Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2016-09-21

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

All times shown according to UTC.

Time Nick Message
00:02 Marbug joined #git
00:02 ap0c joined #git
00:02 autoboxer joined #git
00:05 e14 joined #git
00:06 vigilvindex_ joined #git
00:07 Vampire0 joined #git
00:09 bobdel joined #git
00:16 nettoweb joined #git
00:17 romerocesar joined #git
00:17 evanwang joined #git
00:17 NeXTSUN joined #git
00:18 ToxicFrog joined #git
00:19 jason237 joined #git
00:26 RedNifre_ joined #git
00:28 Marbug joined #git
00:28 inflames joined #git
00:29 dmpeters63 joined #git
00:34 e14 joined #git
00:43 JamesBaxter joined #git
00:43 medeirosthiiago joined #git
00:44 rbern joined #git
00:44 dendazen joined #git
00:45 dreiss joined #git
00:47 souther joined #git
00:49 ShapeShifter499 joined #git
00:52 e14 joined #git
00:55 linuxmint joined #git
00:57 mgoodwin joined #git
00:57 snthsnth joined #git
00:57 raijin joined #git
00:58 mgoodwin What's the best way to take an old branch or stash and "tell if applying them will actually change any code" diff ... and .. don't seem to be effective - what I'm trying to ascertain is whether or not all the code from said branch/stash is present
00:58 mgoodwin Against a rebased/squashed master
00:59 boombatower joined #git
00:59 whitby `git diff` should work, no? What does it output?
00:59 autoboxer joined #git
01:00 whitby e.g. `git diff master master` outputs nothing because there's no diff between them
01:00 cagedwisdom joined #git
01:00 tesuji joined #git
01:00 rbern joined #git
01:01 mgoodwin Right but I don't want the negative diff
01:01 mgoodwin I just want to know if it would add anything
01:01 durham_ joined #git
01:02 mgoodwin because if there are N commits inbetween it will show all the stuff that I've removed/added as well
01:03 mgoodwin http://hastebin.com/uceyisorej.coffeescript this for example is what im looking at
01:03 ochorocho__ joined #git
01:03 mgoodwin I want to know if I "missed" something from branch initial_setup-WIP because after a rebase `git branch -d` will report that it's not merged
01:04 LBV_User joined #git
01:05 whitby Ah, you mean that `git diff` will show you stuff on master that's been added _in addition to_ what's in the -WIP branch
01:05 mgoodwin right
01:05 mgoodwin as well ass ... since the hashes differ
01:05 mgoodwin because of the rebase
01:05 mgoodwin as*
01:07 whitby Yeah... I don't know
01:07 mgoodwin I mean I could sit here and manually look at the "green" changes from 12c4645 and see if they're present, but that seems tedious
01:07 mgoodwin Was just wondering if there's a 'right way' to do it
01:07 cjohnson mgoodwin: I would: copy the branch you want to merge into as branch-copy, merge your target into the copy
01:07 cjohnson and see what it says
01:08 jbu_ joined #git
01:08 whitby You might be able to rebase again or "manually" replay each commit. That should have no effect if their changes are already present.
01:09 whitby But I'd need to test that and I don't have a setup handy
01:10 inflames joined #git
01:10 e14 joined #git
01:11 raijin joined #git
01:12 souther joined #git
01:16 e14 joined #git
01:18 Gsham joined #git
01:18 mingrammer joined #git
01:18 mgoodwin Rebasing master on top of the WIP branch seems to be effective, though I had to do a lot of manual conflict resolution and remember the desired state
01:21 ToBeCloud joined #git
01:23 clr mgoodwin: take a look at 'git patch-id', it would take some scripting but might help
01:25 mattcen joined #git
01:28 drbean joined #git
01:29 jbu__ joined #git
01:30 d^sh joined #git
01:30 rhizome mgoodwin: are you trying to diff the branch against a past commit, or against HEAD (minus $filelist extra stuff)?
01:31 CEnnis91 joined #git
01:32 cqi joined #git
01:32 mgoodwin HEAD, because of squashing and rebasing changing the the position and meaning
01:32 e14 joined #git
01:32 rhizome kinda looks like you want a one-side diff, restricted to the files in branch that have differences with HEAD?
01:33 raijin joined #git
01:33 mgoodwin right
01:33 rhizome commandline diff has left/right-only options, trying to find it in git diff
01:34 iamthad joined #git
01:35 sargas joined #git
01:35 Vampire0_ joined #git
01:35 nettoweb joined #git
01:36 rhizome mgoodwin: how's this look http://stackoverflow.com/a/6879568/135845
01:37 romerocesar joined #git
01:38 safe joined #git
01:39 arescorpio joined #git
01:39 PaulCapestany joined #git
01:40 ayjay_t joined #git
01:46 Gsham joined #git
01:49 ilbot3 joined #git
01:49 Topic for #git is now Welcome to #git, the place for git help and aquarium keeping | 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: 2.10.0 | Getting "cannot send to channel"? /msg gitinfo .voice | Keep good care of your git-shell and your inner snail will thank you
01:50 duderonomy joined #git
01:52 MattMaker joined #git
01:53 harish joined #git
01:54 theskillwithin joined #git
01:54 mgoodwin For the stash it seems like just checking out the master to a copy branch and applying the stash w/conflicts and then diff master..master-copy is best
01:54 fstd joined #git
01:55 mgoodwin need like a.... `git diff <master>...<master_after_commitish_applied>`
01:55 * mgoodwin shrug
01:55 mgoodwin (in a single command)
01:57 ap0c joined #git
02:00 random_numbers joined #git
02:00 random_numbers Hey, where do git-hook scripts run by default  (pwd) ? The topmost git project directory or the .git/hooks dir?
02:01 pks joined #git
02:01 random_numbers I'm essentially trying to get post-merge to run 'make $programname'.
02:06 morenoh149 joined #git
02:08 ap0c_ joined #git
02:10 Goplat joined #git
02:11 sargas joined #git
02:13 evanwang joined #git
02:13 mingrammer joined #git
02:19 axisys joined #git
02:20 Vampire0 joined #git
02:26 mizu_no_oto joined #git
02:26 Thorn__ joined #git
02:31 Rikairchy joined #git
02:32 marlinc joined #git
02:32 dreiss joined #git
02:35 ojdo joined #git
02:36 Guest29441 joined #git
02:38 mozzarella joined #git
02:39 cdg joined #git
02:40 cdg joined #git
02:44 snthsnth joined #git
02:45 danslo joined #git
02:49 ASOLAr joined #git
02:49 zivester joined #git
02:51 whomp joined #git
02:52 lagothri1 joined #git
02:53 tubaman joined #git
02:54 peepsalot joined #git
02:55 hexagoxel joined #git
02:55 chachasmooth joined #git
02:57 cyphase joined #git
02:58 GandalfTheWizard joined #git
02:59 noelthermo joined #git
03:01 SwiftMatt joined #git
03:02 raijin joined #git
03:02 aavrug joined #git
03:03 aavrug joined #git
03:04 ochorocho__ joined #git
03:05 re1 joined #git
03:07 rchavik joined #git
03:11 raspado joined #git
03:15 noelthermo left #git
03:15 f3r70rr35f joined #git
03:15 QwertyDragon joined #git
03:15 red82 joined #git
03:24 ExeciN joined #git
03:30 romerocesar joined #git
03:30 Lunatrius joined #git
03:37 Lunatrius joined #git
03:46 mehola joined #git
03:46 danslo joined #git
03:53 Andrew_K joined #git
03:53 Guest9 joined #git
03:55 morenoh149 joined #git
03:56 rsukla joined #git
03:57 raspado joined #git
04:00 dermoth joined #git
04:00 harish joined #git
04:01 mingrammer joined #git
04:05 romerocesar joined #git
04:06 SwiftMatt joined #git
04:08 clemf joined #git
04:09 laumann joined #git
04:10 MattMaker joined #git
04:14 cxreg2 left #git
04:15 cxreg joined #git
04:16 polyzen joined #git
04:17 phanimahesh joined #git
04:17 nydel07 joined #git
04:18 dodobrain joined #git
04:20 daey_ joined #git
04:21 nedbat joined #git
04:23 SwiftMatt joined #git
04:27 nilg joined #git
04:33 theskillwithin joined #git
04:35 mrx88 joined #git
04:40 SwiftMatt joined #git
04:41 Raging_Hog joined #git
04:44 ap0c joined #git
04:44 cqi joined #git
04:45 Muzer joined #git
04:46 tristanp joined #git
04:46 danslo joined #git
04:47 blunaxela joined #git
04:48 bocaneri joined #git
04:48 blackwind_123 joined #git
04:51 SwiftMatt joined #git
04:52 Cabanossi joined #git
04:55 mingrammer joined #git
04:58 subhojit777 joined #git
04:58 alexsystemf____ joined #git
05:01 dec0n joined #git
05:03 bmahe joined #git
05:03 tmz joined #git
05:04 tcurdt joined #git
05:05 ochorocho__ joined #git
05:05 vdl joined #git
05:06 justan0theruser joined #git
05:06 t0by joined #git
05:21 pijiu joined #git
05:22 rsukla joined #git
05:23 pksadiq joined #git
05:24 satya4ever joined #git
05:24 m1m3-50 joined #git
05:24 pksadiq is there a short way for 'git rebase -i' and selecting 'edit'? like 'git rebase --edit' so that I can create an alias for that
05:25 Akimb joined #git
05:30 SwiftMatt joined #git
05:33 TSS_ joined #git
05:33 QwertyDragon joined #git
05:33 ayjay_t joined #git
05:34 red82 joined #git
05:35 ayjay_t joined #git
05:38 mangolisk_ joined #git
05:44 morenoh149 joined #git
05:44 lindenle joined #git
05:44 a3Dman joined #git
05:47 tristanp joined #git
05:47 danslo joined #git
05:50 mingrammer joined #git
05:50 qt-x joined #git
05:51 TSS_ joined #git
05:53 glebihan_ joined #git
05:55 Mobutils joined #git
05:59 mingrammer joined #git
05:59 rbr joined #git
06:00 m1dnight_ joined #git
06:02 jknetl joined #git
06:03 circ-user-myOtS joined #git
06:04 UTAN_dev_ joined #git
06:04 niels__ joined #git
06:06 cyphase joined #git
06:06 rscata joined #git
06:07 lb1c joined #git
06:07 moestevens joined #git
06:08 JeroenT joined #git
06:09 JeroenT joined #git
06:09 johnny56_ joined #git
06:10 chele joined #git
06:15 cqi joined #git
06:16 Silenced joined #git
06:18 glebihan_ joined #git
06:20 jaafar joined #git
06:20 glebihan_ joined #git
06:21 phanimahesh joined #git
06:22 dan2k3k4 joined #git
06:23 roelmonnens joined #git
06:24 rendar joined #git
06:25 dreiss joined #git
06:25 huggy joined #git
06:26 _ikke_ pksadiq: piping it to a sed script that replaces pick with edit?
06:27 GandalfTheWizar1 joined #git
06:28 pksadiq _ikke_: is that possible? let me try
06:29 GandalfTheWizard joined #git
06:30 eduardas_m joined #git
06:32 Ravana joined #git
06:32 lordjancso joined #git
06:33 dhollinger joined #git
06:35 red82 joined #git
06:35 zeroed joined #git
06:35 Sceorem joined #git
06:36 AndreSomers joined #git
06:39 apotry joined #git
06:39 dan2k3k4 joined #git
06:40 cdown joined #git
06:41 vuoto joined #git
06:43 SwiftMatt joined #git
06:44 cdown_ joined #git
06:46 ronskisika joined #git
06:47 ahmedelgabri joined #git
06:48 snthsnth joined #git
06:48 danslo joined #git
06:58 pitastrudl joined #git
06:59 Doginal_ joined #git
07:02 ochorocho__ joined #git
07:02 TomyWork joined #git
07:02 danslo joined #git
07:05 niels__ joined #git
07:06 niels___ joined #git
07:08 dhollinger joined #git
07:08 diogenese joined #git
07:09 jknetl joined #git
07:09 overlord_tm joined #git
07:10 psprint How to have git log --date=relative --pretty=format:'%cd %h %s'   output date aligned in one column, i.e. the commit IDs should be aligned because date is correctly padded with spaces
07:10 Nitesh joined #git
07:11 Nitesh Why we cannot see the new files we have added to git project, on doing a git diff ?
07:13 ThomasLocke joined #git
07:13 ThomasLocke joined #git
07:13 Fuzzmz Nitesh, what does git status tell you? Have you added them via git add before trying the diff?
07:15 satifant joined #git
07:16 submitnine joined #git
07:19 ome joined #git
07:21 laumann joined #git
07:21 PCatinean joined #git
07:25 Balliad joined #git
07:25 ayjay_t joined #git
07:26 BadSnappa_ joined #git
07:27 tchan joined #git
07:30 viran joined #git
07:31 ramsub07 joined #git
07:31 viran Hey, I have a repo that have a git submodule in it, i want to add that submodule to the parent repo configuration so if anyone clones it.. it automaticly clones the submodule as well, is it possible?
07:32 Alienpruts joined #git
07:32 morenoh149 joined #git
07:33 danslo viran: yeah, tell the people that clone your repo to use the --recursive flag when cloning
07:33 danslo otherwise, no
07:33 viran danslo: cool, thanks...
07:33 viran same goes for pulling?
07:34 danslo viran: the manual for git-pull tells me there's a bug with that right now
07:35 danslo BUGS
07:35 danslo Using --recurse-submodules can only fetch new commits in already checked out submodules right now. When e.g. upstream added a new submodule in the just fetched commits of the superproject the submodule itself can not be fetched, making it impossible to check out that submodule later without having to do a fetch again. This is expected to be fixed in a future Git version.
07:36 viran ok, thanks
07:36 JeroenT joined #git
07:37 vuoto joined #git
07:37 gispka joined #git
07:37 inflames joined #git
07:38 vuoto joined #git
07:39 vuoto joined #git
07:40 vuoto joined #git
07:40 Mimicize joined #git
07:41 cyan__ joined #git
07:42 bobdobbs joined #git
07:43 _ng joined #git
07:43 Silenced joined #git
07:43 monbro joined #git
07:44 psprint Is it possible to embed \f (^L), \1 (^A), etc. in log output?
07:44 bobdobbs I've been managing some files with git. I then had to create a new project in folder higher up the file tree, so in my new git project I created a gitignore and made the new project ignore the subdirectory tree in which the original project lives.
07:45 bobdobbs I just tried to add a new file to my older project. But the older project seems to have inherited the gitignore of the new project
07:45 bobdobbs So I'm unable to add the file.
07:46 bobdobbs So, how do I ringfence git projects when one lives in the subdir of another one?
07:47 kurkale6ka joined #git
07:47 QwertyDragon joined #git
07:47 tristanp joined #git
07:48 lindenle joined #git
07:49 overlord_tm joined #git
07:50 Macaveli joined #git
07:51 HappyLoaf joined #git
07:53 rory left #git
07:54 MrWoohoo joined #git
07:55 heftig psprint: sure
07:56 mcfdez joined #git
07:59 AndreSomers joined #git
07:59 mcfdez Hello. I have this situation. I have a git repo (in bitbucket) and in my local pc the repo cloned for development. In a server I have exactly the same files that I have in the local pc but there is not under git (bisically git folder not exists). Exists any option to syncronizate this files with the repo?
08:01 Milijus joined #git
08:01 Milijus git clone -b  <--what does the -b  stand for?
08:01 buzzLightBeer joined #git
08:02 selckin man git-clone
08:02 gitinfo the git-clone manpage is available at http://jk.gs/git-clone.html
08:02 groundnuty hey, I would know that I can use a pre-commit hook to perform some actions on files before things get commited
08:02 groundnuty however I only would like to act when *.mp3 files are commited
08:02 groundnuty .gitattributes comes to mind
08:02 encod3 joined #git
08:02 groundnuty but I'm not sure how to add a script that will get executed there
08:03 groundnuty also at the proper moment
08:03 psprint heftig: you mean, when I do e.g. in Zsh, $'\1' and place that in --pretty=format:? But is it possible to obtain directly via Git?
08:03 mcfdez Milijus, are you talking with me?
08:03 Milijus Yes...
08:03 sonikspin joined #git
08:03 psprint heftig: ah, %x
08:04 mcfdez Milijus, the problem is that if I execute in the server git clone <repo> it says that the folder is not empty
08:04 ramsub07 joined #git
08:05 Milijus mcfdez, interesting...
08:05 azv5 joined #git
08:11 m1m3-50 joined #git
08:12 monbro joined #git
08:13 matleh joined #git
08:14 clmsy joined #git
08:14 tlaxkit joined #git
08:15 _ADN_ joined #git
08:17 JeroenT_ joined #git
08:19 Fuzzmz mcfdez, on the server do you already have a folder named as the git repo?
08:19 Fuzzmz if yes, that's why it can't clone
08:19 Fuzzmz either rename/delete that folder or do git clone <newfolder>
08:24 Balliad joined #git
08:25 azv4 joined #git
08:26 bencevan_ joined #git
08:26 Balliad joined #git
08:26 fedepad joined #git
08:27 chll_ joined #git
08:27 mithenks joined #git
08:28 DelphicOkami joined #git
08:30 mischat joined #git
08:30 rorro joined #git
08:31 star_prone joined #git
08:32 AndreSomers joined #git
08:32 RedNifre_ joined #git
08:33 sagittarian joined #git
08:37 ferr joined #git
08:39 vuoto joined #git
08:41 ap0c joined #git
08:41 djb-irc joined #git
08:43 h0bbit joined #git
08:43 vuoto joined #git
08:45 vuoto joined #git
08:47 netj joined #git
08:50 vuoto joined #git
08:50 star_prone joined #git
08:51 vuoto joined #git
08:56 mikecmpbll joined #git
08:59 cdown_ joined #git
08:59 Nitesh Fuzzmz: I could not see them by just doing git diff. But on doing git add and git diff --staged , I could see.
09:03 ochorocho__ joined #git
09:03 dhollinger joined #git
09:04 cdown_ joined #git
09:09 notebox joined #git
09:10 achadwick joined #git
09:14 Serpent7776 joined #git
09:14 MattMaker joined #git
09:14 ap0c_ joined #git
09:14 mischat joined #git
09:19 JeroenT joined #git
09:21 morenoh149 joined #git
09:23 circ-user-myOtS joined #git
09:25 Repox joined #git
09:32 dminuoso joined #git
09:32 dminuoso Hi. Im editing a hunk from git add -p with my editor being vi. When I :wq the hunk, the change is not added to the index.
09:33 dminuoso (Instead git just stops there, as if I had never edited the hunk)
09:33 jorj joined #git
09:34 phanimahesh joined #git
09:34 MattMaker joined #git
09:34 selckin did you ctrl-c the next hunk question?
09:35 dminuoso Nope.
09:36 dminuoso Ahh wait, so the working file is never modified?
09:36 pro_metedor joined #git
09:36 selckin you have to finish all the questions, or gracefully exit the git add -p
09:36 selckin if you go yes/no a bunch and then ctrl-c, it doesn't do anything
09:37 dminuoso selckin: I don't abort the hunk question..
09:37 AndreSomers joined #git
09:37 selckin the next one?
09:41 star_prone joined #git
09:41 inflames joined #git
09:41 niels__ joined #git
09:43 courrier joined #git
09:43 bruce_lee joined #git
09:43 bruce_lee joined #git
09:44 MattMaker joined #git
09:45 ap0c joined #git
09:48 iob- joined #git
09:49 lindenle joined #git
09:51 ahr3n joined #git
09:51 King_Hual joined #git
09:51 mischat joined #git
09:53 harish joined #git
09:55 compeman joined #git
09:55 compeman joined #git
09:55 compeman joined #git
09:55 compeman joined #git
09:57 fuchstronaut joined #git
09:59 Akimb joined #git
09:59 rorro joined #git
10:00 MattMaker joined #git
10:02 autoboxer joined #git
10:03 ochorocho__ joined #git
10:07 Thorn__ joined #git
10:15 phanimahesh joined #git
10:16 acetakwas joined #git
10:16 acetakwas When naming a commit with several changes, how you title it?
10:16 acetakwas For example...
10:17 acetakwas If my commit made changes a, b, c and d. Should I just write "View changelog" as the title message
10:17 osse What do the changes have in common?
10:18 acetakwas They may not have much in common
10:18 JeroenT joined #git
10:18 osse in that case make four commits
10:18 stamina joined #git
10:18 acetakwas Like, I added some try/catch in some places, changed some config...
10:18 acetakwas What if I wanted one?
10:20 MattMaker joined #git
10:20 Remram[m] joined #git
10:21 osse If you want one you have a title problem
10:22 osse write a title that suggests there more to the message. write a long one
10:23 Monpt joined #git
10:24 phanimahesh joined #git
10:24 ahr3n joined #git
10:24 raijin joined #git
10:29 troyt joined #git
10:29 ahmedelgabri joined #git
10:31 tlaxkit joined #git
10:32 grawity joined #git
10:34 dut- joined #git
10:35 MattMaker joined #git
10:36 Akimbas joined #git
10:43 GandalfTheWizar1 joined #git
10:44 dut- left #git
10:46 avar Anyone here have a suggestion for a regex to match a git tag?
10:46 * avar reading refs.c trying to make sense fo it..
10:46 PioneerAxon joined #git
10:46 dut- joined #git
10:46 osse avar: ^refs/tags/
10:46 osse ez
10:47 dut- left #git
10:47 avar I mean the output you'd get from "git tag -l"
10:47 avar (?<!-)\S+(?!\.) <- shitty attempt so far
10:48 osse !xy
10:48 gitinfo Woah, slow down for a bit. Are you sure that you need to jump through that particular hoop to achieve your goal? We suspect you don't, so why don't you back up a bit and tell us about the overall objective...
10:48 moritz avar: .*, maybe?
10:48 osse I don't follow
10:48 osse moritz: hah, that was my idea as well
10:49 jbitdrop joined #git
10:49 avar I'm patching the bit in gitweb.cgi that formats log messages now: https://github.com/git/git/blob/​master/gitweb/gitweb.perl#L2035
10:49 fuchstronaut joined #git
10:50 avar Sort of works, won't get non-ascii refnames and whatnot though: https://gist.github.com/avar/02​29d15e6eabbed0b8937c9c16a1054a
10:50 avar But I wonder if anyone cares about stuff not matching [A-Za-z0-9.-]+ in practice
10:52 grawity joined #git
10:52 osse i can't tell what the input is
10:52 mischat joined #git
10:52 avar the log message as shown by "git show"
10:52 avar i.e. the commit message body
10:53 ShekharReddy joined #git
10:54 MattMaker joined #git
10:54 quesker left #git
10:55 doppo joined #git
10:58 tcurdt joined #git
10:59 star_pro_ joined #git
11:00 ahmed_elgabri joined #git
11:00 JeroenT joined #git
11:02 mkoskar joined #git
11:03 rominronin joined #git
11:03 Lomex joined #git
11:08 bremner joined #git
11:08 bremner joined #git
11:10 morenoh149 joined #git
11:12 bencevan_ joined #git
11:13 harish joined #git
11:14 m1m3-50 joined #git
11:14 mingrammer joined #git
11:14 a3pq51 joined #git
11:15 m1m3-50 joined #git
11:16 raijin joined #git
11:21 leeN joined #git
11:22 Andrew_K joined #git
11:23 star_prone joined #git
11:24 fahadash joined #git
11:27 Jellyg00se joined #git
11:28 mingrammer joined #git
11:28 ayjay_t joined #git
11:29 ayjay_t joined #git
11:30 MattMaker joined #git
11:31 JeroenT joined #git
11:32 subhojit777 joined #git
11:33 eihabarabia joined #git
11:34 mizu_no_oto joined #git
11:36 nuck joined #git
11:39 Sceorem joined #git
11:41 red82 joined #git
11:41 dan2k3k4 joined #git
11:42 jimi_ joined #git
11:44 rbr joined #git
11:45 MattMaker joined #git
11:45 fuchstronaut joined #git
11:46 circ-user-myOtS joined #git
11:46 Prisoner-2460_1 joined #git
11:50 nanga joined #git
11:50 nanga joined #git
11:51 lindenle joined #git
11:52 mkoskar joined #git
11:53 DolphinDream joined #git
11:54 raijin joined #git
11:55 rnsanchez joined #git
11:56 maret joined #git
11:59 maret Hi everyone, If I created pull request on feature branch on github and it got conflict. What is best way to resolve it , if there are few other feature branches? Let say I create two branches ->(I am on master now) git checkout -b featureA; git checkout -b featureB; now I create commmits in both of them , meanilwhile some other branches are merged into master. NowI create pull request for both branches and in both of them I have a conflict
11:59 maret I resolve both of them by merging master into them and create new merge commit in each of branches
12:00 ibasaw joined #git
12:00 Fuzzmz maret, how i'd handle it is like this
12:00 i7c like how
12:00 maret should I just push both of those branches to remote and they should be both merged easily to master?
12:00 Lomex joined #git
12:00 PaulCapestany joined #git
12:00 Fuzzmz set up the repository you're trying to get your PR as another remote (like upstream)
12:00 mischat joined #git
12:01 Fuzzmz git pull upstream master to get the latest changes from master
12:01 i7c maret, after you merged master into your branch, the conflicts should not happen anymore, yes.
12:01 pwnawannab[NE] joined #git
12:01 maret Fuzzmz: I did that in order to update my local(forked) repo master branch
12:01 Fuzzmz great
12:01 i7c maret, it would now be a fast-forward merge.
12:01 Fuzzmz now go to featureA and do a git rebase master
12:01 i7c Alternative is rebase.
12:01 i7c !merge_vs_rebase
12:01 gitinfo [!merge_or_rebase] For a detailed discussion of the dis-/advantages of merge and rebase, and when to best use which, see: https://coderwall.com/p/ny1hia -- and here's how Linus explains which one to use: http://www.mail-archive.com/dri-devel​@lists.sourceforge.net/msg39091.html
12:02 NeXTSUN joined #git
12:02 Fuzzmz I'd rebase instead of merging, it keeps the history look cleaner, but the maintainer of the original repo might not care
12:03 Fuzzmz in either case you'll have to fix conflicts
12:03 ibasaw left #git
12:03 i7c I'd also prefer to rebase pull requests
12:04 maret i7c: fast forwar medge for feature A merged with master , but if I maintainer will merge feature B merged with master there would be another conflict possibly
12:04 i7c simply because its easier to see what changed
12:04 ochorocho__ joined #git
12:04 i7c maret, that is an issue which you cannot solve with magic.
12:04 maret Fuzzmz: you would rebase it when? when resolving pull request or instead of merging master into feature branch?
12:05 MattMaker joined #git
12:05 mrpackethead joined #git
12:05 i7c If both pull requests change some parts you will always have to deal with conflicts.
12:05 Fuzzmz maret, the flow would be like this: update master, change to feature branch, git rebase master, resolve conflicts on that feature branch, git push --force
12:05 Fuzzmz that updates your pull request so it's mergeable
12:06 Fuzzmz and yeah, after your featureA gets merged you'll need to do the same thing for featureB
12:06 Fuzzmz basically every time your destination branch updates you need to update (rebase) all the feature branches targeting it
12:06 maret isn't having merge request better in this case , especially because some of conflicted files contains changes by 3rd developer?
12:06 Tmp_ joined #git
12:07 Fuzzmz I don't think I understand. You'd still have the merge request
12:07 ColdKeyboard joined #git
12:07 Fuzzmz you're updating your featureA branch to start from the latest changes from master
12:07 autoboxer joined #git
12:07 Fuzzmz which, after the push, automatically updates the pull request
12:08 maret Fuzzmz: well if I rebase master with feature there wont be merge commit would it.
12:08 Fuzzmz when merging your feature/pull-request into master? depends on how that repository is set up
12:09 Fuzzmz it can be set so it always does a merge commit
12:09 Fuzzmz I think that's the default on GitHub
12:09 maret yes it is but you did mention using rebase  which doesn't have merge commit by default AFAIK
12:10 Fuzzmz yes, you won't get a merge commit on your feature branch
12:10 i7c rebase does not merge, so there is no merge commit.
12:10 Monpt joined #git
12:10 Fuzzmz what rebase does is it gets all your changes and re-applies them on top of the branch you gave as an argument
12:11 synthroid joined #git
12:11 maret although github does support squashing of commmits when  merging pull request
12:11 Fuzzmz don't know how it works on github tbh
12:12 maret so it seems that it doesn't really matter if I have merge commit or not, because the maintainer can sqaush it if he wish to do so. Just never tried it so not sure if it helps if I don't have merge commits
12:12 Fuzzmz in general most projects I've dealt with prefer rebasing over merging master into your branch
12:12 dsdeiz joined #git
12:12 dsdeiz joined #git
12:12 Fuzzmz but yes, it's up to the maintainer
12:13 bruce_lee joined #git
12:13 bruce_lee joined #git
12:13 i7c rebase can be very tiring for long-lived branches that live in parallel.
12:13 ash_workz joined #git
12:13 i7c I'm working on a project with a branch that doesn't get merged for over 6 months. Rebasing would be hell.
12:14 Fuzzmz heh, yeah, in case you need to share that branch with others you can't really do much besides merging
12:14 Fuzzmz at least if you don't want to make the other devs hate you
12:15 i7c master gets merged into this branch from time to time to keep the code kind of up-to-date.
12:15 i7c It’s a lot easier than rebasing hundreds of commits every time. :)
12:16 Fuzzmz wait, why would it matter? as long as there aren't any conflicts there's no difference between doing git merge master or git rebase master
12:16 Fuzzmz ignoring the speed difference I guess :)
12:16 i7c But there are conflicts.
12:16 Fuzzmz but then don't you have to fix them anyway?
12:16 i7c And come keep recurring during rebases
12:16 Fuzzmz enable git rerere
12:16 i7c Sure I have to fix them.
12:16 i7c Yeah
12:17 i7c but still
12:17 Fuzzmz in the end, whatever works better
12:17 ThiefMaster how does git pick a secret key to sign commits? i have a key in my keychain that - among other emails - contains the email i commit with, but it doesn't pick it automatically. it does, however, work fine if i have another key that has just the email i commit with
12:17 bdonnahue joined #git
12:17 bdonnahue hello. is there a way to change a commit message after the fact?
12:18 osse bdonnahue: git commit --amend
12:18 ThiefMaster bdonnahue: fyi, this rewrites history and needs a force push if you had already pushed the commit with the wrong message
12:18 JeroenT joined #git
12:19 bdonnahue ok thanks
12:19 ahmed_elgabri joined #git
12:20 Fuzzmz ThiefMaster, I think that has more to do with gpg than git itself
12:20 Fuzzmz not exactly sure how gpg picks the default key if not specified by git
12:20 ThiefMaster ah ok, so i'll just specify the key id explicitly
12:20 Fuzzmz via the user.signingkey variable
12:20 MattMaker joined #git
12:20 Fuzzmz you can set it via git config (--global) user.signingkey keyid
12:21 Fuzzmz you can also do a git config commit.gpgsign true so you won't always have to add the -s flag
12:21 Fuzzmz -S, sorry
12:25 crose joined #git
12:25 buffal0 joined #git
12:25 cdown joined #git
12:25 ShekharReddy joined #git
12:25 fmcgeough joined #git
12:26 vuoto joined #git
12:27 jimi_ joined #git
12:27 yoh joined #git
12:28 phanimahesh joined #git
12:28 mischat joined #git
12:29 mlittlej1 joined #git
12:31 eihabarabia joined #git
12:31 romerocesar joined #git
12:32 dan2k3k4 joined #git
12:32 andlabs joined #git
12:34 actinide joined #git
12:35 maret hmm so I've tried rebase and I notice one important difference. With git merge I don't see commits made on feature A branch I see just commits from master + merge commit. Am I missing something? With rebase I see all of master commit with feature a commits on top of it
12:35 MattMaker joined #git
12:36 GodGinrai joined #git
12:37 ash_workz joined #git
12:37 Fuzzmz maret, you should rebase looks good
12:37 Akimb joined #git
12:37 Fuzzmz not sure about merge
12:38 PCatinean joined #git
12:39 Zilenc3 joined #git
12:39 inflames joined #git
12:40 maret Fuzzmz: I think I am doing something wrong there with the merge because It doesnt make sense for other commits to dissapear
12:40 Fuzzmz I'd expect to see your commit followed by a merge commit
12:42 Prisoner-2460_1 joined #git
12:44 garethdaine joined #git
12:44 maret yeah ou well I will use rebase anyway. BTw you did use git push -f which I suppose I have to use because I've rebased with master so older commits from feature branch A went on top of new commits on master there for the history has changed, correct?
12:45 Fuzzmz yes, you need to use -f (--force) because by rebasing you've basically changed the history of your feature branch
12:45 sdothum joined #git
12:45 maret and because I am only one working on feature A this doesn't really matter and it doesn't have an effect when the maintainer will merge pull request
12:45 Fuzzmz exactly
12:45 maret because from his point of view there are new commits on top of master
12:45 maret ok
12:46 maret ok and now  I will do the same for feature branch B, but what should I rebase it with? If I rebase it with master then there will be another conflict between master + feature A and master and feature B. What if I would rebase it with feature A + master branch instead of master?
12:47 mischat_ joined #git
12:47 shinnya joined #git
12:47 Fuzzmz I'd wait for feature A to get merged
12:48 maret ok and then resolve possible conflicts hmm there seems to be a lot of room for conflicts. They way we work is that feature branches are small and I always created them from master
12:48 johnmilton joined #git
12:49 maret and some of feature branches do related to same files
12:51 Fuzzmz there's unfortunately no way around that as long as you work on the same files :(
12:51 Fuzzmz unless you touch different parts of the file, in that case you shouldn't get a merge conflict
12:52 subhojit777 joined #git
12:54 Brando753-o_O_o joined #git
12:54 Tmp_ left #git
12:54 MattMaker joined #git
12:55 buffffal00 joined #git
12:55 johnmilton joined #git
12:56 tobiasvl !eek
12:56 gitinfo [!eekaconflict] Merge conflicts are a natural part of collaboration. When facing one, *don't panic*. Read "How to resolve conflicts" in man git-merge and http://git-scm.com/book/ch3-​2.html#Basic-Merge-Conflicts then carefully go through the conflicts. Picking one side verbatim is not always the right choice! A nice video explaining merge conflicts: https://www.youtube.com/watch?v=zz7NuSCH6II
12:56 dan2k3k4 joined #git
12:56 Triq joined #git
12:58 isysd joined #git
12:59 griffindy joined #git
12:59 morenoh149 joined #git
12:59 dcpc007 joined #git
13:01 mlittlej joined #git
13:01 dolgov joined #git
13:01 rchavik_ joined #git
13:02 Triq Hi, could anyone help me with a problem I'm having migrating a svn repository to git? Everything seems to work fine up to the step where tags/branches are suppose to be moved, but all my tags and branches are located in a different folder .git/svn/ref... I tried moving those but got some errors.
13:03 cdg joined #git
13:05 ThiefMaster hm, apparently commit.gpgsign=true also signs other people's commits when i cherrypick them. is that actually intended and is there a way to avoid it?
13:05 ThiefMaster it doesn't seem to make much sense to sign someone else's commits
13:07 Remram[m] joined #git
13:07 Remram[m] joined #git
13:08 OnOut2k16 joined #git
13:09 M-shine joined #git
13:09 M-meznak joined #git
13:09 M-mistake joined #git
13:09 aviraldg joined #git
13:09 Ineentho[m] joined #git
13:09 mkoskar joined #git
13:09 danslo joined #git
13:09 MattMaker joined #git
13:10 lb1c joined #git
13:13 Fuzzmz ThiefMaster, well, it sort-of-does, as basically you signing says "I applied this commit and not some other person"
13:13 cdown joined #git
13:14 danslo joined #git
13:14 Fuzzmz the committer and author can be different people
13:15 maret small issue when resolveing conflict while rebasing I fixed first conflict then run continue then another small conflict appear where basicly only the lines surrounding the conflict were an issue. like this https://www.pastiebin.com/57e287a1e8fb4. Problem is that if I removed them there is nothing to add and I can't run git rebase --continue
13:16 bsanford joined #git
13:16 maret I think the problem was that I've updated this file on that line during the first conflict because I didn't realize that fix will be applied by later commit when rebasing
13:18 psprint Universal way to get current branch, handling various cases of unborn branches or other like ?
13:18 Fuzzmz maret, not sure how you can fix that without git rebase --abort, starting over and not fixing the file in the first conflict
13:18 dwsr joined #git
13:18 maret ok I will just do it agian
13:18 dwsr For example, I am building a collection of scripts. I found someone else's that I like, but it's in a similar configuration (collection of scripts). I know that you can do submodules in Git, but is it possible to say "I only want this one file"?
13:18 dwsr Is it possible to do this ^?
13:19 mlittlej joined #git
13:19 Fuzzmz dwsr, nope
13:19 Fuzzmz it's all or nothing
13:19 saml how can I git pull  some other directory?   git pull --rebase -- /path/to/repo
13:20 dwsr I figured as much, but doesn't hurt to ask
13:20 dwsr Fuzzmz: Thanks.
13:20 MattMaker joined #git
13:20 saml do I need to cd into /path/to/repo first always?   trying to write a script that pulls multiple repo
13:20 Fuzzmz dwsr, no problem
13:20 encod3_ joined #git
13:21 Fuzzmz saml, git --git-dir=/path/to/repo pull --rebase should work
13:21 sangy joined #git
13:21 dan2k3k4 joined #git
13:21 saml ah thanks
13:22 saedelaere joined #git
13:22 saml git --git-dir=./foo/.git  --rebase    error: Cannot pull with rebase: You have unstaged changes.
13:22 Fuzzmz saml, or better git -C /path/to/repo pull --rebase
13:22 CEnnis91 joined #git
13:22 Fuzzmz saml, you first want to either add and commit your changes or remove them
13:22 Fuzzmz before doing a pull --rebase
13:23 saml all repo is clean clone.   what's -C?  can't find it in git help pull
13:23 i7c man git
13:23 gitinfo the git manpage is available at http://jk.gs/git.html
13:24 Fuzzmz -C isn't a flag for pull, it's a flag for git (available in git 1.8.5 or newer)
13:24 saml ah thanks
13:24 Fuzzmz are you sure that the repo is clean?
13:24 Fuzzmz Triq, haven't migrated from svn to git, but give http://www.sailmaker.co.uk/blog/2013/05/05/migrati​ng-from-svn-to-git-preserving-branches-and-tags-3/ a look
13:24 jeffreylevesque joined #git
13:25 saml yup
13:25 saml -C works   --git-dir does not work as expected
13:25 Zilenc3 joined #git
13:26 Triq Fuzzmz: I will, thank you
13:26 ExoUNX joined #git
13:26 i7c saml, with --git-dir you probably must also set --work-tree=
13:26 thadtheman joined #git
13:26 i7c Not sure tho.
13:27 Fuzzmz yeah, --git-dir need to point to the .git folder and you also need to do --work-tree to be able to do any changes
13:28 Fuzzmz -C is a better option if you have a new enough version of git
13:28 Fuzzmz my bad on the above
13:28 ajhansen joined #git
13:28 rjack2 joined #git
13:28 i7c And for some reason --git-dir=~/some/thing does not expand ~
13:29 i7c Silly option.
13:29 ajhansen left #git
13:29 osse usually that's the shell's job
13:29 i7c I wonder why it doesn't in this case.
13:30 i7c Neither bash nor zsh.
13:30 ilmari because ~ is only expanded at the beginning of a word
13:30 i7c I see.
13:30 i7c Leave out the = and it works.
13:31 CheckDavid joined #git
13:31 sylario joined #git
13:31 JeroenT joined #git
13:32 buffal0 joined #git
13:32 ilmari http://pubs.opengroup.org/onlinepubs/969991​9799/utilities/V3_chap02.html#tag_18_06_01
13:33 raspado joined #git
13:34 osse i7c: for zsh you can do  setopt magicequalsubst
13:35 eka_ joined #git
13:35 MattMaker joined #git
13:36 Prisoner-2460_1 joined #git
13:37 rorro joined #git
13:37 garethdaine joined #git
13:40 JeroenT joined #git
13:40 buzzLightBeer joined #git
13:40 mingrammer joined #git
13:41 niels__ joined #git
13:42 wrksx joined #git
13:42 wrksx hey
13:42 PaulCapestany joined #git
13:42 i7c osse, I see.
13:42 wrksx I only used one computer for my project dev, and created a lot of tags to mark versions.
13:43 rnsanchez_ joined #git
13:43 nydel07 joined #git
13:43 wrksx today, I checkedout my remote from a new computer, and tags are not there.
13:43 i7c git push --tags
13:43 wrksx i7c: ha
13:43 i7c :)
13:44 Fuzzmz you can also do git push --all if you want to push both commits and tags at the same time
13:44 i7c I hardly ever want to push all branches.
13:44 i7c And that’s what --all also does.
13:45 Fuzzmz crap, you're right
13:45 AaronMT joined #git
13:46 leeN joined #git
13:46 raspado joined #git
13:46 bsanford joined #git
13:47 LionsMane joined #git
13:47 _rnsanchez_ joined #git
13:47 raspado joined #git
13:48 yottabyte joined #git
13:48 wrksx i7c: will try that tomorow then, ty =)
13:48 yottabyte on windows with git bash, I think git diff will always start up vi/vim in cmd. how can I change that to use notepad or atom?
13:49 PaulCapestany joined #git
13:49 osse yottabyte: it starts 'less'
13:49 osse a pager
13:49 tristanp joined #git
13:50 osse yottabyte: you can change the pager to be notepad or atom, but i don't know how well it works
13:50 Darren__ joined #git
13:50 MattMaker joined #git
13:50 f3r70rr35f joined #git
13:50 yottabyte how can I find out how to use less
13:50 i7c There’s also difftool. (No idea of git on windows, I might add)
13:51 i7c yottabyte, it doesn’t have many features other than scrolling and searching.
13:51 osse yottabyte: man less
13:51 osse ah windows
13:51 i7c Setting marks.
13:51 i7c https://linux.die.net/man/1/less
13:51 osse yottabyte:  less --help
13:52 kpease joined #git
13:52 buzzLightBeer left #git
13:52 ramsub07 joined #git
13:52 lindenle joined #git
13:53 tgies joined #git
13:55 tgies left #git
13:57 le_melomane joined #git
13:58 danemacmillan joined #git
13:59 jnavila joined #git
13:59 jason237 joined #git
13:59 faux___ joined #git
13:59 JamesBaxter joined #git
13:59 lightsoda joined #git
14:03 red82 joined #git
14:04 NeverDie joined #git
14:05 ochorocho__ joined #git
14:05 MattMaker joined #git
14:06 ayjay_t joined #git
14:07 ekinmur joined #git
14:09 rbr joined #git
14:09 mingrammer joined #git
14:09 ResidentBiscuit joined #git
14:10 PCatinean joined #git
14:11 marc_v92 joined #git
14:13 mycyber joined #git
14:14 raijin joined #git
14:16 ocbtec joined #git
14:19 dmpeters63 joined #git
14:19 MattMaker joined #git
14:20 ajhansen joined #git
14:21 sargas joined #git
14:21 sargas joined #git
14:23 vuoto joined #git
14:23 jaafar joined #git
14:25 phaleth joined #git
14:28 mkoskar joined #git
14:29 Pyrogerg joined #git
14:30 Hanumaan joined #git
14:31 noodle joined #git
14:33 DnlXz joined #git
14:34 MattMaker joined #git
14:35 pullphinger joined #git
14:35 koovec joined #git
14:36 nedbat I want to see how a branch has grown beyond a particular tag, so I use:   git log the_branch ^the_tag    # which works, but I need to run it on many repos, some of which don't have that branch or tag. Can I make it be silent if they don't exist in the repo?
14:36 notebox joined #git
14:36 morenoh149 joined #git
14:37 madewokherd joined #git
14:40 LBV_User joined #git
14:40 rsukla joined #git
14:40 mischat joined #git
14:41 cbreak nedbat: you can check if it contains them before querying
14:41 cbreak git rev-parse can do that
14:42 ekinmur joined #git
14:42 mischat joined #git
14:42 rbr joined #git
14:44 terminal_echo joined #git
14:44 crose joined #git
14:45 MattMaker joined #git
14:45 nedbat cbreak: can you show me an example?  git rev-parse the_branch  # fails in the same way, noisily
14:46 dminuoso joined #git
14:46 osse it will only fail if the_branch doesn't exist
14:46 cbreak nedbat: it will fail, so check the return value
14:47 nedbat osse, cbreak: ok, I need to supply --verify -q.   just -q doesn't make it quiet.
14:47 nedbat thanks.
14:47 cbreak indeed
14:47 nedbat although, --verify -q  still outputs the answer if the branch does exist....
14:47 thiago_ joined #git
14:48 boombatower joined #git
14:48 cbreak yes, that's the point
14:48 cbreak if you don't want the reply, just ignore it
14:49 snowkidind joined #git
14:49 patarr joined #git
14:50 inflames joined #git
14:52 nedbat cbreak: yup, ok
14:53 peepsalot i'm not sure i understand the difference between git reset  and git reset --soft.   the soft option is like doing an extra "git add ."?
14:53 chelly joined #git
14:54 cbreak peepsalot: no
14:54 cbreak peepsalot: the --soft option tells git "do not touch the index at all"
14:54 cbreak without it, the default is --mixed
14:55 peepsalot whats the index
14:55 nedbat peepsalot: the staging area (also called the cache.... :(   )
14:55 cbreak which tells git "... and change the index to match the state that you reset to"
14:55 nedbat peepsalot: not sure if this visualization helps or not, but it's at least a different way to look at it: http://ndpsoftware.com/git-cheatsheet.html
14:56 cbreak for example: git reset HEAD~ changes the current branch to point at the parent commit, and changes the index to match
14:56 cbreak with --soft, it wouldn't do the index changing
14:56 lindenle joined #git
14:56 Pyrogerg joined #git
14:56 Limes left #git
14:56 Pasi joined #git
14:57 gargola joined #git
14:58 mangolisk joined #git
14:59 peepsalot so is it the other way around then?  "git reset --mixed" is like doing "git reset --soft; git add ."
15:01 cbreak peepsalot: git reset --mixed is like git reset --soft and git read-tree HEAD
15:01 cbreak it is unrelated to what ever you have littering in your working directory
15:01 InfoTest joined #git
15:01 cbreak so git add is not involved
15:01 xcesariox joined #git
15:02 hobodave joined #git
15:02 matsaman joined #git
15:03 cdown joined #git
15:03 robotroll joined #git
15:03 star_prone joined #git
15:04 rominronin joined #git
15:04 MattMaker joined #git
15:04 timthowtdi joined #git
15:06 ShapeShifter499 joined #git
15:06 mischat joined #git
15:08 lindenle joined #git
15:09 Silenced joined #git
15:09 clemf joined #git
15:10 joki joined #git
15:10 shinnya joined #git
15:10 JeroenT joined #git
15:12 ejb joined #git
15:14 overlord_tm joined #git
15:14 King_Hual` joined #git
15:14 tyreld joined #git
15:16 eivarv joined #git
15:17 antlarr_away joined #git
15:17 ouac joined #git
15:18 dminuoso joined #git
15:20 MattMaker joined #git
15:20 happy-dude joined #git
15:20 giacomino joined #git
15:20 satbyy joined #git
15:22 rivarun joined #git
15:29 GodGinrai joined #git
15:29 clemf joined #git
15:30 MattMaker joined #git
15:31 rbr joined #git
15:32 brendyn joined #git
15:33 bencevan_ joined #git
15:35 mingrammer joined #git
15:37 clemf joined #git
15:38 subhojit777 joined #git
15:38 zeroed joined #git
15:38 zeroed joined #git
15:39 derpsauce__ joined #git
15:39 Doginal joined #git
15:40 derpsauce__ I've got a repo that I've made changes to, but I'd like to revert to my last commit
15:42 _ikke_ derpsauce__: Did you push that commit already?
15:43 duderono_ joined #git
15:44 hashpuppy joined #git
15:45 fahadash joined #git
15:46 mingrammer joined #git
15:46 actinide joined #git
15:47 derpsauce__ no
15:47 derpsauce__ wait, yes
15:47 WayToDoor joined #git
15:47 Limes joined #git
15:48 _ikke_ Is there a reasonable chance someone else has fetched that commit?
15:48 derpsauce__ nope, it's a personal repo
15:48 _ikke_ ok
15:48 _ikke_ git reset --keep HEAD~
15:49 _ikke_ git push --force-with-lease origin <branch>
15:49 gargola joined #git
15:49 r00twrh joined #git
15:49 MattMaker joined #git
15:49 peepsalot what is the "tree"? is that what the same as the local repo?   http://ndpsoftware.com/git-​cheatsheet.html#loc=index;
15:50 derpsauce__ ugh, I'm getting an "untracked working tree file .blah blah blah"
15:50 _ikke_ peepsalot: Where do you see tree in there?
15:50 peepsalot _ikke_, as in read-tree
15:51 _ikke_ It's an internal object that stores the file and directory structure
15:51 FilipNor1ic joined #git
15:52 _ikke_ read-tree is a low-level command btw
15:52 actinide joined #git
15:52 peepsalot it was mentioned earlier, as the diffrence between git reset --mixed and --soft
15:52 _ikke_ Right
15:53 peepsalot still just trying to wrap my head around it, just got off a conference call
15:53 _ikke_ Ok
15:53 vassagus joined #git
15:53 _ikke_ So the difference between --mixed and --soft is that --mixed also updates the index
15:53 Doginal joined #git
15:54 ahmedelgabri joined #git
15:55 clemf joined #git
15:56 sdothum joined #git
15:57 peepsalot with --mixed, the changes between HEAD (before the reset) and what you reset to, are left staged at the end of it all?
15:57 dan2k3k4 joined #git
15:57 peepsalot and with --soft?  they are where?
15:58 sdothum joined #git
15:58 peepsalot just in the workspace?
15:58 _ikke_ peepsalot: with --soft, reset only changes the branch where HEAD points to, and nothing else
15:59 bencevans joined #git
15:59 _ikke_ is that clear?
16:00 peepsalot ok
16:00 Ravana joined #git
16:01 peepsalot but does what i said about --mixed make sense?
16:01 dminuoso joined #git
16:01 flaviodesousa joined #git
16:02 _ikke_ with --mixed, git updates the index to the commit you reset to, so that basically *unstages* the changes
16:02 dreiss joined #git
16:02 howdoi joined #git
16:03 _ikke_ with git reset --soft HEAD~, you move HEAD back a commit, but the changes are left in the index
16:03 _ikke_ (ie. staged)
16:03 _ikke_ with git reset [--mixed] HEAD~, HEAD is moved back a commit, and the changes from the previous HEAD are unstaged
16:04 thiago joined #git
16:04 rominronin joined #git
16:04 vassagus_ joined #git
16:04 MattMaker joined #git
16:05 sdothum joined #git
16:06 vassagus joined #git
16:06 ochorocho__ joined #git
16:06 danslo joined #git
16:07 peepsalot ah ok.  i guess i was sorta picturing it wrong in my head, like when you commit, i thought was picturing the index as getting emptied or cleared.  so in reality the index is just another pointer to a node in the tree?
16:08 _ikke_ The index is a special 'structure'
16:08 _ikke_ A single file, but it contains caching information, and pointers to blobs
16:08 vassagus_ joined #git
16:09 whomp joined #git
16:09 _ikke_ and other information git uses
16:11 Silenced joined #git
16:13 compeman joined #git
16:13 nitric joined #git
16:15 peepsalot ok so from what you're saying, its very much like i thought at the very beginning.  "git reset --soft" is similar to doing "git reset --mixed; git add ." (assuming your working directory is clean)
16:15 enufn joined #git
16:16 compeman joined #git
16:16 _ikke_ yeah, that last assumption is important
16:17 compeman joined #git
16:19 MattMaker joined #git
16:21 EvilPeng1 joined #git
16:22 blackwind_123 joined #git
16:22 Pasi joined #git
16:24 sdothum joined #git
16:25 Pyrogerg joined #git
16:26 Zamyatin joined #git
16:26 Dong__ joined #git
16:26 star_prone joined #git
16:29 Dong__ Got an git internal question, when you clone a very small repo, all the objects was still sent by the smart protocol as a pack stream, yet git client did not save it in pack file rather expand into loose objects, where is the threshold defined for this behavior?
16:29 Ubulo joined #git
16:30 drupald joined #git
16:30 romerocesar joined #git
16:32 ansimation joined #git
16:35 Electrometro joined #git
16:35 MattMaker joined #git
16:35 crose joined #git
16:35 roger_raymond joined #git
16:35 mycelium joined #git
16:36 mischat joined #git
16:37 perrier-jouet joined #git
16:38 grawity fetch.unpackLimit
16:41 rnsanchez_ joined #git
16:42 synthroid joined #git
16:44 Dong__ grawity: genius! Thanks!
16:46 jaafar joined #git
16:49 synthroid joined #git
16:49 Dong__ grawity: mind if I ask a question on thin pack?
16:49 grawity mind if I ask you to just ask the actual question?
16:49 whomp joined #git
16:49 MattMaker joined #git
16:50 Dong__ sorry. When a thin pack is sent over, suppose it is ofs_delta inside, then what would be the offset mean?
16:50 durham joined #git
16:51 Dong__ i haven't actually saw a thin pack yet, no idea about what it looks like.
16:54 ruby_ joined #git
16:55 jrnieder joined #git
16:55 ochorocho__ joined #git
16:56 nikivi joined #git
16:56 whomp joined #git
16:56 mikecmpbll joined #git
16:57 dminuoso left #git
16:58 clemf joined #git
17:00 perrier-jouet joined #git
17:01 blunaxela joined #git
17:01 Kopitaja joined #git
17:02 nd_ joined #git
17:02 rorro joined #git
17:03 PocketKiller joined #git
17:03 PocketKiller left #git
17:03 derpsauce__ so, my git status says that I deleted some files but I would like to revert to a point where those files weren't deleted
17:03 rominronin joined #git
17:03 grawity check out those files from HEAD
17:04 Vampire0_ joined #git
17:05 MattMaker joined #git
17:05 wogi joined #git
17:06 wogi_ joined #git
17:06 derpsauce__ that failed
17:06 derpsauce__ since I've renamed my repo since I last pulled, how do I checkout/pull to a particular filename?
17:07 grawity the repo's name doesn't matter
17:07 grawity the entire commit history is kept right here in .git/
17:08 grawity so it should be just a matter of using the right checkout command
17:08 grawity do you have any *other* changes that you'd like to keep
17:08 shinnya joined #git
17:08 derpsauce__ Everything I'm doing is failing and making things worse. I just want the repo as it was at the last commit in origin
17:09 grawity git reset --hard origin/master
17:09 derpsauce__ this is possibly exactly what I was looking for
17:10 grawity though `git reset --hard HEAD` might have worked, too
17:10 thiago joined #git
17:11 Darren__ joined #git
17:12 alex1a joined #git
17:13 whomp joined #git
17:15 dan2k3k4 joined #git
17:15 fuchstronaut joined #git
17:17 autoboxer joined #git
17:17 aard_ joined #git
17:18 felixjet joined #git
17:18 wavis joined #git
17:19 MattMaker joined #git
17:19 vrishab Hi.
17:20 cagmz joined #git
17:20 vrishab I clone a remote git read-only repo. Make modifications on the local master and commit them in local master. Sync from origin/master and merge the changes.
17:21 vrishab Now my local master has local + origin changes. I am planning to work on a new feature / bug. I changes that I have commited in my local master are conflicting with them. How can I get a exact copy of origin/master to start my work ( assuming I don't need to do a git clone )
17:21 thebloggu joined #git
17:23 vrishab "I changes I have commited" -> "The changes I have commited"
17:25 _ikke_ What do you mean with getting an exact copy of origin/master ?
17:25 vrishab _ikke_, meaning the current origin/master without any of my local changes, since it is totally new feature.
17:26 _ikke_ well, origin/master already should not contain your changes
17:27 vrishab yes, I don't want to do a 'git clone' ( if that's what you mean ). Is there a way to avoid cloning the repo ?
17:27 jaafar joined #git
17:27 _ikke_ There is no need to do a git clone
17:27 star_prone joined #git
17:28 _ikke_ But tell me, why do you think origin/master is not exactly what you fetched last from the remote?
17:28 _ikke_ Or are you asking how you can get an up-to-date version?
17:29 Mobutils joined #git
17:30 vrishab _ikke_, I am making a change which will affect a lot of files. Some of these files have already been modified as part of another feature and the patches sent for review. It will take time for them to be reviewed ( some may need to be reworked )
17:31 nidr0x joined #git
17:31 vrishab _ikke_, If I make modifications on top of my local changes, my patches will not be valid in origin/master
17:34 laumann joined #git
17:34 vrishab I just need a local branch which exactly the same as 'git clone'
17:35 psprint_ joined #git
17:35 vrishab without any of my commits to my local repo.
17:35 Pasi joined #git
17:35 MattMaker joined #git
17:37 Vampire0 vrishab, git checkout -b feature origin/master
17:38 hanthings joined #git
17:39 Sasazuka joined #git
17:40 vrishab Vampire0, exactly what I needed. thanks !
17:41 ahmedelgabri joined #git
17:41 vrishab _ikke_, am new to git. so .....
17:42 ShekharReddy joined #git
17:43 GodGinrai joined #git
17:45 dviola joined #git
17:46 crayon joined #git
17:47 mrx88 joined #git
17:47 duderonomy joined #git
17:50 jerematic joined #git
17:50 jbitdrop joined #git
17:50 Vampire0 vrishab, yw
17:51 elastix joined #git
17:53 jason237 joined #git
17:55 MattMaker joined #git
17:55 Noldorin joined #git
17:57 inflames joined #git
17:58 multi_io joined #git
17:59 morenoh149 joined #git
18:01 fedepad joined #git
18:02 leeN joined #git
18:02 Pasi joined #git
18:02 derpsauce__ So, if I wanted to add a file to a different branch, how would I do that?
18:03 whomp joined #git
18:04 fuchstronaut joined #git
18:05 _ikke_ derpsauce__: Can you give more context?
18:05 derpsauce__ I added a file to a branch, I want to add it to another one
18:06 _ikke_ Most common way is to merge that one branch into the other branch
18:07 leeN joined #git
18:07 derpsauce__ But then I'd need to ignore a whole host of files that I didn't want to add
18:07 realz joined #git
18:09 MattMaker joined #git
18:10 jimi_ joined #git
18:10 _ikke_ Ok
18:10 ExoUNX how crazy it to run a database through git?
18:10 alex1a joined #git
18:10 ExoUNX that typical or not recommended
18:11 _ikke_ another option is git checkout <branch> && git checkout <otherbranch> <file>
18:11 _ikke_ ExoUNX: I would not recommend it
18:12 Zilenc3 joined #git
18:12 madewokherd "database" can mean almost anything, I could see using git for some text-only "databases"
18:13 madewokherd it really depends on your use case
18:14 ahmedelgabri joined #git
18:15 cdg joined #git
18:17 djb-irc joined #git
18:19 netcarver joined #git
18:19 glebihan_ joined #git
18:19 jeffl35 joined #git
18:20 roger_raymond joined #git
18:20 Zodd joined #git
18:21 glebihan_ joined #git
18:22 Xam is branch a type of reference with its name as an identifier? and does it always sit on the latest commit of its branch?
18:22 aard_ joined #git
18:22 Xam is branch a type of reference with its name as an identifier? and does it always sit on the latest commit of its path?
18:23 grawity yes, and usually
18:23 crayon joined #git
18:23 grawity (among refs, branches are special in that Git always updates them when you commit)
18:23 grawity but other than that, *all* refs (including branches) point to a single specific commit – not to "latest"
18:24 Zamyatin joined #git
18:24 Xam so the only way (afak) to move branch identifier (eg. master) is to use reset?
18:24 Xam since moving it to previous commits mean you will eventually move to a different commit than the successive commits?
18:25 MattMaker joined #git
18:25 mikecmpbll joined #git
18:25 Xam -->(c1)-->(c2)-->(c3-HEAD)
18:26 Xam git reset head^^ --hard
18:26 wavis left #git
18:26 Xam -->(c1-HEAD)-->(c2)-->(c3)
18:26 _ADN_ joined #git
18:27 Xam so c2 and c3 are no longer part of the main branch?
18:27 Xam abandoned
18:27 Xam ?
18:28 Xam but if you checkout any commit using its hash
18:28 grawity there are other ways to update a branch, or a generic ref
18:29 guampa joined #git
18:29 perrier-jouet joined #git
18:29 grawity regarding your example,
18:29 grawity -->(c1)-->(c2)-->(c3 <- master <- HEAD) would be more correct
18:29 mingrammer joined #git
18:29 jeffl35 joined #git
18:29 grawity a `git reset` would change it to -->(c1 <- master <- HEAD)-->(c2)-->(c3)
18:30 SteffanW2 joined #git
18:30 Xam i see your point
18:30 grawity but a `git checkout` would only result in -->(c1 <- HEAD)-->(c2)-->(c3 <- master); i.e. a "detached HEAD"
18:30 Xam it both checks out two steps and moves the head
18:31 Xam what happens to when you checkout a previous commit?
18:31 canuk joined #git
18:31 Xam does it insert the commit after it and before the successive commit
18:31 grawity no
18:31 grawity the history is not a straight line, it's a graph
18:32 Xam so what happens?
18:32 grawity if you check out an old commit, and start committing new stuff from there, you get a second branch
18:32 Xam what will it name be?
18:32 Xam what its name will be I mean
18:32 canuk Is it possible to do a git diff between repo1/some-folder and repo2/some-folder?
18:32 grawity it's an unnamed one (an anonymous head), which can later be named (using `git branch foo`) or just left as is (at risk of getting garbage-collected)
18:32 cagedwisdom joined #git
18:32 synthroid joined #git
18:33 Xam now i understand what branch does
18:33 madewokherd HEAD is always a name for whatever's currently checked out
18:33 Xam it doesn't create a path per se, it creates an identifier that takes a different path than the other one on the same node
18:34 grawity I suppose
18:34 _ikke_ It builds a structure
18:34 grawity branches, and refs in general, are nothing more than names for (pointers to) a specific commit in the graph
18:35 HappyLoaf joined #git
18:35 Xam how valuable is git experience?
18:35 _ikke_ Many companies use git nowadays
18:35 canuk as a programmer/dev? almost mandatory
18:36 grawity well, half of it is general DVCS experience, which (IMHO) is required for any developer
18:36 _ikke_ right
18:36 grawity [or, well, *just* VCS experience for that matter.]
18:36 mikecmpbll joined #git
18:36 _ikke_ Workflow and habbits
18:36 Xam I see
18:36 grawity the specific tools, be it Git or Hg or Plastic or whatever, aren't as important IMHO
18:37 Xam thanks everyone and especially grawity
18:37 Meera joined #git
18:38 grawity canuk: you can `git fetch` repo2 into repo1
18:38 grawity canuk: then diff master:some-folder FETCH_HEAD:some-folder, for example
18:40 rahul-kumi joined #git
18:40 stamina joined #git
18:42 psprint What cool action can be done on a tracked file, besides Diff?
18:43 _ikke_ You can checkout a commit
18:43 _ikke_ s/commit/file
18:43 _ikke_ you can blame a file
18:43 _ikke_ you can remove a file
18:43 _ikke_ you can (un)stage a file
18:43 KevinMGranger you can slice a file! you can dice a file! All for just 19.99!
18:44 _ikke_ You can commit changes to a file
18:44 _ikke_ You can hash a file
18:44 Repox joined #git
18:45 MattMaker joined #git
18:46 psprint It's a rational question I have a cool Zsh plugin https://asciinema.org/a/0idkrbui455ubyhrcg0d7snnm
18:47 psprint _ikke_: how about "HEAD~1_Diff" action, diffing to previous commit, sensible?
18:48 _ikke_ sure
18:48 rnsanchez_ joined #git
18:48 _ikke_ psprint: Looks cool
18:48 psprint blame is nice
18:49 psprint thanks. You set "CurrentProject" and have that WallOfKnowledge under Ctrl-O Ctrl-U
18:49 dan2k3k4 joined #git
18:49 psprint any more cool buttons would be great
18:50 _ikke_ So it basically generates the commands to execute a certain action, right?
18:50 psprint yes, first does diff --name-only
18:50 Fuzzmz mmm, git bisect + auto testing to see which commit broke something is awesome
18:51 NeverDie joined #git
18:51 ig0r_ joined #git
18:52 raijin joined #git
18:52 cagedwisdom joined #git
18:53 _ikke_ psprint: Do you use any kind of zsh 'framework' ?
18:54 psprint _ikke_: it's my 1.3 year code base, most of all it uses curses module, so I can draw textual UI. As for framework I recommend my own, Zplugin :]
18:54 _ikke_ ah ok
18:54 _ikke_ I'm still using oh-my-zsh
18:54 psprint there's my plugin there, zsh-navigation-tools
18:54 psprint also Zcurses
18:56 clemf joined #git
18:56 _ikke_ They're not bundled with omz?
18:56 psprint Zcurses is binary zsh module
18:57 psprint by "Also Zcurses" I meant: ZNT is also using Zcurses
18:57 psprint git diff HEAD\~1 a-file doesn't necessarily return diff, If file changed e.g. HEAD~2 away ..
18:58 satifant joined #git
18:59 LebedevRI joined #git
19:00 MattMaker joined #git
19:00 cagedwisdom joined #git
19:00 LebedevRI hi. some time ago i asked about git-gui regression, that it no longer signs new commits. was there any changes in this regard, any workaround to make it sign?
19:00 Electrometro joined #git
19:02 rorro joined #git
19:02 psprint git log -1 -p shows the diffs for single file
19:02 stompleee joined #git
19:02 sergiodj joined #git
19:03 psprint _ikke_: by "You can commit changes to a file" you maybe meant "you can commit just single file without staging, unstaging, a short path"? Rather not, but..
19:04 stompleee joined #git
19:04 Groscheri joined #git
19:04 _ikke_ psprint: You can, but it's not often that I use that
19:05 _ikke_ psprint: Note that you were asking thing you could do to a file
19:05 _ikke_ So I just listed things that could be done to a file
19:05 hweaving joined #git
19:05 hweaving Can anyone tell what git GUI this is just from a screenshot?  http://gitready.com/images/squash1.png
19:05 fuchstronaut joined #git
19:05 psprint yes and thanks! so there is a short-path to commit a file without whole staging-unstaging process? but indeed, it's rather not often needed
19:06 hweaving My best guess is SmartGit, but I'm not sure.
19:06 rahul-kumi joined #git
19:06 _ikke_ psprint: correct
19:06 _ikke_ git commit <file> would ignore the current index, and just commit that file
19:07 _ikke_ hweaving: I haven't used that many GUIs, so I couldn't tell
19:07 star_prone joined #git
19:07 TinkerTyper joined #git
19:09 Pasi joined #git
19:09 durham joined #git
19:09 Reujea0_ joined #git
19:10 Reujea0_ Hey I am tring to cheery pik this https://android.googlesource.com/kernel/commo​n/+/bc84b4adb1469e3d05ad76c304a4c545feaf1f88 into my repository https://github.com/Packmaa​n7144/kernel_nvidia_s8515 ; how do I do that?
19:10 stomplee joined #git
19:11 _ikke_ dfsdfsdfsdfps
19:11 DonRichie joined #git
19:11 _ikke_ ~,.
19:11 _ikke_ ~>
19:11 _ikke_ sorry for the noise :-(
19:11 Reujea0_ ?
19:11 hweaving Cat on keyboard?
19:11 _ikke_ No, connection broken
19:14 wogi joined #git
19:14 wogi_ joined #git
19:15 MattMaker joined #git
19:15 circ-user-f7FIN joined #git
19:16 ramsub07 joined #git
19:18 psprint _ikke_: to not just generate commands, I could do some processing and dig up something useful about a file, but no idea what could it be
19:19 vrishab Reujea0_, http://stackoverflow.com/questions/9339429/wh​at-does-cherry-picking-a-commit-with-git-mean
19:20 rhizome is there a way to create a branch with its own remote, and no history from master?
19:20 rhizome or should it simply be a separate repo?
19:20 _ikke_ You can create an orphan branch
19:20 mkoskar joined #git
19:21 _ikke_ and add a different remote which it pushes to by default
19:21 rhizome i've branched off current master and set up the remote for the branch, so i guess my remaining question is about history. this is for opensourcing part of a closed codebase.
19:22 NeXTSUN joined #git
19:22 _ikke_ !orphan
19:22 gitinfo To create an orphaned branch(unrelated to existing history): `git checkout --orphan newbranchname`. This will leave your index/worktree as-is(use `rm .git/index; git clean -dfx` to delete EVERYTHING). If you get 'error: unknown option `orphan`' see !orphan_old. For an empty/null commit see !orphan_null.
19:24 mingrammer joined #git
19:24 rhizome nice, thanks. if I can ask about strategy, it seems i should get master (or an arbitrary branch?) into the state the orphan should start at, branch the orphan, then reset master back to a commit restoring its previous state?
19:24 _ikke_ I think so
19:25 rhizome cool, thanks a lot.
19:26 mikecmpbll joined #git
19:26 Fuzzmz in all honesty, I'd just use a separate repo for the open source part
19:27 Fuzzmz this way it's harder to accidentally leak future parts of the closed codebase
19:27 rhizome i've thought of that. the closed stuff is really just additions as an implementation of the codebase, i.e. content, styling, etc.
19:28 rhizome so should be easy to keep separate (i hope, in relative ignorance). plus i'm the only dev
19:28 Fuzzmz oh, ok
19:29 rhizome essentially an open-core model, if you're familiar with that emerging buzzword
19:29 Fuzzmz yeah
19:30 MattMaker joined #git
19:30 bmahe joined #git
19:31 AnarchyAo joined #git
19:31 kpease joined #git
19:31 dsantiago joined #git
19:32 DonRichie joined #git
19:34 patarr joined #git
19:34 pigeon joined #git
19:36 ringzer0 joined #git
19:36 kpease joined #git
19:37 johnmilton joined #git
19:37 tang^ joined #git
19:38 jaequery joined #git
19:38 hweaving left #git
19:39 Vampire0_ joined #git
19:40 ochorocho__ joined #git
19:40 MattMaker joined #git
19:41 lolmac joined #git
19:42 qws-user-1228 joined #git
19:43 Gitzilla joined #git
19:44 theskillwithin joined #git
19:46 qws-user-1229 joined #git
19:49 geoid joined #git
19:49 satbyy joined #git
19:50 stamina joined #git
19:50 sergiodj hi there.  is there a way to autosquash ("autofix", actually, since I don't care about the commit message) a range of commits non-interactively?  I have a huge repository that I want to shrink to just the last 100 commits, and doing that interactively is impractical
19:50 hashpuppy joined #git
19:51 Gsham joined #git
19:51 patarr joined #git
19:51 _ikke_ sergiodj: Did you create fixup commits, or are these just normal commits?
19:52 sergiodj _ikke_: normal commits
19:52 sergiodj I didn't know about the fixup commits back then
19:53 patarr joined #git
19:53 sanketdg joined #git
19:53 _ikke_ ok
19:53 _ikke_ How do you decide what commits to fixup?
19:54 sanketdg joined #git
19:54 sergiodj _ikke_: I'll get the 101st most recent one, and fixup everything between root+1 and it
19:54 fees joined #git
19:55 MattMaker joined #git
19:55 sergiodj and I'll script that, so that I can "clean up" the repository every week or so
19:55 ramsub07 joined #git
19:55 crayon joined #git
19:56 crayon joined #git
19:59 Pasi joined #git
20:00 cdg joined #git
20:00 ramsub07 joined #git
20:01 Vampire0_ joined #git
20:02 nwkj86 joined #git
20:04 Pasi joined #git
20:05 djb-irc joined #git
20:06 fuchstronaut joined #git
20:10 MattMaker joined #git
20:13 netj joined #git
20:13 _ikke_ sergiodj: EDITOR='awk "{ if (NR > 1 && NR < 5) sub(\"pick\", \"edit\"); print }"' git rebase -i ...
20:13 _ikke_ replace 5 with the number of commits
20:15 sergiodj _ikke_: sweet, let me give it a try.  thanks
20:15 EgS joined #git
20:16 Peng joined #git
20:16 aswen joined #git
20:16 patarr joined #git
20:16 cspotcode joined #git
20:19 mattcen joined #git
20:20 qws-user-1228 joined #git
20:21 daio joined #git
20:21 daio hey all is it possible to commit to two repositories in parallel
20:21 daio say I wanted to make sure my project is hosted in 2 seperated places
20:21 Apteryx joined #git
20:22 Apteryx Hi! Are git submodules tracking the master branch by default now?
20:23 _ikke_ daio: You commit into the local repository
20:23 _ikke_ daio: you can then (in two steps) push to two repositories
20:23 _ikke_ Apteryx: Not that I know
20:23 daio _ikke_, take it I just change the master and then push
20:25 _ikke_ You can commit on any branch and push that branch
20:25 Vampire0_ joined #git
20:25 MattMaker joined #git
20:25 Apteryx _ikke_: OK. Thanks.
20:26 hashpuppy i forked a repos.  i edited the history and did a git push --force (i'm the only one).  now i'm trying to create a pull request, but it says there is no common history.  when i do git rebase -i upstream/develop it seems to create the history twice on top of the original upstream/develop history.  should i be running something else?
20:26 Vampire0 daio, well, you can add an arbitrary number of remote repositories and push to those
20:26 daio Vampire0, perfect thank you
20:27 sergiodj _ikke_: I've had to do some adjustments to your solution, but the idea works fine.  thanks again
20:27 Vampire0 hashpuppy, you probably have to do what the section "how to recover from upstream rebase" at man git rebase tells you to come to a good state again
20:27 gitinfo hashpuppy: the git-rebase manpage is available at http://jk.gs/git-rebase.html
20:27 hashpuppy thanks, Vampire0
20:28 Vivekananda joined #git
20:28 Vampire0 daio, afair you can only push to one remote at a time, but you can easily create either a Git alias or shell alias that does the two pushes as one command sequentially
20:28 javaJake Vampire0: push --all?
20:29 javaJake Oh, nvm, one branch, two rmotes
20:31 Vampire0 javaJake, exactly, --all just pushes all branches to one remote :-)
20:32 daio so I want to really add a master-gitlab master-private
20:32 daio and just push to both one at a time
20:32 _ikke_ sergiodj: How did you solve it?
20:33 Vampire0 daio, yes, or you make an alias that allows you to push to both with issuing one command
20:33 laumann joined #git
20:34 ahr3n joined #git
20:34 rgrinberg joined #git
20:35 sergiodj _ikke_: I've had to use GIT_SEQUENCE_EDITOR (to mess only with the sequence of commits), and this variable expects a script that accepts one argument (the filename of the sequence).  then, I've adjusted awk to do an inplace replacement, and that's it
20:35 sergiodj probably worth a blog post, given the amount of time I've spent looking for a solution
20:37 Vampire0 daio, something like            git config alias.pushall '!pushall() { git push master-gitlab "$@" && git push master-private "$@"; }; pushall'           but that is straigt from my brain untested
20:38 daio Vampire0, I will give it a whirl thank you :)
20:38 Gsham joined #git
20:38 mrx88_ joined #git
20:40 MattMaker joined #git
20:40 mingrammer joined #git
20:43 derpsauce__ how to I fix HEAD detatched from origin/MASTER
20:45 _ikke_ sorry
20:45 _ikke_ Connection is flaky
20:45 moritz derpsauce__: in the easiest case, git checkout master
20:46 jaafar joined #git
20:47 Vampire0 derpsauce__, !zomb
20:47 gitinfo derpsauce__: [!zombie] Brainzz! More brainzzzz!  For information about detached heads, see !detached
20:47 nilg joined #git
20:48 gfixler joined #git
20:49 mrx88 joined #git
20:49 Gsham joined #git
20:49 raijin joined #git
20:49 snthsnth joined #git
20:50 derpsauce__ I have "uncommitted changes"
20:52 aidalgol joined #git
20:52 likewhoa joined #git
20:53 Vampire0 joined #git
20:54 cjohnson derpsauce__: well, those changes you have
20:54 cjohnson are you trying to commit them or blow them away
20:54 Samual joined #git
20:55 nikivi joined #git
20:55 MattMaker joined #git
21:00 bmahe joined #git
21:01 sin_ joined #git
21:03 sergiodj left #git
21:04 _ng joined #git
21:04 nb_bez___ joined #git
21:05 Blkt joined #git
21:06 _ikke_ joined #git
21:06 MattMaker joined #git
21:07 fuchstronaut joined #git
21:07 fuchstronaut joined #git
21:08 cdg joined #git
21:09 navidr joined #git
21:09 raijin joined #git
21:10 Juzzika joined #git
21:11 flopsec joined #git
21:12 archaic joined #git
21:13 zivester joined #git
21:15 mingrammer joined #git
21:15 mozzarella joined #git
21:17 autoboxer joined #git
21:19 dodobrain joined #git
21:20 isysd joined #git
21:24 rendar joined #git
21:30 dodobrain joined #git
21:30 ujjain joined #git
21:30 ujjain joined #git
21:31 RichiH joined #git
21:34 Atlantic777 joined #git
21:34 Atlantic777 joined #git
21:36 DonRichie joined #git
21:38 jaequery joined #git
21:39 Bombe joined #git
21:40 clemf joined #git
21:44 psprint _ikke_: found something interesting: git show some-branch:some-file.js. Shows a file as it is in other branch. Could be used to do diff current_branch - selected_branch
21:44 _ikke_ Why not gi diff currentbranch selectedbranch <file>?
21:45 a3Dman joined #git
21:45 psprint that's better
21:46 psprint could use what you wrote for diff, and show as a checkout > tmp_file ; $EDITOR tmp_file
21:47 _ikke_ Why would you want to do the latter?
21:47 ramsub07 joined #git
21:47 _ikke_ To see the file in an editor?
21:48 psprint yes.. a quick "side-checkout"
21:51 pullphinger joined #git
21:52 olalonde joined #git
21:53 olalonde how to squash all commits  that haven't been pushed to origin yet into a single commit?
21:54 tokkee joined #git
21:55 MadHatter42 joined #git
21:56 Vampire0 olalonde, !perf
21:56 gitinfo olalonde: [!postproduction] So, you want to make your commit history look pretty before pushing? http://sethrobertson.github.com/GitPostProduction talks you through how to use 'rebase -i' to do this.
21:58 Lor_ joined #git
21:58 pullphinger joined #git
21:58 dpower joined #git
21:58 canuk joined #git
21:59 Sasazuka joined #git
21:59 nidr0x joined #git
21:59 ShekharReddy joined #git
22:01 NeXTSUN joined #git
22:01 olalonde was hoping for a shorter answer,but thanks :P
22:01 __main__ joined #git
22:02 blr joined #git
22:02 e14 joined #git
22:02 aapaaal joined #git
22:03 cjohnson rebase -i
22:03 cjohnson that's the shorter answer
22:03 cjohnson :)
22:03 psprint _ikke_: this idea is better. I read about a "wipe" operation: Commit All; Reset Hard HEAD^1; – changes are still in Git Object, until gc, in case one changes mind about the. You said about git commit file. So Wipe-File is possible: commit a file, move HEAD back 1 entry. The file is available (cat-file? but rather not on commit's SHA), in case one changes mind. How to move HEAD to ^1? Soft reset?
22:03 Vampire0 olalonde,      git rebase -i @{u}         then    s/^pick /s/    and you're done
22:04 Sna4x8 joined #git
22:04 Zodd joined #git
22:05 tyreld joined #git
22:06 Zodd does rebase alter commits between dev and master?
22:07 Sna4x8 In SVN, when there's a conflict there is an option to postpone.  This leaves the user with a series of files to resolve.  For example: file.js.r220, file.js.r223, file.js.r228, file.js.mine.  I found this very easy to resolve, because I could vimdiff two files at a time, from oldest to newest, until all files are identical.
22:07 meLon joined #git
22:07 meLon joined #git
22:07 Sna4x8 Is there a similar way to resolve conflicts with git?  I'm using vimdiff as my mergetool, and right now I get four buffers (and I understand what each is and the workflow for merging).
22:08 fuchstronaut joined #git
22:08 therealGent joined #git
22:09 justanotheruser joined #git
22:10 m0viefreak joined #git
22:10 danimal2 joined #git
22:11 sangy joined #git
22:11 avar So svn will present you with all versions from when you divirged for conflict resolution?
22:11 jeffreylevesque joined #git
22:11 Sna4x8 Correct.
22:12 avar Doesn't that produce a lot of falso positives where some previous version of the file may have conflicted with your latest version, but subsequent changes may have resolved it?
22:12 osse Sna4x8: you can produce the files easy enough
22:12 andlabs joined #git
22:12 Sna4x8 osse: True, I can just save each buffer to a file.
22:12 Zodd joined #git
22:13 Sna4x8 avar: I haven't had that experience, no.
22:13 avar In any case, git doesn't do this natively because for that reason it's usually a bad idea, but you can just rebase/merge in a loop against each commit that your tree diverges from
22:13 Sna4x8 avar: I think SVN would resolve that situation automatically.
22:13 osse Sna4x8: those won't be the exactly the same though. the difftool will only show you the versions from the base and the branches. not all version in between
22:13 Zodd does rebase just takes the whole branch until the common ancestor commit (without taking it) and append it to master?
22:13 DonRichie joined #git
22:13 avar I.e. if you diverged at A, and there's A B C D E F, git will rebase/merge on top of F, but you can just as easily in a for-loop first rebase/merge against B, then C, then D etc.
22:14 Zodd avar: are you answering my question?
22:14 osse Zodd: the answer is yes
22:15 osse Zodd: it will take all commits in the range master..HEAD
22:15 osse except meregs
22:15 avar Zodd: My message came 2 seconds after yours, I don't type that fast, no :)
22:15 Zodd i got disconnected
22:16 reim joined #git
22:16 jaequery joined #git
22:16 Zodd so the use case for rebase is when you consider the dev a more prominent path than master?
22:17 Zodd when you think of dev as beginning to be more prominent than master/
22:17 Zodd ?
22:17 osse the use case is when you want to pretend all the commits in the branch to the side were at the end
22:17 osse so it becomes a straight line
22:17 Zodd hmm
22:18 Zodd i feel like merge is when you approve features to be added to master whereas rebase is when to paths are competing to assume master
22:18 osse you feel wrongly
22:18 Vampire0 zodd, !perf shows some use-cases for rebase
22:18 gitinfo zodd: [!postproduction] So, you want to make your commit history look pretty before pushing? http://sethrobertson.github.com/GitPostProduction talks you through how to use 'rebase -i' to do this.
22:19 Zodd ouch
22:19 Sasazuka__ joined #git
22:19 osse Zodd: https://raw.githubusercontent.com/Osse​/git-stuff/master/merge-vs-rebase.txt
22:20 Zodd thnx
22:23 Urbany joined #git
22:24 mattcen joined #git
22:24 avar /w 6
22:24 avar WINDOW SIX
22:24 osse RAINBOW SIX?!
22:27 nettoweb joined #git
22:28 BlackPanx joined #git
22:28 jaequery joined #git
22:29 WizJin joined #git
22:29 Zodd REBASE: After git checkout feature; git rebase master:
22:29 Zodd ?
22:30 osse i mean after you run those commands
22:30 LeBlaaanc joined #git
22:30 Zodd oh
22:30 Zodd i got it
22:30 Zodd so where from osse?
22:31 florian_ joined #git
22:33 autoboxer joined #git
22:34 madewokherd joined #git
22:34 buffal0 joined #git
22:35 King_Hual_ joined #git
22:38 osse guy asked me why i help, what i do and where i'm from in private :S
22:38 snthsnth joined #git
22:38 spacelord joined #git
22:39 LeBlaaanc joined #git
22:40 crazycoder joined #git
22:40 crazycoder hello
22:40 gitinfo crazycoder: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.
22:40 crazycoder i have removed the remote tags
22:40 crazycoder but when i try to list them i see https://gist.github.com/anonymous​/ad6dc9e50f84f2a9937a8732d86c3939
22:40 crazycoder how is it possible that they still there?
22:40 qws-user-1228 joined #git
22:42 ningu joined #git
22:43 ningu what command should I use if I want to know if a particular text string occurs in any committed file in any branch in the whole repository?
22:43 ningu git log -S seems like a possibility but not sure
22:43 osse that would do the trick
22:43 ningu ok, cool
22:43 LeBlaaanc joined #git
22:44 osse or git rev-list --all | while read hash; do git grep -e 'particular string' $hash; done
22:44 osse but that might give maaaany matches
22:44 ningu hmm, ok
22:44 ningu I will try that just to be sure
22:44 Sna4x8 avar, osse: Thanks for the help earlier.
22:45 osse crazycoder: that's just how tags work. you have to delete them by hand
22:45 osse crazycoder: the easy thing is to delete all  tags, then fetch all tags
22:45 crazycoder osse
22:45 crazycoder what command can i use to remove all tags
22:45 crazycoder locally and remotly
22:45 crazycoder ?
22:45 inflames joined #git
22:45 crazycoder remotely
22:46 osse git tag -d
22:46 osse you seem to know how to remove them remotely since you said you just did
22:46 crazycoder yah only to confirm
22:46 crazycoder :d
22:47 osse push --delete
22:47 fuzzmz joined #git
22:48 ningu left #git
22:49 bmahe joined #git
22:53 actinide joined #git
22:54 spacelord joined #git
22:59 raijin joined #git
23:00 autoboxer joined #git
23:02 avar Sna4x8: np
23:03 mingrammer joined #git
23:04 reim joined #git
23:05 red82 joined #git
23:05 mizu_no_oto joined #git
23:05 YottaByte joined #git
23:06 YottaByte hi y'all, learning git, if I want to undo my changes and go to source, is "git checkout -- myfile" the right way to do it?
23:06 mablae_ joined #git
23:06 osse yes
23:06 osse git status will tell you
23:06 YottaByte can I do multiple files deliminted by commas?
23:07 osse no
23:07 YottaByte I've also been seeing when I do "git pull" and I have local changes to files, it'll say pull aborted because I need to merge
23:07 osse delimited by whitespace
23:07 YottaByte oh nice
23:07 osse just like everything else on the CLI
23:07 Gitzilla joined #git
23:07 YottaByte what's CLI?
23:07 osse command line
23:08 YottaByte ah
23:08 YottaByte how about the git pull merge thing?
23:08 jeffl35 joined #git
23:08 fuchstronaut joined #git
23:09 osse do what git says
23:09 YottaByte I don't know how though
23:09 YottaByte i thought git pull would merge for me automatically
23:09 YottaByte that's what svn update would do
23:09 YottaByte (unless there were conflicts)
23:09 osse it says "Please, commit your changes or stash them before you can merge."
23:10 ochorocho__ joined #git
23:10 YottaByte oh, so I'd do "git add -a", "git commit -m "whatever I changed"" and then "git pull"?
23:11 osse for example
23:11 YottaByte and that won't get pushed up until I do "git push"?
23:11 osse correct
23:11 YottaByte it's starting to come together...
23:11 YottaByte ty
23:11 Prisoner-2460_1 joined #git
23:11 snthsnth joined #git
23:12 jeffl35 joined #git
23:12 jeffl35 joined #git
23:12 jeffl35 joined #git
23:13 ap0c joined #git
23:16 andygrossman joined #git
23:16 dhollinger joined #git
23:16 red82 joined #git
23:18 Rodya_ joined #git
23:22 geoid joined #git
23:23 ochorocho__ joined #git
23:23 cyan__ joined #git
23:24 Milijus joined #git
23:25 _ADN_ joined #git
23:26 dsdeiz joined #git
23:26 dsdeiz joined #git
23:27 ayjay_t joined #git
23:29 NeverDie joined #git
23:32 btree joined #git
23:33 nikivi joined #git
23:33 ochorocho__ joined #git
23:33 Milijus joined #git
23:36 e14 joined #git
23:37 btree left #git
23:39 MrWoohoo joined #git
23:41 ToBeCloud joined #git
23:43 therealGent joined #git
23:43 harish joined #git
23:44 ochorocho__ joined #git
23:44 faux___ joined #git
23:44 safe joined #git
23:45 buffal0 joined #git
23:46 dreiss joined #git
23:48 raijin joined #git
23:51 Kaisyu joined #git
23:51 jaafar joined #git
23:52 NeXTSUN joined #git
23:53 JamesBaxter joined #git
23:54 RedNifre_ joined #git
23:54 mand0 joined #git
23:57 mingrammer joined #git
23:58 dsdeiz_ joined #git
23:58 ochorocho__ joined #git

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