Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2017-04-17

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

All times shown according to UTC.

Time Nick Message
00:00 TbobbyZ joined #git
00:00 tnecniv joined #git
00:05 hotpancakes joined #git
00:06 hotpancakes I just came across this: "If you find a bug, the worst thing you can do is fix it in the just-released version, and not fix it in master. To avoid it, you always fix forward. Fix it in master, then cherry-pick it into another patch-release branch." Can anyone tell me what's meant by that last sentence?
00:07 hotpancakes What would "cherry-picking it into another patch-release branch" do?
00:09 MattMaker joined #git
00:10 dreiss joined #git
00:11 ToBeCloud joined #git
00:13 durham joined #git
00:15 mizu_no_oto joined #git
00:15 scarabx joined #git
00:15 ali1234 hotpancakes: it means having a stable branch and a developement branch ("master")
00:16 ali1234 you do all development work in the master branch, and then you backport bug fixes only to the stable branch
00:17 ali1234 cherry-pick is a convenience tool for copying a commit to a different branch
00:18 TheSimonator joined #git
00:18 hotpancakes joined #git
00:22 ali1234 hotpancakes: "git cherry-pick" is literally a command, if you didn't realise, so you can read the help to see exactly what it does
00:22 MaxSan left #git
00:28 theiostream joined #git
00:30 sunri5e joined #git
00:32 juvenal joined #git
00:36 tsetair_ joined #git
00:37 Cabanossi joined #git
00:38 alexandre9099 joined #git
00:39 tnecniv joined #git
00:43 mellernoia joined #git
00:44 madewokherd joined #git
00:44 pietv joined #git
00:46 jnewt_ joined #git
00:47 justan0theruser joined #git
00:48 mizu_no_oto joined #git
00:49 telephone joined #git
00:49 telephone joined #git
00:56 fstd_ joined #git
00:56 unplayednamer joined #git
00:57 circ-user-8ns7x joined #git
00:59 Nugget joined #git
01:01 hotpancakes joined #git
01:03 telephone joined #git
01:03 telephone joined #git
01:04 pks joined #git
01:04 MattMaker joined #git
01:08 cdown_ joined #git
01:08 trip joined #git
01:09 cqi joined #git
01:11 pks joined #git
01:13 a3Dman joined #git
01:14 durham joined #git
01:14 dsdeiz_ joined #git
01:15 juvenal joined #git
01:23 hotpancakes joined #git
01:26 Lyqyd_ joined #git
01:27 hotpancakes joined #git
01:28 peterbec` joined #git
01:31 drodger joined #git
01:37 Cabanossi joined #git
01:38 inflames joined #git
01:41 TheSimonator joined #git
01:43 peterbec` joined #git
01:47 hotpancakes joined #git
01:47 Kaisyu joined #git
01:49 ilbot3 joined #git
01:49 Topic for #git is now Welcome to #git, the place for git help. Just git help, apparently. | Git pages at jk.gs offline due to domain glitch, hopefully to be fixed on Monday | Public logs at http://jk.gs/glog | First visit? Read: http://jk.gs/git | Current stable version: 2.12.2 | Getting "cannot send to channel"? /msg gitinfo .voice | This is serious business. https://madeitwor.se/git-jokes
01:50 theiostream joined #git
01:50 jameser joined #git
01:51 CussBot joined #git
01:51 circ-user-8ns7x joined #git
01:55 MattMaker joined #git
01:56 mizu_no_oto joined #git
02:01 trip joined #git
02:03 raijin joined #git
02:03 sunri5e joined #git
02:06 chigang joined #git
02:08 SwiftMatt joined #git
02:10 tmcmahon joined #git
02:12 jameser joined #git
02:14 canoon joined #git
02:15 cqi joined #git
02:15 durham joined #git
02:15 durham joined #git
02:17 durham joined #git
02:20 durham joined #git
02:20 ujjain joined #git
02:20 ujjain joined #git
02:21 durham joined #git
02:24 justanotheruser joined #git
02:25 ojdo joined #git
02:25 durham joined #git
02:29 dreiss joined #git
02:30 wgrant joined #git
02:31 telephone joined #git
02:31 telephone joined #git
02:31 rchavik joined #git
02:36 clemf joined #git
02:38 SwiftMatt joined #git
02:39 juvenal joined #git
02:44 lb1c joined #git
02:45 preyalone joined #git
02:46 pks joined #git
02:47 circ-user-8ns7x joined #git
02:47 aavrug joined #git
02:51 fakenerd joined #git
02:52 Vortex35 joined #git
02:52 safe joined #git
02:53 peterbec` joined #git
02:56 lagothri1 joined #git
02:58 fakenerd joined #git
02:58 TikityTik joined #git
02:59 darkbit joined #git
03:01 kyan joined #git
03:02 TikityTik joined #git
03:02 juvenal joined #git
03:02 chachasmooth joined #git
03:03 hexagoxel joined #git
03:06 peterbec` joined #git
03:12 inflames joined #git
03:12 jameser joined #git
03:14 aidalgol joined #git
03:16 mizu_no_oto joined #git
03:16 Andrew_K joined #git
03:26 nexemacs joined #git
03:29 grayhatter joined #git
03:29 grayhatter joined #git
03:30 grayhatter joined #git
03:30 grayhatter joined #git
03:33 armyriad joined #git
03:35 boombatower joined #git
03:38 TbobbyZ joined #git
03:42 raynold joined #git
03:48 onehrxn joined #git
03:50 a3Dman joined #git
03:51 Cabanossi joined #git
03:52 jost__ joined #git
03:53 circ-user-8ns7x joined #git
03:55 TikityTik joined #git
03:56 Andrew_K joined #git
03:57 ewitz_ joined #git
03:57 pks joined #git
03:57 tnecniv joined #git
04:01 fakenerd joined #git
04:09 jameser joined #git
04:11 peterbec` joined #git
04:13 qt-x joined #git
04:14 svm_invictvs joined #git
04:21 Cabanossi joined #git
04:25 PioneerAxon joined #git
04:26 rosek joined #git
04:26 SwoleGymBro joined #git
04:31 SwoleGymBro joined #git
04:34 SwoleGymBro joined #git
04:39 SwoleGymBro- joined #git
04:40 rosek joined #git
04:41 freddy_ joined #git
04:41 ayogi joined #git
04:47 rosek joined #git
04:49 telephone joined #git
04:49 telephone joined #git
04:53 sbulage joined #git
04:56 qt-x joined #git
04:59 t-mart joined #git
05:00 fakenerd joined #git
05:00 freddy_ joined #git
05:01 finalbeta joined #git
05:05 diogenese joined #git
05:06 cqi joined #git
05:09 MattMaker joined #git
05:09 TikityTik joined #git
05:09 MattMaker joined #git
05:10 MattMaker joined #git
05:13 peterbec` joined #git
05:21 cyberz joined #git
05:21 tsetair joined #git
05:23 pks joined #git
05:29 TbobbyZ joined #git
05:30 svm_invictvs joined #git
05:33 TbobbyZ joined #git
05:33 circ-user-8ns7x joined #git
05:34 bannakaffalatta joined #git
05:35 qt-x joined #git
05:36 fakenerd joined #git
05:36 Cabanossi joined #git
05:38 SwiftMatt joined #git
05:43 shadow-x joined #git
05:50 svm_invictvs joined #git
05:51 qt-x joined #git
05:53 nick123 joined #git
05:54 freimatz joined #git
05:54 qt-x1 joined #git
05:59 fakenerd joined #git
06:03 drodger joined #git
06:04 learning joined #git
06:05 usm4n joined #git
06:07 jameser joined #git
06:08 drodger joined #git
06:08 michael1 joined #git
06:14 pks joined #git
06:17 basiclaser joined #git
06:22 qt-x joined #git
06:25 usm4n joined #git
06:26 circ-user-8ns7x joined #git
06:31 ogny joined #git
06:34 govg joined #git
06:35 ogny joined #git
06:36 rokups joined #git
06:36 Cabanossi joined #git
06:38 fakenerd_ joined #git
06:40 jameser joined #git
06:41 durham joined #git
06:42 fakenerd joined #git
06:42 Skyfinn joined #git
06:46 michael1 joined #git
06:46 overlord_tm joined #git
06:49 murii_ joined #git
06:54 TomyLobo2 joined #git
06:57 dsdeiz joined #git
06:57 dsdeiz joined #git
07:05 mkoskar joined #git
07:08 learning joined #git
07:09 peterbec` joined #git
07:10 Silenced joined #git
07:10 auzty joined #git
07:10 davimore joined #git
07:11 selckin joined #git
07:11 fakenerd_ joined #git
07:16 fakenerd joined #git
07:21 Cabanossi joined #git
07:22 masuberu joined #git
07:27 fakenerd joined #git
07:36 schleppel joined #git
07:37 YuGiOhJCJ joined #git
07:45 SwiftMatt joined #git
07:49 r0ry joined #git
07:51 leah2 joined #git
07:52 mentor joined #git
07:52 mentor joined #git
07:56 alexggor1on joined #git
07:56 oleo joined #git
07:56 kegster joined #git
08:01 marcogmonteiro joined #git
08:01 Silenced joined #git
08:03 navidr joined #git
08:05 dreiss joined #git
08:20 King_Hual joined #git
08:21 leah2 joined #git
08:22 Introoter joined #git
08:22 CheckDavid joined #git
08:23 qt-x joined #git
08:23 jstimm joined #git
08:25 finalbeta joined #git
08:29 nevodka joined #git
08:35 freddy_ joined #git
08:40 peterbec` joined #git
08:41 D4R5C0D3 joined #git
08:46 rominronin joined #git
08:47 ZyX-I joined #git
08:47 ispn joined #git
08:48 ayogi joined #git
08:50 SwoleGymBro- joined #git
08:51 fakenerd joined #git
08:51 robotroll joined #git
08:59 fakenerd joined #git
09:06 freddy_ joined #git
09:09 Skyfinn joined #git
09:10 thesquib joined #git
09:13 mikecmpbll joined #git
09:20 Forlorn joined #git
09:21 Cabanossi joined #git
09:22 Cabanossi joined #git
09:22 telephone joined #git
09:22 telephone joined #git
09:23 ayogi joined #git
09:25 Cabanossi joined #git
09:26 MineCoins joined #git
09:28 cqi joined #git
09:28 jimi_ joined #git
09:28 intellix joined #git
09:34 learning joined #git
09:46 rominronin joined #git
09:47 pietv joined #git
09:49 manuelschneid3r joined #git
09:53 SwiftMatt joined #git
09:55 imack joined #git
09:58 courrier joined #git
09:58 vuoto joined #git
10:02 Taikibansei joined #git
10:05 SwiftMatt joined #git
10:08 fakenerd_ joined #git
10:10 telephone joined #git
10:11 telephone joined #git
10:12 afuentes joined #git
10:16 scoobertron joined #git
10:24 hegge joined #git
10:25 satifant joined #git
10:31 ziombelas joined #git
10:32 PioneerAxon joined #git
10:35 nikow joined #git
10:36 SwoleGymBro joined #git
10:39 circ-user-8ns7x joined #git
10:40 peterbec` joined #git
10:42 zefferno joined #git
10:48 Karazhan joined #git
10:50 netj joined #git
10:58 m0viefreak joined #git
10:58 jozwior joined #git
11:02 TbobbyZ joined #git
11:04 HurricaneHarry joined #git
11:05 aruns joined #git
11:07 fakenerd joined #git
11:11 Tatou_ joined #git
11:12 Tatou_ left #git
11:12 Tatou joined #git
11:12 Tatou Help me understand something
11:12 Tatou If I: git clone https://github.com/MyName/MyRepo, that works
11:13 Tatou But git clone https://bitbucket.org/myname/myRepo fails.. why?
11:13 fakenerd_ joined #git
11:13 selckin go ask bitbucket
11:14 darkbit joined #git
11:14 Tatou grrr
11:14 bremner Tatou: you could start by giving an actual error message.
11:15 bremner it's probably still bitbucket specific, but without any information, it's hard to say.
11:16 Tatou bremner: I think it was quite obvious .. It was clearly a mercurial repo <_<
11:16 rnsanchez joined #git
11:16 bremner not sure if serious...
11:16 idlemind joined #git
11:16 selckin Tatou: i'd like to award you big asshole of the day award
11:16 Tatou This channel is far too aggressive.
11:17 selckin aye go away
11:17 Tatou jeez
11:17 Tatou No wonder svn is beating git :/
11:18 aruns joined #git
11:18 evil-god joined #git
11:19 evil-god i7c: hi my best friend
11:20 bremner Tatou: ok, that's enough. If you have actual questions fine, but otherwise take the ranting elsewhere
11:24 indistylo joined #git
11:25 dreiss joined #git
11:27 miczac joined #git
11:32 telephone joined #git
11:32 telephone joined #git
11:34 Tatou bremner: It was clearly a joke.
11:35 learning joined #git
11:35 Tatou left #git
11:36 Cabanossi joined #git
11:36 peacememories joined #git
11:38 marcogmonteiro joined #git
11:38 incredible joined #git
11:39 scottjl joined #git
11:42 peterbec` joined #git
11:43 oleo joined #git
11:43 scottjl joined #git
11:44 fakenerd joined #git
11:46 King_Hual joined #git
11:47 jameser joined #git
11:52 peacememories joined #git
11:53 kegster joined #git
11:56 leeN joined #git
11:56 peacememories joined #git
11:57 johnflux_ joined #git
11:57 johnflux_ Hey all
12:00 synthroid joined #git
12:01 johnflux_ I'm trying to merge.  How can I automatically resolve conflicts using 'theirs'
12:02 johnflux_ but I want to use theirs for conflicting chunks - not the whole file
12:04 rivarun joined #git
12:06 ogny joined #git
12:07 King_Hual` joined #git
12:08 rorro joined #git
12:15 Nilesh_ joined #git
12:17 sarri joined #git
12:17 sarri joined #git
12:17 earnestly joined #git
12:18 diogenese joined #git
12:18 Acerific joined #git
12:22 robattila256 joined #git
12:22 dave0x6d joined #git
12:23 fakenerd joined #git
12:28 acetakwas joined #git
12:30 s1scha joined #git
12:30 Tobbi joined #git
12:32 Introoter joined #git
12:34 mrcrjs joined #git
12:35 mrcrjs left #git
12:36 perlpilot joined #git
12:38 jeffreylevesque joined #git
12:38 dc joined #git
12:39 aavrug joined #git
12:43 peacememories joined #git
12:43 zerow joined #git
12:44 mentor joined #git
12:48 jagob joined #git
12:48 dsdeiz joined #git
12:48 dsdeiz joined #git
12:48 jonascj_ joined #git
12:51 livingstn joined #git
12:51 cdown joined #git
12:53 SCHAPiE joined #git
12:53 mvensky joined #git
12:54 peacememories joined #git
12:54 TheSimonator joined #git
12:55 ecthiender joined #git
12:56 p4trix joined #git
12:56 dglambert joined #git
12:56 dglambert yo
12:56 ecthiender Hi, can I take a git diff from one repo and then apply the patch file in a different repo? I mean ofcourse its the same code with same directory structure, but two different repos and different history.
12:56 jonascj_ Hi guys, relatively new to anything but "git clone" and "git pull". I've pushed a commit to a remote github repo, but I forgot to pull those changes on another machine before I made some other changes and committed those. Now of course I can not push the second changes.
12:56 BlaXpirit ecthiender, sure, go for it
12:57 jonascj_ Should I just merge (that happens if I now try "git pull"), or is there another more elegant solution? The chanegs are not in conflict, so nothing to merge really.
12:57 BlaXpirit jonascj_, yeah you can merge  and get a merge commit,  or you can run   git pull --rebase
12:57 ecthiender BlaXpirit, cool
12:57 BlaXpirit which will pull those other changes and reapply your commits on top of those
12:58 Topic for #git is now Welcome to #git, the place for git help. Just git help, apparently. | Public logs at https://gitirc.eu/log | First visit? Read: https://gitirc.eu | Current stable version: 2.12.2 | Getting "cannot send to channel"? /msg gitinfo .voice | This is serious business. https://madeitwor.se/git-jokes
12:58 ecthiender BlaXpirit, so when I apply the patch by running git apply, nothing happens.
12:58 jonascj_ BlaXpirit: so with no reason for a merge rebase is probably the prettiest?
12:59 BlaXpirit jonascj_, yeah
12:59 ecthiender it exits without printing any error and doesn't apply any of those changes
12:59 jast testing, ignore me...
12:59 jast man git-status
12:59 gitinfo the git-status manpage is available at https://gitirc.eu/git-status.html
13:00 Tobbi joined #git
13:00 MineCoins joined #git
13:02 topi`_ left #git
13:02 learning joined #git
13:02 jonascj_ Damn it, now I failed to abort the merge which I had initiated by "git pull", so now I have a local merge commit. Can I undo that?
13:03 BlaXpirit hm i'm wary of giving advice, something destructive could happen
13:03 Derperperd joined #git
13:03 BlaXpirit ok I got it -  please run  git log   and save the output somewhere just in case
13:03 marcogmonteiro joined #git
13:04 jonascj_ BlaXpirit: nah, it has not been pushed and the changes are minimal, I have considered just checking out the lastest version from remote and reapplying my changes manually
13:04 BlaXpirit there you can see the latest merge commit,  and you can also see the latest commit with just the changes you want to push
13:04 BlaXpirit you can run  git reset --hard COMMIT_ID    where you can copy commit id of that commit^
13:05 BlaXpirit and if this goes wrong, you can run  git reset --hard COMMIT_ID   copying the commit id of the merge commit - which will at least return you to where you are right now
13:05 jonascj_ alright, I'll give git reset --hard a try, will it leave files unchanged (not that it really matters)?
13:05 jast you don't have to save the log, git does it for you automatically (in the reflog)
13:06 jast reset --hard destroys uncommitted changes and restores the file versions from the commit you specify
13:06 BlaXpirit jonascj_, no, it does not leave files unchanged,  it hard resets everything to what it was at that commit
13:06 BlaXpirit so you can then run  git pull --rebase
13:07 dermoth joined #git
13:07 jonascj_ BlaXpirit: ah yes, thank you, it seemed to work out as wished :)
13:07 jonascj_ it was unchanged in my case since the merge did not do anything to the files I was working on, but of course not unchanged overall
13:08 oskarkv joined #git
13:09 juvenal joined #git
13:10 peacememories joined #git
13:11 mentor On continuing a rebase after resolving a conflict, is there any way to tell git that I don't want to edit the commit message?
13:12 jast when continuing a rebase git doesn't normally even let you edit the commit message
13:13 jast assuming you use 'git rebase --continue' and not 'git commit' (which isn't the "correct" way to continue a rebase)
13:13 mentor Erm
13:13 mentor On continuing a rebase after resolving a conflict in a fixup action
13:13 mentor ?
13:13 jast ah. :) that's a more interesting question...
13:14 jast my guess is that if you do get an editor, it's a bit of an oversight in the implementation of fixup
13:14 jast as I recall, 'fixup' grew out of 'squash' (which regularly opens a commit message editor when it's done)
13:15 mentor Yeah, it's like it turns into a squash
13:16 jast I'll take a quick look at the code
13:16 mentor I'm just doing a test
13:17 whitby To jast's previous comment, you might be able to use git commit --no-edit instead of --continue
13:17 whitby But I'd need to test that
13:17 BlaXpirit i do this often but robably havent noticed this effect. I dont mind the editor, it's just an extra chance to bail out if you notice something wrong
13:17 whitby Even though it's the the "correct" way
13:17 jameser joined #git
13:18 borkr joined #git
13:18 drodger joined #git
13:19 relipse joined #git
13:19 jast man, rebase has so many moving parts
13:20 vkareh joined #git
13:22 mentor Really? I got the impression from using it that it had a clean, elegant implementation…
13:22 jast it has, in principle, but at some point someone started porting some of it to C, and now some of it is shell script and some of it is not
13:23 jast it looks like someone inserted code in sequencer.c for processing interactive rebase that always asks for a commit message except when using the 'edit' action
13:23 jast there is more code for dealing with interactive rebase further down but that new code seems to always deal with it first and then return without letting the other code get a say
13:24 jast no, I misread that bit actually
13:24 DarkPsydeLord joined #git
13:25 jast but yeah, --continue with interactive rebase seems to always spawn an editor except if the entry being processed is of type 'edit'
13:25 gregor2 joined #git
13:25 jast well, actually it just adds another flag for 'edit', the meaning of which isn't completely apparent from the context :)
13:26 CheckDavid joined #git
13:28 jast never mind, all of this code isn't being used in the latest released version yet
13:31 fakenerd_ joined #git
13:31 mentor https://pastebin.com/ebcvkNUG
13:32 concerti joined #git
13:32 mentor Little useful information there
13:33 brit411_ joined #git
13:36 Cabanossi joined #git
13:37 Bennux joined #git
13:37 mvensky joined #git
13:39 rastos2 joined #git
13:40 kegster joined #git
13:44 peterbec` joined #git
13:44 Xenophon1 joined #git
13:47 livingstn joined #git
13:47 shinnya joined #git
13:49 ToBeCloud joined #git
13:49 CrowX- joined #git
13:51 CrowX- when two users work on the same file, and both of them want to make a push, other than one person making a push and the other one pulls those changes and afterwards adds his own changes, is there any other simpler way for GIT to automatically merge changes of two users?
13:52 Ryanar joined #git
13:52 jast you can pull after locally committing your changes, too
13:52 Ryanar joined #git
13:52 jast but, well, the changes do have to be combined at some point :)
13:53 cjohnson CrowX-: somebody has to take responsibility for the merge, even if it can be done automatically
13:53 cjohnson which it can't always
13:53 Es0teric joined #git
13:54 CrowX- cjohnson, so there is no automatic way provided by Git, correct?
13:54 CrowX- or is there?
13:54 Dougie187 joined #git
13:54 jast nothing more automatic than that
13:55 cjohnson no because like I said, merges cannot automatically be done all the time
13:55 cjohnson so when there's a conflict how would such a mechanism work?
13:55 cjohnson I have all my devs use a common set of git aliases and the push alias includes a pull first to "automate" the issue
13:55 jast also, automatic merges can result in invalid code, so it's a good idea to manually review them
13:56 guampa joined #git
13:57 CrowX- so the process is this: 1. user works on code.   2. user copies his code.  3. user makes a pull. 4. user adds back his changes.     5. user makes a push.
13:57 CrowX- correct?
13:57 jast CrowX-: steps 2 and 4 are unnecessary
13:58 jast work on code. commit. pull. push.
13:58 TheSimonator joined #git
13:58 Es0teric joined #git
13:59 bannakaffalatta joined #git
13:59 CrowX- jast, won't pull overwrite local changes?
13:59 govg joined #git
13:59 CrowX- if both users have edited the same file
13:59 jast no, it never overwrites anything. it combines.
14:00 jast and if uncommitted changes are in the way it will not actually do anything, just to be safe
14:00 CrowX- jast, so if two users have edited the same file, git will combine the changes automatically?
14:00 Soni joined #git
14:00 jast if possible, yes
14:00 jast it can't always be done automatically, so occasionally 'git pull' will prompt the user to do some extra manual work
14:01 jast if you're new to using pull/merge and friends, read the output carefully so you don't miss anything important :)
14:01 CrowX- 'possible' being if users edited different lines of the code that do not conflict with each other?
14:01 jast yes
14:01 CrowX- cool
14:01 CrowX- thanks
14:01 jast you're welcome
14:02 Soni can I write a "git remote helper" in sed?
14:03 ayogi joined #git
14:03 jast is it possible in sed? no idea. if it's possible, does Git care what it's written in? not really.
14:04 ExoUNX joined #git
14:05 mentor joined #git
14:08 AaronMT joined #git
14:08 lavalamp joined #git
14:11 unbalancedparen joined #git
14:11 LeBlaaanc joined #git
14:13 lewix joined #git
14:13 cdg joined #git
14:14 Isla_de_Muerte joined #git
14:15 oneliner joined #git
14:15 raijin joined #git
14:17 HardlySeen joined #git
14:18 TikityTik joined #git
14:19 ResidentBiscuit joined #git
14:19 oneliner hello there, my git bash from git-scm.com (win10-64 ver 2.10.1) stopped working one day, I am trying unnistall/reinstall and now install dialog froze (full bar, no message, over five min)
14:20 ski7777 joined #git
14:20 oneliner alt tab shows two windows, i wonder if a dialog box is stuck behing the unmovable git install window
14:22 monjwf_ joined #git
14:22 gws joined #git
14:22 gws left #git
14:22 gws_ joined #git
14:25 gws_ My .gitignore works for (ignores) a pattern if that pattern is a directory, but not if the pattern is a Windows symbolic link. Is there something I can do to my ignore pattern to make git actually ignore the presence of the link?
14:26 selckin whats the pattern
14:27 gws_ just the name of the folder/link
14:27 monjwf joined #git
14:27 jast append a '/'
14:27 gws_ BillingCenter/modules/configuration/gtest/sandbox/suite
14:27 selckin i was thinking to drop a /
14:27 jast oh
14:28 jast I misread your question
14:28 gws_ I tried it both ways : /
14:28 selckin windwos symlinks (junction?) are so special, i don't know
14:28 jast symlinks internally have a .lnk suffix
14:29 jast junctions act like normal directories, so should be fine
14:29 circ-user-8ns7x joined #git
14:29 jast so you may want to add a ..../suite.lnk pattern, too
14:29 gws_ The application won't know what to do with a shortcut, and git is tripping over Windows's specialness
14:29 gws_ Oh, you mean it's showing up from the APIs as suite.lnk? I'll give that a shot.
14:30 jast .lnk is hidden by the windows UI even if you have enabled showing all other file extensions
14:30 jast it takes advanced trickery to unhide it
14:30 gws_ Trickery I'm okay with, but not sure I have the time for the advanced course.
14:30 jast :)
14:32 gws_ No luck. .../suite.lnk and .../suite* still have the link showing up as a new untracked file.
14:32 jast dang
14:33 jast well, who knows what extra magic the emulation layers (msys, cygwin or whatever version you have) are doing...
14:33 gws_ Even running 'git status' as Administrator, which I thought might get around the problem - Windows restricts creating (and presumably manipulating) symlinks by default
14:33 gws_ Just git for Windows
14:33 cjohnson Easy answer: burn the windows installation for the ground
14:33 cjohnson hahaha
14:33 jast yeah, that's msys
14:33 thebope joined #git
14:33 gws_ cjohnson: I heartily agree but it's not my computer to burn, so
14:33 jast oh, you may be talking about a different type of symlink, then
14:34 cjohnson gws_: yeah I'm just kidding I know it's not a helpful suggestion :P But in all seriousness this is why I always suggest doing dev in a linux VM
14:34 jast "Symbolic links are transparent to users; the links appear as normal files or directories, and can be acted upon by the user or application in exactly the same manner."
14:34 jast LIES
14:34 cjohnson classic msft
14:34 gws_ Actually I tried both mklink /d (symbolic link to folder) and mklink /j (hard link to folder, AKA junction point)
14:35 kbeflo joined #git
14:35 gws_ Okay so at this point it sounds like this isn't a well-known issue and I should ask the mailing list
14:35 crose joined #git
14:35 cjohnson just for shits and giggles try making it as a real folder and stuff
14:36 cyberz joined #git
14:36 cjohnson confirm that your pattern is as you expect
14:36 gws_ As a real folder, it's ignored.
14:36 cjohnson ha!
14:36 Mewho joined #git
14:36 cjohnson one last thing, how are you inspecting the created link, via cmd or via explorer?
14:36 cjohnson maybe explorer is hiding some extra parts of the filename
14:36 gws_ So the pattern is right for normal things, but I wondered if there's a magic syntax for links
14:37 jast I don't think so
14:37 raijin joined #git
14:37 rorro joined #git
14:37 cjohnson no because the entire point of links is that they are transparent
14:37 gws_ it looks like a folder shortcut (arrow icon overlay) in Explorer. CMD dir says the name "suite" with the full name of the target in brackets after
14:37 oneliner windows completely screwed up my winbash, cute
14:38 oneliner was perfectly fine
14:38 cjohnson gws_: are you sure it's not actually a shortcut?
14:38 oneliner then one day, poof
14:38 Mewho_ joined #git
14:38 jast cjohnson: way ahead of you here
14:38 gws_ "2017-04-17  09:56    <JUNCTION>     suite [c:\core\guidewire\Release\4.2\PolicyCenter\modules\configuration\gtest\sandbox\suite]"
14:39 jast I thought it was a shortcut first, because all the different names confuse me. it wasn't.
14:39 cjohnson weird
14:39 gws_ I tried it different ways, this is after 'mklink /j'
14:40 Ryanar joined #git
14:41 Mewho_ Hi, I was wondering if an rpm is available for git 2.12
14:41 scottjl joined #git
14:41 gws_ Ultimately this is Perforce's fault. I just want to share some common non-production code between different products in a suite but Perforce won't let me map the same depot path to more than one local folder. The fun stuff always comes when you're trying to come up with a workaround for something else...
14:42 vrach joined #git
14:42 jast Mewho_: there are no official rpms. in any case, there are many different distributions that use RPM with different base dependencies, you're going to have to be more specific. :)
14:42 peacememories joined #git
14:42 Mewho_ Rpm for Rhel/CentOS
14:43 gws_ I do have a fallback plan (script to copy the suite folder from one module to the others), but that's ugly.
14:43 jast which version, though? centos 6 packages aren't gonna work well on centos 5, for instance
14:43 Mewho_ Preferably 6
14:43 hotpancakes joined #git
14:43 oneliner let me fine tune my windows (reboot)
14:44 gws_ "fine-tune"  : p
14:45 peterbec` joined #git
14:46 gws_ thank you selckin, jast, cjohnson for your help!
14:46 cjohnson np, good luck!
14:46 simmel joined #git
14:47 Mewho_ 6.6 to be more specific
14:47 CrowX- joined #git
14:48 jast Mewho_: these folks have packages, but I can't tell you how reputable they are... https://ius.io/
14:48 Mewho joined #git
14:49 Mewho Thanks, I'll check it out
14:50 Es0teric joined #git
14:53 d10n-work joined #git
14:53 simmel What is 'HEAD' or 'HEAD~3' called? I'm looking for a similar shortcut for the first commit which isn't pushed. 'origin/master^' works but I wanna know if there are other 'shortcuts' which are useful.
14:53 jast man gitrevisions has 'em al
14:53 gitinfo the gitrevisions manpage is available at https://gitirc.eu/gitrevisions.html
14:53 selckin man git-rev-parse
14:53 gitinfo the git-rev-parse manpage is available at https://gitirc.eu/git-rev-parse.html
14:54 jast use my link! mine!!!! ;)
14:54 selckin aye
14:54 * cjohnson googles "git revision aliases" and clicks the link from google so jast doesn't get any referral revenue
14:54 jast there isn't a magical syntax only the first commit that isn't pushed
14:55 jast *for only
14:55 jast cjohnson: you're ruining my life here
14:55 * jast shakes fist
14:55 * simmel uses `man gitrevisions`
14:56 cjohnson can you use merge-base and compare master with origin/master ?
14:56 cjohnson or will that not always work
14:56 simmel Thanks jast and selckin!
14:56 orbyt_ joined #git
14:56 jast using the manpage will save you a lot of data traffic, too
14:56 jast cjohnson: sure, but then how do you say "one commit 'after' that"
14:57 jast it's easy to make a range with *all* unpushed commits, though
14:57 al-damiri joined #git
14:57 jast ... and to answer one remaining element of the question: !HEAD
14:57 gitinfo HEAD is a 'pointer' to the currently checked out branch (or commit, if HEAD is !detached). In bare repositories it tells clients which branch to checkout initially after cloning. Unlike commonly believed, HEAD is *not* something that exists separately for every branch. It also is *not* necessarily the newest commit in the repo (that's hard to define in a DVCS, anyway...)
14:58 rorro joined #git
15:01 peacememories joined #git
15:03 Es0teric joined #git
15:04 Darcidride joined #git
15:05 Kulrak joined #git
15:05 Kulrak joined #git
15:06 gws_ ...aaargh! I just created a new repo to show steps-to-reproduce on this Windows symbolic link problem, (can you guess what comes next?)
15:06 Cabanossi joined #git
15:07 bremner you installed linux?
15:07 gws_ (COMBO BREAKER!)
15:07 cjohnson it worked!
15:07 Starky joined #git
15:07 jast the universe imploded?
15:07 cjohnson glad I could solve it for you
15:07 gws_ Of course it works as expected in the new repo.
15:07 cjohnson round of applause for me
15:07 jast hah
15:07 cjohnson lol
15:08 gws_ git init, mkdir bar, echo foo > .gitignore, mklink /d foo bar, git status ... no unchanged files
15:08 cjohnson Just to confirm you were getting an *untracked* file and not a change to an already-tracked file?
15:08 cjohnson that's the only other thing I can think
15:09 gws_ right, because the link/folder didn't exist before
15:10 gws_ I need to go back and look again because I installed the latest git yesterday...Friday...whatever... and something in the notes mentioned changing the config of existing repos WRT symbolic links. I thought that was core.symlink, but maybe not.
15:10 madewokherd joined #git
15:11 durham_ joined #git
15:11 hN joined #git
15:11 publio joined #git
15:12 durham_ joined #git
15:14 r0ry joined #git
15:19 gregor2 joined #git
15:22 gkatsev_ joined #git
15:24 nexemacs joined #git
15:25 EDAKIRI joined #git
15:25 thiago joined #git
15:25 thiago joined #git
15:26 hotpancakes joined #git
15:27 ForeignBiscuit joined #git
15:28 Levin^ joined #git
15:28 Levin^ hi there
15:29 Levin^ I tried several approaches on trying to ban .DS_Store from my repository, but everything failed, it's still are being found to be tracked
15:29 Es0teric joined #git
15:29 Eugene Levin^ - you need to `git rm` them *and* add them to .gitignore
15:30 Eugene http://stackoverflow.com/questions/107701/how-can-i-remove-ds-store-files-from-a-git-repository
15:31 Levin^ hm I see
15:31 Levin^ ok thanks
15:32 darkmist_ joined #git
15:33 hotpancakes joined #git
15:33 Levin^ left #git
15:34 guampa joined #git
15:35 lektrik joined #git
15:35 portal_ joined #git
15:36 lektrik joined #git
15:36 Gustavo6046 joined #git
15:38 lektrik joined #git
15:38 miczac joined #git
15:40 theiostream joined #git
15:40 kegster joined #git
15:43 TikityTik joined #git
15:44 giovanni joined #git
15:47 portal___ joined #git
15:47 i7c evil-god: hallo.
15:47 imack joined #git
15:49 gregor2 joined #git
15:54 re1 joined #git
15:54 todd_dsm joined #git
15:54 svm_invictvs joined #git
15:55 YuGiOhJCJ joined #git
15:55 masuberu joined #git
15:56 mikecmpbll joined #git
15:58 phase3 joined #git
16:00 portal_ joined #git
16:01 synthroid joined #git
16:01 TheSimonator joined #git
16:02 synthroi_ joined #git
16:02 aruns joined #git
16:02 Levin^ joined #git
16:02 Levin^ hi there
16:03 Levin^ just a little question, what is the difference between (.*?) and (.*) ?
16:03 Levin^ they both select "everything", right? but the question mark does anything different?
16:04 qqx joined #git
16:05 preaction Levin^: is that a regex question?
16:05 Mobutils joined #git
16:05 Levin^ preaction: oops, entered a wrong channel
16:06 Levin^ ^^'
16:06 Levin^ left #git
16:07 durham joined #git
16:08 govg joined #git
16:09 nexemacs joined #git
16:10 raijin joined #git
16:12 cagedwisdom joined #git
16:12 nitric joined #git
16:14 sz0 joined #git
16:16 henriknj joined #git
16:16 ncthom91 joined #git
16:17 peepsalot joined #git
16:20 gws_ FWIW, * takes the longest possible match and *? takes the shortest possible match  #themoreyouknow
16:21 hotpanca_ joined #git
16:22 cjohnson "greedy"
16:29 MineCoins joined #git
16:29 peterbecich joined #git
16:30 bannakaffalatta joined #git
16:31 borkr joined #git
16:31 kadoban joined #git
16:32 raijin joined #git
16:32 gws_ Well, I found the problem with windows symlinks. The issue was with my .gitignore. I have a base version that's copied into each version branch of our application (don't ask/blame Perforce) when I set up the working copy and I had added my new ignore pattern to the template.
16:32 tristanp joined #git
16:32 gws_ The reason creating a directory named 'suite' didn't show up any changed files is 'status' doesn't care about empty directories.
16:32 gws_ It was a red herring.
16:33 portal_ joined #git
16:34 cjohnson hahahahahaha
16:34 cjohnson nice
16:34 cjohnson so other than that, it works as expected?
16:36 gws_ yes, now the shared directory is not noticed, which is what I wanted
16:38 gws_ Along the way I thought I maybe had to start with a brand-new repo to get out from under some setting from the previous installed version of git, I inited an empty one and imported all the commits from the original as a remote.
16:38 meLon joined #git
16:38 gws_ That got me all the history I had so far and hasn't shown any ill effects.
16:39 gws_ Didn't fix the real problem, but the technique could be useful to someone in, as Tom Lehrer said, a somewhat bizarre set of circumstances.
16:41 svm_invictvs joined #git
16:42 kadoban There's not that many ways to get your git repo screwed up badly enough that that's necessary, but there's a few obscure commands that require other obscure commands to undo I suppose. If it works, it's all good.
16:42 MACscr joined #git
16:42 acetakwas joined #git
16:47 peterbec` joined #git
16:47 kfoonamalik joined #git
16:49 orbyt_ joined #git
16:51 masuberu joined #git
16:52 seveneve1 joined #git
16:52 tnecniv joined #git
16:57 kegster joined #git
16:58 synthroid joined #git
16:58 pietv joined #git
17:00 shinnya joined #git
17:00 livingstn joined #git
17:01 nd_ joined #git
17:03 evil-god huhu i7c
17:04 TikityTik joined #git
17:05 diogenese joined #git
17:08 tyreld joined #git
17:08 sbeller joined #git
17:11 Es0teric joined #git
17:12 j_kota joined #git
17:14 zerow joined #git
17:17 gugah joined #git
17:19 tnecniv joined #git
17:19 howdoi joined #git
17:20 zerow joined #git
17:20 happycoffeebean joined #git
17:22 boombatower joined #git
17:23 yogesh_ joined #git
17:23 tnecniv joined #git
17:24 gunnaro joined #git
17:25 thiago joined #git
17:25 sq-one joined #git
17:26 Silenced_v2 joined #git
17:29 duderonomy joined #git
17:30 gopar joined #git
17:31 circ-user-8bIZ6 joined #git
17:33 cdown_ joined #git
17:35 Pseudocrat joined #git
17:35 Pseudocrat joined #git
17:37 Discovery joined #git
17:38 marianina8 joined #git
17:38 matoro joined #git
17:38 rasto joined #git
17:39 relipse joined #git
17:40 guampa joined #git
17:41 govg joined #git
17:44 pwnz0r joined #git
17:46 lavalamp joined #git
17:47 jrnieder joined #git
17:47 seveneve3 joined #git
17:47 peterbec` joined #git
17:51 kpease joined #git
17:51 dglambert joined #git
17:51 dglambert yo
17:53 drodger joined #git
17:56 kpease joined #git
17:59 lambada355 joined #git
18:00 NeXTSUN joined #git
18:01 planigan joined #git
18:01 guampa joined #git
18:02 rominronin joined #git
18:02 Sasazuka joined #git
18:06 Cabanossi joined #git
18:10 Mysterious_Light joined #git
18:12 TbobbyZ joined #git
18:12 CheckDavid joined #git
18:13 Mysterious_Light joined #git
18:13 ertesx joined #git
18:13 Mysterious_Light left #git
18:14 Gianormagantrous joined #git
18:15 zerow joined #git
18:15 NeverDie joined #git
18:16 Cavallari joined #git
18:17 _main_ joined #git
18:18 guampa joined #git
18:20 Masber_080 joined #git
18:21 xall joined #git
18:24 prg3 joined #git
18:25 jstimm joined #git
18:28 raijin joined #git
18:28 juvenal joined #git
18:30 gopar joined #git
18:34 mda1 joined #git
18:36 Ryanar joined #git
18:37 dglambert Hey, one of my git config files is hosted on a work network drive, this causes problems I believe when I am away from work
18:37 dglambert Is it possible to duplicated it and change the path to a local location
18:39 rwp dglambert, What do you mean when you say "is hosted on a work network drive"? Do you mean symlink'd to /net/foo or some such?
18:40 dglambert I don't think its a symbolic link
18:40 dglambert git config --list --show-origin returns, file:U://.gitconfig     merge.tool=kdiff3
18:40 dglambert for example.
18:41 dglambert and it is actually hosted there
18:41 ambro718 joined #git
18:42 dglambert So, i think that maybe my global one
18:42 dglambert so I want to move that to my user folder
18:43 Es0teric joined #git
18:44 Endarked joined #git
18:44 ambro718 Hi, can someone suggest a repository strategy for this problem: I have project A, which contains a mixture of generic code (A), specific code (B) and a framework (C). I want to separate the framework (C) into its own repository. But the framework relies on the generic code (A). The framework repository should include the generic code so it can live standalone. How to handle the resulting duplication of the generic code?
18:46 ambro718 I can easily make a script to scan the framework for parts of generic code that it depends on and automatically copy those into the framework repository
18:47 rwp dglambert, You can definitely undo that configuration and then everything would run locally. Because certainly when you are off network that will be unavailable.
18:48 dglambert How do I know for sure its my global config that is pointing there
18:48 dglambert Then, how to I remap my global config
18:49 rwp Sorry but I am not a ms-windows user so I don't know for your situation.  For me it would be ~/.gitconfig.  You will need to look around.  Probably an environment variable?
18:49 dglambert hmmm, I'll check the environment variables
18:51 dglambert No, don't think its an environment variable
18:52 rwp ambro718, Are you familiar with git submodules? !submodule
18:52 gitinfo ambro718: git-submodule is ideal to add subsidiary git repositories to a git superproject when you do not control the subprojects or more specifically wish to fix the subproject at a specific revision even as the subproject changes upstream. See http://www.git-scm.com/book/en/Git-Tools-Submodules
18:53 ambro718 rwp: yes. What specifically do you mean? To separate the generic code and link it as the submodule from the two current repositories (specific code B and framrwork C)?
18:53 finalbeta joined #git
18:53 juvenal joined #git
18:53 rwp Something like that. I don't have a specific plan. But no one else suggested anything. So it seemed like a good thing to ask.
18:54 rwp Sorry. No good ideas from me today.
18:54 ambro718 I want the framework to be as easy to use as possible and the use of the generic library transparent
18:54 tnecniv joined #git
18:55 ambro718 so maybe I can make release packages which bundle only the required parts of the library, but the git repo has a submodule linking to the full library?
18:57 definity joined #git
18:57 rasto joined #git
18:58 portal_ joined #git
18:59 gregor2 joined #git
18:59 tuguldur joined #git
19:01 drodger joined #git
19:02 definity Hi, I have added a file to my local repo, it was over 100mb, then deleted that file, then pushed. Now I no longer have that file in my repo, but every commit I do it still tries to upload the file and i get ' ! [remote rejected] master -> master (pre-receive hook declined)'
19:02 _ikke_ right, the file is still in your history
19:03 _ikke_ it's not enough to just remove it from the last commit, you'd also have to remove it from the previous commits that contain it
19:04 tapout joined #git
19:05 tuguldur joined #git
19:06 MattMaker joined #git
19:06 Ryanar joined #git
19:06 Cabanossi joined #git
19:08 Rodya_ joined #git
19:11 masuberu joined #git
19:11 zerow joined #git
19:12 Atm0spher1c joined #git
19:12 tnecniv joined #git
19:12 Anthaas joined #git
19:12 definity _ikke_: How can I do that?
19:14 Masber_080 joined #git
19:14 drodger joined #git
19:16 happycoffeebean joined #git
19:19 ok91 joined #git
19:19 lmatteis joined #git
19:19 shtrb|laptop joined #git
19:20 MineCoins joined #git
19:20 shtrb|laptop can I do a partial clone from a git repo ? (HEAD - X for example) ?
19:20 monq joined #git
19:21 ali1234 shtrb|laptop: --depth 1 might be what you want
19:21 shtrb|laptop thanks !
19:21 ali1234 or depth n
19:21 ali1234 i expect there's a way to ask for a range of commits too, but i dont know about it
19:22 shtrb|laptop depth would the do the trick for me (didn't wish to clone few GB just :) )
19:22 ferr joined #git
19:23 _ikke_ definity: Do you know when you added that file, was that recently?
19:23 shtrb|laptop yes today
19:24 lewix joined #git
19:25 hashpuppy joined #git
19:25 bongo joined #git
19:26 dglambert just chiming back in if any Windows guys are on that have had to remap there global config file
19:29 publio joined #git
19:30 definity _ikke_: About 3 weeks ago
19:31 afuentes joined #git
19:32 rwp definity, But you haven't been able to push since you added that file?  And so rewriting it out of your history is okay?
19:32 definity rwp: No I havent
19:32 definity Yeah
19:32 definity I have it backed up somewhere else
19:32 rwp definity, In that case if it were me I would make a branch of all of my current work. Everything will be safe on the branch. Then reset my local master back to before my bad commit
19:33 simmel left #git
19:33 rwp Then I would cherry pick or 'git rebase -i' from my local branch the changes I wanted that isn't the bad commit.
19:35 rwp Find a commit that is before your bad commit that is also on the upstream master branch. Then I would git reset to that commit. That will get your local working copy able to sync with the upstream master again.
19:35 rwp Then walk your local working copy forward with whatever changes you do want to push upstream. Add and commit them locally. Then push those upstream as you would normally.
19:37 raijin joined #git
19:37 suy_ left #git
19:37 suy joined #git
19:37 shtrb|laptop left #git
19:38 definity Okay thanks :)
19:39 stutter joined #git
19:40 stutter i've come to a position where I have to restore data from the .git/objects folder -- does anyone know of any scripts that exist that can facilitate rebuilding the tree + restoring the files within it from the objects folder? or am i looking at writing my own script to do it?
19:40 Eugene stutter - uh..... what? How did you end up with only an objects/ folder?
19:41 stutter monday ate me
19:41 Eugene LIterally just objects/ ? Nothing else in .git/ ?
19:41 stutter there's more, except it's a new repo
19:41 stutter has no commits
19:41 Eugene !fixup
19:41 gitinfo So you lost or broke something or need to otherwise find, fix, or delete commits? Look at http://sethrobertson.github.com/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history!
19:41 stutter i wrote a bunch of code and was creating the first commit
19:42 Eugene Did you `git add` everything at least ?
19:42 stutter i did a git add  ., which added everything to the index (obviously). aborted the add because i wanted to ignore some stuff...and when i composed the git ignore i did a git reset --hard afterwards, without thinking
19:42 stutter and it wiped out pretty much everything :(
19:42 stutter i did add everything, thankfully
19:43 Eugene jast - go pay your registrar
19:43 Eugene Normally I would like you to man git-fsck, but jast forgot to pay for the bot this decade
19:43 gitinfo the git-fsck manpage is available at https://gitirc.eu/git-fsck.html
19:43 bremner lol.
19:43 stutter lol
19:44 Eugene Oho, its already been moved elsewhere
19:44 Eugene stutter - does `git reflog` say anything?
19:44 stutter nope
19:44 stutter it's empty
19:44 stutter :D
19:45 nexemacs_ joined #git
19:45 Eugene This is a fun puzzle. So you need to fidn the tree object, but you don't have its hash, and no commits point to that hash
19:45 Eugene Try: `git fsck --unreachable` ?
19:45 stutter you could call it fun, sure, haha
19:45 stutter i do have a missing tree reference
19:45 Eugene !pastebin
19:45 gitinfo [!paste] Please use a pastebin for snippets longer than one line. https://gist.github.com and https://vomitb.in both work well.
19:45 definity I think i made it worse!
19:46 Eugene So if you haven't already you should `cp` what you have to a backup file
19:46 stutter https://gist.github.com/jimrubenstein/1ebff38e8f4bdbc68f3a8b54caf88196
19:46 stutter i cp'd a few times
19:46 stutter ;o
19:46 stutter there is no 4b825dc642cb6eb9a060e54bf8d69288fbee4904 in the objects folder
19:46 stutter maybe that's why i'm getting the 'missing tree'
19:46 stutter ;)
19:47 Eugene How about `git rev-list --objects --all`? Might be a bit spammy
19:47 stutter nothing
19:47 Eugene Well then your object database is empty
19:47 StephenS joined #git
19:48 StephenS what is the actual issue with this error: insufficient permission for adding an object to repository database .git
19:48 stutter Eugene sounds like the way my whole monday has gone
19:48 Eugene StephenS - usually its UNIX permissions from using a shared repo. Are you trying to do git-over-SSH with multiple users?
19:48 StephenS nope, simple git pull from my own shell
19:49 StephenS I could run this as a root, but thats a workaround
19:49 Eugene stutter - the best I can say is to paw through .git/objects/*/ by hand. The git object format is pretty trivial to pull data out of, man gitrepository-layout I think
19:49 gitinfo stutter: the gitrepository-layout manpage is available at https://gitirc.eu/gitrepository-layout.html
19:49 stutter Eugene yeah, i've used cat-file to list the contents of a bunch of the files
19:49 Eugene Or `git show` the ID
19:49 peterbec` joined #git
19:49 StephenS can't I just chown whole .git or any other idea?
19:49 stutter and writing a script to parse it and restore should be relatively "easy"
19:49 Eugene I would write a shell script to walk through objects/ and cat-file each one to a new working dir
19:49 stutter just gotta find a place to start, haha
19:49 Eugene Then figure ito ut by hand
19:50 Eugene StephenS - you don't end up with a permissions problem if you're using a sane umask and not doing group stuff. `ls -lha` the repository & pastebin
19:50 Eugene Hint: did you `sudo git` at some point?
19:51 ahrs joined #git
19:52 StephenS solved it
19:52 StephenS chowning .git to the local user fixes the issue
19:52 StephenS I saw some objects with root perms though, so I guess in some point pulling as root was issued
19:53 StephenS I did chown user:user -R .git/objects if anyone is interested :)
19:56 dvaske joined #git
19:57 Eugene Yeah, don't do what blogs say. `sudo git` is wrong 99% of the time
19:57 stutter indeed
19:58 Eugene Especially `git pull`, which is broken two ways: it doesn't use your SSH identity(root should not have SSH privkeys set up, wtf just no) and it breaks checkout
20:01 Hello71 sudo cd
20:02 TbobbyZ_ joined #git
20:06 theiostream joined #git
20:10 lewix joined #git
20:12 MrTrustor joined #git
20:13 theiostream joined #git
20:14 govg joined #git
20:19 peepsalot joined #git
20:19 stutter victory
20:23 yqt joined #git
20:24 cdown joined #git
20:26 MineCoins joined #git
20:26 shgysk8zer0 joined #git
20:28 lewiz joined #git
20:30 Es0teric joined #git
20:35 lewix joined #git
20:35 aidalgol joined #git
20:36 Cabanossi joined #git
20:37 TbobbyZ joined #git
20:38 ResidentBiscuit joined #git
20:40 dave0x6d joined #git
20:41 hashtagstrashtag joined #git
20:44 bodkan joined #git
20:49 bannakaffalatta joined #git
20:51 peterbec` joined #git
20:51 dreiss joined #git
20:52 scottchiefbaker joined #git
20:53 SwoleGymBro- joined #git
20:53 scottchiefbaker If I clone a repo it starts on master... if I want to work with the origin/foobar branch what's the best way to do that? I've been doing "git checkout origin/foobar; git checkout -b foobar"
20:53 scottchiefbaker seems like that should be doable in one command?
20:54 cbreak scottchiefbaker: git checkout foobar
20:54 zeroed joined #git
20:54 zeroed joined #git
20:57 Eugene scottchiefbaker - man git-clone, it accepts a --branch argument to override the default(which is HEAD, not master; in most repos it happens to be a symlink). In any case, it doesn't matter because you still get all the branches: `git branch -vva` and then checkout as above.
20:57 gitinfo scottchiefbaker: the git-clone manpage is available at https://gitirc.eu/git-clone.html
20:57 ali1234 joined #git
20:58 oleo joined #git
20:59 cdown_ joined #git
21:00 cdg joined #git
21:01 Rodya_ joined #git
21:06 ExeciN joined #git
21:06 Rodya_ joined #git
21:13 jeffreylevesque joined #git
21:18 alexggordon joined #git
21:21 Cabanossi joined #git
21:29 dsdeiz joined #git
21:29 dsdeiz joined #git
21:29 mattcen joined #git
21:29 ToBeCloud joined #git
21:30 daynaskully joined #git
21:31 Vampire0 joined #git
21:32 NeXTSUN joined #git
21:38 cdg joined #git
21:40 ozcanesen joined #git
21:43 juvenal joined #git
21:44 courrier joined #git
21:47 raijin joined #git
21:54 bannakaffalatta joined #git
21:57 paws_ joined #git
22:03 drodger joined #git
22:08 lavalamp joined #git
22:08 gopar joined #git
22:09 HoierM joined #git
22:10 tango_ I don't suppose there's a way to mix git request-pull, git format-patch and git send-email, is there?
22:11 HoierM joined #git
22:11 scottchiefbaker I'm trying to find the commit where a specific function (line) was added. What's the best way to to go back through history of this one file looking for when "function foobar()" was added
22:11 foreal joined #git
22:14 Dougie187 left #git
22:19 cjohnson git log -S 'function foobar'
22:19 cjohnson I think
22:22 scottchiefbaker cjohnson, Actually that finds the FIRST reference of that string not all of them
22:23 cjohnson not sure what you mean, when I run git log -S 'some-string' it lists every commit whose contents contain some-string
22:23 scottchiefbaker the man page says: use the feature iteratively to feed the interesting block in the preimage back into -S
22:23 cjohnson what specifically did you run and what did you get as a result?
22:24 scottchiefbaker git log -S 'account_group' rats/plugins/email.inc
22:24 scottchiefbaker and I get ONE commit back
22:24 scottchiefbaker but there are at LEAST two that have account_group as modified
22:25 drodger joined #git
22:25 cjohnson how do you know?
22:25 cjohnson they're both in the current branch?
22:26 cjohnson do you have both commit hashes? I would git show each one that you suspect shoudl match, probably there is some misunderstanding
22:26 cjohnson or difference
22:27 scottchiefbaker Oh I wonder if it's a diff branch
22:27 scottchiefbaker is there a way to do all branches?
22:27 cjohnson Hrm, not sure
22:27 Acerific joined #git
22:27 cjohnson man gitrevisions
22:27 gitinfo the gitrevisions manpage is available at https://gitirc.eu/gitrevisions.html
22:27 oldschool^ left #git
22:28 cjohnson I don't see anything listed here that mentions "all branches"
22:28 cjohnson oh --all ?
22:28 cjohnson or --branches
22:28 cjohnson man git-log
22:28 gitinfo the git-log manpage is available at https://gitirc.eu/git-log.html
22:30 juvenal joined #git
22:31 justanotheruser joined #git
22:33 madewokherd joined #git
22:34 acetakwas joined #git
22:36 Cabanossi joined #git
22:39 scottchiefbaker cjohnson, https://www.perturb.org/tmp/gitlog.png
22:40 scottchiefbaker Two different commits in the same file that modify that account_group line
22:40 scottchiefbaker but git log -S "account_group" only shows the first one
22:40 scottchiefbaker Is it supposed to show more than one?
22:41 cjohnson only if they're in the same branch
22:41 cjohnson run it again with --branches
22:41 cdg_ joined #git
22:41 scottchiefbaker git log --branches -S 'account_group' rats/plugins/email.inc
22:41 scottchiefbaker same thing... only one. WEIRD!
22:42 scottchiefbaker I did --all too and still only one
22:42 cjohnson Yeah there's something else afoot, because that technique definitely works well for me
22:43 cjohnson I even confirmed it scans across branches with --branches
22:43 scottchiefbaker Since I know those two commits... is there a way to show which branch each landed on?
22:45 scottchiefbaker Is this a git version thing? I'm running 2.9.0
22:51 cjohnson Honestly that is one thing I have never quite had a good grasp on how to determine
22:51 cjohnson oh I'm dumb
22:51 cjohnson http://stackoverflow.com/questions/2706797/finding-what-branch-a-git-commit-came-from
22:51 cjohnson git branch --contains <commit>
22:52 scottchiefbaker Oh... looks like -S only looks at the CHANGED part (not the whole line)
22:52 scottchiefbaker I was having account_group = 1 and then the next commit account_group = 2
22:52 peterbec` joined #git
22:52 scottchiefbaker i.e. the account group part never changed
22:52 scottchiefbaker If I did it with -G it works like you explained
22:55 anuxivm joined #git
22:57 gopar joined #git
22:58 scottchiefbaker cjohnson, That was very helpful thanks. I documented it too: https://www.perturb.org/display/1233_Git_Finding_all_commits_that_match_a_given_line.html
22:58 cjohnson scottchiefbaker: interesting! good to know
22:58 cjohnson I thought about that but then assumed it would include the whole line for changed lines, and it just wouldn't include the surrounding lines
22:58 cjohnson good to know it's literally only the delta
22:59 scottchiefbaker cjohnson, From what I can tell -S is the Delta and -G is the whole line
23:03 gopar joined #git
23:04 cjohnson fantastic
23:04 TheSimonator joined #git
23:04 cjohnson thanks scottchiefbaker
23:05 re1_ joined #git
23:15 portal_ joined #git
23:15 TikityTik joined #git
23:17 hashpuppy joined #git
23:18 juvenal joined #git
23:28 TheSimonator joined #git
23:28 KI6ESH joined #git
23:30 TinkerTyper joined #git
23:32 scoobertron joined #git
23:33 KI6ESH left #git
23:33 orbyt_ joined #git
23:34 nexemacs joined #git
23:35 menip joined #git
23:36 jameser joined #git
23:37 jones- joined #git
23:38 telephone joined #git
23:38 telephone joined #git
23:38 Starky joined #git
23:40 askb joined #git
23:42 thebope joined #git
23:44 ToBeCloud joined #git
23:48 raynold joined #git
23:52 gopar joined #git
23:52 Guest56014 joined #git
23:53 peterbec` joined #git
23:54 sunri5e joined #git
23:55 dsdeiz joined #git
23:56 johnflux_ joined #git
23:59 gopar joined #git

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