Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2016-10-27

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

All times shown according to UTC.

Time Nick Message
00:03 BenderRodriguez joined #git
00:03 drmt joined #git
00:04 Electrometro joined #git
00:04 watersoul_ joined #git
00:05 Gonzo89_ joined #git
00:05 mizu_no_oto joined #git
00:08 royalt213 joined #git
00:09 Derperperd joined #git
00:10 nettoweb joined #git
00:12 alex1a_ joined #git
00:12 jimi_ joined #git
00:17 _28_ria joined #git
00:17 anuxivm left #git
00:19 e14 joined #git
00:25 _28_ria joined #git
00:26 d0nn1e joined #git
00:30 railssmi_ joined #git
00:33 d0nn1e joined #git
00:36 f3r70rr35f joined #git
00:37 fuchstronaut joined #git
00:37 `slikts joined #git
00:37 Gsham joined #git
00:39 d0nn1e joined #git
00:40 cyphase joined #git
00:40 nettoweb joined #git
00:44 fstd_ joined #git
00:47 d0nn1e joined #git
00:50 Dohplay joined #git
00:51 fakenerd joined #git
00:52 theskillwithin joined #git
00:52 f3r70rr35f joined #git
00:52 Spacew00t joined #git
00:52 Mutter joined #git
00:54 lindenle joined #git
00:56 ojdo joined #git
00:56 royalt213 joined #git
00:56 royalt213 joined #git
00:57 irqq joined #git
00:57 Mutter joined #git
00:57 d0nn1e joined #git
01:00 jaguarmagenta joined #git
01:01 jaguarma_ joined #git
01:02 Electrometro joined #git
01:02 d^sh_ joined #git
01:03 raijin joined #git
01:03 d0nn1e joined #git
01:07 ochorocho__ joined #git
01:08 lindenle joined #git
01:08 finalbeta joined #git
01:08 dodobrain joined #git
01:13 ki0 joined #git
01:14 thiago joined #git
01:16 cdown joined #git
01:16 f3r70rr35f joined #git
01:20 cagmz joined #git
01:21 WhiskerBiscuit joined #git
01:22 ResidentBiscuit joined #git
01:23 duarteper joined #git
01:24 hck89 joined #git
01:24 warthog9 joined #git
01:26 LukeLR joined #git
01:30 warthog9 joined #git
01:30 Dougie187 joined #git
01:36 sonikspin joined #git
01:36 mellernoia joined #git
01:37 f4 joined #git
01:38 fuchstronaut joined #git
01:39 nawwmz joined #git
01:40 juvenal joined #git
01:41 akurniawan joined #git
01:41 gfixler joined #git
01:46 Vampire0_ joined #git
01:46 warthog9 joined #git
01:48 ilbot3 joined #git
01:48 Topic for #git is now Welcome to #git, the place for git help and public-access television programming | Please take the git survey: https://jk.gs/gsurv | Public logs at http://jk.gs/glog | First visit? Read: http://jk.gs/git | Current stable version: 2.10.1 | Getting "cannot send to channel"? /msg gitinfo .voice | git-bisect: No, not like that. Grow up.
01:50 e14 joined #git
01:54 MrWoohoo joined #git
01:58 bruce_lee joined #git
02:03 Goplat joined #git
02:05 fakenerd joined #git
02:07 aidalgol joined #git
02:15 liuzhen joined #git
02:16 lindenle joined #git
02:17 hexagoxel joined #git
02:18 kille joined #git
02:19 sunri5e joined #git
02:21 jaguarmagenta joined #git
02:23 mizu_no_oto joined #git
02:23 ResidentBiscuit joined #git
02:24 jaguarma_ joined #git
02:33 ulkesh_ joined #git
02:34 ayjay_t joined #git
02:36 dsantiago joined #git
02:39 fuchstronaut joined #git
02:41 ResidentBiscuit joined #git
02:42 lagothri1 joined #git
02:43 e14 joined #git
02:43 nettoweb joined #git
02:44 inflames joined #git
02:46 Xuefer joined #git
02:48 chachasmooth joined #git
02:51 diogenese joined #git
02:52 Jameser joined #git
02:52 howdoi joined #git
02:52 mehola joined #git
02:55 Rodya_ joined #git
02:55 Dougie187 left #git
02:59 Lunatrius joined #git
03:00 Orbitrix joined #git
03:01 Orbitrix joined #git
03:03 diogenese joined #git
03:05 aavrug joined #git
03:05 f3r70rr35f joined #git
03:06 PioneerAxon joined #git
03:08 warthog9 joined #git
03:08 mangolisk joined #git
03:14 diogenese joined #git
03:14 Vortex35 joined #git
03:16 sssilver joined #git
03:17 buddyq joined #git
03:18 sangy joined #git
03:24 offtopik joined #git
03:27 offtopik any suggestions for handling program version numbers while using git (php web apps)? I've essentially abandoned them because it seems like overkill to have a commit just for incrementing the number yet I want to keep ability for git revert
03:28 joki joined #git
03:28 ojacobson git-describe is often a good place to start
03:28 kadoban offtopik: You could just tag with versions, unless it needs to be in text itself?
03:30 offtopik kadoban so any given web app might have up to a dozen installations, and each one could be on a different commit from master and it previously was helpful for a user to reference something back to me, so I changed from incrementing version to a time/date stamp with first part of sha1
03:30 offtopik this shows in the corner for the user to see
03:31 kadoban Seems fine to me
03:32 xall_ joined #git
03:33 offtopik it works ok for now, but every other project I look at uses major.minor.patch scheme and right now you can't really look at one app and know it is ahead or behind another install based on sha1; the fact that I show timestamp could help, but if I roll back an install the timestamp updates anyway
03:33 roue_ left #git
03:33 dreiss joined #git
03:34 kadoban If you want that, I'd just use tags and git-describe
03:34 davidosomething joined #git
03:36 offtopik ojacobson kadoban thanks; git-describe looks promising
03:39 fuchstronaut joined #git
03:42 robotroll joined #git
03:44 Prawnzy joined #git
03:45 midnightmagic joined #git
03:46 User458764 joined #git
03:52 svm_invictvs joined #git
04:03 a_thakur joined #git
04:05 Andrew_K joined #git
04:09 milki hbai: i gotta check with my parents. they mentioned grave visiting
04:09 milki df
04:09 milki er..
04:16 User458764 joined #git
04:17 phanimahesh joined #git
04:23 ToBeCloud joined #git
04:25 Peng joined #git
04:29 Cabanossi joined #git
04:35 cdown joined #git
04:38 daey_ joined #git
04:40 fuchstronaut joined #git
04:40 fuchstronaut joined #git
04:41 a_thakur joined #git
04:43 evie_hammond joined #git
04:43 a_thakur joined #git
04:52 kazantip joined #git
04:54 a_thakur_ joined #git
04:55 cyan__ joined #git
04:57 freimatz joined #git
04:59 tax joined #git
05:04 nilg joined #git
05:06 ayjay_t joined #git
05:07 mykespb joined #git
05:10 watabou joined #git
05:13 brandonw joined #git
05:13 benbergman left #git
05:14 star_prone joined #git
05:14 Kaisyu joined #git
05:15 subhojit777 joined #git
05:16 dsdeiz joined #git
05:16 dsdeiz joined #git
05:18 User458764 joined #git
05:23 Darren_ joined #git
05:24 finalbeta joined #git
05:26 davidosomething joined #git
05:28 satya4ever joined #git
05:32 barteks2x joined #git
05:34 User458764 joined #git
05:34 cyphase joined #git
05:42 fuchstronaut joined #git
05:42 offtopik joined #git
05:43 vlt joined #git
05:45 _28_ria joined #git
05:46 liuzhen joined #git
05:47 barteks2x_ joined #git
05:48 dling joined #git
05:48 raphink joined #git
05:50 Kaisyu joined #git
05:50 _cyril_ joined #git
05:54 futabachannel joined #git
05:58 Raging_Hog joined #git
06:01 Anja joined #git
06:03 User458764 joined #git
06:05 davidosomething joined #git
06:08 LucaTM joined #git
06:09 nilg joined #git
06:09 eivarv joined #git
06:10 hahuang65 joined #git
06:10 rscata joined #git
06:11 eivarv joined #git
06:16 sssilver joined #git
06:16 freimatz joined #git
06:17 jnavila joined #git
06:18 sunri5e joined #git
06:18 lindenle joined #git
06:19 chele joined #git
06:20 knittl I've just encountered a weird problem
06:20 knittl I did a clone --bare of a --mirror repository with only one local branch. Then I created new branches in the mirror and ran "git fetch" in the bare clone. No new branches were updated. Am I missing something?
06:21 lordjancso joined #git
06:21 dan2k3k4 joined #git
06:21 knittl fetch --all did not help either
06:22 knittl hmmmm. there seems to be the fetch = +refs/heads/*:refs/remotes/origin/* line missing
06:22 knittl is this by design?
06:22 barteks2x joined #git
06:23 _ikke_ Doesn't it have a fetch refspec at all?
06:23 knittl nope
06:23 _ikke_ that version is for non-bare branches
06:24 _ikke_ bare repositories*
06:24 knittl ok. then +refs/heads/*:refs/heads/*
06:24 knittl but there is no fetch= line in the config
06:24 _ikke_ right
06:24 cdown joined #git
06:25 knittl shouldn't there be? _ikke_
06:25 _ikke_ there should
06:25 _ikke_ Can you try to clone again and see if it happens agian
06:25 _ikke_ again
06:26 knittl I already cloned twice
06:26 _ikke_ and same result?
06:26 knittl it's on the same filesystem, could that be a problem? (since hardlinks of object db)
06:26 brk_ joined #git
06:26 _ikke_ I don't think that should make a difference in the refspec being used
06:27 knittl git clone --bare repo.git/ missing-fetch
06:27 knittl Cloning into bare repository 'missing-fetch'...
06:27 knittl done.
06:27 knittl cat missing-fetch/config # no fetch, only remote.origin.url = ...
06:27 _ikke_ What version of git do you have?
06:27 knittl git version
06:27 knittl git version 2.10.0.304.gf2ff484
06:28 _ikke_ That's new enough
06:28 knittl it's custom built from next
06:28 _ikke_ right
06:28 dodobrain joined #git
06:32 dreiss joined #git
06:34 Derperperd joined #git
06:35 Xuefer joined #git
06:36 bmahe joined #git
06:38 Sceorem joined #git
06:42 JeroenT joined #git
06:42 fub joined #git
06:42 fuchstronaut joined #git
06:45 bocaneri joined #git
06:47 star_prone joined #git
06:48 User458764 joined #git
06:49 hariel joined #git
06:49 JeroenT joined #git
06:50 elect joined #git
06:51 k200 joined #git
06:51 k200 can i / should I change branch name after it has a major code changed and previous commit merged to master-> praduction?
06:52 _ikke_ You could just create a new branch
06:53 NimeshNeema joined #git
06:53 zulutango joined #git
06:53 kimonorouge joined #git
06:56 acetakwas joined #git
06:56 corsibu joined #git
06:56 futabachannel left #git
06:59 lordjancso joined #git
07:00 NimeshNeema joined #git
07:01 k200 ok
07:02 qt-x joined #git
07:03 _28_ria joined #git
07:06 alcohol joined #git
07:07 mkoskar joined #git
07:08 ki0 joined #git
07:08 ki0 joined #git
07:09 vuoto joined #git
07:09 alcohol joined #git
07:11 _28_ria joined #git
07:12 TomyWork joined #git
07:12 manuels joined #git
07:12 _ng joined #git
07:14 maret joined #git
07:15 ki0 joined #git
07:18 pks joined #git
07:19 knittl will git-filter branch create new hashes when I use the same commit message and tree?
07:20 GavinMagnus joined #git
07:21 _ikke_ knittl: The hash will change if any of (commit msg, author/comitter, parent(s), tree) changes
07:21 knittl I mistakenly fetched branches and now I have to run filter-branch again :(
07:21 knittl 1 hr
07:21 _ikke_ Why?
07:21 knittl which sucks
07:21 armyriad joined #git
07:21 _ikke_ If you fetch branches, only remote tracking branches are updated
07:21 knittl fetch will not update the reflog, will it?
07:22 knittl it's a --mirror repo
07:22 _ikke_ knittl: reflog is disabled by default for bare repositories
07:23 knittl oh no. it looks good. I have filter-branch: rewrite in the reflog for that specific branch
07:23 knittl yisss
07:25 laumann joined #git
07:26 overlord_tm joined #git
07:26 finalbeta joined #git
07:28 PCatinean joined #git
07:28 Alienpruts joined #git
07:28 _ikke_ yippie
07:31 dsdeiz_ joined #git
07:32 NitroLux joined #git
07:33 bongjovi joined #git
07:34 NitroLux|2 joined #git
07:37 GavinMagnus joined #git
07:38 JeroenT_ joined #git
07:40 rchavik joined #git
07:40 VladGh joined #git
07:40 User458764 joined #git
07:42 wrouesnel joined #git
07:43 JeroenT joined #git
07:43 fuchstronaut joined #git
07:44 Peng joined #git
07:44 halbomin joined #git
07:45 ahmedelgabri joined #git
07:46 GavinMagnus left #git
07:50 dan2k3k4 joined #git
07:54 armyriad joined #git
07:54 geoid joined #git
07:54 afuentes joined #git
07:54 offtopik joined #git
08:01 NitroLux joined #git
08:02 jbitdrop joined #git
08:03 NitroLux|3 joined #git
08:03 fees joined #git
08:04 Balliad joined #git
08:04 mikecmpbll joined #git
08:05 star_prone joined #git
08:07 GavinMagnus joined #git
08:07 Fissure joined #git
08:07 chll_ joined #git
08:07 encod3 joined #git
08:07 leehambley joined #git
08:08 cyan__ joined #git
08:10 sea-gull joined #git
08:14 JeroenT joined #git
08:15 maret joined #git
08:15 King_Hual joined #git
08:16 fub joined #git
08:17 djb-irc joined #git
08:18 JeroenT_ joined #git
08:18 mrpackethead joined #git
08:19 ufk joined #git
08:19 ufk hi!
08:19 ufk I want each commit to be applied on both github and my local git ssh server. is that possible ?
08:19 corsibu joined #git
08:20 ManiacTwister joined #git
08:21 tobiasvl ufk: what do you mean? when you push to one of them it should automatically push to the other?
08:22 tobiasvl https://git-scm.com/book/en/​v2/Customizing-Git-Git-Hooks
08:23 Orbitrix joined #git
08:23 moritz ufk: you can also configure your local git to push to two remotes by default
08:23 mischat joined #git
08:23 ufk whoo lots of options here
08:24 ufk okok.. I want my local git to push whenever something is pushed to it
08:25 ufk so i need to create a hook for post-receive right ?
08:26 corsibu joined #git
08:26 User458764 joined #git
08:29 marcogmonteiro joined #git
08:30 Guest54183 joined #git
08:31 a_thakur joined #git
08:33 tvw joined #git
08:34 ocbtec joined #git
08:37 raijin joined #git
08:41 acetakwas joined #git
08:41 jfr joined #git
08:43 djb-irc joined #git
08:43 dangerousdave joined #git
08:44 aidalgol joined #git
08:44 skunkz joined #git
08:44 fuchstronaut joined #git
08:47 Macaveli joined #git
08:50 eivarv joined #git
08:50 djb-irc joined #git
08:51 jab416171 joined #git
08:51 clmsy joined #git
08:52 p4trix joined #git
08:52 wget_ joined #git
08:52 davimore joined #git
08:54 davimore__ joined #git
08:54 Anja joined #git
08:54 a_thakur joined #git
08:55 fub left #git
08:56 skunkz_ joined #git
08:57 ayjay_t joined #git
08:58 Jimmio92 joined #git
08:59 Jimmio92 Hey all. I've got a git repo with multiple submodules. Can I add a CMakeLists.txt to their folders without modifying those projects for building?
09:00 brk_ joined #git
09:01 Jimmio92 and if I change a submodule's CMakeLists.txt (so it no longer builds unneeded options..), will those changes be recognized by the main repo and handled or not?
09:01 moritz they'll be marked as changes in the submodule
09:01 j416 does anyone know what the state of Gogs is? have they gotten to their senses and fixed their issues?
09:03 jknetl joined #git
09:03 dangerousdave Morning
09:03 drbean_ joined #git
09:03 aswen joined #git
09:05 dangerousdave I have just become aware of four commits in our repository that I can't account for, the auto message for each is "Merge branch 'develop' of https://bitbucket.org/foo/cymap into develop"
09:06 dangerousdave looking at the graph, it appears these commit have been made on an unnamed ghost branch of develop
09:07 dangerousdave any ideas on whats gone wrong please?
09:08 qt-x joined #git
09:08 stemid joined #git
09:08 stemid is it possible to have alternate remotes for submodules just like for any repo?
09:11 t4nk592 joined #git
09:13 maret joined #git
09:13 tobiasvl stemid: yes
09:15 Jimmio92 moritz: Any way to tell git I want to track these files in the root repo?
09:15 _ikke_ no
09:15 _ikke_ the parent repo ignores everything in a submodule directory
09:16 phanimahesh joined #git
09:20 Jimmio92 Well... this is annoying. I wonder how the heck I'm going to change CMakeLists.txt in one of the submodules without forking that project...
09:20 Jimmio92 Thanks all.
09:20 deanosaur joined #git
09:24 ferr joined #git
09:25 ahmed_elgabri joined #git
09:25 JeroenT joined #git
09:25 irqq joined #git
09:27 aelgabri joined #git
09:35 dangerousdave joined #git
09:37 qpdb joined #git
09:37 star_prone joined #git
09:37 jab416171 joined #git
09:40 User458764 joined #git
09:40 corsibu joined #git
09:40 troulouliou_div2 joined #git
09:40 troulouliou_div2 joined #git
09:42 davimore_ joined #git
09:43 Ravana joined #git
09:44 turci[m] joined #git
09:45 fuchstronaut joined #git
09:45 fuchstronaut joined #git
09:45 corsibu joined #git
09:47 hck89 joined #git
09:48 turci[m] left #git
09:50 LukeLR joined #git
09:51 rorro joined #git
09:51 _cyril_ joined #git
09:53 finalbeta joined #git
09:54 jfr joined #git
09:56 dan2k3k4 joined #git
09:59 ieee joined #git
09:59 rnsanchez joined #git
10:02 freimatz joined #git
10:02 jfr joined #git
10:04 nettoweb joined #git
10:04 pijiu joined #git
10:06 SwiftMatt joined #git
10:06 mrpackethead joined #git
10:07 drbean joined #git
10:09 kimonorouge joined #git
10:15 mehola joined #git
10:19 davimore__ joined #git
10:22 Prawnzy joined #git
10:24 brk_ joined #git
10:24 drmt joined #git
10:25 User458764 joined #git
10:27 moop joined #git
10:27 NimeshNeema joined #git
10:27 GavinMagnus left #git
10:27 thebigj joined #git
10:27 dpower joined #git
10:27 eduardas_m joined #git
10:28 bluezone joined #git
10:28 Lor_ joined #git
10:29 thebigj I am using Ubuntu 16.04. I configured vimdiff here is my `.gitconfig` file: http://dpaste.com/0DXBHFX
10:30 k200 joined #git
10:30 k200 shall i directly editing master for just one new css class adding commit?
10:30 thebigj When I run `git difftool -t vimdiff` I am unable to see the diff in vimdiff. The output is same to `git diff`.
10:31 thebigj I already installed `vim` package. And when I run `vimdiff prog_1 prog_2` I am able to see nice visual diff in vimdiff in my computer.
10:31 thebigj Can anyone tell what can be the reason for that?
10:31 qpdb joined #git
10:34 robattila256 joined #git
10:37 brk__ joined #git
10:39 Macaveli joined #git
10:41 liuzhen joined #git
10:41 barteks2x joined #git
10:43 solf1re joined #git
10:45 mischat joined #git
10:46 fuchstronaut joined #git
10:48 _noblegas joined #git
10:48 jnewt joined #git
10:49 pijiu2 joined #git
10:49 freimatz joined #git
10:50 ahr3n joined #git
10:51 nikivi joined #git
10:51 pijiu2 joined #git
10:52 _ikke_ thebigj: I have no config at all, and git difftool -t vimdiff works for me
10:52 pijiu2 joined #git
10:53 Darcidride joined #git
10:53 thebigj _ikke_ Thanks for replay. Actually I figured out the problem. I was rebasing the branch and there was conflict.
10:53 thebigj I was using difftool for observing that conflict
10:53 thebigj But then I managed to configure mergetool from http://www.rosipov.com/blog/u​se-vimdiff-as-git-mergetool/
10:54 pijiu2 joined #git
10:54 thebigj _ikke_: Still there is one proble. According to http://www.rosipov.com/blog/u​se-vimdiff-as-git-mergetool/ I should get `Local, Base and Remote` window in top but, I only get three `Local` window.
10:54 thebigj * windows
10:55 nikivi joined #git
10:55 pijiu2 joined #git
10:55 fragMental joined #git
10:55 Xuefer_CBXIj joined #git
10:56 ahmedelgabri joined #git
10:57 pijiu2 joined #git
10:57 JeroenT joined #git
10:57 quackgyver joined #git
10:57 JeroenT joined #git
10:58 juvenal joined #git
10:59 Ediz joined #git
11:01 encod3 joined #git
11:02 ravenx joined #git
11:03 ravenx how can i remove commits that i accidentally did (and pushed)?
11:03 ravenx so i would liek to rollback, and then push it, without polluting the git history
11:03 ravenx (i accidentally commited a virtual environemtn in python)
11:05 lb ravenx !revert
11:05 gitinfo That's a rather ambiguous question... options: a) make a commit that "undoes" the effects of an earlier commit [man git-revert]; b) discard uncommitted changes in the working tree [git reset --hard]; c) undo committing [git reset --soft HEAD^]; d) restore staged versions of files [git checkout -p]; e) move the current branch to a different point(possibly losing commits)[git reset --hard $COMMIT]?
11:05 moritz ravenx: you can do an interactive rebase and drop those commits
11:05 moritz ravenx: but it means rewriting public history, which comes with its own slew of problems
11:05 moritz !public
11:05 gitinfo [!rewrite] 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
11:10 ravenx thanks guys.
11:11 Serpent7776 joined #git
11:12 GavinMagnus joined #git
11:12 Darcidride joined #git
11:13 HiddenPanda joined #git
11:15 User458764 joined #git
11:16 leehambley joined #git
11:16 mehola joined #git
11:22 alex1a joined #git
11:23 rorro joined #git
11:24 drbean_ joined #git
11:27 drbean- joined #git
11:29 vuoto joined #git
11:33 qpdb joined #git
11:33 dsdeiz joined #git
11:33 dsdeiz joined #git
11:36 stux|RC-only joined #git
11:37 netj joined #git
11:37 danslo joined #git
11:37 GavinMagnus joined #git
11:42 geoid joined #git
11:44 User458764 joined #git
11:45 netj joined #git
11:46 mizu_no_oto joined #git
11:46 LionsMane joined #git
11:47 fuchstronaut joined #git
11:50 nawwmz joined #git
11:53 lindenle joined #git
11:53 deanosaur joined #git
11:56 SuperSeriousCat joined #git
11:57 fuchstronaut joined #git
11:57 fuchstronaut joined #git
11:58 LionsMane joined #git
12:00 kraut joined #git
12:03 mischat joined #git
12:03 roercik joined #git
12:07 ShekharReddy joined #git
12:08 skunkz_ joined #git
12:11 GavinMagnus left #git
12:12 HiddenPanda_ joined #git
12:13 johnmilton joined #git
12:13 HiddenPanda_ joined #git
12:13 synthroid joined #git
12:14 gamelaster joined #git
12:15 davimore_ joined #git
12:16 gamelaster Hello, Im using GitHub and I have two repos. App and UIframework. In App I want to include the UIframework repository. But the whole work I need do in App repository. I tried use submodule, but I cannot push a changes from App repository to UIframework, any ideas?
12:18 jeffreylevesque joined #git
12:18 HiddenPanda joined #git
12:18 moritz gamelaster: to push changes in a submodule, you first have to change directory to the submodule, then check out a branch, apply your changes, commit, push, go to the main repo, and record the changed sha1 of the subrepo
12:19 moritz gamelaster: developing in submodules is generally a pain, because they are designed to be in detached HEAD state
12:19 BSaboia joined #git
12:19 moritz gamelaster: if repo size isn't a concern, git subtree might be a viable option
12:19 moritz !subtree
12:19 gitinfo The subtree merge method is great for incorporating a subsidiary git repo into your current one with "unified" history. Read http://git-scm.com/book/en/v2/Git-To​ols-Advanced-Merging#_subtree_merge for more info, see also !git-subtree and !git-stitch-repo.
12:20 offtopik joined #git
12:20 gamelaster moritz: so another way doesnt exists? Because I doing all work on my App repository, where I want to publish my UIframework to other people, so separate repository, but still I need to do all work in App repository. Subtree can do that?
12:20 moritz gamelaster: yes
12:20 selckin why not just have 2 repos
12:21 moritz that's the other option
12:21 moritz if there's relatively loose coupling between the versions
12:21 gamelaster I have 2 repos, but I need a UIframework repository "inside" of App repository because all of work I do in App repository
12:21 gamelaster And moritz thanks, im gonna try it
12:21 selckin so put it there?
12:22 fling joined #git
12:23 Prawnzy joined #git
12:24 acetakwas joined #git
12:25 netj joined #git
12:26 ProbabilityMoon joined #git
12:27 Noldorin joined #git
12:27 waveclaw joined #git
12:28 rnsanchez joined #git
12:29 star_prone joined #git
12:29 tvl joined #git
12:31 JeroenT joined #git
12:32 hanthings_ joined #git
12:33 mischat joined #git
12:33 lvns joined #git
12:34 vuoto joined #git
12:35 netj joined #git
12:36 xaa joined #git
12:37 kimonorouge joined #git
12:38 Endarked joined #git
12:38 xaa joined #git
12:39 BlackMaria joined #git
12:39 rkazak joined #git
12:40 JeroenT_ joined #git
12:41 waveclaw joined #git
12:41 shinnya joined #git
12:42 brk_ joined #git
12:50 _cyril_ joined #git
12:50 JeroenT joined #git
12:52 Derperperd joined #git
12:53 gamelaster moritz: Im trying it but I can update only from the original UIframework repository to App UIframework branch (to subtree). But I need from App subtree push changes to App UIframework branch and then to original UIframework repository. Is this possible?
12:53 corsibu joined #git
12:54 PCatinean joined #git
12:54 eindoofus_ joined #git
12:55 nilg joined #git
12:55 _eddy joined #git
12:55 Probabil_ joined #git
12:57 ash_workz joined #git
12:57 mischat_ joined #git
12:57 mischat_ joined #git
12:58 DeltaHeavy joined #git
13:00 styler2go joined #git
13:01 waveclaw joined #git
13:03 akurniawan joined #git
13:03 jimi_ joined #git
13:03 dan2k3k4 joined #git
13:05 PaulCapestany joined #git
13:06 d0nn1e joined #git
13:06 scorphus joined #git
13:07 LionsMane joined #git
13:07 waveclaw joined #git
13:07 Ryanar joined #git
13:08 PioneerAxon joined #git
13:08 kimonorouge joined #git
13:10 rnsanchez joined #git
13:11 Infamous joined #git
13:11 paws_ joined #git
13:12 waveclaw joined #git
13:12 wrouesnel2 joined #git
13:13 GavinMagnus joined #git
13:14 nfk joined #git
13:14 bongjovi joined #git
13:17 rivarun joined #git
13:18 ExoUNX joined #git
13:20 maret joined #git
13:20 CEnnis91 joined #git
13:20 cdg joined #git
13:23 IrishPrime joined #git
13:23 kcija joined #git
13:23 IrishPrime joined #git
13:24 zeorin_ joined #git
13:25 hck89 joined #git
13:25 anonymuse joined #git
13:26 plos joined #git
13:28 plos joined #git
13:29 plos joined #git
13:29 tm512 joined #git
13:30 buddyq joined #git
13:30 LionsMane joined #git
13:33 mischat joined #git
13:34 f3r70rr35f joined #git
13:35 mindfart joined #git
13:39 barteks2x joined #git
13:41 sdothum joined #git
13:42 mrpackethead joined #git
13:43 zeorin_ left #git
13:44 davimore joined #git
13:45 Es0teric joined #git
13:46 bongjovi Maybe not the place to ask but... does anyone encounter problems when running git commands on an external drive?
13:47 bongjovi i'm backing some things up from our GitLab account to some NAS but git clone sits idle for about 2 minutes before it does anything
13:47 mischat joined #git
13:48 bongjovi quite possible not a git issue though
13:48 bongjovi possibly*
13:48 Dougie187 joined #git
13:49 nettoweb joined #git
13:49 mischat_ joined #git
13:50 LionsMane joined #git
13:51 tobiasvl yeah, pretty slow here over nfs too
13:51 holgersson_ joined #git
13:52 bongjovi took roughly 5 minutes to clone a 380KB repo
13:55 kimonorouge joined #git
13:56 Shentino joined #git
13:57 boombatower joined #git
13:57 Renku joined #git
13:57 Renku hi
13:58 osse my name is
13:58 tobiasvl huh?
13:58 Renku is it possible to have a client hook in git triggering when i do "git pull", however the pull is made with the --rebase option or not?
13:58 kimonorouge joined #git
13:59 a3Dman joined #git
13:59 Renku i tried with the post-rewrite and post-merge
13:59 Renku post-merge is triggered when doing git pull without git --rebase
13:59 LucaTM joined #git
13:59 Renku but post-rewrite isn't triggered when i do git pull --rebase
13:59 Renku (because i don't have any commit to "rewrite" probably)
13:59 GavinMagnus left #git
14:00 Renku the best would have been a post-receive on the client side, but it seems not to be effective either :/
14:00 Renku (wooh... it doesn't seem to be effective* sorry about that)
14:00 osse Renku: post-rewrite is triggered by git rebase
14:03 Renku yep, but if there is nothing to "replay", nothing is triggered
14:04 rmichnik joined #git
14:05 StrongJava joined #git
14:05 cyan__ joined #git
14:05 akurniawan joined #git
14:05 Renku i mean, there is changes on the server side, but no new commit on my side, and then nothing is "rebased", do you see what i mean? :/
14:05 Jameser joined #git
14:05 tobiasvl you mean if it's a fast-forward? wouldn't the same be true for a regular non-rebase pull?
14:05 osse yes
14:06 osse also yes
14:06 osse your best best then is to make an alias for pull that does what you want
14:06 osse i wonder what that is, btw. maybe there's a different way
14:07 alhariel joined #git
14:08 AaronMT joined #git
14:09 anonymus_ joined #git
14:13 JeroenT_ joined #git
14:14 crose joined #git
14:16 CtrlC How do I push to the server if somebody did a push there?
14:16 Renku CtrlC: you have to pull first
14:16 CtrlC Git suggest pulling. I just want to make sure nothing gets lost.
14:17 CtrlC Renku, No data loss?
14:17 Renku if you pull nothing will be lost
14:17 CtrlC Thank you.
14:17 Renku a merge will be done :)
14:17 Renku if there is conflict, you'll have to resolve it manually
14:17 Renku so no risk
14:18 davimore_ joined #git
14:18 corpsicle what would make git fail to recognise the same code in the same files in two branches when i do git cherry?
14:18 Renku tobiasvl and osse: yes, we could compare it to a fast-forward, and i don't see what kind of alias would do the trick
14:18 ravenx joined #git
14:18 corpsicle several times ive noticed this when trying to align two branches
14:18 CtrlC Renku, I do the push this way: git push -u origin master how would the pulling be?
14:18 tobiasvl CtrlC: git goes to great lengths to prevent data loss. unless you run commands with --force or --hard flags you don't need to worry about it.
14:18 osse Renku: an alias that first does pull, then does whatever you want
14:19 corpsicle its not the same commits but afaik git is supposed to be smart enough to see that its the same code
14:19 Renku CtrlC: git pull origin master (if you already pushed once, you can just do "git pull" and "git push" ^^)
14:19 CtrlC tobiasvl, That's awesome. Thank uoi.
14:19 CtrlC Thanks.
14:19 tobiasvl corpsicle: can you explain better what you mean? recognize the same code how?
14:20 Xuefer_CBXIj left #git
14:20 Xuefer joined #git
14:20 encod3 joined #git
14:20 corpsicle tobiasvl: in the manfile it says that its supposed to regonise that even if its not the same commit, its the same code applied in both branches
14:21 corpsicle "The equivalence test is based on the diff, after removing whitespace and line numbers. git-cherry therefore detects when commits have been "copied" by means of git-cherry-pick(1), git-am(1) or git- rebase(1).
14:21 corpsicle "
14:21 _ikke_ see man git patch-id
14:21 gitinfo the git-patch-id manpage is available at http://jk.gs/git-patch-id.html
14:21 Renku osse: i'll explain what i would like to do: we have files in a folder (folder named "db") containing update of the database scripts. Problem, often, developpers when pulling forget to apply the patch of this section, so I would like to make a warn after a pull if and only if this folder had changes in its files... How, if I first do a pull then do something else, I would know the old version of the folder?
14:22 osse Renku: you can use the reflog of HEAD
14:23 osse but all developers must be told to use this alias then...
14:23 osse same goes for the hook
14:23 Renku the use of an alias should not be a problem
14:23 Renku but i would prefer a simple hook i install on their computers
14:23 tobiasvl Renku: also see !config
14:23 gitinfo Renku: [!configfiles] It is recommended to store local configuration data in a file which is not tracked by git, but certain deployment scenarios(such as Heroku) may require otherwise. See https://gist.github.com/1423106 for some ideas
14:24 moop joined #git
14:24 osse how about a check for whether the patch is already applied in build.sh or run.sh or make or whatever
14:24 Prawnzy joined #git
14:24 inflames joined #git
14:25 oncall-pokemon joined #git
14:25 PioneerAxon_ joined #git
14:25 Renku unable to load page :o
14:25 Renku oh, sorry
14:26 Renku my os is too old x)
14:26 tobiasvl your os?
14:27 e14 joined #git
14:27 Renku operating system
14:27 danslo_ joined #git
14:27 Renku i'm on an old debian >.> but that's not important^^ thanks for the link, i'll check heroku
14:27 tobiasvl well I know what "os" means, just wondering what you mean by your os being too old to load gist
14:27 jnavila joined #git
14:27 encod3 joined #git
14:27 tobiasvl old os implies old web browser?
14:28 Renku aaaand i'll try to see how to use reflog in shell x)
14:28 Renku in my case yep
14:28 Renku i'm on a debian squeeze
14:28 jozwior joined #git
14:28 Renku firefox is not able to work on it anymore (it crashes everytime), i don't even talk about chrome
14:28 Renku so i'm stuck with that good old epiphany...
14:28 tobiasvl ouch
14:28 Renku yep, ouch ^^ love my job.
14:28 Renku >.>
14:28 tobiasvl squueze is out of LTS too
14:28 Renku i know
14:29 Renku in fact, i have a virtual machine on my squeeze
14:29 Renku just to run a decent OS x)
14:29 Renku because i have to stay on squeeze for the installed applications (you know how it is, some airports still are on Windows 3.1 <3)
14:29 kpease joined #git
14:30 Renku anyway, thanks a lot for all theses informations, i'll check that up later
14:30 Renku see you :)
14:30 Renku left #git
14:31 Darren_ joined #git
14:32 JeroenT joined #git
14:34 User458764 joined #git
14:34 sssilver joined #git
14:35 dranve joined #git
14:36 encod3 joined #git
14:37 Jameser joined #git
14:39 jbeeze joined #git
14:40 bittyx joined #git
14:42 Es0teric joined #git
14:43 leeN joined #git
14:43 qpdb joined #git
14:44 JeroenT joined #git
14:44 matsaman joined #git
14:44 AlHafoudh joined #git
14:45 al-damiri joined #git
14:46 Jameser joined #git
14:51 Learath2 joined #git
14:52 dsdeiz joined #git
14:53 ProbabilityMoon joined #git
14:55 mjanssen left #git
14:56 WakiMiko_ joined #git
14:56 ych joined #git
14:56 GingerGeek joined #git
14:56 GingerGeek joined #git
14:56 grawity joined #git
14:56 fahadash joined #git
14:56 bocaneri joined #git
14:56 JeroenT joined #git
14:57 VagabondAzulien joined #git
14:57 Jameser joined #git
14:58 kedare joined #git
14:58 fmcgeough joined #git
14:59 star_prone joined #git
14:59 emattiza joined #git
15:00 nickthetait joined #git
15:00 ahf joined #git
15:00 kedare joined #git
15:00 stfn joined #git
15:01 nd joined #git
15:01 Zumochi joined #git
15:01 Zumochi joined #git
15:02 JeroenT joined #git
15:04 e14 joined #git
15:04 JeroenT joined #git
15:04 russoneto joined #git
15:04 JeroenT joined #git
15:05 Xethron joined #git
15:05 mischat joined #git
15:05 Tykling joined #git
15:08 corsibu joined #git
15:09 e14 joined #git
15:09 LiENUS joined #git
15:11 crayon joined #git
15:11 spaceone hi
15:11 m0viefreak joined #git
15:11 spaceone i want to cherry-pick all commits from abcdef until the HEAD of a branch
15:12 spaceone how do i specify this revision range?
15:12 aarobc joined #git
15:12 danslo_ left #git
15:12 danslo joined #git
15:12 danslo spaceone: apparently since 1.7.2 you can specify a range
15:13 danslo git cherry-pick A..B
15:14 spaceone danslo: so i can just specify the commits?
15:14 spaceone and is there some dry-run which shows all commits which cannot be cherry-picked without manual merging?
15:14 cdown joined #git
15:15 canton7 no: git doesn't know until it tries whether it can do a cherry-ick
15:16 canton7 although you can easily abort the cherry-pick (git cherry-pick --abort), or undo it after the fact, so there's no harm in just trying
15:16 midgetspy joined #git
15:17 blippyJon joined #git
15:17 spaceone well okay
15:18 gfixler joined #git
15:20 aard_ joined #git
15:20 juvenal joined #git
15:20 subhojit777 joined #git
15:20 GavinMagnus joined #git
15:21 nettoweb joined #git
15:23 chris2 joined #git
15:23 le_melomane joined #git
15:25 sangy joined #git
15:25 pijiu joined #git
15:25 Random832 if it's a lot of commits you might want to do a rebase instead
15:26 nettoweb1 joined #git
15:26 pijiu joined #git
15:26 JamesBaxter joined #git
15:26 GavinMagnus joined #git
15:27 GavinMagnus left #git
15:28 cagedwisdom joined #git
15:29 jasondockers joined #git
15:29 tlaxkit joined #git
15:31 mrpackethead joined #git
15:31 rorro joined #git
15:32 Jameser joined #git
15:33 gareppa joined #git
15:33 kimonorouge joined #git
15:35 danslo joined #git
15:35 robotroll joined #git
15:35 buddyq joined #git
15:36 NimeshNeema joined #git
15:37 ahmed_elgabri joined #git
15:37 pokalyis joined #git
15:39 le_melomane joined #git
15:39 eduardas_m joined #git
15:41 blippyJon left #git
15:43 Lor_ joined #git
15:43 brokensyntax joined #git
15:44 King_Hual joined #git
15:45 mar77i can I commit an empty commit?
15:45 osse mar77i: yes with --allow-empty
15:46 thebloggu joined #git
15:47 Ryanar joined #git
15:49 Dougie187 Quick question, why do people use empty commits?
15:49 Dougie187 For notes and stuff? or is there something else that I'm missing?
15:50 synthroid joined #git
15:50 canton7 it used to be useful for the first-ever commit, as you couldn't rebase the first-ever commit
15:50 canton7 that's changed now, though
15:51 e14 joined #git
15:53 osse I once created an empty commit so I could tag it to bump the version. Poor dependency handling lead to no actual changes in the version controlled code
15:53 osse I could have just made a new tag on the same commit. Can't remember why I didn't do that.
15:53 ochorocho__ joined #git
15:55 netj joined #git
15:56 lindenle joined #git
15:56 barteks2x joined #git
15:58 chandu20 joined #git
16:00 spaceone Random832: well, no i can't rebase because the branches are very different :/
16:00 spaceone but it worked nice
16:02 _cyril_ joined #git
16:02 canton7 spaceone, you can use rebase to grab a load of commits and replay them on top of another branch, like cherry-pick with a range. They do much the same thing
16:02 canton7 this is the rebase syntax which use --onto
16:02 jccn joined #git
16:03 sssilver_ joined #git
16:04 spaceone canton7: ah okay, i will have a look at this
16:04 crayon joined #git
16:05 cyan__ joined #git
16:07 mischat joined #git
16:07 kimonorouge joined #git
16:08 tubaman joined #git
16:08 dreiss joined #git
16:08 Derperperd joined #git
16:09 tyreld joined #git
16:09 Derperperd left #git
16:10 ochorocho__ joined #git
16:10 VagabondAzulien joined #git
16:14 imMute osse: probably code that can't handle two tags on one commit - it might get confused as to which is "latest".
16:15 acetakwas joined #git
16:15 einnjo joined #git
16:16 osse imMute: ah, that might be it. the code in question was perhaps 'git describe' :P
16:17 madewokherd joined #git
16:19 Noldorin joined #git
16:20 buddyq_ joined #git
16:21 elsevero joined #git
16:22 Oebele joined #git
16:22 navidr joined #git
16:22 Es0teric joined #git
16:24 anonymuse joined #git
16:24 DopeMeston joined #git
16:25 Prawnzy joined #git
16:26 durham joined #git
16:27 oskarkv joined #git
16:28 Gsham joined #git
16:28 converge joined #git
16:28 converge joined #git
16:30 sea-gull_ joined #git
16:32 algun joined #git
16:32 MaximusColourum joined #git
16:33 tlaxkit joined #git
16:34 svm_invictvs joined #git
16:36 rnsanchez joined #git
16:36 diogenese joined #git
16:37 aelgabri joined #git
16:39 Walex joined #git
16:39 red82 joined #git
16:39 elsevero_ joined #git
16:40 ResidentBiscuit joined #git
16:41 Gsham joined #git
16:42 hariel joined #git
16:45 elsevero joined #git
16:45 tchan joined #git
16:45 tchan joined #git
16:45 Macaveli joined #git
16:46 jozwior joined #git
16:46 buddyq Git status gives me: nothing added to commit but untracked files present (use "git add" to track)
16:46 buddyq I want to ignore all those files listed.
16:46 barteks2x joined #git
16:46 buddyq what do I need to do?
16:46 Es0teric joined #git
16:47 osse buddyq: add their names to .gitignore
16:47 osse you can also put globs there, like *.jpg
16:47 buddyq they are added to gitignore
16:47 osse wrongly in that case
16:47 buddyq my gitignore ignores everything unless specified
16:48 overlord_tm joined #git
16:49 sophiag joined #git
16:49 buddyq how do I retrieve a remote branch that does not exist on local?
16:50 osse fetch
16:50 buddyq i tried git fetch origin <branch>
16:50 ocbtec joined #git
16:50 sonOfRa Why won't git add -p show me the "s" (split this hunk) option when the hunk is this large: https://gist.github.com/sonOfRa/​ccafcafe5f1955f5570b6daf00c8549f ?
16:51 a_thakur joined #git
16:51 osse sonOfRa: because there's no way to split it. that option is only offered when there are context lines in the middle of two "mini-hunks"
16:51 buddyq result of: git fetch origin payments = * branch            payments   -> FETCH_HEAD
16:51 buddyq now what?
16:52 buddyq afterwards, git branch does not show the brach
16:52 jimi_ joined #git
16:52 osse git branch -r lists remote branches
16:53 buddyq okay...please stick with me while I explain what I did..... ;)
16:53 sophiag hi quick git noob question just to be sure since i've ended up wasting half days after screwing things like this up before... i've been working on a local master i should have surely branched first and now would like to branch it, revert back to a previous commit in one of the branches, clean that one up a bit, push it to the remote repo, then cont
16:53 sophiag inue working on the one i'm working on now as the master so that when i've *finally* debugged these major changes i can push it to remote. does that make sense as an explanation? and then from there i just wan to make sure the commands i have in mind are correct so as to not waste time...
16:54 hahuang65 joined #git
16:55 buddyq My host already had another branch named "develop". I created a branch also called "develop" on my local machine and sent it up to the master/origin (github). I then fetch any changes from github to my website to make site changes. However, now my github has "develop" which I renamed to "payments" and my website I think still thinks "payments" is the old "develop".
16:55 cdown joined #git
16:58 sonOfRa osse: thanks, that makes sense :)
16:59 armyriad joined #git
17:01 osse buddyq: not sure what that has to do with your issue
17:01 Balliad joined #git
17:02 sophiag i guess if i'm to go into what i was thinking it would be: git branch -l target location, then on the NEW branch i would find the commit i want in the log and do git reset xyz, then (this is the part i'm unsure about) push that remote
17:02 osse sophiag: what's your suggested solution?
17:02 osse sophiag: "part 1" I guess: https://raw.githubusercontent.com/Osse/gi​t-stuff/master/git-reset-explanation.txt
17:02 Vampire0 joined #git
17:03 sophiag osse: first of all, do you understand my explanation of the situation?
17:03 sophiag just so we're on the same page
17:03 robotroll joined #git
17:04 osse sophiag: almost. you say "revert back to a previous commit in one of the branches" but which one?
17:04 osse master makes sense to me
17:05 sophiag well, i guess i'm uncertain about the terminiologically
17:05 sophiag so maybe you should tell me the best means of approaching it
17:05 osse buddyq: anyway, git branch -r lists remote brances (the ones that git fetch updates). I suggest you do 'git fetch --all --prune' to make sure all remote stuff is "clean"
17:05 sophiag i'm currently working on a major revamp that's not ready yet
17:06 sophiag and i want to push a minor version based on changes made on a previous commit: the same one i currently have remote
17:06 sophiag so i'm not actually sure which is wise to keep as master in this case
17:06 nitric joined #git
17:06 sophiag (i did say i'm a noob...)
17:08 mischat joined #git
17:08 sophiag oh, actually one thing i said was incorrect
17:08 osse that I can't help you with
17:08 sophiag no?
17:08 osse I don't know anything about your code or how you work
17:08 osse that's what should determine what should be master
17:09 sophiag well, let's say i would prefer to keep the one i'm working on right now master
17:09 sophiag and just branch it in order to push some incremental changes today since the major ones aren't finished yet
17:09 sophiag which also involves reverting that branch to a previous commit, something i do know how to do
17:10 freimatz joined #git
17:10 sophiag the one part where i think it gets sketchy for me is pushing that branch to remote given it's not the master
17:10 sophiag is that something you can help me with?
17:10 osse sophiag: git push theremote thebranch
17:10 sophiag oh, it's fine to do that and won't return fatal?
17:11 osse why should it?
17:11 SteffanW joined #git
17:11 sophiag not sure, i just thought it might
17:11 osse there's only one way to find out
17:12 sophiag true
17:12 osse nothing bad will happen in either case
17:12 sophiag ok, i suppose the other thing is tbh i've also had issues with reattching the head even tho it seems very simple
17:12 sophiag once i revert the commit is there anything i could be doing wrong?
17:12 ehynds joined #git
17:13 sophiag i just look in the log and do git reset xyz
17:13 sophiag with xyz being the full string of the commit in the log
17:14 osse that's not a revert in the git sense
17:14 osse !revert
17:14 gitinfo That's a rather ambiguous question... options: a) make a commit that "undoes" the effects of an earlier commit [man git-revert]; b) discard uncommitted changes in the working tree [git reset --hard]; c) undo committing [git reset --soft HEAD^]; d) restore staged versions of files [git checkout -p]; e) move the current branch to a different point(possibly losing commits)[git reset --hard $COMMIT]?
17:14 sophiag right, i was doing --hard
17:14 Balliad joined #git
17:14 sophiag i think eventually
17:15 sssilver joined #git
17:15 sophiag because i'll want to make commits from that point
17:16 f4 joined #git
17:16 f4 joined #git
17:17 chamunks joined #git
17:17 f4 joined #git
17:17 sophiag was that right or wrong? git reset --hard $COMMIT  if my intention is to lose all future commits, attach the head there, and make future commits there?
17:17 f4 joined #git
17:17 osse Nothing in particular can go wrong unless you've already pushed with branch before, but it sound like you're creating a new one
17:17 f4 joined #git
17:18 osse You are right
17:18 f4 joined #git
17:18 i_want_dinner joined #git
17:19 sophiag ok great
17:19 mrpackethead joined #git
17:19 sophiag i think the problem was simply that i didn't do that initially
17:20 sophiag oh, and idk why i said git branch -l target location
17:21 sophiag that makes no sense now that i'm reading the man page
17:21 osse Correct!
17:21 sophiag it would seem in this case i don't need options
17:21 freekevin joined #git
17:21 sophiag ok, :)
17:21 sophiag just skittish after wasting some hours when a noob you know?
17:22 sophiag better than working on someone else's project and force pushing to delete their work i suppose :p
17:22 rorro joined #git
17:25 sophiag ah! i forgot i need to migrate the git dir too :p luckily that's really easy
17:25 alhariel joined #git
17:26 leehambley joined #git
17:28 cdown_ joined #git
17:30 hariel joined #git
17:34 elsevero joined #git
17:35 davimore__ joined #git
17:35 dpower joined #git
17:35 shpew12 joined #git
17:37 sophiag osse: i'm feeling stupid now...if i create a branch and check it out (didn't realize i could have done that with one command) that doesn't create a new project, right? i feel sketchy working on two branches from one project like i'm likely to make a mistake with the code that way. how do i have it so they're two projects almost as if i cloned it?
17:37 VagabondAzulien joined #git
17:37 chachasmooth joined #git
17:37 bluezone joined #git
17:37 dviola joined #git
17:37 osse sophiag: git doesn't use the term "project".
17:37 osse so I don't understand
17:37 buddyq joined #git
17:38 sophiag repo
17:38 sophiag directory
17:38 osse in that case, no
17:38 sophiag ah you just can't do that
17:38 cdown joined #git
17:38 sophiag i'm super confused as to how it even works if it's one directory
17:38 osse sophiag: everything is contained in the hidden .git dir. git checkout switches out what you see on disk
17:39 sophiag huh
17:39 mikecmpbll joined #git
17:39 sophiag so i had better make sure i check it out between every edit?
17:39 sophiag if i'm going to switch back and forth?
17:40 sophiag (i mean, i'm unlikely to...more likely just edit the one branch today then just be working on the master)
17:40 sophiag i guess let me try it...
17:42 davimore_ joined #git
17:42 thiago joined #git
17:43 sangy joined #git
17:43 sophiag woah...ok what if one of the commits involved moving the git directory?
17:43 Faylite joined #git
17:43 sophiag if i revert back from there given i did git add -A
17:43 osse sophiag: you cannot move that dir
17:43 sophiag ok, so that will stick right?
17:44 osse yes
17:44 sophiag even with the hard reset?
17:44 sophiag ok cool
17:44 sophiag that's what i want
17:44 grayjoc joined #git
17:45 sophiag yikes...moment of truth
17:45 sophiag i checkout the branch
17:45 sophiag got a message it switched
17:46 osse OMG
17:46 sophiag you're probably laughing that i'm nervous
17:46 sophiag haha
17:46 osse i would be if I didn't have a headache
17:46 osse maybe
17:46 sophiag well i can relate
17:46 sophiag have you heard of chronic migraine?
17:46 sophiag i means you have a migraine 24/7
17:46 sophiag i'm waiting to get botox treatments
17:47 sophiag they inject from your face down to your shoulders
17:47 Jimmio92 joined #git
17:47 osse oh
17:47 sophiag right now i'm on a shitload of meds...doesn't make coding too easy you know?
17:47 osse in that case i cannot relate. I'm just feeling slightly under the weather.
17:47 osse but I have had botox, for other reasons
17:47 sophiag i prefer functional stuff so i don't have to manage state in my brain
17:48 Spacew00t joined #git
17:48 sophiag and i should get better with version control too. i bet that would help
17:48 sophiag these minor annoyances will pay off
17:48 osse !video
17:48 gitinfo [!talks] Some good video talks about Git: [yt] http://goo.gl/z72s (Linus Torvalds: History&Concepts); [yt] http://goo.gl/R9H2q (Scott Chacon: Git basics, live examples); http://vimeo.com/35778382 (Randal Schwartz: Git basics, descriptional); http://vimeo.com/46010208 (Jessica Kerr: Git basics, descriptional)
17:48 osse i recommend the second and fourth
17:48 sophiag cool thanks
17:49 sea-gull_ joined #git
17:49 dreiss joined #git
17:49 sophiag personally i enjoyed that email from...was it raymond? to torvalds about how he had to make his code manageable to other people and even himself at later points. lol
17:50 sophiag http://lwn.net/2000/0824/a/esr-sharing.php3
17:50 fees joined #git
17:51 Limes joined #git
17:51 e14 joined #git
17:52 leehambley joined #git
17:52 Sasazuka joined #git
17:52 davimore__ joined #git
17:53 sophiag yeah i screwed something up in the course of this
17:54 sophiag by moving that git dir
17:54 sophiag now the two originally tracked files i didn't move manually are in the wrong directory
17:54 osse how did that happen?!
17:54 tlaxkit joined #git
17:55 sophiag no idea
17:55 sophiag but after the hard reset
17:55 sophiag couldn't think of emjoi for my face rn
17:56 mibr0 joined #git
17:56 freekevin joined #git
17:56 osse why do yolu think it's moved?
17:56 osse that  sounds super weird
17:56 sophiag it's not the right file either...
17:56 armyriad joined #git
17:56 sophiag too far back
17:56 e14 joined #git
17:56 sophiag oh no it's right
17:56 sophiag i just didn't commit in between
17:57 sophiag that's ok
17:57 sophiag it's just...idk how to migrate now that i have to checkout to switch between them?
17:57 sophiag i guess i'll just copy the code somewhere
17:57 sophiag but not sure about this location bullshit
17:58 sophiag i guess i just move them and then since the head is where it's at it's whatever the fuck it is?
17:58 perlpilot joined #git
17:59 osse migrate what?
18:00 sophiag i just mean i need to copy some changes to the code i never commited
18:00 sophiag and it's rough because i didn't realize i'd need to checkout the branches in between
18:00 osse sophiag: you can also use checkout to obtain specific files from a different branch
18:00 sophiag so what i'll do is check out to master and then copy the contents of the file to a new place to work from
18:00 osse and put them into the working directory
18:01 osse no, that's the noob loser way of doing it
18:01 sophiag it really is
18:01 sophiag ok what's the smart headacher way?
18:01 osse git checkout master; git checkout otherbranch -- files you need
18:01 sophiag and it won't overwrite them?
18:02 sophiag or...
18:02 ych joined #git
18:02 osse ok you need to be more specific in what you want to do
18:02 FroMaster joined #git
18:02 sophiag i guess since i didn't commit in between i have a choice of working backwords or forwards
18:02 sophiag given what you're saying...probably backwards is easier
18:03 saedelaere joined #git
18:04 sophiag what i mean by that is i'll just be deleting things pertaining to new features, many of which are in other files that won't be there
18:04 sophiag which will be easy in emacs
18:04 sophiag oh wait...
18:04 sophiag also...
18:04 sophiag the damn reverted branch has a dir that shouldn't be there
18:04 sophiag it's all screwed up
18:04 sophiag wtf
18:05 sophiag maybe i used a wrong commit?
18:05 sophiag i don't think so though...
18:05 ayjay_t joined #git
18:05 sophiag i think it only reverted the files tracked on that date
18:05 sophiag which makes sense
18:06 sophiag i figured it would have deleted new files added since
18:06 sophiag whereas it left them
18:06 fuchstronaut joined #git
18:07 osse reset --hard will also delete files
18:07 sophiag i did use --hard
18:07 dsantiago joined #git
18:07 sophiag so that it would reattach the head
18:08 osse in that case the new files were never committed in the first place
18:08 sophiag not on that date
18:08 WayToDoor joined #git
18:08 K1ck3dByC0ns0l3 joined #git
18:08 sophiag they were committed later
18:08 sophiag and i figured that meant they would be deleted
18:08 osse yes they would be
18:09 chandu20 joined #git
18:09 sophiag anyway, i'm going to pull in the file i need to this branch, delete the old shit...make a commit...check that my master is fine...then breathe...
18:09 sophiag that should be ok i think
18:09 sophiag not ideal
18:10 sophiag but i didn't do this ideally to start with
18:10 laumann joined #git
18:10 sophiag ugh, i moved the main file and now getting an error trying to check out to master
18:10 sophiag what an ass pain
18:11 talayhan joined #git
18:11 sophiag i'll move it back
18:11 VagabondAzulien joined #git
18:12 osse you can only switch branches when you either no uncommitted changes, or the files which do have uncommitted changes do not differ between the branches
18:12 leeN joined #git
18:12 elastix joined #git
18:13 ahr3n joined #git
18:13 btree joined #git
18:13 btree left #git
18:13 sophiag ugh it's apparently telling me the file i'm trying to checkout to the other branch is "not known"
18:14 einnjo joined #git
18:14 sophiag yet i did a commit just before branching today
18:14 sophiag maybe i didn't add all
18:16 sophiag yet i just did and told nothing to commit
18:17 sophiag ah!!!! i just stupidly used a path
18:17 sophiag git doesn't work with paths, right? just tracks files...
18:18 ToxicFrog sophiag: er, what? Files are identified by path.
18:18 ToxicFrog It doesn't track directories, if that's what you mean.
18:19 sophiag well when trying to checkout a file from one branch to another you don't enter the path, right?
18:19 sophiag or at least that seemed to have tripped me up right now
18:21 sophiag or it just didn't work?
18:21 sophiag this is bizarro land
18:21 sophiag this is why i came on this channel to ask and now it is actually hours later!
18:22 sophiag what's going on with my repo?
18:22 chamunks joined #git
18:22 blitheBoudica joined #git
18:22 TomyLobo joined #git
18:22 bunniez joined #git
18:25 sophiag so i moved the git dir and made a commit before branching
18:25 sophiag apparently doing that has made the branch worthless in that none of the changes seems to have an effect
18:26 sophiag it won't checkout files from the master, for example
18:26 Prawnzy joined #git
18:26 gnac joined #git
18:26 sophiag a hard reset kept files that weren't present at that point
18:26 sophiag and it moved filed to a different directory
18:27 sophiag osse: what do you think? i may just delete this branch...
18:28 sophiag but is there a way to examine it and figure out why it's doing this when i create new ones?
18:28 gnac I've got a setup where I'm not getting updates from the remote repository. I originally cloned a remote repository to a local bare  ( git clone --bare //remote local.git )
18:29 gnac Then I created a local working repository ( git clone local.git working )
18:29 gnac Now, after I fetch into the bare from the original remote, my working directoy never updates using git pull. It always thinks it up to date.
18:31 ToxicFrog sophiag: to check out a file from another commit, it's "git checkout <commit id> -- <path/to/file>"
18:31 ToxicFrog E.g. 'git checkout feature-branch -- test/ui/interact.lua'
18:32 nickthetait joined #git
18:32 sophiag toxicfrog: ah, that explains it...
18:33 sophiag also probably explains why what i just did was checkout my entire master to my other branch....yikes
18:33 ToxicFrog You can also checkout a directory, which checks out everything inside it (git checkout feature-branch -- test/ui/)
18:34 sophiag yeah...can i undo that somehow cuz i fucked up bad...
18:34 ToxicFrog What exactly did you do that needs undoing?
18:35 emattiza joined #git
18:35 ToxicFrog It sounds like you did something like "git checkout feature-branch; git checkout master -- ." to switch branches and then overwrite your working directory with the contents of master?
18:35 sophiag i did what you just said: checked out my entire master to another branch
18:36 sophiag yup
18:36 sophiag you typed faster
18:36 t0by joined #git
18:36 ToxicFrog Ok, so you just want to roll back the working directory to the latest commit on your current branch?
18:36 ToxicFrog git reset --hard; git clean -f
18:37 User458764 joined #git
18:37 sophiag ah well i knew how to do a hard reset
18:37 PioneerAxon joined #git
18:37 ToxicFrog This will reset the index and working tree to match the latest commit, and then delete all untracked unignored files
18:37 sophiag just never committed the checkout
18:37 ToxicFrog Use 'clean -fx' to delete ignored files as well
18:37 User458764 joined #git
18:37 ToxicFrog Since you never committed the checkout, this should be sufficient
18:37 sophiag ok let's try
18:38 jbitdrop joined #git
18:38 ToxicFrog Note that this will nuke all uncommitted changes
18:38 sophiag i had some super weird behavior just creating the working branch before you came into this convo too
18:38 sophiag i'm fine with that
18:38 sophiag was ready to nuke the whole branch
18:38 Achylles joined #git
18:39 ToxicFrog (the reason for the "git clean" is that if you have files in the working tree that aren't tracked in the current branch, "git reset --hard" won't delete them for you)
18:39 sophiag right
18:40 sophiag now one thing that was super weird and i think i have to do the shitty manual way is when i created the branch from the last commit it kept files that were untracked then and MOVED the tracked ones up to where i moved the git dir
18:41 sophiag i'm not sure if that explanation actually makes sense
18:41 sophiag i have the git dir two dirs down
18:41 sophiag moved it up and commited
18:42 sophiag THEN made the working branch
18:42 sophiag and did a hard reset back a bit
18:43 sophiag two weird things happened: 1. it moved two tracked files i didn't move manually up to where the git dir was (this didn't happen in the master). 2. it kept all the files added in the master since the commit i reset it to instead of deleting them
18:44 sophiag can any of that be explained? i'm ready to just fix it all by hand then do add -A and commit from there
18:45 blackwind_123 joined #git
18:45 sophiag and not sure if i should handle this BEFORE checking out the file i need to edit, commit, and FINALLY put to remote :p
18:47 ToxicFrog What do you mean by "moved the git dir up"?
18:48 ToxicFrog You moved the .git directory, or the directory containing it?
18:48 bunniez hello, i have a very simple question... i would like to view the changes in my modified, unstaged files. i've been googling but all the results seem to be talking about using diff to view the changes in STAGED files. how do i view changes in modified but not yet staged files? trying to figure out if this old old work needs to be scrapped. thx very much in advance.
18:48 disi bunniez: git diff will view unstaged changes
18:48 disi bunniez: git diff --cached will view staged changes
18:48 ToxicFrog bunniez: "git diff" by default diffs staged vs. unstaged. "git diff --cached" diffs HEAD vs staged. If you want HEAD vs unstaged, I don't recally how to do that offhand.
18:49 ToxicFrog But it sounds like staged vs. unstaged is what you want anyways?
18:49 ToxicFrog sophiag: see question about moving git dir
18:49 sophiag i moved .git
18:49 bunniez no, i don't have anything staged. i want to view the changes between modified, unstaged files, and their versions from the last commit.
18:49 sophiag which i've done before
18:49 bunniez (coming back to an old project after a long time away)
18:49 ToxicFrog sophiag: so, git tracks all paths relative to the .git dir, not relative to /
18:50 ToxicFrog sophiag: if you moved .git but not the rest of the repo, from git's perspective you just moved everything *else* and have a lot of delete/create or move/rename changes pending
18:50 sophiag well i've done this before and been fine, just never had to branch
18:50 sophiag ah damn
18:50 zivester joined #git
18:51 hck89 joined #git
18:51 sophiag wait...that makes sense. i moved it relative to .git. then i commited.
18:51 ToxicFrog (which should be clearly evident when you run "git status" or look at "git gui")
18:52 manuels joined #git
18:52 sophiag i'm not sure that what you're saying explains this
18:52 ToxicFrog I think a transcript of what you did would be really helpful here :/
18:52 sophiag unless i forgot to do something
18:52 sophiag git status from the master?
18:53 sophiag i've been asking how to get something like that
18:53 ToxicFrog My best guess is that you moved .git around, committed the resulting file moves, then created a new branch from an earlier commit that predates the move, and when you checked out the new branch it undid the move (because from that branch's perspective it hasn't happened yet)
18:53 ToxicFrog But that's just a guess
18:53 star_prone joined #git
18:53 ToxicFrog What do you mean by "something like that"?
18:53 ToxicFrog By "a transcript", I meant "a lot of all the git commands you ran to get here, starting from the last state that 'made sense', ideally accompanied by their output"
18:53 ToxicFrog *a log of
18:54 sophiag yes, a transcript....i keep asking how do i diagnose all of this
18:54 sophiag but git log just gets commits
18:54 emptynine left #git
18:54 sophiag you're saying status from the master should be what i'm looking for?
18:55 bmahe joined #git
18:55 bunniez left #git
18:55 VagabondAzulien joined #git
18:55 sophiag and what i'm saying is that the orginally tracked files are NOT in their original place. they've moved to the same place as .git. that is the weird behavior
18:55 ToxicFrog That question doesn't make any sense.
18:56 tobiasvl sophiag: ToxicFrog means that they would like to see the commands you ran in your terminal
18:56 tobiasvl copy and paste a transcript of your terminal, or your bash history or whatever if you've closed terminal windows between
18:56 sophiag +tobiasvl: i've said about five times i'd like to know how to get that history
18:57 sophiag a transcript of my terminal would involve man pages and such
18:57 tobiasvl sophiag: just use your mouse and copy text from your terminal window?
18:57 tobiasvl well just pick the relevant stuff obviously
18:57 tobiasvl or look in your ~/.bash_history
18:57 offtopik joined #git
18:57 sophiag are you kidding? there's no history of changes to git other than commits?
18:58 ToxicFrog sophiag: ...no? Why would there be?
18:58 ToxicFrog The whole point is to record commits. Intermediate states of the index, runs of "git help", etc are just noise.
18:59 sophiag i assumed because you kept intimating that there was a way to retrieve something to that effect
18:59 ochorocho__ joined #git
18:59 ToxicFrog No, I just wanted the history of commands you entered
18:59 tobiasvl bash already has that functionality
18:59 sophiag ok, let's just step back a bit
18:59 sophiag i don't need to figure this out
18:59 LukeLR joined #git
18:59 sophiag i didn't do this right from the beginning
18:59 sophiag i'm backtracking now for that reason
19:00 sophiag so why don't i just checkout the file i need since you've been so helpful with that
19:00 davimore joined #git
19:00 sophiag and then manually fix those issues and commit from there
19:00 sophiag then in the future i know to just create my working branch...uh...when i actually start working
19:01 sophiag and generally educate myself more when i'm not in the middle of a project
19:01 sophiag because unless it keeps moving files around like that i think i'm cool
19:02 dermoth joined #git
19:03 alhariel joined #git
19:04 sophiag so to refresh... to checkout a file from master to working it's git checkout <working_branch> <commit> <filename>
19:04 sophiag right?
19:05 ProjektGopher joined #git
19:06 tobiasvl no
19:06 tobiasvl what's a working branch?
19:06 tobiasvl to checkout a file from master it's git checkout master filename
19:06 elsevero joined #git
19:07 sophiag oh, from inside the working branch
19:07 Groscheri joined #git
19:07 sophiag i switched to master
19:07 sophiag that was the mistake
19:07 mrpackethead joined #git
19:07 sophiag gotcha
19:07 Prawnzy joined #git
19:07 bmahe joined #git
19:07 fuchstronaut joined #git
19:07 sophiag i think i was told another way of doing it as well tho
19:07 sophiag but this is the way i actually want
19:08 converge joined #git
19:08 converge joined #git
19:11 Rodya_ joined #git
19:15 Rodya_ joined #git
19:16 sophiag ok...seemed everything was alright...switched back to the master to check (nervous noob) before editing the working branch and there are directories missing
19:17 afuentes joined #git
19:17 ToxicFrog sophiag: how (what exact command) did you use to switch, and in what way are they missing?
19:18 ToxicFrog Present in working but not in master? Or you remember commiting them to master earlier but they aren't there now?
19:18 sophiag i'm been using add -A
19:18 nilg joined #git
19:18 sophiag so yes i commited them today to master
19:19 sophiag shouldn't be the hugest deal to recreate, but still...
19:19 sophiag i only see that one commit in the log
19:20 sophiag it's that and one from way far back...the one i did the hard reset from in the working branch
19:20 ToxicFrog Are you sure you reset the right branch?
19:20 ToxicFrog And, again, what command did you use to switch branches?
19:20 sophiag git checkout maser
19:20 sophiag *master
19:20 ToxicFrog What does "git reflog master" say?
19:21 ToxicFrog That's a log of all operations that changed what the "latest commit" on master is.
19:21 ToxicFrog So it includes not just commits, but also resets, rebases, commit --amend, and suchlike.
19:21 kimonorouge joined #git
19:21 sophiag matches regular log
19:22 ToxicFrog If you accidentally threw away a bunch of commits by using "git reset --hard <old commit>", for example, that'll be in the reflog, along with the commit id of what the branch head was before you did that, so you can get it back.
19:22 sophiag it's a whole lot cleaner than the log tho so useful command
19:22 sophiag but no, same
19:22 osse reflog cleaner than log? o_O
19:22 osse some fuckery is going on
19:23 ToxicFrog Yeah that's super weird and unexpected
19:23 sophiag no i just mean appearance
19:23 sophiag disregard comment
19:23 sophiag it matches it literally
19:23 ToxicFrog Are you sure you made the commits you're thinking of on master, then?
19:23 nwkj86 joined #git
19:23 ToxicFrog Perhaps "git reflog" (not "git reflog master") can help here?
19:24 finalbeta joined #git
19:24 Gsham joined #git
19:24 conrad_ joined #git
19:25 sophiag ok, what am i supposed to be seeing here?
19:25 osse sophiag: that can't be the case. for one the default formats are different. git log prints a full hash and author, reflog doesn't
19:25 sophiag actually wait same
19:25 osse if it matches *literally* then you're either always running log or always running reflog
19:26 sophiag reflog just shows commits for both branches
19:26 sophiag so i made one commit in the the working branch and i see that too
19:26 sophiag that's all
19:26 stomplee joined #git
19:27 Derperperd joined #git
19:27 sophiag i'm saying there aren't additional commits that would explain any directories deleted
19:28 shpew12 joined #git
19:28 sophiag not that it looks the same
19:28 sophiag i don't have my email address in reflog for example
19:29 irqq joined #git
19:29 ToxicFrog sophiag: reflog should also list any resets or similar that you did
19:30 ToxicFrog If you made commits in that repo earlier today that are no longer showing up in "git log", they should still show up in "git reflog"
19:30 sophiag there's only one reset, which was in the working branch
19:30 Es0teric joined #git
19:31 sophiag oh, you're saying i should be able to just reset back?
19:31 osse git log --all -- path/to/missing/dir     this will list all commits touching that dir (creation, change, removal)
19:31 sophiag with the short hash?
19:31 ochorocho__ joined #git
19:31 ki0 joined #git
19:31 ToxicFrog sophiag: and the commit(s) on master that you remember making earlier today don't show up in "git reflog" at all?
19:31 sophiag no it does
19:32 sophiag it was in the log too
19:32 sophiag everything is here
19:32 ToxicFrog ...why didn't you say so when I asked about that earlier?
19:32 sophiag i didn't see you ask. i'm sorry
19:32 ToxicFrog Ok, so, find the abbreviated commit sha in the reflog for the commit that should be the most recent one on master
19:32 ToxicFrog And then:
19:32 ToxicFrog git checkout master
19:32 ToxicFrog git reset --hard <commit id>
19:33 eihabarabia joined #git
19:33 sophiag ok, and no effect on the other branch. i mean, obvious question...but you can see why i'm asking stupid questions now yeah?
19:34 ToxicFrog (I didn't ask for that as such, but I did ask what the reflog looked like, and you said that it "matched regular log", which I took to mean that the commits you wanted to recover didn't appear in either the log or the reflog)
19:34 ToxicFrog And yeah, it won't have any effect on your other branches, it'll only move master.
19:34 sophiag i just figured it was meaningless if it looked like the log was all
19:35 sophiag anyway, let's do this
19:35 de-vri-es joined #git
19:36 sophiag didn't do it
19:36 sophiag i may have to go back one furhter
19:36 sophiag that commit is surely the fucking problem
19:36 sophiag explanation for all this weird behavior
19:36 codepython777 joined #git
19:37 sophiag but i've manually moved git dirs before! that's what i'm told you do!
19:37 sophiag wait no
19:37 sophiag i can't
19:37 sophiag i can't move it past that
19:37 sophiag i didn't commit in between
19:37 sophiag it's this or nothing at this poitn
19:37 sophiag ah shit
19:38 sophiag i have a backup
19:38 sophiag when version control fails :(
19:39 sophiag i'll manually fix it with the backup and commit now
19:39 sophiag fucking fuck
19:39 sophiag why would moving the git dir do this?
19:39 ToxicFrog I'm really surprised that someone told you move the .git dir around manually instead of moving the files *in* the git dir to where you wanted them
19:39 sophiag wait, that's what you normally do?
19:39 osse i have a sneaky suspicion that vigilant use of 'git status' will be good medicine in the future
19:40 ToxicFrog It shouldn't, but I'm kind of flailing in the dark here because I still have only the vaguest of ideas of what you actually did and what state your repo is in
19:40 osse git itself will never move .git
19:40 sophiag well ideally i init in the right place to start
19:40 sophiag and just don't be a fuckup
19:40 ToxicFrog ...yes? If I want to move files in my git repo into a subdirectory, I mkdir the subdirectory and then "git mv" the files
19:40 sophiag osse: git status please?
19:40 ToxicFrog I don't move the .git directory around leaving the files where they are, that's *really* easy to get wrong.
19:41 ToxicFrog (possibly they thought you were asking about moving the directory that *contains* the .git directory, which is entirely safe)
19:41 osse sophiag: huh?
19:41 vdl joined #git
19:41 sophiag ToxicFrog: this was much more stupid.. i had only wanted to post a couple files to github and not the project file so the reverse order of what you're saying
19:42 sophiag osse: well i just used git status...it tells you if your branches match slash if they match remote?
19:42 ToxicFrog what?
19:43 ToxicFrog No, git status gives you the current status of your index/worktree relative to your last commit
19:43 ToxicFrog I.e. it tells you what branch you're working on, which files have changes staged for commit, which files have changes not staged for commit, and which files are present but untracked
19:43 osse it may also tell you how your branch compare to a certain remote branch (based on config)
19:43 ToxicFrog It also tells you if you're in the middle of some longer operation (rebase, conflicted merge, bisect, etc) and gives you some help text.
19:44 gfixler joined #git
19:44 ToxicFrog And if your branch is based on a remote branch it'll also tell you how many commits ahead/behind you are (as of the last time you fetched).
19:45 Guest31014 joined #git
19:45 sophiag right so i 'm one commit behind rn and the working dir is clean
19:46 sophiag (working dir meaning nothing to commit, not the branch i created)
19:46 finalbeta joined #git
19:46 ToxicFrog Anyways, osse's recommendation wasn't "what is the output of git status right now", it was "in the future, whenever you do anything confusing and/or complicated, use "git status" early and often to orient yourself"
19:47 osse ToxicFrog: you know me so well <3
19:47 dodobrain joined #git
19:47 sophiag ah ok
19:47 VagabondAzulien joined #git
19:47 sophiag but tbh i'm not sure that would have helped here
19:47 sophiag i guess the answer is do not move the .git
19:47 sophiag i've heard conflicting info on that
19:48 sophiag and also had it work fine often
19:48 shpew12 joined #git
19:49 fscala joined #git
19:49 osse git status would have told you whether some files you thought was committed weren't, for example
19:50 osse it would tell you whether you have changes you've forgotten to commit
19:50 akanksha__ joined #git
19:50 osse let me unconflict it for you: don't move .git unless you move the repository as a whole
19:50 cyan__ joined #git
19:50 gkatsev if you run `git commit --amend` does the original authorship still exist or does it change to be me?
19:50 ToxicFrog sophiag: you can move the *entire* thing -- the directory containing .git and the entire work tree -- around as much as you like, and it'll work fine, and git doesn't care
19:50 gkatsev or does that only happen with `--reset-author`?
19:51 ToxicFrog sophiag: if you move *just* .git and leave the worktree alone, this is equivalent to moving the entire worktree somewhere without touching .git, and will, at best, result in git expecting a commit that moves/renames/deletes/adds *everything*.
19:51 ToxicFrog Sometimes that's actually want you want to do, but usually not.
19:52 cagedwisdom joined #git
19:52 ToxicFrog For uses cases like "I only want to publish this subdirectory to external source control", there's tools specifically for that (git subtree, git filter-branch)
19:52 gkatsev ok, just verified the author is intact
19:52 gkatsev thanks
19:52 kimonorouge joined #git
19:54 leehambley when calling something like "$ git cat-file -t 423a59cecffb92512ce6da66a26176b2e1dc4da3" and getting back the word "commit", does Git infer the type by the attributes in the beginning of the file, e.g if it has a line like `parent <sha1>`, then it must be a commit ?
19:54 tax joined #git
19:54 ochorocho__ joined #git
19:55 ToxicFrog leehambley: all git objects (including blobs) start with a header informing git of the content type
19:55 _ikke_ leehambley: Each object contains a header with the type and the length of the object
19:55 leehambley except `blob`, right ?
19:55 leehambley ohh, right, sorry - *even* blob, ok wow
19:55 leehambley didn't know that
19:55 ToxicFrog If you make a blob that looks like a git commit, commit it, and then use cat-file -t on it, you'll get "blob", because the header identifies it as such -- it's a blob with the contents of a git commit.
19:56 cdg joined #git
19:57 sophiag ToxicFrog: the process i've been doing is to move the .git dir and then git add -A and do a commit so everything is in its new relative place. i'm not sure if that's what you're describing
19:57 sophiag also...i just cleaned up everything in the master from my backup...
19:57 sophiag guess what?
19:57 sophiag now they're present as untracked files in the working branch i created
19:58 sophiag wtf
19:58 leehambley with `git hash-object` and `git cat-file` I can't make it print the `blob` header, but I'll keep experimenting, thansk
19:58 sophiag why can't i have them just be in the master?
19:58 xall_ joined #git
19:58 ToxicFrog sophiag: did you actually commit them in master?
19:58 ToxicFrog If not, they'll "float" into other branches until you actually commit them
19:58 ToxicFrog !float
19:58 gitinfo 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.
19:58 sea-gull joined #git
19:59 ToxicFrog Again, use "git status" before and after "git commit" so that you know exactly what you're about to commit and what your state is afterwards.
19:59 leehambley (you're right ToxicFrog , thanks - sorry for doubting you, they're have a `blob <content length>\0` header
19:59 ToxicFrog And now, I have to run. Good luck!
20:00 nedbat joined #git
20:00 eblake joined #git
20:00 eblake question on git send-email
20:01 eblake I was recently forced to switch which outgoing SMTP server I use, due to tighter SPF rules on the domain I wanted as my From: address
20:02 eblake I used to be able to spoof my from address and use my ISP's SMTP, where I had to have .gitconfig containing sendemail.smtpuser
20:02 eblake but now to send with my work's domain, the mail HAS to go through work's servers to avoid an SPF reject
20:02 eblake but work's servers are behind a VPN, where it is assumed that you can't even connect to the server unless you have already authenticated VPN,
20:03 eblake and therefore work's SMTP server does not recognize the AUTH command,
20:03 eblake leading to this cryptic failure:
20:03 eblake Command unknown: 'AUTH' at /usr/libexec/git-core/git-send-email line 1390.
20:03 eblake until I figured out that I had to delete sendemail.smtpuser from my .gitconfig
20:03 eblake but WHY does send-email hang up so drastically?
20:03 eblake couldn't it easily be taught that if the SMTP server it is talking to rejects the AUTH command, to just keep trying on the grounds that authorization is not required?
20:04 eblake and why is there no easy way from the command line to override the .gitconfig to turn off smtpuser?
20:05 eblake you can easily _change_ the username, with --smtp-user=, but there doesn't seem to be an easy way to NOT try authentication if .gitconfig has already turned on smtpuser
20:05 sophiag ToxicFrog: yes, I commited them in the master
20:06 Lor_ joined #git
20:06 sophiag sorry for the delay...multitasking
20:06 Sasazuka joined #git
20:07 ki0 joined #git
20:08 dan2k3k4 joined #git
20:09 fuchstronaut joined #git
20:09 Starky joined #git
20:10 andlabs joined #git
20:10 andlabs Should a git add called by a pre-commit hook modify the current commit or the next one?
20:10 andlabs If the latter, how do I change the current commit in a pre-commit hook?
20:11 ytti left #git
20:12 Seveas andlabs: you don't.
20:12 andlabs then what should I do instead
20:13 satifant joined #git
20:13 Seveas depends on what you're trying to achieve
20:13 andlabs have some files that should be tracked by version control be regenerated on each commit
20:14 Seveas no, you don't.
20:14 Seveas those files don't belong in version control
20:14 Seveas eblake: is Authen::SASL installed?
20:14 LabMonkey joined #git
20:14 buddyq joined #git
20:16 UTAN_dev_ joined #git
20:16 sophiag ToxicFrog: what i'm saying is git satus is now telling me i must commit all those files i already committed to master to the working branch as well and that makes no sense to me...
20:17 sophiag it's as if i should have just cloned it instead  :/
20:17 mischat joined #git
20:17 sophiag ah wait
20:17 sophiag wait
20:17 sophiag not every single file
20:18 sophiag could it be it knows they're dependencies of the one file i checked out from the master?
20:18 krs93 joined #git
20:23 sophiag anyway...i HAVE wasted half a day now despite seeking the aid of git experts before just going ahead with a simple branch and reset :p
20:23 lbdr joined #git
20:24 sophiag doesn't seem to be any solution except to go ahead and do my work instead of wasting time, then i guess since i'm the sole dev i can delete the garbage deps after i push
20:24 sophiag wtf tho
20:24 chris123456 joined #git
20:25 eblake Seveas: on a Fedora system, perl-Authen::SASL is not installed
20:25 lbdr left #git
20:25 eblake could it be a broken package dependency, where git-email-2.7.4-3.fc24.noarch should be dragging it in?
20:26 * eblake goes and retries with the package installed, to see if behavior changes...
20:26 rorro joined #git
20:27 eblake hmm, that's not how the package is named :(
20:27 jstimm joined #git
20:28 BSaboia joined #git
20:28 eblake oh, it's already installed, but spelled perl-Authen-SASL-2.16-7.fc24.noarch
20:29 WhiskerBiscuit joined #git
20:29 nettoweb1 Hello guys, how I bring the changes made in a old commit, to my working directory?
20:30 BigDru joined #git
20:30 nettoweb1 Because I made a commit, then reverted, and after months I decide to bring that first commit changes so I can change in my working directory
20:30 _ikke_ nettoweb1: You could revert the revert
20:31 eblake nettoweb1: try 'git cherry-pick oldid'
20:31 _ikke_ That's also possible
20:31 nettoweb1 lemme try
20:31 rivarun nettoweb1: i think you're looking for `git revert -n  <hash>`
20:31 andlabs thanks
20:31 andlabs left #git
20:31 Seveas eblake: hmm, actually looking at the error/code now instead of the knee-jerk response (usually auth issues are caused by that package missing), why would git ignore auth failures when you configure it to authenticate? That seems wrong to me.
20:31 skunkz joined #git
20:32 nettoweb1 thanks eblake and everybody
20:32 eblake Seveas: the problem was that I _used_ to have to authenticate, but now I don't, because I switched SMTP servers (no thanks to SPF reject rules)
20:32 nettoweb1 that was the command
20:32 Seveas eblake: so you hould remove the authentication config
20:32 eblake but it took me a lot of head-scratching to figure out why send-email was failing, because I didn't realize that smtpuser HAS to be omitted with the new server
20:33 eblake yes, I got it working when I removed smtpuser, but my question is why can't git just do the right thing, and TRY sending the email instead of giving up when the server rejects AUTH as an unknown command
20:33 eblake and why is there no easy way to undo smtpuser settings from the command line to override .gitconfig settings, for one-off testing
20:33 Seveas that would be the wrong thing
20:34 Seveas you're telling it to use authentication, continuing when authentication fails would most definitely not be 'the right thing' then
20:34 johnmilton joined #git
20:34 eblake but I didn't KNOW I was telling it to use authentication
20:34 Endarked joined #git
20:34 eblake the .gitconfig setting was entered years ago
20:34 eblake I was trying to play with new git send-email --smtp... options to talk to my new server
20:34 Seveas you can override .gitconfig settings with git -c
20:35 eblake but until I figured out that I _had_ to erase sendemail.smtpuser from .gitconfig, the old server settings were interfering
20:35 eblake and it seems like authentication to a server that lacks AUTH support is trivial: assume that you are authenticated, rather than erroring out
20:36 Seveas nope. Assume you're not talking to who you're talking to and error out.
20:36 eblake the error message could at least be made less cryptic
20:36 eblake even a 'server doesn't understand AUTH; are you sure you need smtpuser?' would have saved me time
20:37 Seveas unfortunately, parsing error messages from smtp servers is quite error-prone. There's no standard format/error
20:37 a3Dman joined #git
20:38 Seveas so it's quite understandable that git doesn't even try
20:38 eblake so SMTP doesn't have a reliable different error code for auth failed vs. auth not understood?
20:38 Seveas correct
20:38 jozwior joined #git
20:38 rasto joined #git
20:39 theskillwithin joined #git
20:41 emattiza joined #git
20:42 krs93_ joined #git
20:44 kimonorouge joined #git
20:45 zulutango joined #git
20:48 netj joined #git
20:48 wrouesnel joined #git
20:50 joki joined #git
20:51 emattiza joined #git
20:52 jasondockers joined #git
20:53 Guest15796 joined #git
20:53 elisa87 please suggest fixes http://pastebin.com/9HB2TDYW
20:55 mrpackethead joined #git
21:00 aidalgol joined #git
21:03 Prawnzy joined #git
21:04 HoierM joined #git
21:05 urda joined #git
21:05 tonyj joined #git
21:05 osse elisa87: sounds very weird that master and origin/master have no common commits
21:05 elisa87 osse: can you please help me figure out a solution?
21:06 osse git pull tells you the solution
21:06 elisa87 I did git pull but still got error osse
21:06 osse i know, I saw the paste
21:06 osse see lines 46 and 49 specifically
21:07 osse tho lines 51 and 53 is a more permanent solution you probably want (and probably is want you expected)
21:09 HiddenPanda joined #git
21:09 fuchstronaut joined #git
21:10 gnac I can't get updates from my remote repository. I created a bare repo via git clone -bar //remote local, but now, after I have recently pushed files to the remote from another workspace, git fetch doesn't get the latest.
21:11 gnac after a fetch, git log --all --branches doesn't show the recent commits.
21:12 gnac I have verified that they are on the remote repository with git log.
21:12 aarobc joined #git
21:14 jozwior joined #git
21:15 kimonorouge joined #git
21:16 Achylles joined #git
21:17 HoierM joined #git
21:18 converge joined #git
21:18 codepython777 joined #git
21:18 amway joined #git
21:19 DuckHuntDog joined #git
21:21 jstimm joined #git
21:21 fscala joined #git
21:21 grayjoc joined #git
21:26 fuchstronaut joined #git
21:27 jbeeze joined #git
21:29 Gonzo89_ joined #git
21:32 sophiag joined #git
21:33 Xethron joined #git
21:36 diogenese joined #git
21:37 red82 joined #git
21:39 e14 joined #git
21:41 tax joined #git
21:42 anuxivm joined #git
21:43 polyzen joined #git
21:43 a3Dman_ joined #git
21:47 diogenese joined #git
21:49 nilg joined #git
21:49 col joined #git
21:50 col joined #git
21:54 strugee joined #git
21:56 ki0 joined #git
21:57 daynaskully joined #git
21:58 e14 joined #git
22:00 fenris_kcf joined #git
22:01 kadoban joined #git
22:06 e14 joined #git
22:06 nilg` joined #git
22:06 fenris_kcf joined #git
22:06 moei joined #git
22:06 fenris_kcf * "for repo in $(ls) ; do pushd ${repo} && git status && popd ; done"
22:07 ojzc joined #git
22:07 sssilver joined #git
22:08 akurniawan joined #git
22:08 red82 joined #git
22:09 expatriot joined #git
22:12 fenris_kcf can this nevertheless be covered with git submodules?
22:12 expatriot Hello, lots of users but quiet.  I am going to ask a question.  Are you lurkers listening?
22:13 sbeller fenris_kcf: "git submodule foreach expression" or git submodule status/summary
22:13 cyan__ joined #git
22:13 sbeller expatriot: ask, don't ask to ask
22:14 expatriot There are two repositories with the same name and one is the fork of the other or they are distinct forks from a single one.  I want to create a pull-request on both forks.  I have committed my changes on one and I want to put a fork request on the other.
22:14 Vinnie_win joined #git
22:15 sarbs joined #git
22:15 fenris_kcf sbeller: ah, interesting. should this also be used to pull all the repos (git submodule foreach git pull)?
22:15 grawity joined #git
22:15 __name__ joined #git
22:15 xehbit joined #git
22:16 nettoweb joined #git
22:16 gniourf joined #git
22:16 sbeller fenris_kcf: newer versions of Git can do "git pull --recurse-submodules={all/check/none}" IIRC, see the man page
22:17 elsevero joined #git
22:17 expatriot Basically I want to use 'hg export' but for git.
22:17 cdown joined #git
22:17 sssilver joined #git
22:17 sbeller expatriot: If they are distinctively similar i.e. forked a long time ago, maybe make a patch and apply that on the other one locally and do a pull request as usual
22:18 expatriot Yes, okay, I want to export a patch from commits I have already made.
22:18 sbeller hg export rev is like git show rev accordign to https://www.mercurial-scm.org/wiki/GitConcepts
22:19 sarbs am i constantly dropping and rejoining?
22:19 gtrmtx joined #git
22:19 durham sbeller: probably closer to git format-patch, since they're both meant to produce patches for emailing and applying
22:19 gtrmtx can you set gogs up to automatically add new commits to your repository?
22:20 gtrmtx or set git to automatically push to gogs?
22:20 gniourf joined #git
22:21 xehbit joined #git
22:21 danslo joined #git
22:21 fenris_kcf sbeller: my version indeed has that option, but on using "git pull --recurse-submodules=yes" it says that i didn't specify i remote repository
22:21 fenris_kcf * a remote ...
22:23 fenris_kcf "git submodule update --recursive" seems to work
22:26 sarbs joined #git
22:27 red82 joined #git
22:29 fuchstronaut joined #git
22:30 expatriot Thanks sbeller, this should make things easier.
22:31 star_prone joined #git
22:32 jozwior joined #git
22:32 aarobc joined #git
22:32 Muzer joined #git
22:32 Dougie187 left #git
22:34 hmsimha joined #git
22:36 hmsimha If I have a filename and its md5sum, is there a git command that can help me find out what commit introduced the version of that file which generates that md5sum?
22:37 Starkythefox joined #git
22:39 leehambley the separator, in git objects between headers and content, is it simply \n\n ? … if so how does the parser know if the *body* includes \n\n somewhere ?
22:42 emattiza joined #git
22:42 fuchstronaut joined #git
22:42 fuchstronaut joined #git
22:44 railssmith joined #git
22:46 matthiaskrgr hi, is there a way to make git count the actual amount of needed objects (like before git gc) but without actually compressing anything after that?
22:46 matthiaskrgr objects == internal deltas,  or whatever these are
22:46 sangy joined #git
22:47 Achylles joined #git
22:49 gnac I can't get updates from my remote repository. I created a bare repo via git clone -bare //remote local, but now, after I have recently pushed files to the remote from another workspace, git fetch doesn't get the latest.
22:57 Vampire0 gnac, !dont
22:57 gitinfo gnac: [!dont_summarize] Please do not summarize what you did; !pastebin _exactly_ what commands you ran, what the output was, and why you think that this is wrong. Otherwise you may hide the real issue. See also !obfuscate and !repro
22:58 e14 joined #git
23:00 bmahe joined #git
23:02 alex1a joined #git
23:04 Prawnzy joined #git
23:05 Sasazuka joined #git
23:06 mischat joined #git
23:06 tax joined #git
23:08 gnac Vampire0: well thanks at least for responding. I'd given more information in past posts that went by quietly. Here's the pastebin http://paste.opensuse.org/87967914, but its not really that complicated.
23:10 pijiu joined #git
23:11 gnac I can't paste all of the output of git log due to proprietary issues, and if I run git fetch again, it won't do exactly what it did the first time since the commit on the remote.
23:12 Vampire0 gnac, !snip
23:12 gitinfo gnac: [!obfuscate] Please do not elide, obfuscate, or otherwise hide information from !repro pastes. This invariably leads to pastes that hide the actual issue or (worse) are not consistent, sending the investigation down false paths. See !secret.
23:12 e14 joined #git
23:13 raphink joined #git
23:13 phroa gnac: try `git pull`
23:14 Jameser joined #git
23:14 gnac Vampire0: what do you want me to do? I can't recreate things exactly as they were the first time I ran the commands.
23:14 gnac phroa: this is a bare repo, only git fetch is meaningful.
23:15 phroa oh, I read that as the remote was the bare one
23:17 gnac phroa: I created a bare repo from the remote because that is the only lcoation exposed to the remote on the network. I then use that bare repo as a remote inside a different network that has access to the bare repo, but not the original remote.
23:17 gnac However, getting updates from the bare repo to the internal network is not an issue, just getting updates from the remote remote.
23:17 einnjo joined #git
23:18 dodobrain joined #git
23:19 ToxicFrog Kainxxnx”©%∆¶∆][. MNsyu:
23:20 DoctorMonocular joined #git
23:20 DoctorMonocular hey everyone, I'm trying to run a post-receive hook and it's apparently failing
23:21 DoctorMonocular is there a place I see the error output, or a way to output that manually?
23:22 re1 joined #git
23:22 Jameser joined #git
23:25 tax joined #git
23:26 codepython777 joined #git
23:27 tax joined #git
23:27 xall_ joined #git
23:27 tax joined #git
23:29 jccn joined #git
23:32 Vampire0 gnac, well, I didn't say you need to have how things are the first time you ran something
23:32 Vampire0 gnac, but give a proper transcript of what you do now, meaning the exact fetch command AND its output
23:33 Vampire0 gnac, what makes you thing the fetch does not work
23:34 Jameser joined #git
23:34 Ryiya__ joined #git
23:35 gnac Vampire0: I think the fetch is not working because I do not see known changes made in the remote in the change log of the local.
23:35 gnac Vampire0: I'm recreating a new local repo from scratch so I can repeat the process by making changes to the remote after the local has been created ...
23:38 pokalyis joined #git
23:39 madewokherd joined #git
23:39 Vampire0 Are you sure the changes are not there, or are you mistakenly look at the local branches while a fetch only updates the remote tracking branches
23:42 Dohplay joined #git
23:43 fuchstronaut joined #git
23:45 inflames joined #git
23:46 dsdeiz joined #git
23:46 dsdeiz joined #git
23:49 plasticle joined #git
23:49 justan0theruser joined #git
23:51 gnac Vampire0: I'm pretty sure. I don't see any evidence of them in the change log of the bare repo. downstream repos dont get the changes when they pull. Here is a pretty complete recreation of the issue, starting with a new bare repo cloned from the remote:
23:51 Fat_Gandolf joined #git
23:51 gnac http://paste.opensuse.org/28223805
23:51 gnac In this case, commit 8e5b7c10918f3cfc377d0955ef83ad64aac5a4c0 was the one that I wasn't getting on my other repo when fetching.
23:52 e14 joined #git
23:53 gnac you can see in the log of the local, I am not getting 8e5b7c1..4772659 after the fetch.
23:55 railssmith joined #git
23:56 ctcrnitv joined #git
23:56 ctcrnitv has anyone here ever had to do branch forcing?
23:56 Vampire0 gnac, well, I see that you use git log ... -n 1. This does only show one commit, so you might miss your commit in the output even if it is in the repo
23:57 ctcrnitv I'm learning about it but I campt really imagine a use case
23:57 Vampire0 ctcrnitv, never heard of branch forcing. What do you mean exactly?
23:57 ctcrnitv for example, force master branch to move 3 commits behind, etc
23:57 gnac Vampire0: it also shows the last commit. If I pasted every commit the pastebin would be thousands of lines long.
23:57 ctcrnitv something like 'git branch -f master HEAD~3'
23:57 Vampire0 ctcrnitv, sure, I often move around branch heads
23:58 ctcrnitv Vampire0, don't you move the branch head with checkout though?
23:58 Vampire0 ctcrnitv, no, checkout does not move a branch head. It just moves HEAD or updates files from a commit-ish or the index
23:59 cdesai joined #git
23:59 thiago joined #git
23:59 eroussel joined #git
23:59 Vampire0 ctcrnitv, to move a branch head I use reset or branch -f, depends on mood and use-case
23:59 ctcrnitv hm ok, but what is a common use case for you using branch -f?

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