kuukunen: yes
03:56 msafi Kuukunen: there are like 30 forks
03:57 Kuukunen msafi: the thing is, if you rewrite history in a repo other people have pulled from, it can cause some trouble
03:58 msafi All these bad commits happened about an hour or two ago
03:58 msafi If I rewrite them no, hopefully the damage will be minimal, if any
03:58 msafi no = now
03:59 Kuukunen just so you know, if someone DID pull in the last hour, then in the next day they write some code on top of the commit, then pull again, their work will be deleted
03:59 Kuukunen ...at least I think that's still how it happens
03:59 Kuukunen and last time I checked, they don't even get a warning about it
04:00 Peuc joined #git
04:00 mingrammer joined #git
04:00 dsantiago joined #git
04:02 navidr joined #git
04:02 Kuukunen msafi: so basically you want to be pretty damn sure no one has pulled :P
04:02 msafi I can't. I'm just gonna rebase. Push now. And hope for the best...
04:03 rominronin joined #git
04:03 dreiss joined #git
04:03 thecomedian joined #git
04:04 daey_ joined #git
04:04 tHEMtZ joined #git
04:04 Kuukunen msafi: in that case what you probably want to do is make a new branch out of the current master, then reset --hard master to the last commit you know is good, then rebase --interactive the new branch on top of master
04:04 Kuukunen and then use the text editor to remove and squash the commits
04:04 Spec-Chum joined #git
04:05 Kuukunen msafi: but just so you know, the correct way to fix messes is to commit a new commit that fixes it, not rewriting history
04:06 mehola joined #git
04:06 Remram[m] joined #git
04:06 Ineentho[m] joined #git
04:06 M-mistake joined #git
04:06 M-meznak joined #git
04:07 msafi Okay, I'm about to push. If I don't post again, you people will know there was a mutiny and I got murdered...tell my family I love them...
04:08 msafi Looks like it worked, woohoo! https://github.com/text-mask/text-mask/commits/master
04:09 msafi The new line looks nice and clean! https://i.imgur.com/uHSASnx.png
04:11 msafi The magic of SourceTree. Out of desperation I tried to drag and drop commits to rearrange them and it actually worked!
04:11 Kuukunen right :V
04:11 Kuukunen msafi: I've no idea how to do it with sourcetree, but if you want to keep your history clean, use pull --rebase instead of just pull
04:11 jnewt when i do a git rebase --i to edit a commit in history, how do i get all the child branches to go with the edit instead of getting a copy of whatever branch i am on?
04:11 alansaul joined #git
04:16 dermoth joined #git
04:17 SpecChum joined #git
04:17 kadoban jnewt: You need to then rebase the other branches on the new, rewritten branch, if that's what you want to do. I don't think there's an automatic way to do that.
04:17 rscata joined #git
04:17 jnewt kadoban: that's what i usually do, but it gets tedious if there's several branches or lots of commits and all you want to do is split a single commit.
04:17 anonymuse joined #git
04:17 nd joined #git
04:17 KeelOfSteel joined #git
04:17 Kuukunen jnewt: what do you mean by that?
04:17 Kuukunen jnewt: you mean, you have multiple branches that have gone off in lots of different directions and you somehow want to rebase all of them on top of something?
04:17 kisonecat joined #git
04:17 arand_ joined #git
04:18 kadoban They have a commit reachable from N branches, they want to modify that commit in all branches (or some subset maybe)
04:19 Cabanossi joined #git
04:20 cworth joined #git
04:21 jnewt yeah, rebase splits out a new branch, when what i really want is to just split a commit, and leave everything else in the same structure
04:22 Kuukunen I don't know if it's easily possible, but to me it sounds like it's very much against the whole design philosophy of git
04:22 Kuukunen like, maybe if you write a script to automate it or something...
04:23 LUMIA930 joined #git
04:25 fakenerd joined #git
04:27 alansaul joined #git
04:27 zrts joined #git
04:30 subhojit777 joined #git
04:31 subhojit777 joined #git
04:33 mohabaks joined #git
04:35 danslo joined #git
04:39 justanotheruser joined #git
04:48 Puffball joined #git
04:51 a_thakur joined #git
04:51 Sasazuka joined #git
04:59 _ng joined #git
05:02 hahuang65 joined #git
05:06 cdown joined #git
05:07 dmto joined #git
05:09 cdown_ joined #git
05:14 warthog9 joined #git
05:15 Hink joined #git
05:17 mangolisk joined #git
05:17 cqi joined #git
05:24 spb joined #git
05:26 stuh84 joined #git
05:29 Repox joined #git
05:31 bmahe joined #git
05:32 jguillen joined #git
05:33 iskander_work joined #git
05:36 danslo joined #git
05:44 ThomasLocke joined #git
05:45 cdown joined #git
05:45 anttim joined #git
05:46 anonymuse joined #git
05:47 anttim joined #git
05:49 newbiee joined #git
05:49 qt-x joined #git
05:50 newbiee I have made some changes to some code and i want to push these changes to a different git server along with all the previous history. How to do it?
05:50 d0nn1e joined #git
05:51 newbiee i can make a single commit and push it to git server but then i will lose the history. How to preserve the history?
05:52 ojacobson Make a single commit and push it to the new server.
05:52 settermjd joined #git
05:52 ojacobson Git will do the right thing.
05:53 ojacobson (Generally, when you push a name, git will also push any objects necessary for that name to make sense. For a name that refers to a commit, like a branch, that's all of the commit's files, plus all of its ancestors - the entire history of that branch, back to the beginning/s of time.)
05:54 thiago every commit includes the entire history leading up to it
05:54 newbiee ojacobson:let me test
05:56 mingrammer joined #git
05:56 dmto I was gonna git push origin master, but then I got high
05:56 tvw joined #git
05:57 Kuukunen I was gonna rewrite history too, but then I got... sober
06:00 Juri_Agarin joined #git
06:00 ytl joined #git
06:02 ochorocho__ joined #git
06:03 Raging_Hog joined #git
06:04 mathu joined #git
06:06 scorphus joined #git
06:09 bocaneri joined #git
06:10 elect joined #git
06:11 omegak joined #git
06:13 Kicer86 joined #git
06:16 slayerjain joined #git
06:17 sanscheese joined #git
06:20 yuhlw_ joined #git
06:20 JeroenT joined #git
06:27 jaguarmagenta joined #git
06:29 govg joined #git
06:31 fakenerd joined #git
06:32 mingrammer joined #git
06:34 Haliucinas joined #git
06:35 Sceorem joined #git
06:37 danslo joined #git
06:39 SwiftMatt joined #git
06:39 satya4ever joined #git
06:39 eltucan joined #git
06:39 PioneerAxon joined #git
06:40 harish__ joined #git
06:41 dreiss joined #git
06:41 eltucan trying to run "arc version" and get "Exception The current working directory is not part of a working copy for a supported version control system (Git, Subversion or Mercurial). (Run with `--trace` for a full exception trace.)"
06:41 eltucan sorry if i seem noobish im new to this
06:42 jceb joined #git
06:42 danslo joined #git
06:42 t0by joined #git
06:42 eltucan ...reading intro
06:42 Macaveli joined #git
06:43 rominronin joined #git
06:43 submitnine joined #git
06:46 thiago eltucan: please explain the relationship to Git
06:47 eltucan i cloned a git repository and am using the arc command to do code revisions on phabricator
06:47 thiago arc is not git
06:47 anonymuse joined #git
06:47 thiago if you want to use that tool, please search for its proper forums and channels for help
06:47 thiago we can help you with git
06:48 eltucan i get that but i figured since its complaining about something to do with some vcs that its related
06:48 thiago I'm going to ask you a question you can't answer:
06:48 thiago what git command did you run?
06:48 eltucan thnx i'll look for #arc lol
06:48 maret joined #git
06:48 eltucan git clone "blah_path"
06:48 thiago and did you cd into the dir after that?
06:49 maret hi everyone I run git rebase and one merge commit is missing from list of commits how is that possible?
06:49 eltucan yes
06:49 thiago eltucan: and what command did you run that complained?
06:49 thiago eltucan: what git command, I mean
06:49 eltucan u got me thiago
06:49 maret I didn't remove it
06:50 thiago maret: rebase linearises history, so merge commits disappear
06:50 maret ok automatically hmm
06:50 maret is that safe for sure?
06:50 eltucan left #git
06:50 maret what if that merge commit would have conflicts resolution?
06:50 thiago maret: for sure it'sn ot
06:50 thiago it's not safe
06:50 thiago do not rebase if you have merge commits
06:52 Kuukunen maret: there IS --preserve-merges tho
06:53 Kuukunen maret: but overall it's a very bad idea to do actual work in merge commits, other than doing the absolute minimum work for merging
06:53 thiago that's when you're merging something from outside the rebased commits
06:54 Kuukunen thiago: preserve-merges?
06:54 thiago yes
06:54 thiago if the commits on both sides of the merge have been rebased, the merge disappear
06:54 thiago disappears
06:54 star_prone joined #git
06:55 maret so its not recommend to doing that? because I want to use it to remove one commit and rename few commits
06:55 pijiu joined #git
06:55 thiago it's not recommended to rebase merge commits
06:56 maret ok so what about to run git rebase -i --preserver-merges?
06:56 thiago still not recommended
06:56 ysh joined #git
06:57 maret why exactly?
06:57 DieguezZ joined #git
06:57 thiago because rebasing merges is a bad idea in any case
06:57 kadoban maret: It's confusing at best -i --preserve-merges, the manpage explains a bit of why
06:57 thiago --preserve-merges doesn't always work and even when it does, it may screw things up
06:57 thiago also, I don't think you can do interactive and preserve-merges at the same time
06:58 maret what should I do then? I have one commit which I need to throw out and few which I need to rename
06:58 thiago give up and move on
06:59 thiago or do a git rebase (no -i, no --preserve-merges) to linearise history first, then git rebase -i
06:59 thiago that is, drop all merges
07:00 maret well I run git rebase but that just said that current branch is up to date
07:00 inflames joined #git
07:01 thiago -f
07:02 duderonomy joined #git
07:02 PioneerAxon joined #git
07:02 maret ok did merge commit still there
07:02 hahuang65 joined #git
07:02 stux|RC-only joined #git
07:03 Kuukunen thiago: I dunno, quick test would indicate that --preserve-merges does work with --interactive and it does preserve merges in the history
07:03 maret hmm I think this is too risky
07:03 Kuukunen ...not that it's a good idea of course
07:04 ljc joined #git
07:04 leeN joined #git
07:05 danslo joined #git
07:05 Kuukunen maret: the correct way of handling it is admitting you made a mistake and the mistake will be in the history :P
07:06 maret hmm what if I would create another branch starting before the wrong commit then cherry pick rest the proper commits
07:06 gynter joined #git
07:06 thiago Kuukunen: like I said, it depends on whether only one of the sides of the merge is rebased or all of them
07:06 gynter Hello, is it possible to change commiter name and e-mail for existing commit?
07:06 maret yes
07:06 thiago maret: you can do that
07:06 gynter I somehow managed to do that with rebase few days ago, but can't recall it anymore
07:06 thiago gynter: that's a history !rewrite
07:06 gitinfo gynter: Rewriting public history is not recommended. Everyone who has pulled the old history will have to do work (and you'll have to tell them to), so it's infinitely better to just move on. If you must, you can use `git push --force-with-lease <remote> <branch>` to force (and the remote may reject that, anyway). See http://goo.gl/waqum
07:07 Kuukunen maret: or you could revert the commit?
07:07 gynter thiago: I'm aware of that, not a public repo
07:08 Kuukunen speaking of upstream rebases, was that feature added to git that would detect and warn about it?
07:08 JeroenT joined #git
07:08 maret Kuukunen nah its github fork before pull request so it make sense to fix it now
07:08 Kuukunen I saw some mentions of something like that potentionally incoming
07:08 thiago gynter: then see man git-filter-branch
07:08 gitinfo gynter: the git-filter-branch manpage is available at http://jk.gs/git-filter-branch.html
07:08 Muzer joined #git
07:08 Kuukunen maret: well sure, you can just cherry-pick all the commits you want, essentially doing interactive rebase by hand
07:09 maret you know what I will just try git rebase solution , there are unit tests so any problems should be covered
07:09 maret if not that that we need to fix unit tests anyway
07:10 Kuukunen well as long as you do all your testing in a separate branch without touching any of the others, you should be just fine
07:10 Kuukunen and it shouldn't be dangerous to test things out
07:10 cbarbieri joined #git
07:10 maret just to be sure, this is github repo which was forked and commits I want to rebase are those pushed to my github repo if I rebase to last commit of original - forked repo, its commits should NOT be affected by rebase right?
07:10 rbr joined #git
07:11 maret then I force push to my repo(there is one other programmer how is aware of this, who has pushed all changes)
07:11 Darcidride joined #git
07:11 Kuukunen as long as you don't rebase or cherry pick any of the original commits, they're not affected
07:15 ronskisika joined #git
07:15 JeroenT joined #git
07:16 TomyWork joined #git
07:17 Darcidride joined #git
07:18 ljc joined #git
07:18 cyan__ joined #git
07:19 IanV0rn2341 joined #git
07:20 jceb joined #git
07:20 jguillen joined #git
07:20 Raging_Hog joined #git
07:21 a_thakur_ joined #git
07:25 Balliad joined #git
07:26 Hounddog joined #git
07:27 garethdaine joined #git
07:28 a_thakur joined #git
07:32 danslo_ joined #git
07:34 star_prone joined #git
07:34 dodobrain joined #git
07:34 ahmedelgabri joined #git
07:35 danslo joined #git
07:38 quetz joined #git
07:39 star_prone joined #git
07:43 encod3 joined #git
07:45 ytl joined #git
07:45 v60r__ joined #git
07:49 fees joined #git
07:51 t0by joined #git
07:51 jceb joined #git
07:51 Neon joined #git
07:52 cqi joined #git
07:52 alansaul joined #git
07:55 Neon In my .gitignore I exclude js files like so: /**/js/**/*.js . Now when I try to reinclude a specific file with !/templates/foo/js/bar.js it just doesn't show up in the status and I can't add it either, not even with git add -f. Ideas for a nice and clean solution (I know I could temporarily remove the ignore entry, add the file, then reignore the folders, but that's dirty)?
07:56 jast FWIW for one-off cases you can bypass the ignore rules: git add -f
07:56 Neon The thing is I have a deep directory structure with js folders that may contain deep structures with js files that should be ignored.
07:56 jast of course it would be better to figure out why the unignore rule doesn't work
07:56 Neon jast: Well, git add -f didn't work.
07:57 jast then something else is going on
07:57 Neon :(
07:57 jast is any of the subdirs a git repository of its own, perhaps?
07:57 ASOLAr joined #git
07:58 Neon Ah, I found it out.
07:58 garyserj joined #git
07:58 Neon There was a .gitignore in the templates/foo dir. Although it didn't contain and js-related rules, it probably blocked the reinclusion of my outer .gitignore.
07:58 Neon *any js-related files
07:59 Neon *rules
07:59 Neon :S good morning brain
07:59 Seveas Neon: that's not very likely
07:59 cqi joined #git
08:00 Seveas a .gitignore in a subdir does not 'block' a higher level .gitignore
08:00 Seveas though individual rules may clash
08:00 oleg joined #git
08:01 jast yeah, that's the thing, though. does the unignore rule in a top-level .gitignore supersede the ignore rules in another .gitignore file? I've never had to think about that case
08:01 Seveas you can check which rule is responsible with git check-ignore -v path/to/file
08:01 jast either way, 'git add -f' should be bypassing all of them in any normal situation
08:01 Neon Yeah, I'm confused now. See, I'm using Atom, if you know it. It's a code editor. It displays a file tree and colors files in different colors according to their state in git. As I deleted the inner gitignore the bar.js lost its ignore color and gained a neutral color that tells me it's added and committed.
08:02 jast files that are already tracked aren't actually affected by ignore rules, no matter what atom tells you
08:02 Neon Yes, it appears the file has already been added in the past.
08:02 Neon It was just displayed incorrectly.
08:02 jast right
08:02 jast well, that's another mystery solved
08:03 ochorocho__ joined #git
08:03 Neon An Atom mystery actually. Maybe should be reported.
08:03 jast yeah, maybe
08:03 jast not that I care. atom didn't really excite me when I tried it... :)
08:04 Neon I didn't mean you should report it. :P
08:04 Neon What are you using?
08:04 jast that's what I thought. I just wanted to do some gratuitous bashing. :P
08:04 jast currently, vim
08:05 afuentes joined #git
08:05 jast but at the end of the day, all software still sucks
08:05 marcogmonteiro joined #git
08:05 Neon vim is my choice on x-less systems.
08:05 gynter left #git
08:06 Neon Where vim sucks in my opinion is the setup until it suits you perfectly.
08:06 jast well same for atom ;) it didn't really do anything interesting out of the box
08:07 someniatko joined #git
08:07 jast though I'm generally willing to take the time for setup
08:07 Neon But atom at least uses languages I know. :P I never got into the vim config language really.
08:07 jast I spent a week or two getting into emacs recently, too
08:07 Neon Is that worth it?
08:08 jast I'm not really sure
08:08 danslo joined #git
08:09 jast I used a starter kit with vim bindings (spacemacs) and fairly quickly got to a point where it became quite usable
08:09 forrestv is there some way to tell git of a custom tool to use to merge some types of files?
08:09 jast but I think the real benefit only comes when you get into customizing it yourself, emacs lisp and all
08:10 forrestv e.g. i have a python script that can semi-intelligently 3-way merge .svg files - how can i get git to use it?
08:10 Neon Yeah. That's probably the case for all the bigger editors.
08:10 jast forrestv: have a look at man git-mergetool, which you can configure for custom tools, too
08:10 gitinfo forrestv: the git-mergetool manpage is available at http://jk.gs/git-mergetool.html
08:11 Neon The perfect editor has yet to be written.
08:11 jast true
08:11 jast the perfect *, in fact
08:12 forrestv jast, thanks, i had found that and was looking into it; however, it seems that it is always manually called (git won't call it to attempt to merge files that it can't merge)
08:12 Neon I take it as my mission. Program the perfect *.
08:12 Kartagis joined #git
08:12 Kartagis hello
08:12 forrestv instead, git just provides a handy shortcut to start the merge tool
08:12 jast forrestv: in that case you may want to look into merge drivers, see man gitattributes
08:12 gitinfo forrestv: the gitattributes manpage is available at http://jk.gs/gitattributes.html
08:12 Kartagis I'm on a Mac and I have homebrew
08:12 Kartagis how do I add git completion?
08:13 jast I don't know homebrew, I assume it's using bash for its shell?
08:13 jast in the git source distribution there's completion code for various shells in contrib/completion
08:13 jast with bash I believe all you have to do is source that file in your .bashrc file
08:14 hellyeah-senpai joined #git
08:14 forrestv jast, ah, that's exactly what i need. thanks! strangely, that was seemingly impossible to find via google...
08:14 jast forrestv: well that's what we're here for ;)
08:15 danslo joined #git
08:16 anonymuse joined #git
08:19 jceb joined #git
08:19 jangeador joined #git
08:21 Oebele joined #git
08:21 kozie Kartagis: Perhaps this can help you - https://gist.github.com/trey/2722934
08:21 kozie ;)
08:22 kozie if you are using bash, perhaps you might want to install bash 4 as well using brew
08:22 kozie or perhaps even zsh or fish. I remember not having to do anything for the completion to work with zsh
08:23 sharkz_ joined #git
08:23 courrier joined #git
08:24 Rish joined #git
08:24 chll_ joined #git
08:24 selckin can i find a stash i dropped? doesn't seem to be in reflog
08:25 Kartagis kozie: this is not autocomplete
08:25 alexsystemf__ joined #git
08:26 kozie well, i see `source /usr/local/git/contrib/completion/git-completion.bash` being mentioned
08:26 kozie or i just don't know what you are aiming for
08:26 Tee_Pee joined #git
08:27 Kartagis oh, that means I shouldn't really look at the first several lines
08:28 nights joined #git
08:28 Spec-Chum joined #git
08:28 kozie well, i guess you should.. it's like a step-by-step guide of thing
08:28 kozie i think you also need the bash-completion thing if you're on bash
08:29 xaa joined #git
08:29 twisted` joined #git
08:29 moop joined #git
08:30 Kartagis kozie: which I have
08:31 Kartagis ow, I have that file
08:31 Kartagis cool
08:31 Kartagis let me add that to .bash_profile
08:31 Kartagis LET ME DAMNIT
08:32 iota_pi joined #git
08:32 ggherdov joined #git
08:32 ELLIOTTCABLE joined #git
08:33 luke_penn joined #git
08:33 mintux joined #git
08:34 kozie haha ^_^
08:34 ribasushi joined #git
08:34 jast selckin: that's because stashes are implemented as a reflog. now the only way to find it is to look for dangling commits.
08:35 jknetl joined #git
08:35 dsantiag_ joined #git
08:36 King_Hual joined #git
08:36 King_Hual joined #git
08:36 ysh joined #git
08:37 Fissure joined #git
08:39 selckin jast: right, thanks, found my stash! :)
08:39 dsantiago joined #git
08:39 LeMike joined #git
08:40 achlys joined #git
08:40 ggherdov joined #git
08:41 ljc joined #git
08:42 zzaiek joined #git
08:42 cstrahan joined #git
08:42 JackMc joined #git
08:43 bedouin joined #git
08:44 felowe joined #git
08:45 noahsussman joined #git
08:45 ulkesh joined #git
08:45 etcinit joined #git
08:45 yaewa joined #git
08:45 ytl joined #git
08:45 mehdi__ joined #git
08:46 angular_mike joined #git
08:47 quackgyver joined #git
08:47 Tee_Pee joined #git
08:47 teadrop_ joined #git
08:49 dsantiago joined #git
08:50 bedouin joined #git
08:50 Mekkis joined #git
08:51 gajus I have create a new repository. Committed bunch of stuff to master branch. However, I need to somehow add an 'empty' commit in at the beginning that would allow me to create master branch in github and then raise PR with the rest of the changes
08:51 gajus any suggestions how to do it?
08:52 someniatko left #git
08:52 gajus I was thinking of create a commit that creates an empty README.md, then use rebase to move this commit to the beginning, then push that code to the repository, then push the remaining code under a different branch
08:52 osse gajus: the premise sounds wrong. what does an "empty" commit have to do with PRs ?
08:52 masak you already have a branch and you would like to re-do it from scratch with an empty commit at the start?
08:52 star_prone joined #git
08:53 gajus masak well, thats the only way I see how to enable me to create a PR will all of the other changes
08:53 tobie joined #git
08:53 gajus these branches need to share a common ancestor
08:53 mintux i connected network drive via ssh and clone a repo there but when i try to push it returns this error: http://codepad.org/0VMmijJC
08:55 gajus or ... I could rename the current branch to something else than master, push that to the repository
08:55 mintux also i try to pull it from local to server in server just cursor blank and nothing happened
08:56 gajus but then how would I create an empty master branch that would allow me to merge the former branch?
08:56 flopsec joined #git
08:56 refried_ joined #git
08:56 jast gajus: why do you want to use PR in a fresh repo in the first place?
08:56 masak gajus: if you don't want to change history after the fact, you could simply make a commit that removes all the files. (save them somewhere else when you do that.)
08:56 gajus jast to enable people to review the code (git flow)
08:57 masak though "somewhere else" could very well be "it's already in the Git history" :)
08:57 jast in an empty repo???
08:57 gajus jast whats surprising about requiring to review code in a new project?
08:57 mintux any idea?
08:58 jast well, you could just, you know, not use a PR for that
08:58 jast since there is no actual merging that needs to happen
08:58 tobie I've carried out a large change on a file that's similar to an HTML to Markdown conversion (i.e. content is pretty much the same, presentation changed dramatically). Is there anything I can do to preserve history across such changes so git blame still returns meaningful info?
08:58 gajus jast and how do you perform code review with that?
08:58 Silenced_v2 joined #git
08:59 ronskisika joined #git
08:59 jast gajus: I like to call it "looking at the commits"
08:59 Silenced_v2 joined #git
08:59 gajus ..
08:59 jast I really don't understand what's so complicated about that
09:00 cyan__ joined #git
09:00 gajus anyway, how do I rebase to the first commit-1?
09:00 gajus suppose that the first commit sha is 9d2d4fc12a6381fa6ce69387c0ddd084c1937357
09:01 solf1re joined #git
09:01 jast tobie: as always, people can just look at the blame output in different revisions and draw conclusions... but there's no additional deep magic that I know of
09:01 solf1re joined #git
09:01 edward joined #git
09:01 kadoban gajus: You're likely looking for --root option to rebase
09:02 achlys joined #git
09:02 gajus Thank you
09:02 anonymuse joined #git
09:03 hahuang65 joined #git
09:03 Rish joined #git
09:04 tobie jast: right so there's no manual mapping that could be added, for example.
09:05 mintux i removed git and use rsync instead
09:06 danslo joined #git
09:07 sonOfRa joined #git
09:11 luke_penn joined #git
09:11 daumie254 joined #git
09:11 ahmedelgabri joined #git
09:12 lss8 joined #git
09:13 lb1c s/git/rsync/g
09:14 WizJin joined #git
09:15 Andrew_K joined #git
09:16 ertes joined #git
09:17 danslo joined #git
09:18 ertes hi…  can i mark certain commits as "never push"?  i'd like to push everything up to, but not including, one of those
09:18 masak ertes: just to be clear, that'd mean you can never push later commits either
09:19 masak ertes: if you push a commit, all its ancestor commits are also pushed
09:19 ertes masak: that's fine, it fits my workflow (commit a lot, interactive rebase, push)
09:19 tobiasvl ertes: !xy
09:19 gitinfo ertes: Woah, slow down for a bit. Are you sure that you need to jump through that particular hoop to achieve your goal? We suspect you don't, so why don't you back up a bit and tell us about the overall objective...
09:19 achlys_ joined #git
09:20 * masak was going to mention the pre-push hook, but perhaps `!xy` is a better answer... :)
09:20 bket joined #git
09:20 ertes again, it fits my workflow:  i make a lot of small commits with mostly for-myself messages, and at the end (just before pushing), i reorder, recombine and reword them
09:21 ertes but some of them i like to keep local, because they are mostly workarounds to make the project build for me
09:22 masak ertes: the way you express that makes me think you don't quite grok what it means that pushing a commit also by *necessity* pushes all its previous history
09:22 al-damiri joined #git
09:23 masak ertes: to the point that if you really wanted to exclude some commits, you'd have to create a new branch, rebase-interactive there to exclude some commits, and then push that
09:25 xaa joined #git
09:25 ertes pre-push hook sounds interesting, if i can rewrite *which* commit i want to push with it…  thanks for the pointer
09:26 masak *sigh*
09:26 masak shoulda gone with !xy...
09:26 gitinfo Woah, slow down for a bit. Are you sure that you need to jump through that particular hoop to achieve your goal? We suspect you don't, so why don't you back up a bit and tell us about the overall objective...
09:26 thomas_25 joined #git
09:26 bauruine joined #git
09:26 ertes (as for the "you don't think right" stuff: sorry, but i really understand git much better than you assume)
09:26 thomas_25 what's the best way to clone a repo as a template/starting point (without the .git directory) ?
09:26 ertes (i should just learn to communicate it properly)
09:26 masak best of luck, then
09:26 jceb joined #git
09:26 tobiasvl thomas_25: uuuh, clone it and rm -r .git ?
09:27 tobiasvl thomas_25: what do you mean exactly?
09:27 thomas_25 i meant that actually -- but curious is there specifically a workflow for that
09:27 thomas_25 maybe a git command just for that
09:27 thomas_25 apparently not
09:27 ertes masak: to ease your mind, my workflow *never* requires push --force ;)
09:27 osse thomas_25: you could download an archive
09:28 thomas_25 yeah, that's one of the options mentioned on the web too
09:28 masak ertes: oh, I'm not so worried about that. `push --force` has its uses.
09:28 tobiasvl well, if you use github or something, then you can download the archive that provides, yes
09:28 masak ertes: though if you're in a team pushing to the same branch, check out `push --force-with-lease`
09:28 ochorocho__ joined #git
09:29 ertes masak: that's what i'd do, but the rebase stuff is entirely local…  nobody ever sees it
09:29 bauruine_ joined #git
09:30 ljc joined #git
09:31 Munt joined #git
09:32 ljc joined #git
09:33 anonymuse joined #git
09:34 mintux left #git
09:34 eduardas_m joined #git
09:36 bourbaki joined #git
09:36 jaguarmagenta joined #git
09:37 bourbaki Is there any way to determine if since some commit a merge with some other branch has happened?
09:37 bourbaki I would like to write a script that if a merge with another branch has happened leads to a rebuild on the buildserver instead of an incremental build
09:37 mkoskar joined #git
09:39 Corneliu_ joined #git
09:42 tobiasvl perhaps: git rev-list --min-parents=2
09:42 tobiasvl to see if the sequence of commits received contains a commit with 2 parents
09:44 archaic joined #git
09:44 JeroenT joined #git
09:46 SwiftMatt joined #git
09:47 pijiu2 joined #git
09:53 bourbaki lets see
09:53 quite t
09:55 bourbaki tobiasvl: as a last parameter i need to provide the commit i last build on the server then ?
09:56 mar77i joined #git
09:58 bourbaki Er... there are a lot of commits that are on develop and origin/develop ... what does that mean?
09:59 justrohu joined #git
10:00 mingrammer joined #git
10:00 justrohu Hello on my local machine I have done change to my files and pushed it to the master branch but when I am doing git pull on server I am getting partial changes
10:00 justrohu any suggestions why
10:01 selckin !nonbare
10:01 gitinfo [!bare] A bare repository is used to push/fetch(useful for running a git server), and contains only the contents of .git/ from a "normal" repo. Read more: http://bare-vs-nonbare.gitrecipes.de/
10:01 cdown joined #git
10:02 i7c justrohu, are you sure you added and committed all the changes before pushing?
10:02 justrohu selckin, sorry I am not getting you properly
10:03 selckin yaewa: yeah its no the right trigger probably !deploy
10:03 gitinfo yaewa: Git is not a deployment tool, but you can build one around it (in simple environments) or use it as an object store(for complex ones). Here are some options/ideas to get you started: http://gitolite.com/deploy.html
10:03 i7c Run git status in the repository where you made the changes.
10:03 justrohu i7c, yes
10:03 jceb_ joined #git
10:03 selckin in short, pushing will not update the working directory
10:03 i7c justrohu, compare the commit ids then (with git log). Are you on the same commit in both repositories?
10:04 i7c selckin, if he's pushing into a non-bare I don't get the part about "git pull".
10:04 selckin his way to try to solve the problem
10:04 justrohu i7c, I dont have different repositories I am working on single repository .. I am working locally doing a push to bitbucket and then I do a pull origin production server
10:05 selckin i7c: but you cuold be right too, not a clear question
10:06 jimi_ joined #git
10:06 i7c justrohu, on both machines, do `g rev-parse HEAD' and compare hashes.
10:07 tobiasvl bourbaki: a range, git rev-list --min-parents=2 OLD..NEW
10:07 justrohu i7c, ok
10:07 tobiasvl bourbaki: you need to check just the payload you receive
10:07 i7c justrohu, g=git, sorry.
10:07 i7c Old habit because I have an alias :)
10:08 mar77i joined #git
10:09 plutonas joined #git
10:09 justrohu i7c, I have checked the hash on production and local ... both of them are same
10:10 plutonas i have git configured for a repository on my laptop. How can I figure out what username I am using so that I can also clone the repository on my desktop?
10:10 ysh joined #git
10:10 plutonas this was configured a long time ago and I don't remember the details
10:10 i7c Then if git status says your working tree is clean, all tracked files are identical, justrohu. Not sure what you mean by partial changes.
10:11 justrohu ok .. i7c .. could you please help me understand the difference between git merge and git rebase
10:11 jast plutonas: git remote -v will show the URLs used, that may already be all the information you need. if that's not sufficient, the next step depends on what protocol you're using.
10:12 i7c !merge-vs-rebase
10:12 i7c mhmm
10:12 jast !merge_vs_rebase
10:12 gitinfo [!merge_or_rebase] For a detailed discussion of the dis-/advantages of merge and rebase, and when to best use which, see: https://coderwall.com/p/ny1hia -- and here's how Linus explains which one to use: http://www.mail-archive.com/dri-devel@lists.sourceforge.net/msg39091.html
10:12 i7c Thank you, jast.
10:13 plutonas ok thanks guys it worked
10:14 plutonas one more question, how can I store username and password on my desktop now, so that I don't have to type them in every time i do something on the repo?
10:14 magicturtle joined #git
10:15 jast plutonas: depends, are you using git via HTTP or SSH?
10:18 linduxed joined #git
10:19 plutonas jast: i https is how the repository address starts
10:19 plutonas i guess this means http?
10:19 _ikke_ es
10:19 _ikke_ yes
10:21 jast plutonas: yeah. well, there are two ways to do it. both will store your credentials in an unencrypted file in your home dir, just so you know
10:22 jast the easy way is to do 'git config credential.helper store' (you can use 'cache' instead of 'store' to save it in memory for a limited time, more secure but you have to re-type your password periodically)
10:22 jast gotta go, back later
10:22 linduxed joined #git
10:23 ertes the pre-push hook doesn't seem to help =/
10:23 yonatankoren joined #git
10:23 ertes i'll use branches and fast-forwarding for now, but it's really inconvenient
10:24 pro_metedor joined #git
10:24 ertes actually it doesn't even help, so nevermind
10:25 tobiasvl why doesn't it help?
10:26 ertes tobiasvl: unless i'm overlooking something it doesn't allow me to override the commit to push
10:26 tobiasvl override how?
10:27 _ikke_ You can only prevent pushing things, not change what is pushed
10:27 ertes remote is at A, i'm at A -> B -> C…  what i'd like to do is:  if C's message starts with, say, "DEV", i'd like to push the parent B instead
10:28 kurkale6ka joined #git
10:28 j416 ertes: I only read your last message, but it sounds like a flawed workflow to me
10:28 tobiasvl yeah, I gave the whole "XY" spiel earlier
10:28 j416 ertes: push whatever to throwaway branches, merge what is stable
10:28 j416 tobiasvl: ah :)
10:29 ertes can we please skip this?
10:29 j416 ok :)
10:29 ertes thanks =)
10:29 * j416 goes to bake cookies
10:29 i7c Skipped.
10:29 tobiasvl !next
10:29 gitinfo Another satisfied customer. NEXT!
10:29 i7c Now let’s talk about ponies.
10:32 nettoweb joined #git
10:32 aapaaal joined #git
10:33 vrishab joined #git
10:42 anonymuse joined #git
10:42 vassagus joined #git
10:42 aapaaal joined #git
10:43 iob- joined #git
10:45 magikid joined #git
10:48 flavi0 joined #git
10:49 ash_workz joined #git
10:49 Kartagis left #git
10:51 lb1c i7c what's your favorite? mine is Twilight Sparkle
10:51 lb1c isn't it beautiful :) https://en.wikipedia.org/wiki/My_Little_Pony:_Friendship_Is_Magic
10:52 flavi0 so https://help.github.com/articles/remove-sensitive-data/ describes that i can use git gc --prune to get rid of unreferenced commits in the local repository. man i also get rid of refs in the github repo if things have already been pushed there? or would I need to recreate the repo server side and push the cleaned content again?
10:52 flavi0 s/man/can/
10:52 selckin you can't control what github does
10:52 selckin so thats probably the safe thing todo
10:52 i7c lb1c: http://vignette3.wikia.nocookie.net/mlp-gameloft/images/c/cf/Snips_vector.png/revision/latest?cb=20140103095847
10:53 lb1c flavi0 if man git push --delete is sufficient for you
10:53 gitinfo the git-push manpage is available at http://jk.gs/git-push.html
10:53 lb1c because deleting your repo serverside would in the case of github also delete wiki/issues etc
10:54 lb1c i7c just what i had in mind
10:54 lb1c i7c looks like a 4chan pony :D
10:54 flavi0 oh, so to get rid of a single commit (it's really only about repo size, no sensitive data there) i could just push --delete <hash> ?
10:54 i7c :)
10:56 star_prone joined #git
10:56 lb1c flavi0 no you can only push refs
10:57 aapaaal joined #git
10:57 JeroenT joined #git
10:57 flavi0 lb1c: so to get rid of the unreferenced commit in order to decrease repo size i'd have to delete the repo and push the purged one anew?
10:57 rgrinberg joined #git
10:58 lb1c flavi0 no, a git push --delete origin <branch/ref>  would delete that ref from the remote repo
10:58 lb1c flavi0 no need for deleting the whole repo
10:58 flavi0 lb1c: ok wait. a commit is also a ref or not?
10:58 lb1c flavi0 and after that a normal git push origin <ref>
10:59 lb1c no
10:59 i7c He’s more worried about the orphaned commit, I think.
10:59 flavi0 yes.
10:59 lb1c ahh
10:59 danslo joined #git
10:59 flavi0 and the space it occupies in the repo. lets say someone accidentally pushed a 1G video file. deleting the refs to that commit won't help.
11:00 flavi0 (in this case it's a 1Mb pdf but it still annoys me)
11:00 danslo joined #git
11:00 i7c Git gc should take care of it eventually, remotely. Unless the remote git is configured not to do gc, but then it’s not your problem really.
11:00 lb1c flavi0 you just have to wait for the normal garbage collection
11:00 govg joined #git
11:00 i7c And when fetching/cloning the orphaned commit shouldn’t be downloaded, I think.
11:01 flavi0 ahh so garbage collection will take place automatically in a github repo at some point?
11:01 lb1c flavi0 from their help page i seems you have to request it manually: You can't do anything about existing clones or forks of your repository, but you can permanently remove all of your repository's cached views and pull requests on GitHub by contacting GitHub Support.
11:01 flavi0 that's actually sufficient i was only worried about increased cloning/fetching size.
11:01 lb1c source: https://help.github.com/articles/remove-sensitive-data/
11:02 lb1c flavi0 cloning/fetching would not be an issue
11:02 ertes flavi0: if you worry about a 1 MiB PDF, you will go crazy, if you consider that it may have already made it into a backup =)
11:02 lb1c flavi0 no unreferenced object would be transfered
11:03 flavi0 ok i'm fine with that then. i think the github servers can live with that additional Mb of storage without me needing to deal with it further.
11:03 ertes flavi0: in other words: if it didn't contain sensitive information, don't bother…  if it did, learn from the experience =)
11:03 flavi0 ertes: yup, alrighty
11:03 flavi0 thanks all for the help!
11:04 hahuang65 joined #git
11:04 lb1c flavi0 http://stackoverflow.com/questions/15261880/does-github-garbage-collect-dangling-commits-referenced-in-pull-requests it seems github does extra refs so the commit wont be gc because it's referenced somewhere
11:05 lb1c flavi0 if it's urgent(sensitive) contact support, if it's just about tranfser bandwith, don't worry
11:05 tlaxkit joined #git
11:05 flavi0 lb1c: only about bandwidth, so all fine. thanks
11:06 mizu_no_oto joined #git
11:08 inflames joined #git
11:13 leftriver26 joined #git
11:13 fahadash joined #git
11:14 rahtgaz joined #git
11:17 HoierM joined #git
11:18 plutonas thanks jast for the help, it all works as i wanted it
11:19 ojdo joined #git
11:21 rnsanchez joined #git
11:23 i7c joined #git
11:24 star_prone joined #git
11:25 Munt joined #git
11:26 i7c joined #git
11:26 mingrammer joined #git
11:27 sdothum joined #git
11:28 Darcidride_ joined #git
11:29 johnmilton joined #git
11:29 ochorocho__ joined #git
11:31 wxg joined #git
11:31 raijin joined #git
11:31 mehola joined #git
11:34 str joined #git
11:37 DieguezZ joined #git
11:37 jaguarmagenta joined #git
11:39 jceb joined #git
11:42 kyan joined #git
11:43 vassagus joined #git
11:43 plutonas left #git
11:44 ysh joined #git
11:47 tlaxkit joined #git
11:50 danslo joined #git
11:51 ahr3n joined #git
11:52 ysh joined #git
11:52 linduxed joined #git
11:53 ahr3n_ joined #git
11:56 tomog999_ joined #git
12:00 alexsystemf___ joined #git
12:00 moei joined #git
12:00 star_prone joined #git
12:00 JeroenT joined #git
12:01 iota_pi joined #git
12:01 dpower joined #git
12:02 ahmed_elgabri joined #git
12:02 lmatteis joined #git
12:02 Gurkenglas joined #git
12:03 Sceorem joined #git
12:03 justrohu i7c, Thank you for your help
12:03 i7c You’re welcome. :)
12:04 achlys joined #git
12:05 moop joined #git
12:05 synthroid joined #git
12:08 mohabaks joined #git
12:09 mingrammer joined #git
12:13 Darcidride__ joined #git
12:16 ysh joined #git
12:19 alexsystemf___ joined #git
12:20 ayjay joined #git
12:21 ysh joined #git
12:24 anonymuse joined #git
12:25 dede joined #git
12:27 Jellyg00se joined #git
12:28 daniel_rose joined #git
12:30 ysh joined #git
12:32 eijk_ joined #git
12:33 mizu_no_oto joined #git
12:33 mehola joined #git
12:34 JanC joined #git
12:36 ysh joined #git
12:37 vbgunz On Fedora 24, is there a reason why git <tab> will autocomplete most subcommands but not ls-tree or ls-files? I was looking for these and these names are obvious but don't autocomplete or show up as suggestions. Anyone know why?
12:38 moritz vbgunz: probably because they are seen as plumbing, not porcelain
12:38 crose joined #git
12:38 vbgunz moritz: damn, I feel like an idiot, thanks!
12:38 moritz vbgunz: but the same thing has annoyed me too
12:38 moritz mostly with ls-files
12:42 Noldorin joined #git
12:44 vassagus joined #git
12:48 andlabs joined #git
12:48 newbiee joined #git
12:50 duga joined #git
12:50 beyertech joined #git
12:53 achlys_ joined #git
12:53 ffualo joined #git
12:54 Gamecubic joined #git
12:54 shinnya joined #git
12:55 i7c What do they patch tho? The completion scripts that git ships?
12:55 rchavik joined #git
12:55 synthroid joined #git
12:55 star_prone joined #git
12:56 moritz might it ship old completion scripts?
12:56 JamesBaxter joined #git
12:56 moritz I have the same behavior on Debian
12:57 i7c Right, same on Arch, I just checked. Unless you use zsh.
12:58 duga Hi I have a problem with converting subversion repo to the git one - using this script https://bitbucket.org/snippets/duga/koyy8 unfortunately one of projects differs from the original (f.e. master branch differs from svn's trunk) How can I handle with this?
12:58 dsdeiz joined #git
12:59 moritz duga: if the SVN repo uses $Id$ markers, differences are to be expected
12:59 moritz duga: if not, it would be interesting what exactly the differences are
12:59 moritz newline stuff?
13:00 duga moritz: how can I check this (first one)
13:00 duga moritz: one - quite large and old
13:00 moritz duga: like, look into the files
13:00 timvishe` left #git
13:00 moritz duga: make a diff -u file-in-svn file-in-git
13:00 duga moritz: differs f.e. test directory (maven scheme) and impl dir in the sources
13:01 duga in git is not existing and in svn there is
13:01 timvisher joined #git
13:01 moritz duga: are these empty directories?
13:01 duga yes
13:01 duga simple ommited
13:01 moritz duga: git doesn't track directories, it simply creates them when it needs to add a file
13:02 moritz duga: a typical workaround is to put a .gitignore or a .dummy file into such directories, and commit them
13:02 duga moritz: sorry
13:02 duga I didn't understand moritz
13:02 duga moritz: there aren't
13:02 moritz duga: git cannot deal with empty directories
13:02 duga moritz: yes, I didn't understand your question
13:03 duga moritz: those dirs are not empty
13:03 danslo_ joined #git
13:03 duga test dir is not empty -> test case files
13:03 duga impl dir also not empty
13:03 moritz well, you did answer "yes" to whether they were empty
13:03 moritz then I have no idea, sorry
13:03 duga moritz: ok
13:05 hahuang65 joined #git
13:06 sanketdg joined #git
13:06 cdg joined #git
13:06 duga moritz: sometimes people are tired
13:06 cdg joined #git
13:07 jason237 joined #git
13:07 duga if they have to f.ck with shitty old corpo repos
13:08 jhass left #git
13:08 nurey joined #git
13:08 jast I don't have to do that and I'm still tired
13:08 bongjovi joined #git
13:09 alansaul joined #git
13:10 AbstractPaint joined #git
13:11 Ediz joined #git
13:12 achlys joined #git
13:13 * moritz has small kids and is always tired
13:13 raijin joined #git
13:13 ExoUNX_ joined #git
13:16 alansaul_ joined #git
13:18 tvw joined #git
13:19 dsdeiz joined #git
13:19 Guest70874 left #git
13:21 flavi0 left #git
13:23 achlys joined #git
13:24 sangy joined #git
13:25 Guest81 joined #git
13:26 anonymuse joined #git
13:27 ysh joined #git
13:28 Junior joined #git
13:28 std45 joined #git
13:28 frobnic things don
13:29 frobnic 't get better with older kids...
13:29 jimi_ joined #git
13:29 ayjay joined #git
13:30 zivester joined #git
13:30 ochorocho__ joined #git
13:30 moritz they become different :-)
13:31 mycyber joined #git
13:33 l4v2 joined #git
13:34 anonymuse joined #git
13:35 str joined #git
13:36 GodGinrai joined #git
13:37 jhasse How can I "undo" a `git submodule update --init`?
13:38 rnsanchez_ joined #git
13:38 jaguarmagenta joined #git
13:38 Kulrak joined #git
13:38 Kulrak joined #git
13:39 stfn jhasse: git submodule deinit
13:40 SpeakerToMeat joined #git
13:40 anonymuse joined #git
13:41 daniel_rose joined #git
13:41 jhasse stfn, thx, I think it worked :)
13:42 Gsham joined #git
13:42 cdown joined #git
13:42 Dougie187 joined #git
13:43 achlys joined #git
13:44 vassagus joined #git
13:44 nurey_ joined #git
13:45 kpease joined #git
13:46 maret joined #git
13:48 dave0x6d joined #git
13:50 bouncy joined #git
13:51 griffindy joined #git
13:52 ysh joined #git
13:53 libertas joined #git
13:54 Ggci joined #git
13:54 libertas Hi, previously I was using iso-8859-1 encoding in my filesystem and in ~/.gitconfig
13:55 libertas now that I changed OS to a utf-8 encoding, do I need to convert my local repositories?
13:55 libertas Or do I just need to change the commitencoding in ~/.gitconfig to UTF-8?
13:57 sveinse joined #git
13:59 sveinse I am going to switch origin on an existing clone. I've renamed origin to old, setup the new remote. What do I do now? How do I switch local master branch to track the new origin/master ?
13:59 Remram[m] joined #git
13:59 Remram[m] joined #git
13:59 fstd_ joined #git
14:00 AaronMT joined #git
14:00 anonymuse joined #git
14:02 antranigv hi all! anyone can help me to do stashing? :))
14:02 antranigv I did the changes in a wrong branch, heh
14:02 realz joined #git
14:02 ysh joined #git
14:04 jast antranigv: first, try !float. if that doesn't work in your case, what specifically are you having trouble with?
14:04 gitinfo antranigv: If you have made a change in your working directory and have NOT YET COMMITTED, you may "float" that change over to another (`git checkout oldbranch`) or new (`git checkout -b newbranch`) branch and commit it there.  If the files you changed differ between branches, the checkout will fail.  In that case, `git stash` then checkout, and `git stash apply` and go through normal conflict resolution.
14:07 foist joined #git
14:08 danslo joined #git
14:09 sybariten i know i should really ask this in svn but there seems to be a bit of a calm there. There really is no svn equivalence to gits way of dealing with file-ignorations is there?
14:10 alansaul joined #git
14:10 jast IIRC svn does it with its properties system, can't really remember the specifics
14:10 Oebele joined #git
14:11 tlaxkit joined #git
14:11 perlpilot joined #git
14:11 sveinse sybariten: Yup, the svn:ignore property
14:11 madewokherd joined #git
14:12 JamesBaxter joined #git
14:12 nurey joined #git
14:12 phaleth joined #git
14:13 tlaxkit left #git
14:14 sybariten sveinse: i'm obvioulsy doing something wrong...       svn propset -R svn:ignore "*.tga" .   ;   svn add *NTOC*    --->  lots of tgas seem to be added
14:17 clemf joined #git
14:17 Guest81 joined #git
14:18 BSaboia joined #git
14:20 oxsyn joined #git
14:21 hobodave joined #git
14:24 bsanford joined #git
14:24 neurodrone joined #git
14:26 star_prone joined #git
14:26 synthroid joined #git
14:26 perlpilot joined #git
14:29 ayjay joined #git
14:32 morenoh149 joined #git
14:33 std45 joined #git
14:33 dmto joined #git
14:34 Limes left #git
14:34 ResidentBiscuit joined #git
14:34 Limes joined #git
14:34 diogenese joined #git
14:35 daniel_rose joined #git
14:38 refried_ joined #git
14:38 thiago joined #git
14:40 alansaul_ joined #git
14:40 noteness joined #git
14:42 vassagus joined #git
14:46 DJ_Spies joined #git
14:47 vassagus_ joined #git
14:48 Silenced joined #git
14:48 Silenced joined #git
14:49 satbyy joined #git
14:50 pullphinger joined #git
14:50 vrishab joined #git
14:51 eijk_ joined #git
14:53 pullphinger joined #git
14:54 duderonomy joined #git
14:56 antranigv jast: yes it worked by just checking out :) thanks! :)
14:59 zonality joined #git
14:59 danslo joined #git
15:00 dreiss joined #git
15:01 vassagus joined #git
15:02 std45__ joined #git
15:02 daniel_rose joined #git
15:04 fakenerd joined #git
15:05 hahuang65 joined #git
15:06 moneylotion joined #git
15:11 star_prone joined #git
15:11 zacts joined #git
15:12 SJr I created a merge commit from release branch X to release branch Y (the next version) , for some reason [ not sure why ], all the changes introduced into branch Y were marked as having been done  in that commit, and the commit wasn't marked as a merge. This caused a bunch of conflicts with other people who had other merges to do. We eventually decided to just do a reset before the commit and redo the operation. We created a CI job to watch and m
15:12 SJr ake sure no one reintroduced that commit. Well someone just on a very long feature branch didn't bother to follow the instructions, and now I have a giant tree of commits that are 'tainted'. What is the best way to merge those over.
15:12 SJr I looked at cherry picking but then for each merge commit I have to decide which is the parent, I'm not sure if there is an automated way of doing it.
15:13 SJr Alternatively maybe I should just make a patch based on the difference.
15:14 johnmilton joined #git
15:15 daniel_rose joined #git
15:15 watabou joined #git
15:15 preyalone joined #git
15:16 alansaul joined #git
15:16 Guest81 joined #git
15:19 vassagus joined #git
15:20 chrisshattuck joined #git
15:20 snowkidind joined #git
15:21 alansaul_ joined #git
15:22 zeroed joined #git
15:22 jaafar joined #git
15:22 zivester joined #git
15:23 Munt joined #git
15:23 govg joined #git
15:25 _W_ joined #git
15:25 mingrammer joined #git
15:26 mingrammer joined #git
15:27 moneylotion joined #git
15:28 Endarked joined #git
15:29 ochorocho__ joined #git
15:30 tyreld joined #git
15:32 bsanford joined #git
15:34 refried_ joined #git
15:35 vassagus joined #git
15:36 pijiu2 joined #git
15:37 InfoTest joined #git
15:38 traxo-xx joined #git
15:39 jaguarmagenta joined #git
15:41 cjohnson SJr: When I get in deep like that, that's my approach. make a diff and patch it in
15:41 cjohnson lose the history
15:41 hahuang65 joined #git
15:41 cjohnson keep the original branch with the tainted history around for a couple weeks/months until you're sure you got it all
15:42 hahuang61 joined #git
15:42 tvw joined #git
15:44 cyphase joined #git
15:47 ruby_ joined #git
15:48 a_thakur joined #git
15:50 sssilver joined #git
15:50 notebox joined #git
15:50 madewokherd joined #git
15:51 LeBlaaanc joined #git
15:51 sssilver joined #git
15:53 tjone270_ joined #git
15:56 vassagus joined #git
15:57 romerocesar joined #git
15:58 jlebar joined #git
15:58 e14 joined #git
15:59 traxo-xx joined #git
16:03 synthroid joined #git
16:03 traxo-xx joined #git
16:04 fakenerd_ joined #git
16:05 warthog9 joined #git
16:05 andlabs joined #git
16:07 skift joined #git
16:07 eijk joined #git
16:08 Gsham joined #git
16:08 fahadash joined #git
16:10 vrishab joined #git
16:11 irqq joined #git
16:11 dave0x6d joined #git
16:11 jeffreylevesque joined #git
16:11 vassagus joined #git
16:13 tabrez joined #git
16:14 nikivi joined #git
16:15 nitric joined #git
16:16 phanimahesh joined #git
16:20 EvilPenguin joined #git
16:22 govg joined #git
16:22 VagabondAzulien joined #git
16:23 danslo joined #git
16:25 Emperor_Earth joined #git
16:26 fakenerd joined #git
16:26 durham joined #git
16:29 durham joined #git
16:30 perlpilot joined #git
16:31 a_thakur_ joined #git
16:32 Puffball joined #git
16:34 SwiftMatt joined #git
16:34 e14 joined #git
16:36 raijin joined #git
16:42 ruby_ joined #git
16:42 fakenerd_ joined #git
16:42 ruby_ joined #git
16:43 hahuang61 joined #git
16:43 terminal_echo joined #git
16:45 MacWinner joined #git
16:45 netj joined #git
16:46 ojacobson sybariten: if you pass an ignored filename to `svn add` explicitly, then svn will assume you meant to add that file, iirc. Git has a similar mode (you need git add --force, but it'll do it)
16:46 ojacobson the other piece of the puzzle is that glob patterns like *NTOC* are expanded by your _shell_, not by subversion, and all subversion sees is the resulting list of filenames (as arguments)
16:47 vassagus joined #git
16:51 Mobutils joined #git
16:52 otl joined #git
16:52 pothepanda joined #git
16:52 saedelaere joined #git
16:53 Catachan_ joined #git
16:54 ansimation joined #git
16:56 ash_workz joined #git
16:57 miklcct joined #git
16:58 zacts joined #git
16:59 LionsMane joined #git
17:00 joki joined #git
17:01 timmu joined #git
17:03 MattMaker joined #git
17:04 kadoban joined #git
17:05 noteness left #git
17:06 moneylotion joined #git
17:08 miklcct joined #git
17:09 tjone270__ joined #git
17:09 raijin joined #git
17:09 netj joined #git
17:11 vassagus joined #git
17:15 Faylite joined #git
17:16 duderonomy joined #git
17:19 nejni-marji joined #git
17:19 nejni-marji How can I update a local git repository to the latest version on github?
17:20 nurey joined #git
17:21 tjone270 joined #git
17:21 osse nejni-marji: have you made changes to it since you cloned it?
17:21 nejni-marji osse: no
17:21 osse nejni-marji: git pull
17:22 thiago joined #git
17:24 BSAlb joined #git
17:25 Sasazuka joined #git
17:26 irco joined #git
17:27 spudowiar joined #git
17:27 marc_v92 joined #git
17:29 vassagus_ joined #git
17:30 stomplee joined #git
17:30 shurnormal http://gist.github.com/uprego
17:32 thiago joined #git
17:33 LUMIA930 joined #git
17:33 ahmed_elgabri joined #git
17:33 garethdaine joined #git
17:34 hellyeah joined #git
17:34 hellyeah i do git clone blah
17:34 Groscheri joined #git
17:34 hellyeah i dont need git init anymore right?
17:34 osse right
17:37 hellyeah there is a command after git clone
17:38 HoierM joined #git
17:38 hellyeah ah
17:38 hellyeah git config --list --local
17:42 tabrez joined #git
17:43 kab joined #git
17:44 anonymuse joined #git
17:45 roxtrongo joined #git
17:45 vassagus joined #git
17:47 Akien joined #git
17:47 Akien Hi there
17:47 gitinfo Akien: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.
17:48 Akien How can I delete (i.e. clean them off the history/no longer accessible in detached HEAD state) commits in git?
17:49 Akien I'm trying to fork a GPLv2 project that just went proprietary, and I'd like to strip off the proprietary commits
17:49 grawity if they're not on any branch nor tag, then it doesn't matter; dangling commits don't get pushed nor cloned
17:49 grawity that said, on your local repo, there's `git gc` and `git prune`
17:50 osse my super-naïve interpretation of the question has the answer  git reset --hard X  where X is the latest non-proprietary commit
17:51 govg joined #git
17:51 dreiss joined #git
17:51 dtrott_ joined #git
17:51 Akien grawity: Thanks, I'll have a look at this. Next question would be how to remove dangling commits from a GH remote that was force pushed, but I'm not sure it's even feasible. I might just end up recreating the remote.
17:52 grawity wait until the next scheduled Github garbage collect
17:52 grawity AFAIK they're more frequent nowadays but not something you can control
17:52 grawity I'd guess a few days
17:53 chris2 other people cant fetch unreachable objects anyway afaiu?
17:53 daivyk joined #git
17:54 grawity they cannot
17:54 chris2 hm, if B forks A, A commits, A force pushes HEAD^, B pushes a ref to the conjectured hash of A, can B fetch?
17:54 grawity mmmmaybe
17:54 Akien grawity: Thanks, I guess I can wait then, I probably won't be sued for having proprietary upstream commits in my fork for a few days :)
17:54 chris2 would be an interesting thing but i'm too lazy to make a second account now
17:54 grawity well, Github's own web UI makes dangling commits accessible for a while too
17:54 grawity by commit ID, that is
17:54 chris2 the caching layer?
17:54 chris2 of their web stuff?
17:55 multi_io joined #git
17:55 Mobutils joined #git
17:55 grawity but ... if it's a real fork, then that's no different from accessing the *upstream* commits
17:55 grawity I remember when I tried forking torvalds/linux, pushed some garbage to my fork, and it was accessible via https://github.com/torvalds/linux/commit/<id>
17:55 chris2 Akien: by putting stuff on github you grant permission to fork by their conditions
17:55 grawity it's a single repo after all
17:55 chris2 grawity: neat :>
17:56 chris2 (to public repos ofc)
17:57 chabil joined #git
17:59 raijin joined #git
17:59 KevinMGranger Are you sure about that? I could have sworn it was view, not fork
17:59 Syn_Don_ joined #git
17:59 chris2 there is no way to disable the fork button
18:00 KevinMGranger Nope nevermind you're right
18:00 chris2 "You must not transmit any worms or viruses or any code of a destructive nature."
18:00 chris2 so i can't host rm.c ;)
18:00 KevinMGranger Allowing the button to work doesn't necessarily mean it's legal. The CFAA is pretty loosey-goosey. But you're right, their help page says "you allow others to view and fork your repository within the GitHub site."
18:01 chris2 now is the queston whether view+fork implies clone ;)
18:03 shaggycat joined #git
18:04 blackwind_123 joined #git
18:04 SwiftMatt joined #git
18:04 tabrez joined #git
18:05 elisa87 joined #git
18:05 elisa87 hey can you have a look at this error? not sure how to fix it? http://pastebin.com/gxDNW3GF/
18:06 robotroll joined #git
18:06 kadoban elisa87: You haven't commited, because you didn't add anything to the index/staging area first.
18:07 Syn_Don joined #git
18:09 axsuul joined #git
18:10 duderonomy joined #git
18:10 Syn_Don joined #git
18:10 elisa87 kadoban: why do I get this error? jalal@klein:~/computer_vision/py-faster-rcnn$ git add -A
18:10 elisa87 fatal: Not a git repository: caffe-fast-rcnn/../.git/modules/caffe-fast-rcnn
18:11 foist joined #git
18:12 irqq joined #git
18:12 dpower joined #git
18:12 kadoban elisa87: You seem to be inside of the .git directory of a repo? Not sure why you'd be in there.
18:12 moop joined #git
18:12 iota_pi joined #git
18:12 alexsystemf___ joined #git
18:12 lmatteis joined #git
18:13 elisa87 I have cloned a repo from git and have done some changes but already had deleted their .git file and did a git init to create my own .git folder kadoban
18:13 LeBlaaanc joined #git
18:13 chabil joined #git
18:13 kadoban elisa87: Sounds like you did something wrong, at a guess. I don't understand at all that path it's showing you above ^
18:14 elisa87 http://pastebin.com/QRd7jDmT kadoban
18:14 chrisshattuck joined #git
18:15 d4rklit3 joined #git
18:15 vF3hNGxc47h8 joined #git
18:16 kadoban Oh, yeah I dunno, maybe there was submodules in the original that you've broken? I don't really know submodules anymore.
18:17 vassagus joined #git
18:19 MattMaker joined #git
18:20 peepsalot joined #git
18:21 MattMaker joined #git
18:22 dreiss joined #git
18:22 mingrammer joined #git
18:24 dmto joined #git
18:26 vassagus joined #git
18:26 danslo joined #git
18:29 m1dnight1 joined #git
18:29 moneylotion joined #git
18:31 mohabaks joined #git
18:32 rjack2 joined #git
18:34 bruce_lee joined #git
18:36 raijin joined #git
18:37 chabil joined #git
18:38 vassagus joined #git
18:43 daniel_rose joined #git
18:43 Mikerhinos joined #git
18:44 daniel_rose joined #git
18:46 dreiss joined #git
18:46 WayToDoor joined #git
18:47 thm joined #git
18:48 cdown joined #git
18:48 vassagus joined #git
18:49 thm did I understand it right, that running got filter-branch after git replace makes the replacements permanent in the sense that --no-replace-objects doesn't have an effect afterwards?
18:50 _ikke_ thm: yes, that's true
18:50 ItSANgo joined #git
18:51 thm ok
18:52 Guest81 joined #git
18:54 zacts joined #git
18:57 rxo joined #git
18:58 vassagus joined #git
18:58 kyan joined #git
18:59 atomicb0mb joined #git
18:59 Guest81 joined #git
19:02 danimal2 joined #git
19:05 anonymuse joined #git
19:06 vassagus joined #git
19:08 romerocesar joined #git
19:11 axsuul joined #git
19:11 cagedwisdom joined #git
19:14 star_prone joined #git
19:15 star_prone joined #git
19:17 mingrammer joined #git
19:17 thiago joined #git
19:18 avtobiff joined #git
19:20 LeBlaaanc joined #git
19:20 vassagus joined #git
19:21 Macaveli joined #git
19:22 cdown joined #git
19:23 daniel_rose joined #git
19:24 Guest81 joined #git
19:24 nwkj86 joined #git
19:27 thiago joined #git
19:27 danslo_ joined #git
19:29 zacts joined #git
19:32 mda1_ joined #git
19:32 spudowiar joined #git
19:36 LeBlaaanc joined #git
19:37 satbyy joined #git
19:37 satbyy joined #git
19:37 vassagus joined #git
19:37 Balliad joined #git
19:38 courrier joined #git
19:39 EduardoMartins joined #git
19:40 editshare-ericf joined #git
19:40 jaguarmagenta joined #git
19:41 LeBlaaanc joined #git
19:42 dave0x6d joined #git
19:42 LionsMane joined #git
19:43 perlpilot joined #git
19:43 eycsigfy joined #git
19:43 LeBlaaanc joined #git
19:43 synthroid joined #git
19:44 eycsigfy joined #git
19:45 Guest81 joined #git
19:46 al-damiri joined #git
19:46 Macaveli joined #git
19:47 AaronMT_ joined #git
19:50 SwiftMatt joined #git
19:50 vassagus joined #git
19:51 kyan joined #git
19:51 ahr3n joined #git
19:52 dmiles joined #git
19:52 dmiles I would like to put into my code comments a woed liek NEVER_COMMIT if i attept to comment while such work us in the file.. can i make git commit fail?
19:53 _ikke_ precommit hook
19:53 netcarver joined #git
19:53 dmiles so its not POLICY right?
19:54 _ikke_ You can bypass a pre-commit hook, yes
19:55 dmiles is there an example of someone doign this i can look at?
19:55 LeBlaaanc joined #git
19:56 perlpilot git normally puts a pre-commit hook sample in .git/hooks with extensive commentary on how it should work.
19:56 _ikke_ I'm not aware of any
19:56 _ikke_ right
19:57 ruby_ joined #git
19:57 cdown joined #git
19:58 LeBlaaanc joined #git
19:58 drizztbsd joined #git
19:58 cdg joined #git
20:00 romerocesar joined #git
20:00 LeBlaaanc joined #git
20:01 dmiles https://github.com/niden/Git-Pre-Commit-Hook-for-certain-words
20:01 drizztbsd joined #git
20:01 m0viefreak joined #git
20:02 dsantiago joined #git
20:03 kadoban dmiles: That won't work all that well with many usages of the index
20:03 elisa87 how can I fix this issue? :( http://pastebin.com/E11zj9tE
20:03 kadoban dmiles: For example, what if I change a file, add the banned word, stage that change, and then remove the banned word (but keep other changes)?
20:04 _ikke_ elisa87: rewrite the commit and all commits following that introduce this file
20:04 _ikke_ elisa87: !bfg can help with that
20:04 gitinfo elisa87: A tool designed to remove large files, or passwords from history: https://rtyley.github.io/bfg-repo-cleaner/ (!rewrite applies)
20:05 dmiles aha, so i need like a pre-staging hook then right?
20:05 _ikke_ There is no hook like that
20:05 dmiles (there will probly be a way to sneak arround whatever script we come up with .. mainly i am just trying to police myself)
20:05 _ikke_ pre-commit hook is good enough
20:05 _ikke_ That's exactly what pre-commit hooks are for
20:06 kadoban dmiles: It's likely good enough, if you trust yourself a little bit. As an exercise a while back I came up with this, which takes care of many many more cases, but is quite a bit more complicated too: https://www.emptypath.com/git_pre_commit_testing (the end is the result, the main bulk goes through reasoning and step by step)
20:06 _ikke_ You just need to check what's in the index
20:07 AaronMT joined #git
20:07 chrisshattuck joined #git
20:08 elisa87 hey why do you think I get this error? http://pastebin.com/i361C9YV for git lfs
20:09 sssilver joined #git
20:09 LeBlaaan_ joined #git
20:09 _ikke_ elisa87: means that git-lfs is not in $PATH (no clue whether it should be)
20:09 dmiles kadoban: thank you this is excelent
20:09 Akien left #git
20:10 chrissha_ joined #git
20:10 chabil joined #git
20:10 elisa87 turns out I should have first done this (not mentioned in tutorial :[ ) jalal@klein:~/git-lfs-1.4.1$ sudo ./install.sh
20:10 elisa87 [sudo] password for jalal:
20:10 elisa87 Git LFS initialized.
20:10 nitesh` joined #git
20:11 star_prone joined #git
20:11 _ikke_ alright
20:12 duderonomy joined #git
20:12 Guest55 joined #git
20:13 LeBlaaanc joined #git
20:13 boombatower joined #git
20:14 Kioa joined #git
20:15 saml joined #git
20:15 saml is / special in branch name?
20:15 saml git checkout -b feature/foobar;  git commit ...;    git -b feature/foobar/test
20:16 bremner right. think about the branches as pathes (which they are)
20:16 _ikke_ saml: yes, in the sense that / is a path separater
20:16 saml to collaborate on feature/foobar, I want to create a branch off feature/foobar   ,  feature/foobar/test that adds some tests
20:16 _ikke_ that's not possible currently
20:16 _ikke_ You cannot have a dir and a file with the same name
20:16 inflames joined #git
20:16 saml ah okay
20:17 saml feature/foobar-test will do
20:17 waht joined #git
20:17 _ikke_ They are working on different ref-backends
20:17 _ikke_ (though it looks to be stalled)
20:18 qws-user-1228 joined #git
20:18 vassagus joined #git
20:20 MattMaker joined #git
20:21 satifant joined #git
20:23 LeBlaaanc joined #git
20:24 jimi_ joined #git
20:26 mkoskar joined #git
20:27 chrisshattuck joined #git
20:28 LeBlaaan_ joined #git
20:28 danslo_ joined #git
20:29 nikivi joined #git
20:30 Classic joined #git
20:31 [cust] joined #git
20:31 LeBlaaanc joined #git
20:31 star_prone joined #git
20:32 saml i cloned a repo.  then I forked the repo (github).  how can I set remote origin to the forked repo?
20:32 grawity git remote set-url
20:32 saml changing remote origin to different
20:32 grawity or even better, git remote add
20:32 grawity you can have the upstream *and* the fork as remotes
20:32 qws-user-1228 joined #git
20:33 saml i did git remote rm and git remote add origin.   but don't know how to set upstream so that pull and push would work with origin  (forked repo)
20:33 Classic help, i'm getting fatal: not a git repository from my bare repo... note that i accidentally had done a checkout at some point to the bare repo, but i was able to revert it back to bare repo, config has bare.repo set to true, worktrees pointing to the repo work, but commands don't work on the repo itself... operating in windows with 2.9.3
20:33 ahmed_elgabri joined #git
20:34 saml git branch --set-upstream-to=origin/master master   # this worked
20:34 raijin joined #git
20:34 LeBlaaan_ joined #git
20:35 _ikke_ yup
20:36 dodobrain joined #git
20:37 _ikke_ Do you have a file called HEAD in there?
20:37 _ikke_ Classic: ^
20:37 LeBlaaanc joined #git
20:39 elisa87 please have a look at here
20:39 elisa87 http://stackoverflow.com/questions/39258605/git-lfs-not-working-properly-for-files-larger-than-100mb
20:40 _ikke_ elisa87: That you started using git lfs does not mean the old history does not contain the large files anymore
20:41 elisa87 _ikke_:  can you please explain a little more?
20:41 elisa87 and how can I overcome this problem? _ikke_
20:44 cdown joined #git
20:44 Classic ikke: no i do not have a HEAD, I do have fetch_head and head.old
20:45 Typ7 joined #git
20:45 _ikke_ Classic: Ok, that's why the repo is not recognized
20:45 _ikke_ git symbolic-ref HEAD refs/heads/master should fix that
20:45 _ikke_ not sure if that works if git does not recognize the repo though
20:46 Classic thanks! that seems to work
20:46 jeffreylevesque joined #git
20:47 _ikke_ elisa87: posted an answer
20:48 Classic ikke: actually, i just set head.old back to head though, which contained the proper ref, you set me on the right track though
20:48 _Vi joined #git
20:48 skift joined #git
20:48 ahr3n joined #git
20:49 _ikke_ right
20:49 exarch joined #git
20:49 _Vi How do I disable rerere just for one merge (not just rerere-autoupdate, but any modifications). I want to play with merge tools and rerere stays in the way.
20:49 sssilver joined #git
20:50 moei joined #git
20:50 _Vi Can I override config for one git command from argv or environment?
20:50 Classic i have another question.... on a different repo, i removed the remote repository thinking it would protect me since the remote repo completely rewrote history... however all of the remote branches disappeared, I know they are still in my repo, how do I access them?
20:51 _Vi Probalem resolved: `git -c rerere.enabled=false merge ...`
20:51 kyan joined #git
20:51 skift if i have say project "foo", with history, is there a way I can take project "bar" with its own history, and place that into a subdirectory of "foo", so "/foo/bar" and merge their histories into one repository? Where the history of "bar" would essentially only be applied to the now "/foo/bar" directory". Sort of like a subtree merge but i want to have a
20:51 skift single git repo in the end.
20:51 _ikke_ _Vi: yup, was about to suggest that
20:51 vassagus joined #git
20:51 hahuang61 joined #git
20:52 mingrammer joined #git
20:55 Salander27 joined #git
20:56 dermoth_ joined #git
20:56 moei joined #git
20:56 anttim joined #git
20:56 KevinMGranger skift: git subtree add ?
20:59 MeltedLux joined #git
21:00 LeBlaaan_ joined #git
21:03 vuoto joined #git
21:04 skift KevinMGranger: yea thats what i was looking at. but the idea is I am permanantly merging the projects. once it is in a subdirectory, i wont me managing the merged in repo anymore as my goal is to have one master repo. i was just hoping that I could some how pull in the history of "bar" into "foo" or something.
21:05 skift basically, once its merged in, i would actually delete the github repo for "bar" that goes into "/foo/bar"
21:05 vuoto joined #git
21:06 vuoto joined #git
21:06 garethdaine joined #git
21:07 d0nn1e joined #git
21:07 vuoto joined #git
21:07 eijk joined #git
21:08 tvw joined #git
21:08 cjohnson skift: are they forks? or just two totally distinct things
21:09 jgjorgji joined #git
21:09 jgjorgji i know binary diffing is a problem for storing binary diffs, but how about binary files that never change? would that be somehow a problem?
21:10 grawity probably not
21:10 grawity I don't think the packfiles are particularly text-optimized
21:10 skift totally different repos. but they are related. im cleaning up some bad decisions from the last person that used to work here.
21:10 skift everything was its own repo, to the point of it being a pain in the rear
21:11 Jeroen joined #git
21:11 LeBlaaanc joined #git
21:12 l4v2 joined #git
21:12 jeffreylevesque joined #git
21:12 kadoban jgjorgji: You're usually only in a *really* bad spot if both they're huge and change often (with varying definiitons of those depending on how big and how often), or if they change in branches that then need to be merged.
21:13 jgjorgji no i just want to store source .tar.gz for making RPMs and i would be handy to have them in git repo
21:13 jgjorgji they will never change for the same name
21:15 moneylotion joined #git
21:18 cjohnson skift: clone the other repo to somewhere safe and keep it for when you need history
21:18 cjohnson You aren't really going to be able to import an entirely separate history tree like that though maybe I'm wrong
21:18 vassagus joined #git
21:21 sssilver joined #git
21:23 GT4066 joined #git
21:25 FilipNor1ic joined #git
21:26 skift cjohnson: yea, i didnt think i would be able to do this. i just thought i would ask.
21:26 LeBlaaan_ joined #git
21:27 jeffreylevesque joined #git
21:29 dviola joined #git
21:29 danslo_ joined #git
21:29 AnarchyAo joined #git
21:31 Shentino joined #git
21:31 LeBlaaanc joined #git
21:33 AnarchyAo joined #git
21:34 AnarchyAo joined #git
21:35 AnarchyAo joined #git
21:35 dodobrain joined #git
21:35 msonntag joined #git
21:37 LeBlaaan_ joined #git
21:37 maret joined #git
21:37 KevinMGranger skift: unless I'm misunderstanding, why won't that work for you?
21:38 KevinMGranger "Create the <prefix> subtree by importing its contents from the given <commit> or <repository> and remote <ref>. A new commit is created automatically, joining the imported project’s history with
21:38 KevinMGranger your own
21:40 Dougie187 left #git
21:40 compeman joined #git
21:40 cjohnson KevinMGranger: what commands would let you do that? you do mean importing an entire history tree from one unrelated repo to another?
21:41 KevinMGranger Yes, isn't that what git subtree add does?
21:41 jaguarmagenta joined #git
21:42 skift yea, im going to have a clone of the repo somehwere to keep the history. but i only want one resulting git repo.
21:42 moneylotion joined #git
21:43 skift subtree add though seems like its still dependent on the other repo being around. we ultimately want to merge the two and deprecate or delete the old repo.
21:44 KevinMGranger It doesn't need to keep the other one around. It has features to assume you still want to work with the other one, but you don't have to
21:44 skift ooh.
21:44 skift ok.
21:45 KevinMGranger I can't remember the last time I used it, though. So test it out first in case I'm wrong
21:46 duderonomy joined #git
21:46 mingrammer joined #git
21:48 duderonomy joined #git
21:49 skift yea. will do
21:50 LeBlaaanc joined #git
21:51 mehola joined #git
21:52 afuentes joined #git
21:53 str joined #git
21:54 Puffball joined #git
21:54 vassagus_ joined #git
21:55 skift ok so i added the old repo as a subtree into the new one, and its basically like i just copied the files into the subdirectory.
21:55 skift which, i suppose is ok. since i will have the history in the other branch that pertains to the old repo.
21:56 mostlybadfly joined #git
21:57 wrouesnel1 joined #git
21:59 compeman joined #git
22:01 KevinMGranger the history should be part of the current branch, shouldn't it?
22:01 KevinMGranger Hmm, I guess "joining the imported project's history" doesn't necessarily imply that
22:01 andlabs joined #git
22:02 KevinMGranger I guess you could also script out a bunch of patch files from the orginal and apply them with a prefix but that might take longer and it would be a linear history
22:03 atomicb0mb joined #git
22:05 Raed|Mobile joined #git
22:07 skift heh yea im not THAT worried about it. the history is there, just in the branch for that other repo.
22:07 skift which is ok
22:08 PioneerAxon joined #git
22:09 p4trix joined #git
22:10 justanotheruser joined #git
22:10 moneylotion joined #git
22:10 shaggycat joined #git
22:12 romerocesar joined #git
22:14 hahuang61 joined #git
22:14 KnightsOfNi joined #git
22:15 AndrewAlexMac joined #git
22:15 cjbrambo joined #git
22:15 spudowiar joined #git
22:17 cdg joined #git
22:17 rkazak joined #git
22:18 LeBlaaan_ joined #git
22:18 ahr3n joined #git
22:21 astrofog joined #git
22:25 sssilver joined #git
22:26 eijk joined #git
22:27 mattcen joined #git
22:27 cagomez joined #git
22:29 LeBlaaanc joined #git
22:31 danslo_ joined #git
22:32 madewokherd joined #git
22:32 sssilver joined #git
22:32 chrisshattuck joined #git
22:35 Jon28_ joined #git
22:35 sssilver joined #git
22:36 e14 joined #git
22:36 romerocesar joined #git
22:38 moneylotion joined #git
22:38 shinnya joined #git
22:43 ResidentBiscuit joined #git
22:43 cagomez joined #git
22:49 thomas_25 joined #git
22:49 axsuul joined #git
22:54 l4v2 joined #git
22:56 daniel_rose joined #git
22:56 ToBeCloud joined #git
22:57 Guest55 joined #git
22:59 BSaboia joined #git
22:59 dreiss joined #git
23:00 mizu_no_oto joined #git
23:01 ochorocho__ joined #git
23:03 moneylotion joined #git
23:04 elsevero joined #git
23:05 ruby_ joined #git
23:08 fahadash joined #git
23:11 ruby_ joined #git
23:14 JeremyD joined #git
23:14 JeremyD joined #git
23:17 CaneToad joined #git
23:18 CaneToad folks, is it possible to configure git so that when you do a clone, a message appears showing some pre-configured message configured against the branch describing what you've just cloned?
23:19 zerox2a joined #git
23:20 WayToDoor joined #git
23:20 Caelum joined #git
23:22 diogenese joined #git
23:23 osse CaneToad: You mean something you can configure that others will see when they clone your repo? No.
23:24 sangy joined #git
23:24 osse You're free to create a wrapper script or similar for your own use, but  vanilla git does not provide anything
23:25 GavinMagnus joined #git
23:30 romerocesar joined #git
23:32 danslo_ joined #git
23:33 bremner of possible interest: man git-branch --edit-description
23:33 gitinfo the git-branch manpage is available at http://jk.gs/git-branch.html
23:33 bremner doesn't mention clone.
23:34 SleePy joined #git
23:34 mingrammer joined #git
23:37 cagomez joined #git
23:42 anonymuse joined #git
23:42 jaguarmagenta joined #git
23:43 dsdeiz joined #git
23:50 sssilver joined #git
23:51 skylite joined #git
23:51 e14 joined #git
23:52 mehola joined #git
23:52 e14 joined #git
23:53 Wilbur joined #git
23:54 e14 joined #git
23:54 orville joined #git

| Channels | #git index | Today | | Search | Google Search | Plain-Text | summary