Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2016-08-16

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

All times shown according to UTC.

Time Nick Message
00:02 dreiss joined #git
00:06 rpd nullp0inter: It depends on if the "bootstrapping process" needs to be done by the user or if the repo will always provide the right set of files.
00:07 rpd For composer, npm, pip, etc, if you want to vendor the files the bootstrapping process produces, you might just include a static copy. Of course, those tools also let you provide a version snapshot in the JSON configuration file
00:08 rpd For something more complex, like `make`, you usually just want to include the Makefile and have the user generate those files based on their system
00:09 livingstn joined #git
00:11 Derperperd_ joined #git
00:12 bignose left #git
00:13 PalTale joined #git
00:15 tlaxkit joined #git
00:17 mingrammer joined #git
00:19 rchavik joined #git
00:26 hahuang61 joined #git
00:27 johnny56 joined #git
00:28 rkazak joined #git
00:29 mellernoia-work joined #git
00:29 MrTrick left #git
00:30 Lyqyd joined #git
00:30 fuchstronaut joined #git
00:32 Literphor joined #git
00:34 iomari joined #git
00:36 dreiss joined #git
00:37 texinwien joined #git
00:37 eka joined #git
00:40 scott joined #git
00:40 Mayumi joined #git
00:40 Mayumi um, does git reset --hard <somerevision> effect all branches or just current branch?
00:43 bremner just the current
00:43 pigeon joined #git
00:43 rewt that moves the current branch pointer to the specified commit and makes your working tree look like the commit
00:46 cdg joined #git
00:46 nocbreak joined #git
00:49 Dougie187 joined #git
00:51 texinwien joined #git
00:52 madewokherd joined #git
00:54 rj1 joined #git
00:55 tesuji_ joined #git
00:55 iomari joined #git
00:56 ytl joined #git
00:58 mingrammer joined #git
01:00 cqi joined #git
01:00 FuzzySockets joined #git
01:01 pi__ joined #git
01:02 lrgtesbyozacmgot joined #git
01:03 scott__ joined #git
01:05 weox joined #git
01:07 ash_workz joined #git
01:08 a3Dman joined #git
01:11 veilig2000 joined #git
01:12 scott__ joined #git
01:12 d^sh joined #git
01:12 pks joined #git
01:15 ash_workz left #git
01:15 cdg joined #git
01:16 therealGent joined #git
01:16 therealGent joined #git
01:17 Mannn joined #git
01:18 dling joined #git
01:19 thiago joined #git
01:19 rossome joined #git
01:19 circ-user-7lGUN joined #git
01:21 Rodya_ joined #git
01:25 energizer joined #git
01:25 ASOLAr_ joined #git
01:31 ToxicFrog joined #git
01:31 cjohnson @SporkWitch
01:32 netj joined #git
01:32 cjohnson just testing ;P
01:35 dodobrain joined #git
01:38 shinnya joined #git
01:47 hahuang61 joined #git
01:49 Goplat joined #git
01:51 shgysk8zer0 joined #git
01:52 dodobrain joined #git
01:52 vassagus joined #git
01:54 Vampire0_ joined #git
01:54 tgunr joined #git
01:57 ahr3n joined #git
01:58 HappyLoa- joined #git
02:03 harish_ joined #git
02:05 dsantiago joined #git
02:12 livingstn joined #git
02:13 levelxplane joined #git
02:17 FuzzySockets joined #git
02:19 Manuel15 joined #git
02:19 Manuel15 joined #git
02:20 sangy joined #git
02:26 chachasmooth joined #git
02:27 anuxivm left #git
02:28 dan joined #git
02:32 fuchstronaut joined #git
02:35 ash_workz joined #git
02:35 EduardoMartins joined #git
02:35 fahadash joined #git
02:36 Dougie187 left #git
02:36 zacts joined #git
02:36 lagothri1 joined #git
02:37 hexagoxel joined #git
02:37 jimi_ joined #git
02:40 Rodya_ joined #git
02:42 rwb joined #git
02:45 govg joined #git
02:49 circ-user-QC3mE joined #git
02:50 BenderRodriguez joined #git
02:54 ShekharReddy joined #git
02:56 skylite joined #git
02:58 thiago joined #git
02:58 ahr3n joined #git
03:01 flaviocamilo joined #git
03:03 aavrug joined #git
03:05 aavrug joined #git
03:05 csd_ joined #git
03:06 fstd_ joined #git
03:08 watabou joined #git
03:14 Andrew_K joined #git
03:16 thiago joined #git
03:18 RustyShackleford joined #git
03:18 RustyShackleford is force pushing after rebasing a feature branch a bad practice to get into?
03:18 RustyShackleford I have a coworker who's really anti rebasing
03:19 RustyShackleford as long as nobody is branching off my feature branch, I should be completely safe right?
03:25 satbyy joined #git
03:28 steven_a_s joined #git
03:29 eciohc joined #git
03:32 neersighted joined #git
03:33 fuchstronaut joined #git
03:33 auzty joined #git
03:40 SporkWitch RustyShackleford: rebasing on a repo that someone else commits to as well is VERY bad; rebasing on a repo others fork from is slightly less bad, but still bad and you should warn them (it will make things go wonky on their end, requiring action on their part.)  If your rebase is completely isolated, e.g. your branch, B, was split from main, M.  You work on B, do your stuff, M gets patched, and you
03:41 dlaboratory joined #git
03:41 SporkWitch merge that into B.  If you rebase B before it ever leaves your control, this is "fine"; no one else ever saw your history or what's going on, and to be honest they don't really have a reason to care.  As far as they're concerned, you're submitting a pull request to merge B into M.  What matters is what you did, now the order or timeline, because at the end of the day it's the cumulative changes
03:41 SporkWitch that matter.  Just make sure your comments after the rebase are still complete and detailed, so that the sum of modifications are correctly documented.
03:42 SporkWitch (anyone else, feel free to correct me if my assessment or advice has flaws).
03:42 mehola joined #git
03:43 RustyShackleford SporkWitch, so I've been working on my feature branch which branches off master. I do push my changes to remote to back them up
03:43 RustyShackleford then a few times a week I've been rebasing master into my feature branch and force pushing
03:43 SporkWitch RustyShackleford: is the remote in question a repo that anyone else pulls from?
03:43 klean joined #git
03:44 SporkWitch or is this remote YOUR fork?
03:44 RustyShackleford nobody else works on my branch, but yes there are others working on the same repo
03:44 SporkWitch that's just there for you to work on, no one else pulls from it or works on it
03:44 RustyShackleford four of us on one repo, no forks
03:44 ASOLAr joined #git
03:44 SporkWitch RustyShackleford: because the repo itself is shared, I would say that the rebases should be discouraged
03:45 SporkWitch RustyShackleford: what you do on your own fork that no one pulls from is one thing, but this is a shared repository, which means that even though you're the only one pulling from that fork, someone else COULD pull from that fork.  You shouldn't rewrite history in a shared repository.
03:46 steven_a_s joined #git
03:46 SporkWitch what you typically should be doing in this situation is forking your own copy of the repository, doing your stuff in yours, and then submitting a pull request to merge your COMPLETED patch into the main repo on the correct branch (or into master, as the case may be)
03:47 SporkWitch *pulling from that branch, someone else COULD pull from that branch
03:47 RustyShackleford the forking model seems unnecessarily complicated
03:47 RustyShackleford i'm on a different project too that just recently migrated to git
03:47 RustyShackleford keeping your fork up to date is a hassle
03:47 SporkWitch it's not complicated at all.  It keeps things isolated so no one breaks the central master (though you'll get some people that don't like this hierarchical concept with regard to git, reality is it's how things are normally handled)
03:48 cmoneylulz joined #git
03:48 SporkWitch not any moreso than keeping your branch up to date
03:49 SporkWitch have two remotes: origin (your fork to work on) and upstream (the repo you forked from).  you simply pull from upstream to keep your copy sync'd up.
03:49 SporkWitch (all a remote is, in the git client, is an alias that maps to an address; git doesn't care if you fetch from this repo or that)
03:49 howdoi joined #git
03:50 RustyShackleford do you still create feature branches on your own fork?
03:50 RustyShackleford in this project all the devs fork the repo, commit their work on the release branch, then submit a pull request
03:51 howdoi is there a way to autoupdate submodules?
03:51 RustyShackleford i've had problems where I make a commit on that release branch, then try to fetch upstream changes. I've started branching off that feature branch so I don't run into merge conflicts
03:53 SporkWitch RustyShackleford: you're going to have to resolve the merge conflicts eventually anyway.  Really, the release branch shouldn't be getting a lot of changes to it, so you shouldn't have to be pulling from upstream all the time.  The big one will be at the end before you merge your completed feature back in.  Up to that point, you only really need to pull from upstream if there are significant change
03:53 SporkWitch that would affect how you write the rest of your module
03:53 RustyShackleford ah well i'm responsible for the daily build this sprint
03:54 RustyShackleford so i've been doing it every day. But I see your point, don't need to pull those changes into my branch
03:54 SporkWitch if you're seeing a lot of merge conflicts, this implies that your feature branches aren't properly modular; the whole point of such a model is to minimize conflicts, by having people working on separate components
03:56 dtrott joined #git
03:58 a_thakur joined #git
03:58 RustyShackleford so maybe I'll start working on my own fork
03:58 RustyShackleford do my rebasing there
03:59 RustyShackleford the rebase vs merge seems to be a controversial subject online
03:59 a_thakur joined #git
04:00 howdoi submodules anyone?
04:01 thecomedian joined #git
04:01 zincrokx joined #git
04:03 SporkWitch RustyShackleford: again, the main thing is just a matter of whether it's exposed.  Rewriting history before anyone sees it? End result is just a matter of how well commented and documented things are.  Rewrite history where people can SEE that you rewrote history?  That's a big faux pas, as a matter of culture and principle.
04:03 scott_sd joined #git
04:05 steven_a_s joined #git
04:05 root____ joined #git
04:07 mingrammer joined #git
04:09 RustyShackleford SporkWitch, depends on the culture of where you work I guess
04:10 RustyShackleford one coworker says that at a previous place, all the feature branches got squashed down to a single commit
04:11 RustyShackleford they were all working on the same repo, no forks. Idk I see how it could be risky, but if you make sure your branch is isolated I don't see what can go wrong
04:11 SporkWitch RustyShackleford: that's not unusual prior to merge into master, it's everything leading UP TO that point where it's a problem
04:12 SporkWitch RustyShackleford: the idea is typically that you don't touch history until the end.  You clean it up before the merge back into master
04:19 anohigisavay joined #git
04:20 anohigisavay hi. i'm trying to make a wrapper program of git
04:20 anohigisavay and i'm playing with git config
04:21 anohigisavay when i create a config key, say `git config foo.bar "hello"`
04:21 anohigisavay and unset it `git config --unset foo.bar`
04:22 zincrokx joined #git
04:22 anohigisavay it leaves an empty section in .git/config file. and when i doo `git config foo.bar "hello"` once again, it ends up with two section headers
04:23 anohigisavay [foo]\n[foo]\n    bar = hello
04:23 Cabanoss- joined #git
04:24 KeelOfSteel joined #git
04:25 dhollinger joined #git
04:26 dlaboratory joined #git
04:27 codebam I have an apikey saved in a file called "TOKEN". I commit the TOKEN file with nothing in it just so it would be there for anyone who clones then added it to my .gitignore. However after modifying it it's shown in the staging area as modified, how can I ignore it?
04:28 ASOLAr joined #git
04:29 anohigisavay codebam: gitignore is for files that you don't want to put in git repo
04:30 anohigisavay codebam: you might want to use `git update-index --assume-unchanged
04:30 rageshkrishna joined #git
04:31 anohigisavay codebam: but everyone needs to set it in their local repos. i don't know of any better option
04:32 codebam oh okay, well I can just remove the file entirely I guess, that command didn't seem to change anything
04:33 ljc joined #git
04:33 fuchstronaut joined #git
04:34 anohigisavay codebam: i learned that command just days ago. it should work. `git update-index --assume-unchanged <your-token-file>`?
04:34 codebam oh I need to mention the file
04:36 codebam oh that worked
04:36 codebam thank you :)
04:36 anohigisavay xD
04:36 anohigisavay cool
04:37 Literphor joined #git
04:37 anohigisavay codebam: also the command that reverts the change is `git update-index --no-assume-unchanged ...`
04:37 codebam thanks, I'll make sure to write that down somewhere
04:42 phanimahesh joined #git
04:45 scott_sd hi
04:45 gitinfo scott_sd: 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.
04:45 kadoban --assume-unchanged isn't really a good idea. There's more standard approaches to that problem, !config
04:45 gitinfo [!configfiles] It is recommended to store local configuration data in a file which is not tracked by git, but certain deployment scenarios(such as Heroku) may require otherwise. See https://gist.github.com/1423106 for some ideas
04:45 kadoban codebam: ^
04:47 codebam oh cool thanks
04:47 unforgiven512 joined #git
04:48 unforgiven512 joined #git
04:48 moei joined #git
04:48 unforgiven512 joined #git
04:49 unforgiven512 joined #git
04:49 clemf joined #git
04:49 unforgiven512 joined #git
04:50 scott_sd I use git-svn, i have one bug. when i open file use vim and then close it, git status tell file modify, all the line is remove and inter into . but nothing is different. our team someone work in windows i am linux, my git config :core.autocrlf=false
04:50 unforgiven512 joined #git
04:51 unforgiven512 joined #git
04:51 scott_sd i don't want git tell this modify
04:51 scott_sd how to ignore this change or close it in git
04:53 unforgiven512 joined #git
04:54 unforgiven512 joined #git
04:54 scott_sd some time my git-svn not work as i will, like i use git reset --hard HEAD , some file modify file can't recover, i am root
04:55 ASOLAr joined #git
04:58 phanimahesh scott_sd: why do you have autocrlf to false?
04:59 scott_sd i dont't want git auto change file
05:01 scott_sd phanimahesh:  some file wirte in window style and some is unit . git clone will auto change
05:02 ericsupreme joined #git
05:02 tom[] joined #git
05:02 lolmac joined #git
05:03 scott_sd phanimahesh: is core.autocrlf=false cause the problem?
05:03 subhojit777 joined #git
05:03 phanimahesh no. vim is normalising your line endings.
05:04 phanimahesh set gitattributes to tell git which files should have crlf and which files should have lf
05:04 phanimahesh why do you want mixed stuff in same repo, btw?
05:07 slayerjain joined #git
05:07 scott_sd phanimahesh: my team before not unify, everyone use different IDE and style ,And i am new one .I must compatibility them
05:08 scott_sd phanimahesh: they work on windows and use svn. i want to work in my linux and use git-svn
05:09 mehola joined #git
05:09 phanimahesh sorry to hear that. set your vim to write crlf
05:11 scott_sd phanimahesh; thx i try it
05:15 stuh84 joined #git
05:20 M-meznak joined #git
05:23 crayon joined #git
05:23 SporkWitch isn't all of this something that should be handled in git?  Sane default should be UNIX EOL in the repo, OS-specific on checkout; seems like it would be the best way to prevent such issues from cropping up in the first place.
05:24 subhojit777 joined #git
05:25 leeN joined #git
05:28 rkazak joined #git
05:34 fuchstronaut joined #git
05:35 darkbit joined #git
05:37 cosmicfires joined #git
05:37 satya4ever joined #git
05:37 cosmicfires I want to replace some files in my working set with some from a previous commit, what's the best way to do it?
05:39 i7c git checkout <previouscommit> -- path/to/file
05:39 qt-x joined #git
05:39 ASOLAr joined #git
05:40 cosmicfires thanks i7c
05:43 LeMike joined #git
05:46 ytl_ joined #git
05:48 zeroed joined #git
05:49 Raging_Hog joined #git
05:51 star_prone joined #git
05:53 khmarbaise joined #git
05:53 khmarbaise joined #git
05:56 LucaTM joined #git
06:03 rscata joined #git
06:04 djb-irc joined #git
06:05 przemoc joined #git
06:05 Novitsh_ joined #git
06:06 rubyonrailed joined #git
06:06 Repox joined #git
06:07 mariuscc joined #git
06:08 ASOLAr joined #git
06:10 roelmonnens joined #git
06:11 tvw joined #git
06:14 Alex_Rose joined #git
06:15 Alex_Rose Hi there, having a problem:
06:15 Alex_Rose My repo was too big, so I used the BFG Repo-Cleaner
06:15 Alex_Rose I was using bitbucket and they were refusing my new commits because I was over the limit
06:15 Alex_Rose but they didn't get back to me for several days so I now have a bunch of new commits (I work 16 hours a day on my project)
06:16 tabrez joined #git
06:16 Alex_Rose so, I cloned, I deleted old commits to get rid of big files, then I pruned and pushed
06:16 Alex_Rose however
06:16 Alex_Rose I just saw the final step is "okay now delete your own copies of the repo and do fresh clones of the new pristine data
06:17 Alex_Rose because I have dirty history
06:17 Alex_Rose but like, I have a bunch of new commits?
06:17 Alex_Rose so I can't just delete it
06:17 Alex_Rose so now my current version of the repo wants to pull 159 and push 163
06:17 selckin can create patches for them and apply, or maybe add it as a remote and cherry pick
06:18 Alex_Rose Can I do that through git?
06:18 selckin yes
06:18 Alex_Rose What should I google?
06:18 Alex_Rose I'm a massive novice at Git
06:19 Alex_Rose don't even know where to begin
06:19 mehola_ joined #git
06:20 Alex_Rose I'm *this close* to just like
06:20 Alex_Rose deleting the entire repot and just uploading a fresh version
06:20 yitz git help diff HEAD
06:20 yitz Erm.  git help diff
06:20 yitz Then: git diff ... HEAD
06:20 yitz With the patch flags
06:22 Alex_Rose so just git diff --patch HEAD ?
06:23 yitz I think so
06:23 Alex_Rose :S, a bit terrified of losing my work
06:23 yitz But I'm not a git expert
06:23 yitz diff is a safe operation
06:23 Alex_Rose but if I do it I'm not sure how to undo it if it's wrong?
06:23 selckin !backup
06:23 gitinfo Worried about your data while trying stuff out in your repo? The repository in its entirety lives inside the .git directory in the root of your work tree so to backup everything `cp -a path/to/workdir path/to/backup` or equivalent will suffice as long as the repo is not modified during backup. See also http://sethrobertson.github.com/GitBestPractices/#backups
06:23 yitz diff doesn't change anything
06:23 yitz It just outputs info to STDOUT
06:24 a_thakur joined #git
06:24 Kartagis left #git
06:24 M-shine joined #git
06:26 Alex_Rose it's only showing the difference between what I've changed locally and where it's up to locally, not the remote
06:27 MrMojito joined #git
06:27 yitz git pull?
06:27 ASOLAr joined #git
06:27 yitz You could make a fresh clone and do a recursive diff or something
06:28 Alex_Rose well I mean
06:28 yitz Or just copy your changed files from the old dir into the fresh clone
06:28 Alex_Rose won't git pull overwrite all my new cahnges?
06:28 Alex_Rose *changes
06:29 Alex_Rose I guess I could form a manual list of my last 4 commits
06:29 yitz pull's not supposed to overwrite local changes ... but if you want your files in git and don't care about local history, you can do a fresh clone and just copy your files from the old dir into the new one
06:29 selckin cd old; git format-patch <sha of start of work you want tokeep>;  cd ../new/; for i in *.patch; do patch -p1 < $i; git add .; git commit -m "$i"; done;
06:29 Alex_Rose well that's the thing I mean
06:29 Alex_Rose there's like 2gb of stuff here
06:29 cbreak selckin: you want to reimplement cherry-pick? :)
06:30 selckin cbreak: then i have to explain how to add remotes and this is already more effort then i'm willing to provide him
06:30 cbreak Alex_Rose: man git-cherry-pick, man git-remote add
06:30 gitinfo Alex_Rose: the git-cherry-pick manpage is available at http://jk.gs/git-cherry-pick.html
06:30 gitinfo Alex_Rose: the git-remote manpage is available at http://jk.gs/git-remote.html
06:30 nanozer2 joined #git
06:31 Alex_Rose The problem is, I don't trust myself to do this with such important data. Like, I don't mind doing things when I'm 100% confident of my knowledge of what I'm doing
06:32 qt-x joined #git
06:32 Alex_Rose I think what I'll just do (which may sound stupid to you but I think will work)
06:32 Alex_Rose is copy and paste the directory in its entirety
06:32 Alex_Rose then git pull
06:32 Alex_Rose reset any changes
06:32 Alex_Rose first write down anything that's changed in my commits
06:32 Alex_Rose and then manually drag all the scripts across and overwrite
06:32 Alex_Rose then push
06:33 jagob joined #git
06:33 Manuel15 joined #git
06:33 Manuel15 joined #git
06:33 Alex_Rose I cannot afford for this to go wrong
06:33 selckin !backup
06:33 gitinfo Worried about your data while trying stuff out in your repo? The repository in its entirety lives inside the .git directory in the root of your work tree so to backup everything `cp -a path/to/workdir path/to/backup` or equivalent will suffice as long as the repo is not modified during backup. See also http://sethrobertson.github.com/GitBestPractices/#backups
06:33 yitz Alex_Rose: make a backup copy then stop worrying so much
06:34 Alex_Rose that's essentially the same as just copying and pasting the directory, right?
06:34 Alex_Rose like, I trust the shell more for some reason
06:34 rads joined #git
06:35 rom1504 cp is copy and pasting
06:35 fuchstronaut joined #git
06:35 Alex_Rose yeah I'm just manually doing that now in explorer
06:36 mingrammer joined #git
06:38 Alex_Rose okay right, I have my copy, I've written down the shas and the contents of every commit so I feel comfortable fucking around with it now
06:38 Alex_Rose so.. seleckin, do I git pull before I do that?
06:39 Alex_Rose or do I do format-patch first?
06:40 selckin at the basis hyou want to get data from your old repo to your new one, so since you're dev i assumed you where confortable with patch files, but since you're on windows probably not
06:40 Bombe Also, this all seems to be completely over the top.
06:40 selckin so you can take the commits from the old into patch files, and then apply each path as a commit in the new
06:40 selckin or you can add the old repo as a remote in the new one, and cherry pick to automate this
06:41 selckin (this all assumes you want to keep the commits and not the latest state of the data)
06:41 Bombe Just “git remote update”, use some viewer to check what’s changing, the update local refs.
06:41 selckin Bombe: he rewrote the history to remove stuff
06:42 Alex_Rose like, to make it clearer everything that's happened. I have the repo for my ps4 game, we moved the soundtrack from oggs to fmod files, so we had a bunch of huge files that pushed our bitbucket repo over the 2gb max
06:43 Alex_Rose in the 4 days since I used bfg repo cleaner and cleaned up the clone of the repo, I made 4 large commits which I couldn't push because we exceeded the limit
06:43 Alex_Rose now the repo is back down again, but I'm still sitting on these 4 commits
06:43 selckin do you have any devs working on your game? ask them
06:43 Alex_Rose what I want to do is commit those 4 without ruining anything, nuke my repo and redownload
06:43 Alex_Rose I am the dev
06:43 Alex_Rose I'm a c# dev though
06:44 Alex_Rose I have no git experience, I use sourcetree like a noob
06:45 a_thakur_ joined #git
06:45 Alex_Rose I can run circles around anyone in unity but I simply have no formal experience in using git, and messing with my game files is not the correct place to experiment and learn
06:45 Sceorem joined #git
06:45 Alex_Rose I'm already having stress overload from having like 5 days of unbacked up work
06:46 selckin if you don't want to learn and try what we are suggesting you want to pay someone to do it for you
06:46 submitnine joined #git
06:46 Alex_Rose I never said that I don't want to do that, I'm just getting conflicting information from 3 people
06:46 a_thakur joined #git
06:46 selckin there are multiple ways to solve the problem as always
06:46 ASOLAr joined #git
06:46 talin hello... i have a very weird problem with git. i have two branches... and sometimes when i check out one of them, i get the normal text, but at teh bottom it says "not a talin.git repo"... alright. things still seem to work. but then later when i git fetch; git rebase, it says "could not deatach from HEAD". the only way i manage to fix it, is by cloning this repo all over again and starting anew... what on eart
06:47 talin h is happening?
06:47 talin git checkout a; git fetch; git rebase; git checkout b; git fetch; git rebase... is something i do a lot
06:48 talin in this case, it went nuts when i hadn't even committed anything
06:48 selckin talin: do you mean the 'fatal: Not a git repository (or any of the parent directories): .git' message with 'not a git repo' ?
06:48 selckin that probably means soemthing is deleting your .git directory
06:48 talin selckin: the second one
06:48 selckin whats the exact error ?
06:49 talin "Not a talin.git repo"
06:49 selckin when running what command?
06:49 talin git checkout master
06:50 selckin that makes no sense to me, never seen that error before sorry can't help
06:50 talin hm, okay. the .git dir is there
06:50 talin and has some content
06:50 talin this is driving me nuts
06:50 selckin what does git status say?
06:51 talin HEAD detached from refs/heads/master
06:51 talin nothing to commit, working directory clean
06:51 selckin !detached
06:51 gitinfo A detached HEAD(aka "no branch") occurs when your HEAD does not point at a branch. New commits will NOT be added to any branch, and can easily be !lost. This can happen if you a) check out a tag, remote tracking branch, or SHA; or b) if you are in a submodule; or you are in the middle of a c) am or d) rebase that is stuck/conflicted. See !reattach
06:51 talin that's after doing a rebase
06:51 selckin then you probably in the middle of a conflict?
06:51 selckin resolve it and git rebase --continue
06:52 talin no conflict. haven't had any conflict either.. not that git has told me abut in git status or anything, anyway
06:52 selckin what does it say when you try continue (or --abort) ?
06:53 talin "No rebase in progress?"
06:53 talin hmm, if i do git merge, it seems to merge succesfully. and doesn't go crazy again until i try another rebase
06:54 u1106 joined #git
06:55 talin okay, this is weird. git checkout master gives me
06:55 talin Switched to branch 'master'
06:55 talin Your branch is up-to-date with 'origin/master'.
06:55 selckin get in the habbit of running git status after everyething you do, should never go into detached
06:55 talin Not a talin.git repo
06:55 selckin unless you have a conflict etc
06:55 talin so it's up to date with ogirin/master
06:55 talin origin...
06:55 talin so if i fetch/rebase, nothing should happen right?
06:55 talin since it's up to date
06:55 Remram[m] joined #git
06:55 Ineentho[m] joined #git
06:55 M-mistake joined #git
06:55 M-meznak joined #git
06:55 selckin so it said not a talin repo after the sucessfull checkout?
06:56 talin yes
06:56 selckin do you have  some hooks enable or funky prompt stuff?
06:56 talin so now if i git rebase, i get: First, rewinding head to replay your work on top of it...  Not a talin.git repo. could not detach HEAD
06:56 talin no hooks, default prompt
06:58 eivarv_ joined #git
06:58 rads joined #git
06:58 talin if there was anything to rebase, i don't get how git status would tell me that it's up to date instead of "you are ahead of origin/master by 1 commit" or something
06:58 talin and if there is nothing to rebase, then why does it detach head
06:59 selckin it should say "current branch x is up to date"
06:59 TomyWork joined #git
07:00 selckin can't grep an error message that is 'not a x repo' from the git source either
07:01 Darcidride joined #git
07:02 jceb joined #git
07:02 harish_ joined #git
07:02 selckin using any gitattributes stuff?
07:03 Alex_Rose So, I did git format-patch to the.. sha adjacent to the commit I wanted to keep
07:04 Alex_Rose although it immediately then in the console displayed.. the one after it and so on
07:04 Alex_Rose I'm assuming I did that wrong and I want to go one back
07:04 selckin Alex_Rose: so now you have a patch file for every commit, and can apply those in the new repo
07:05 Alex_Rose so, I guess I should start a fresh copy of the repo for the "new" one to keep everything clean?
07:06 selckin thats the easiest way
07:06 daurnimator joined #git
07:08 cqi joined #git
07:09 thorongil joined #git
07:09 talin selckin: gitattributes? not that i know of
07:09 rads joined #git
07:09 Rish joined #git
07:10 ki0 joined #git
07:10 selckin talin: at least in latest git source can't get an (error) message like that so i suspect its coming from somewhere else, gitattributes can be used to configure custom differs etc, but maybe my grep-fo is weak, anyway sorry that i can't help
07:10 selckin *can't grep
07:11 talin selckin: right. when i google this error, i find nothing either, only the "fatal: ..." that you mentioned earlier
07:11 talin this is crazy
07:12 selckin probably not but is this a public repo?
07:13 talin selckin: no, it's at work. lots of others using it
07:13 dcpc007 left #git
07:13 talin this problem seems to occur at seemingly random times... so a few times a week i have to start over
07:14 ki0 joined #git
07:14 Alex_Rose selckin, so, when I did format-patch I did it from.. the red box here
07:14 Alex_Rose http://i.imgur.com/agqIdJZ.png
07:15 dsantiago joined #git
07:15 Alex_Rose but it seems to only grab ones more recent
07:15 jguillen joined #git
07:15 Alex_Rose am I right in thinking I should be grabbing the sha from the green box on the right?
07:15 selckin yeah have to do one before it
07:15 Alex_Rose and then the purple chain is the.. revised chain
07:15 Alex_Rose so if I just type the command again is it all good or do I need to undo what I did?
07:16 selckin it just wrote some *.patch files in the current directory, can safely be deleted
07:16 Alex_Rose sweet
07:16 cbreak honestly, just use cherry-pick
07:16 cbreak so much faster
07:16 cbreak more convenient, safer...
07:16 selckin yeah, create a backup branch, reset to origin/master then cherry pick
07:17 selckin but thats a lot more dangerous to explain to someone not even familiar with copy the files
07:17 Alex_Rose haha
07:17 Alex_Rose indeed
07:17 Alex_Rose I feel like the way you've shown seems pretty safe? I mean, I have a backup anyway
07:17 Alex_Rose all I have to do is pull a fresh repo and then apply the patches, right?
07:18 ahmedelgabri joined #git
07:18 Alex_Rose do I need to commit/push them here first or do I literally just grab the patch files and let it do its magic and then completely forget that my old folder existed?
07:18 selckin no need to commit push anything
07:19 talin i did apt-get source git, and grep for "Not an", and there is nothing like that...
07:21 talin selckin: oh, i found something
07:21 talin sudo find / -type f | xargs grep -sI "Not an" | grep ".git"
07:21 talin /home/talin/repos/broken_do_not_delete/tools/git-tools/hooks/post-checkout-build-tools:  echo "Not a talin.git repo"
07:22 talin that's one of the repos that broke earlier
07:22 selckin so you do have git hooks
07:22 talin huh, apparently
07:23 selckin check files with +x in .git/hooks/
07:23 selckin check core.hooksPath init.templatedir and ~/.gitconfig for global hooks i guess
07:24 talin yes, this post-checkout-build-tools is +x
07:24 dcpc007 joined #git
07:24 Balliad joined #git
07:24 selckin problem solves then
07:25 talin selckin: yes, i'll try to figure out what it's trying to do. it seems to be messing stuff up for me. thanks alot for the help :)
07:26 rads joined #git
07:30 Lomex joined #git
07:30 mingrammer joined #git
07:32 Marbug joined #git
07:34 rads joined #git
07:37 rads_ joined #git
07:38 settermjd joined #git
07:39 MatthewAllan93 joined #git
07:41 csd_ joined #git
07:41 star_prone joined #git
07:42 LiohAu joined #git
07:43 Anticom joined #git
07:44 kurkale6ka joined #git
07:46 ASOLAr joined #git
07:47 t0by joined #git
07:49 rchavik joined #git
07:53 xaa joined #git
07:55 Alex_Rose hrm
07:55 Alex_Rose the patches seem to be struggling with .pngs
07:55 Urbany joined #git
07:55 Alex_Rose I'm getitng.. "can't find file to patch at input line x"
07:55 Alex_Rose I'll just manually type in the locations for now if it'll let me
07:58 Alex_Rose ooooh, not.. quite.. sure.. how.. that.. went.. heh, there were like a million issues with any assets that weren't just scripts
07:58 Alex_Rose might be able to locate and add the missing things
07:58 selckin patch -p1 < file.path;  (in the right order)  should apply cleanly except on files you removed with the rewrite
07:59 Alex_Rose I did
07:59 Alex_Rose well, I moved the patches to the new folder
07:59 Alex_Rose and then did
07:59 selckin thats fine
07:59 Alex_Rose for i in *.patch; do patch -p1 < $i; git add.; git commit -m "$i"; done;
08:00 selckin might wanna do it manually, since its only 3 ?
08:00 filcab joined #git
08:00 xpk joined #git
08:00 Alex_Rose it complained a lot about replacing lf for crlf in about 100 scripts which is fine, main thing was pngs I've added and sound files
08:00 Alex_Rose it really did not like those
08:01 Alex_Rose e.g.
08:01 Alex_Rose "can't find file to patch at input line 360221" "Perhaps you used the wrong -p or --strip option?"
08:01 Alex_Rose frequently
08:01 Alex_Rose I'll open it in unity and see what happens
08:03 garethdaine joined #git
08:05 chandu20 joined #git
08:08 MrC joined #git
08:10 GenZai joined #git
08:10 eivarv_ joined #git
08:13 liuzhen_ joined #git
08:13 tomog999_ joined #git
08:14 weox joined #git
08:15 jnewt joined #git
08:15 jnewt remote: error: insufficient permission for adding an object to repository database ./objects
08:16 robotroll joined #git
08:16 jnewt on the remote server (the permisson for the object folder for group "git" is rwx. my user as ssh'd is a member of the git group.  idk what else i'm missing
08:17 jnewt permissions on the objects folder is 770 right now.
08:17 osse jnewt: what about the directories within it?
08:18 moijk joined #git
08:19 Alex_Rose hrrrm, it's all pretty fucked up
08:19 Alex_Rose I think I'll just do it manually, if that doesn't work, fuck it I'll kill the repo and upload my new one over it
08:20 jnewt wel, most are directors with two letter names.  there looks like about a 50/50 split between 770 and 760, with one "fc" being 755.  owners vary
08:20 jnewt osse^
08:20 rasto joined #git
08:21 jnewt info and pack are 770, also group git, owner is 1009 (maybe a user that is no longer around created this i think).
08:22 binfalse joined #git
08:22 Oebele joined #git
08:25 mingrammer joined #git
08:27 f3r70rr35f joined #git
08:29 indistylo joined #git
08:32 osse jnewt: i think those 760s are going to be a problem
08:33 jnewt osse: i am just looking at some of the ones that work, and there's a lot of variation (presumably because of who was committing).  hard to tell.
08:34 bipolar joined #git
08:34 Guest93616 joined #git
08:35 osse jnewt: the thing that matters is the group permissions
08:35 a_thakur joined #git
08:35 abyx joined #git
08:36 ahmedelgabri joined #git
08:36 jnewt does group need +x too?
08:36 mariuscc joined #git
08:36 davisonio joined #git
08:37 fuchstronaut joined #git
08:38 vuoto joined #git
08:38 a_thakur joined #git
08:39 ams__ joined #git
08:41 vuoto_ joined #git
08:41 ASOLAr joined #git
08:48 Diabolik joined #git
08:51 star_prone joined #git
08:53 rads joined #git
08:54 unforgiven512 joined #git
08:55 flobin joined #git
08:59 vuoto joined #git
08:59 ASOLAr joined #git
08:59 flobin left #git
09:00 slugmandrew joined #git
09:00 beagleGuy joined #git
09:04 slugmandrew can anyone offer advice on a complex (for me) merge/rebase? My master branch has been restructured into a maven multi-module setup, while I have another branch that is 23 ahead | 18 behind that still has the single-module setup. Do I need to rebase onto the master branch or should I just merge the two? I have tried both but not much luck :(
09:04 nikivi joined #git
09:08 jor slugmandrew: you can't avoid the complex merge. The rebase eventually does the same, only makes the history a bit more readable afterwards
09:10 chandu20 joined #git
09:10 jor slugmandrew: a step-by-step (per commit) merge might make it easier, could try git-imerge. But in a case like this you can't avoid manually merging
09:10 slugmandrew jor, ok thanks. I'm not bothered about readability. It's just me for now
09:10 slugmandrew oh, can you do that?
09:11 jhass at some point it might be less work to create a new branch based on master and reapply the changes by hand commit by commit, having git show of the old commit open on the side
09:11 slugmandrew That would def help as the conflicts I'm getting is where I have changed names, and then re-uased names for new classes
09:12 slugmandrew so I changed MyClass to OldMyClass in one commit, then in next one I create MyClass again
09:12 slugmandrew stupid really... my rafactoring needs work
09:12 slugmandrew refactoring*
09:13 chll_ joined #git
09:14 slugmandrew so if i check out the first commit on new_branch that diverges from master_copy and do a merge of those
09:14 slugmandrew then repeat for each commit up the chain?
09:15 Rish joined #git
09:15 slugmandrew then a simple fast-forward to master_copy into master
09:15 slugmandrew am I understanding that correctly?
09:15 jknetl joined #git
09:17 stdint joined #git
09:17 f3r70rr35f joined #git
09:17 stdint left #git
09:18 jor slugmandrew: use cherry-pick for each commit
09:23 infra-red joined #git
09:24 steven left #git
09:25 ASOLAr joined #git
09:26 epdae joined #git
09:31 mooru joined #git
09:31 epdae left #git
09:31 binfalse joined #git
09:35 stdint joined #git
09:35 stdint I have some problems with  git grep, as it is linear grep tool
09:36 mahendra_ joined #git
09:37 stdint but I want to search a word start with specific char, but I can't use the ^ to indicate the start of the word
09:37 stdint the ^ would be regard as the start of the line
09:37 stdint is there any think I could to match the word I want
09:37 fuchstronaut joined #git
09:39 jhass git grep -w
09:39 jhass or -P with \b
09:42 irgendwer4711 joined #git
09:42 stdint jhass, I see thank you
09:43 irgendwer4711 hi, I have a problem with a git installation with apache. Suddenly it stopped working, no push is successful. I could not find any cgi config nor dav. this is strange.
09:44 vassagus joined #git
09:44 a_thakur joined #git
09:44 stdint jhass,  git grep -P \bfoo ?
09:45 stdint sorry, miss the ''
09:45 stdint jhass, thank you
09:45 a3Dman joined #git
09:45 Ezzy joined #git
09:45 davisonio joined #git
09:45 irgendwer4711 maybe someone can help.
09:46 johnny56 joined #git
09:46 harish_ joined #git
09:48 vuoto_ joined #git
09:49 chandu20 joined #git
09:50 vuoto joined #git
09:51 texinwien joined #git
09:51 eivarv joined #git
09:53 stamina joined #git
09:53 fuchstronaut joined #git
09:53 a_thakur joined #git
09:53 fuchstronaut joined #git
09:57 johnmilton joined #git
09:57 vuoto joined #git
09:59 Gloomy joined #git
09:59 vuoto_ joined #git
10:00 Sceorem joined #git
10:00 vuoto joined #git
10:01 rahtgaz joined #git
10:03 anohigisavay joined #git
10:03 inflames joined #git
10:04 groundnuty hey, I would like to make sure that there are no ssh keys in the repo I want to publish
10:05 groundnuty can I somewhow list the list of the files in every commit on a specyfic branch?
10:05 groundnuty in one, neat git command :)
10:09 vuoto joined #git
10:10 tvw joined #git
10:10 rawtaz groundnuty: did you google this?
10:11 groundnuty rawtaz: ofc. I think that I figured it out, you can comment if its a good solution
10:11 _sjs joined #git
10:11 groundnuty git filter-branch --tree-filter 'ls -R1'
10:11 rawtaz i dont remember so i just googled it and found a bunch of hits
10:11 rawtaz e.g. http://stackoverflow.com/questions/424071/how-to-list-all-the-files-in-a-commit
10:12 groundnuty I also did, but somehow non of them did waht I wanted
10:12 rawtaz what do you want and what do you get instead?
10:12 groundnuty just what I wrote in  command
10:13 groundnuty the idea was to list all files in all commits in a branhc
10:13 groundnuty this command I pasted *I think* does that
10:13 rawtaz oh ok
10:13 groundnuty I'm new to filter-branch
10:13 rawtaz :)
10:14 joshuaturner joined #git
10:14 notebox joined #git
10:14 dodobrain joined #git
10:19 cyan__ joined #git
10:19 mingrammer joined #git
10:19 mooru joined #git
10:23 vuoto joined #git
10:26 irgendwer4711 left #git
10:26 slugmandrew jor, this is brilliant - thanks so much. I think I'm getting the cherry-picking thing
10:27 GavinMagnus joined #git
10:27 slugmandrew so I created master_copy, and I'm cherry-picking them one by one and making a new commit for each
10:30 vuoto_ joined #git
10:31 jor slugmandrew: good... and in the future, when you do these big refactoring/renames/moves, send all other developers on a short vacation so no branches diverge ;)
10:31 stdint left #git
10:32 vuoto joined #git
10:33 davisonio joined #git
10:37 mingrammer joined #git
10:40 Impaloo joined #git
10:42 slugmandrew jor: it was unintended! My consultant did the multi-module structure without notifying me :S
10:46 Fijit joined #git
10:47 robotroll joined #git
10:48 vassagus_ joined #git
10:49 ShekharReddy joined #git
10:52 afuentes joined #git
10:54 zh_cool1 joined #git
10:54 eivarv joined #git
10:55 zh_cool1 left #git
10:56 zh_cool joined #git
10:56 zh_cool show?
10:56 i7c I don’t know.
10:58 zh_cool s
11:05 zh_cool joined #git
11:05 i7c What a conversation.
11:06 HappyLoa- joined #git
11:10 Orion3k joined #git
11:11 slugmandrew What? A Conversation? ;)
11:15 ShekharReddy How to get the latest chaneges which are done manually on github accoutn
11:16 ShekharReddy changes which are made on git remote origin
11:16 liamonade joined #git
11:18 ojdo joined #git
11:18 sdothum joined #git
11:19 Anupkumar ShekharReddy: you have to sync the fork
11:19 Anupkumar ShekharReddy: https://help.github.com/articles/syncing-a-fork/
11:20 ShekharReddy git pull helped me
11:20 ShekharReddy thanks for responding man
11:20 ShekharReddy Anupkumar:  ^^
11:20 Anupkumar :)
11:21 ShekharReddy actually i don need to sync fork but what i was asking is something similar to http://stackoverflow.com/questions/15691611/how-do-i-pull-changes-from-github-repository-to-my-local-machine
11:21 ShekharReddy anyways thanks :)
11:24 HoierM joined #git
11:26 rangergord joined #git
11:28 Rodya_ joined #git
11:29 rads joined #git
11:30 rangergord How can I download just the contents of the latest master from a git repo? No history or anything, no branches. The smallest download possible to give me the latest source.
11:30 fuchstronaut joined #git
11:30 knittl can I easily skip back "n" commits during interactive rebase?
11:31 knittl i messed up and would like to re-order some commits before continuing the rebase
11:31 ilmari rangergord: git clone --depth=1 --branch=master <repo>
11:31 Bombe knittl, I think you have to abort the rebase for that.
11:31 Bombe knittl, or finish it and then do the second rebase.
11:31 rangergord ilmari: I already read about that, depth=1 reduced my DL from 44MB to 42.5MB :/
11:32 knittl Bombe: that sucks. It's 200+ commits and I already edited some of them (rerere is disabled)
11:32 rangergord let me try with just --branch=master though
11:32 Bombe knittl, then finish that first. Doing the second rebase later shouldn’t mess up anything.
11:33 knittl Bombe: but what should I do with the conflicts then?
11:33 jor rangergord: git archive?
11:34 fees joined #git
11:34 Bombe knittl, they won’t magically disappear if you somehow do a rebase-within-a-rebase.
11:35 rivarun joined #git
11:35 knittl I don't think I understand
11:37 bartosz_ joined #git
11:38 MainB joined #git
11:41 aguimaraes1986 joined #git
11:42 paws_ joined #git
11:42 LiohAu joined #git
11:44 a_thakur joined #git
11:48 nullp0inter joined #git
11:49 joshuaturner joined #git
11:50 jceb joined #git
11:52 mooru joined #git
11:53 knittl hrm. somehow my todo-list had lots of duplicated entries
11:53 knittl this f****d up my rebase
11:53 knittl (obviously)
11:53 fmcgeough joined #git
11:54 Sceorem joined #git
11:54 ItSANgo joined #git
11:57 binfalse left #git
11:57 stasmakarov joined #git
11:58 codeman joined #git
12:00 CheckDavid joined #git
12:02 airdisa joined #git
12:03 zeroed joined #git
12:06 psprint joined #git
12:06 mohabaks joined #git
12:06 psprint Hello. If I would want to do merge commit fully myself (after merge --no-commit), I should provide which tips as parents? HEAD and MERGE_HEAD ?
12:07 grawity no, `git commit` takes them from MERGE_HEAD automatically
12:07 TooLmaN joined #git
12:07 rangergord joined #git
12:08 psprint grawity: that's why I wrote "fully"
12:08 grawity oh, apologies. let me wipe the dust off my crystal ball.
12:08 grawity you're using `git commit-tree` then?
12:08 grawity or something equally low-level?
12:08 jimi_ joined #git
12:08 psprint grawity: actually it's libgit2
12:08 grawity yeah, the first parent is HEAD, the rest are from MERGE_HEAD
12:08 jimi_ joined #git
12:09 psprint cool, thanks
12:09 johnmilton joined #git
12:10 mkoskar joined #git
12:10 algae joined #git
12:13 Maple__ joined #git
12:14 synthroid joined #git
12:14 Guest93616 joined #git
12:16 johnmilton joined #git
12:18 ferr joined #git
12:18 rwb joined #git
12:21 sc30317 joined #git
12:21 sc30317 how do I add a second repo as a subfolder of my main repo?
12:21 sc30317 for instance, my top repository is a dockerfiles repo
12:22 sc30317 one of the subdirectories should be a git clone of a docker repo
12:22 Sonderblade joined #git
12:22 sc30317 but when I try and add the subrepo, I just see an empty folder in github
12:22 selckin !submodules
12:22 gitinfo [!submodule] 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
12:22 anuxivm joined #git
12:23 sc30317 awesome, thanks gitinfo
12:23 gitinfo sc30317: you're welcome, but please note that I'm a bot. I'm not programmed to care.
12:23 _sjs joined #git
12:23 liamonade joined #git
12:24 slugmandrew that's one cold bot right there
12:24 sc30317 seriously :)
12:25 davisonio joined #git
12:26 settermjd joined #git
12:26 linuxmodder- joined #git
12:26 joshuaturner joined #git
12:27 Gamecubic joined #git
12:27 Finnoloid joined #git
12:29 daniel_rose joined #git
12:32 stfn joined #git
12:34 dmpeters63 joined #git
12:35 dsdeiz joined #git
12:35 dsdeiz joined #git
12:35 charlesr joined #git
12:37 Falc joined #git
12:38 f0rblaze joined #git
12:38 timmu joined #git
12:39 UTAN_dev_ joined #git
12:40 ahmedelgabri joined #git
12:41 kyan joined #git
12:43 netcarver joined #git
12:43 tlaxkit joined #git
12:44 sc30317 joined #git
12:44 Guest93616 joined #git
12:44 sanketdg joined #git
12:45 zincrokx joined #git
12:46 zivester joined #git
12:46 zincrokx joined #git
12:46 btree joined #git
12:47 mkoskar joined #git
12:47 Hink joined #git
12:48 btree left #git
12:51 nilg joined #git
12:51 crelix joined #git
12:52 Falc left #git
12:53 byteflame joined #git
12:54 vassagus joined #git
12:55 YourTheBombNow23 joined #git
12:55 veilig2000 joined #git
12:55 mooru joined #git
12:55 sc30317 joined #git
12:57 rubyonrailed joined #git
12:57 kadoban joined #git
12:57 liuzhen joined #git
12:58 veilig2000 joined #git
12:58 HanClinto joined #git
13:00 Guest93616 joined #git
13:00 sc30317 joined #git
13:00 HanClinto_ joined #git
13:05 rgrinberg joined #git
13:08 jimi_ joined #git
13:10 Guest93616 joined #git
13:10 livingstn joined #git
13:11 anuxivm left #git
13:12 darkbit joined #git
13:14 ahmed_elgabri joined #git
13:14 Guest93616 joined #git
13:15 Derperperd joined #git
13:15 cdg joined #git
13:16 Rodya_ joined #git
13:18 veilig2000 joined #git
13:19 RustyShackleford joined #git
13:19 brandonw joined #git
13:20 wrouesnel1 joined #git
13:21 njbair joined #git
13:22 jceb joined #git
13:22 ExoUNX joined #git
13:24 Macaveli joined #git
13:24 Guest93616 joined #git
13:24 saidelike joined #git
13:24 Zoxc joined #git
13:27 diogenese joined #git
13:30 sanketdg joined #git
13:34 fuchstronaut joined #git
13:37 jceb joined #git
13:37 nettoweb joined #git
13:39 al-damiri joined #git
13:39 fuchstronaut joined #git
13:40 airdisa_ joined #git
13:44 Guest93616 joined #git
13:44 codeman joined #git
13:44 tinanoo joined #git
13:45 ash_workz joined #git
13:45 jstein_ joined #git
13:47 kaldoran joined #git
13:48 bongjovi joined #git
13:49 leeN joined #git
13:49 sangy joined #git
13:49 jaafar joined #git
13:51 shaggycat joined #git
13:51 rossome joined #git
13:54 phaleth joined #git
13:55 infra-red joined #git
13:56 thiago joined #git
13:56 thiago_ joined #git
13:58 Rodya_ joined #git
13:58 YourTheBombNow23 joined #git
13:59 osse I just got this error: "error: error in sideband demultiplexer"
13:59 osse Nice!
13:59 Guest5 joined #git
14:00 AaronMT joined #git
14:01 ManiacTwister joined #git
14:01 l4v2 joined #git
14:01 johnny56 joined #git
14:01 i7c Off the top of my head, I’d claim you’re doing it wrong!1
14:02 nullp0inter joined #git
14:03 tabrez joined #git
14:03 davisonio joined #git
14:03 i7c mhmm http://lists-archives.com/git/750639-git-fatal-error-in-sideband-demultiplexer.html
14:04 csd_ joined #git
14:04 mahendra_ joined #git
14:07 kbs joined #git
14:07 dhollinger joined #git
14:08 GodGinrai joined #git
14:09 phanimahesh joined #git
14:10 Dougie187 joined #git
14:12 Freeze16 joined #git
14:13 catag87 joined #git
14:17 vassagus_ joined #git
14:17 ResidentBiscuit joined #git
14:18 notebox joined #git
14:18 ResidentBiscuit joined #git
14:19 osse i7c: heh, the only thing that happened was that i pushed and the guy next to fetched at the same time
14:19 osse he did ^C, eventually I did too. After that my second git push said everything up to date
14:19 osse But I liked the vwjunkies error
14:19 i7c oh
14:19 i7c I wouldn’t have guessed that a race condition can hit git so hard.
14:21 nilg joined #git
14:21 osse maybe it's gitlab's fault
14:22 catag87 joined #git
14:22 abyx joined #git
14:23 courrier joined #git
14:26 jmibanez joined #git
14:27 davisonio joined #git
14:27 phanimahesh joined #git
14:30 Guest93616 joined #git
14:30 docmur joined #git
14:31 vassagus joined #git
14:31 fahadash joined #git
14:31 jast osse: it's always gitlab's fault :}
14:31 cyan__ joined #git
14:31 GodGinrai #blamegitlab ?
14:31 docmur Once I checkout a commit # with git checkout commitid, how do I merge that to the head so on the next git pull on other systems, they pull that commit id and not the newer ones
14:32 rubyonrailed joined #git
14:34 Hink_ joined #git
14:34 kadoban docmur: That doesn't sound like a merge. That also sounds like rewriting public history, which is not a good idea. Still want to continue?
14:35 daniel_rose joined #git
14:35 tabrez joined #git
14:35 cjohnson left #git
14:36 docmur Well what happened was two commits got pushed to the branch that are introducing an issue, so if I go back before those commits, I have the functionality I want, maybe I'm using the wrong terms.  Would it be better or right to merge the checkout to a new branch instead?
14:36 docmur I'd prefer not to modify the history
14:36 javaJake Hey all! Is there a way to have a few git mirrors which proxies all operations to the master mirror, except 'clone' which is handled locally?
14:36 kadoban docmur: Sounds like you want 'git revert' then, that's the only way to "undo" commits without changing history.
14:36 nettoweb joined #git
14:37 docmur kk I'll look that up, haven't used it before
14:38 docmur thanks :)
14:38 Rodya_ joined #git
14:38 kadoban docmur: Sure, anytime. The manpage should be decent, if you're not used to looking there.
14:39 madewokherd joined #git
14:39 docmur The man is pretty good but I just wanted to make sure I did this properly, because I don't want to corrupt the repo by doing the wrong thing
14:40 kadoban !backup always helps when you're not sure what's going on or what you're doing
14:40 gitinfo Worried about your data while trying stuff out in your repo? The repository in its entirety lives inside the .git directory in the root of your work tree so to backup everything `cp -a path/to/workdir path/to/backup` or equivalent will suffice as long as the repo is not modified during backup. See also http://sethrobertson.github.com/GitBestPractices/#backups
14:41 phanimahesh joined #git
14:45 jeffreylevesque joined #git
14:46 vuoto joined #git
14:46 clemf joined #git
14:47 darkpassenger joined #git
14:48 synthroid joined #git
14:49 Rodya_ joined #git
14:50 zeroed joined #git
14:53 tabrez joined #git
14:55 zivester joined #git
14:55 csd_ joined #git
14:56 eivarv_ joined #git
14:58 rgrinberg joined #git
14:58 Guest93616 joined #git
14:59 a_thakur joined #git
15:00 Pinkamena_D joined #git
15:00 Pinkamena_D Is there any web server type of app I can install on a git server so that I can let others see the recent commits that have been made?
15:01 roelmonnens joined #git
15:02 kadoban There's gitlab and another one
15:02 selckin gitweb
15:02 SporkWitch and still others as well, of varying quality, heh
15:03 selckin gitweb is fairly simple and nice, gitlab is like more comlpex then full github
15:03 jeffreylevesque joined #git
15:04 jast gitweb, cgit, gogs and possibly others
15:04 SporkWitch yeah, only reason i've not tried gitlab myself is the ruby dep; ruby drives me up a wall, and is the most annoying thing i've ever tried to install (all kinds of user-specific stuff; what happened to the point of a package manager being to NOT have to deal with that? lol)
15:04 ilmari gitalist, but that's a bit unmaintained
15:04 SporkWitch Pinkamena_D: for what your'e saying, sounds like gitweb would be the right one; gitlab and bitbucket are pretty beefy and definitely overkill if all you want is someplace someone can view checkins
15:05 jast well, actually gogs is a full hosting service thingy, too, unlike gitweb and cgit (which only visualize repo contents but don't assume a manager role over all of your repos)
15:05 abyx joined #git
15:05 Pinkamena_D yep, I just want something I can show a manager what has been changing recently
15:05 SporkWitch Pinkamena_D: you could even do it with a simple post-commit hook, prepending the commit hash + comment to a simple HTML table
15:05 atomicb0mb joined #git
15:06 jast gitweb is fairly simple and is maintained alongside git itself
15:06 SporkWitch (this wouldn't be pretty, but it would be trivial and basically zero overhead)
15:06 VladGh joined #git
15:06 jast in fact, if you want to run the web interface on your own computer on demand, 'git instaweb' takes care of all the details for you
15:06 Pinkamena_D Let me look at gitweb first, then maybe the script. Thanks for the suggestions
15:07 harish_ joined #git
15:07 SporkWitch Pinkamena_D: https://git.wiki.kernel.org/index.php/Gitweb bunch of others mentioned here as well
15:08 SporkWitch Pinkamena_D: i'd probably try git-php myself, since most webservers will already be LAMP/LEMP stacks anyway, fewer potential deps or config requirements
15:08 VladGh joined #git
15:08 SporkWitch nm, that one's a broken link :(
15:11 Peuc joined #git
15:11 johnmilton joined #git
15:12 Raed|Laptop joined #git
15:13 InfoTest joined #git
15:13 dunpeal joined #git
15:14 dunpeal Hello fellow Gitsters!  How do I list all branches whose remote tracking branch is gone?
15:14 rossome joined #git
15:14 subhojit777 joined #git
15:16 tyreld joined #git
15:17 nettoweb joined #git
15:19 fuchstronaut joined #git
15:19 bsanford joined #git
15:19 lightstalker joined #git
15:21 circ-user-RNfpp joined #git
15:25 eivarv_ joined #git
15:28 perlpilot joined #git
15:28 robotroll joined #git
15:28 _ikke_ dunpeal: How do you differentiate between branches that never had a remote tracking branch (local branches)?
15:29 johnmilton joined #git
15:30 Guest93616 joined #git
15:31 a_thakur joined #git
15:31 dreiss joined #git
15:31 BSaboia joined #git
15:31 cloudlinux-dev joined #git
15:33 morenoh149 joined #git
15:34 cloudlinux-dev left #git
15:35 dunpeal _ikke_: good question!  I'm asking because `git status` tells me "Your branch is based on 'origin/foo', but the upstream is gone."
15:36 dunpeal So somehow git does know.
15:38 ahmed_elgabri joined #git
15:38 abyx joined #git
15:39 synthroid joined #git
15:39 shymega joined #git
15:39 daniel_rose joined #git
15:42 darkpassenger joined #git
15:42 _ikke_ ah ok
15:43 _ikke_ Then the tracking information is left over
15:45 Rish_ joined #git
15:48 cdown joined #git
15:49 ahmedelgabri joined #git
15:49 vuoto joined #git
15:50 vuoto joined #git
15:50 cdown_ joined #git
15:51 elosz joined #git
15:51 fractalizator joined #git
15:51 elosz hello. I've a repo on github. Someone sent me a pull request. But I see "The branch has conflicts that must be resolved". How do I go about this?
15:51 elosz (new to git)
15:52 fooobarrr tell them to resolve the conflicts
15:52 elosz not sure why conflict arose in first place since single person made change in a single file
15:52 elosz he just added a single line
15:52 vuoto joined #git
15:53 vuoto joined #git
15:53 BlaXpirit elosz, and then someone made changes in the same file  and it's not so clear anymore
15:54 crayon joined #git
15:54 crayon joined #git
15:55 elosz BlaXpirit: how do i see who made changes? I am sure nobody has except the guy who sent pull request
15:55 davisonio joined #git
15:56 kbs joined #git
15:56 BlaXpirit elosz, the changes have been made independently, to your project's main branchm possibly by yourself
15:56 ice9 joined #git
15:57 ice9 if there is repo1 that points remote  'base_repo' and repo2 that also points to that remote repo;  repo1 has branch 'mybranch' locally;  how can i add 'mybranch' to repo2?
15:57 elosz BlaXpirit: is there any way? also, I see that the guy forked my repo. It shows his branch is 1 commit ahead , 11 commits behind my master
15:58 cdown_ joined #git
15:58 catag87 joined #git
15:58 BlaXpirit elosz, well that's the point, the guy made the change to your project, and then your project moved on with another 11 commits, and you're only now checking out the pull request
15:58 Guest93616 joined #git
15:59 GenZai joined #git
15:59 BlaXpirit so something of these 11 commits causes the conflict.  if you try to actually merge the pull's branch, you'll see what exactly conflicts, but github probably doesnt show it
16:00 [1]slugmandrew joined #git
16:00 codeman777 joined #git
16:01 daniel_rose joined #git
16:01 a_thakur joined #git
16:02 elosz gotcha BlaXpirit . So resolving conflict responsibility lies with person who made pull request, right?
16:02 afuentes joined #git
16:03 BlaXpirit elosz, usually that's the workflow on github (that person should `git rebase master`)
16:03 tvw joined #git
16:03 catag87 joined #git
16:03 madewokherd joined #git
16:03 BlaXpirit but before github i would guess that it was more common for the maintainer to  justresolve the conflict during merging
16:03 elosz okay BlaXpirit. Also, once he submits pull request, how can I grab it on my local machine?
16:03 Remram[m] left #git
16:04 BlaXpirit eh nobody remembers how to do that, it has to be googled.  or use the `hub` tool
16:04 elosz oh okay
16:04 jstein_ joined #git
16:04 BlaXpirit elosz, on github it says merge conflict, but doesn't it also say "view command line instructions" ?
16:04 _ikke_ !github_pull
16:04 gitinfo To easily see Github Pull Requests in your local repo, use this fetchspec trick: https://gist.github.com/piscisaureus/3342247
16:04 elosz BlaXpirit: it does
16:08 Macaveli joined #git
16:08 zacts joined #git
16:09 kbs fwiw - I usually find git ls-remote <github_repo> gives me enough info to know what interesting refs can be pulled locally
16:10 kbs (in case you happen to use both gerrit and github, as the use slightly different namespaces for essentially the same thing.)
16:13 Rish joined #git
16:17 dhollinger joined #git
16:18 maxorator joined #git
16:19 a_thakur joined #git
16:19 Remram[m] joined #git
16:20 blackwind_123 joined #git
16:20 vuoto joined #git
16:20 ice9 joined #git
16:20 fuchstronaut joined #git
16:20 vuoto_ joined #git
16:21 vuoto joined #git
16:21 vuoto joined #git
16:21 a_thakur joined #git
16:22 elosz left #git
16:24 ajf- joined #git
16:24 cdg joined #git
16:25 vuoto joined #git
16:25 pcre_ joined #git
16:26 The_Phoenix joined #git
16:26 blackwind_123 joined #git
16:26 EvilPeng1 joined #git
16:26 mingramm_ joined #git
16:27 Guest5 joined #git
16:31 ytl joined #git
16:34 Guest93616 joined #git
16:34 jeremyhall joined #git
16:35 slide joined #git
16:35 slide joined #git
16:36 mehola joined #git
16:38 Lomex joined #git
16:38 Exagone313 Hi, I want to fork a project, and still keep updates from the upstream. Do you have a tutorial to help me? Especially for what happens if a file I edit get edited in the upstream. Thanks for your help.
16:39 _ikke_ Exagone313: I don't know a tutorail
16:39 _ikke_ Tutorial
16:39 _ikke_ But if that file changes at the same place, then you get a conflict
16:40 slide joined #git
16:40 slide joined #git
16:41 NightStrike joined #git
16:41 Exagone313 Yes, so what do I do if that happens?
16:42 Rodya_ joined #git
16:42 Exagone313 There are some files I may not want to be edited, or some where I want to do so
16:42 slide joined #git
16:42 slide joined #git
16:43 alansaul joined #git
16:43 mkoskar_ joined #git
16:43 dunpeal joined #git
16:43 alansaul Probably a stupid question, but is there a way to get the git commit of repo in a different directory?, without cd'ing into it?
16:43 slide left #git
16:44 alansaul git rev-parse --short HEAD somewhereelse/on/my/computer
16:44 shaggycat joined #git
16:45 daniel_rose joined #git
16:45 alansaul Or do I need to jump into it, call the command and come back? (i want to embed the commit hash of another repo that my own code depends on, but version numbers won't suffice in this case as they don't get consistently changed with every commit)
16:46 KevinMGranger --git_dir
16:46 KevinMGranger git --git_dir=other/repo rev-parse etc...
16:46 KevinMGranger or the GIT_DIR env variable
16:47 KevinMGranger Oh whoops it's --git-dir
16:47 MainB joined #git
16:48 hahuang61 joined #git
16:49 Guest40482 left #git
16:52 alansaul Think I must be making some mistake?
16:52 daynaskully joined #git
16:52 alansaul it rev-parse --git-dir=/some/where/else/ --short HEAD
16:52 alansaul *git
16:52 alansaul gives me the commit at my current repo
16:53 MattMaker joined #git
16:53 thiago alansaul: git --git-dir=/some/where/else/ rev-parse --short HEAD
16:54 infra-red joined #git
16:54 alansaul thiago: Ah.
16:55 rwswitch joined #git
16:55 azerus joined #git
16:55 alansaul thiago: Now it says it isn't a repository, but if i cd to the exact same place it 'git rev-parse --short HEAD' gives me the correct hash?
16:55 KrazyCod3r joined #git
16:55 KevinMGranger /some/where/else/.git
16:55 alansaul Aha
16:55 davisonio joined #git
16:56 alansaul KevinMGranger: Brilliant thanks guys, exactly what I needed
16:56 thiago alansaul: git-dir is *the* .git dir
16:56 KrazyCod3r left #git
16:56 zincrokx joined #git
16:56 KevinMGranger Otherwise it would probably be named  --repo-dir
16:57 thiago there's --worktree for the worktree
16:57 zincrokx joined #git
16:57 Peuc joined #git
16:58 reda joined #git
16:58 zincrokx joined #git
16:58 infra-red joined #git
16:58 KevinMGranger If you only use worktree, will it still look for a git dir in the *current* dir, or in the worktree dir? I assume the former
16:58 alansaul Yeah makes perfect sense.
16:59 zincrokx joined #git
16:59 thiago current dir
16:59 thiago they're separate things
16:59 zincrokx joined #git
16:59 alansaul Is there a similar thing for the branch? Or if not, is it easy to look up a commit hash, from within another branch? I assume so?
17:00 zincrokx joined #git
17:00 zacts_pi joined #git
17:00 thiago alansaul: explain
17:00 zincrokx joined #git
17:01 thiago alansaul: similar to --git-dir?
17:01 alansaul No sorry, similar to rev-parse
17:01 thiago alansaul: what do you want to do?
17:01 Repox joined #git
17:01 alansaul Basically I want to be able to reproduce my current code, in a couple of years, by keeping hashes of all of the important dependencies
17:02 alansaul I guess the hash is sufficient, the branch doesn't really matter if I can find where the hash was in the whole repo
17:02 alansaul That probably isn't any clearer... Sorry
17:03 thiago alansaul: create a branch or a tag to mark the commit
17:03 thiago a tag would be best, actually. It's never supposed to move, once created.
17:03 annoymouse joined #git
17:03 Bizkit joined #git
17:04 yottabyte joined #git
17:06 alansaul Could do yeah
17:06 alansaul I had a plan that everytime I create a figure, I write the git commit as a bit of metadata
17:06 tesuji joined #git
17:06 alansaul so I wouldn't need to create a tag, out of every commit if I make a small change, it will just be automatic
17:08 vassagus joined #git
17:09 alansaul I mostly create figures with ipython notebook, so I will just collect the important git hashes there, and commit the notebook everytime I create new figures
17:10 MattMaker joined #git
17:12 alansaul Its happened too many times where I need to regenerate a figure for a talk and can't figure out which combination of commits I used from different packages :P
17:13 tabrez joined #git
17:13 durham joined #git
17:13 ki0_ joined #git
17:13 vassagus joined #git
17:14 Darcidride joined #git
17:15 leehambley joined #git
17:15 EvilPeng1 joined #git
17:15 stux joined #git
17:16 vassagus joined #git
17:17 mostlybadfly joined #git
17:17 InfoTest1 joined #git
17:18 daniel_rose joined #git
17:18 dunpeal joined #git
17:19 dreiss joined #git
17:19 jnewt i don't really understand the --shared option for git init.  if i do git init --bare --shared=true, or --shared=group, or even just --shared (defaulting to group per the docs) where do i specify the group name which should have access?
17:20 vassagus joined #git
17:21 fuchstronaut joined #git
17:21 eivarv joined #git
17:22 kadoban jnewt: Honestly, I'd just avoid all of that mess and set up gitolite, personally.
17:22 dunpeal left #git
17:23 Rodya_ joined #git
17:23 SteffanW joined #git
17:25 dhollinger joined #git
17:25 UTAN_dev joined #git
17:25 Lipp joined #git
17:26 vassagus joined #git
17:26 jnewt kadoban, i have and do use gitolite.  that is not the question. i'm looking to understand the --shared switch. can you explain how it determines the "group"? or even any information beyond whats in the manual?
17:26 leehambley joined #git
17:27 HanClinto__ joined #git
17:28 pcre_ joined #git
17:29 kbs jnewt: haven't used it, but the man seems to just say it's more like a umask setting? (i.e., the groups to which the creating user belongs have write access) is that how you read it as well?
17:29 kadoban I've used it before, not really all that interested in dredging up the memories.
17:29 thiago joined #git
17:30 Guest5 joined #git
17:30 nitric joined #git
17:31 davisonio joined #git
17:32 dmpeters63 joined #git
17:32 jnewt kbs: it uses umask settings when --shared=false.  i am thinking it uses the user who creates the repo's primary group (i don't see anything to support or contradict this though).  so if i'm user "bob" and my primary group is "bob", but also belong to "git_users", if i git init --bare --shared-true, i think that only users belonging to "bob" will be able to use it.
17:34 jnewt i tried --shared=git_users, but i get fatal: bad numeric config value 'git_users' for 'arg': invalid unit
17:36 mingrammer joined #git
17:36 thorongil joined #git
17:38 thorongil joined #git
17:40 thorongil joined #git
17:42 thorongil joined #git
17:42 catag87 joined #git
17:44 thorongil joined #git
17:45 thorongil joined #git
17:47 vassagus joined #git
17:47 thorongil joined #git
17:48 Faylite joined #git
17:52 vassagus joined #git
17:52 g105b joined #git
17:52 zacts joined #git
17:52 morenoh149 joined #git
17:53 DolphinDream git script question.. how do you find if a repo is in detached head ?
17:53 Xizor joined #git
17:53 davisonio joined #git
17:54 TomyLobo joined #git
17:54 JanC joined #git
17:55 satifant joined #git
17:58 KevinMGranger I guess you could see if .git/HEAD is a hash instead of a ref, but maybe there's a better way
17:59 osse KevinMGranger: git symbolic-ref HEAD
17:59 osse oh, i should address that to DolphinDream
17:59 KevinMGranger Right, just tried that. It'll error out with "fatal: ref HEAD is not a symbolic ref" if you're detached
17:59 DolphinDream osse :) just found that same one
18:00 Remram[m] DolphinDream: http://stackoverflow.com/a/1593487/711380
18:00 DolphinDream branch=$( git symbolic-ref HEAD --short -q)
18:00 cfxd joined #git
18:00 LeBlaaanc joined #git
18:01 kpease joined #git
18:03 eivarv joined #git
18:03 jason237 joined #git
18:05 cdg joined #git
18:05 clemf joined #git
18:08 jost joined #git
18:13 joshuaturner joined #git
18:14 eivarv joined #git
18:17 HoierM joined #git
18:17 yottabyte hi guys, on windows you get git bash or something, and now I can use ls and dir and run .sh scripts in windows cmd, how does that work? what is it?
18:19 Remram[m] It's just that, a version of bash compiled for Windows, with all the standard coreutils like 'ls', 'cp', 'mv'...
18:19 njbair joined #git
18:19 ToxicFrog yottabyte: it's not "just" bash, it's bash plus a bunch of shell utilities, perl, etc
18:20 axelitus joined #git
18:20 ToxicFrog Which are all windows native builds and thus can be called from cmd.exe as well as from bash.
18:20 ToxicFrog I think it may be based on MSYS, but don't quote me on that.
18:20 tgunr joined #git
18:20 yottabyte but what is it called
18:20 yottabyte is it mingw?
18:21 Munt joined #git
18:21 GodGinrai no
18:21 GodGinrai is this default on windows?
18:21 GodGinrai and which windows?
18:22 GodGinrai because you are either talking about windows 10, or the "ubuntu shell" feature that microsoft is toting
18:22 ToxicFrog GodGinrai: they're talking about "git bash for windows"
18:22 ToxicFrog Which is neither of those things
18:22 _28_ria joined #git
18:22 GodGinrai s/windows 10/powershell
18:22 GodGinrai well he said "now I can"
18:22 GodGinrai as if he could do this with command line instead of git bash
18:22 ToxicFrog GodGinrai: yes, because GB4W comes with coreutils, and puts them in $PATH
18:22 GodGinrai when he previously couldn't
18:23 ToxicFrog Remram[m] covered this already
18:23 mingrammer joined #git
18:23 fuchstronaut joined #git
18:23 GodGinrai I also thought I was in a different channel, which lead to some of my confusion here <.<
18:23 sagerdearia joined #git
18:23 davisonio joined #git
18:23 jimi_ joined #git
18:24 clemf joined #git
18:24 jagob joined #git
18:25 rgrinberg joined #git
18:26 HappyLoa- joined #git
18:26 weckl joined #git
18:27 jason237 joined #git
18:27 kyan joined #git
18:27 weckl_ joined #git
18:27 weckl joined #git
18:32 kulelu88 joined #git
18:33 JamesBaxter joined #git
18:34 Derperperd joined #git
18:35 diegoviola joined #git
18:39 DolphinDream is there a dryrun option for pull ?
18:39 clemf joined #git
18:39 preaction DolphinDream: that's just "fetch"
18:40 KevinMGranger Or do you mean "see what the remote ref is pointing to"
18:40 CussBot joined #git
18:40 DolphinDream i know.. but if i run git pull --dry-run .. will it work as just git fetch ? (script commands)
18:40 preaction what do you want a dry run to do?
18:40 KevinMGranger No, because that's not an option
18:40 khmarbaise joined #git
18:41 DolphinDream preaction, emulate the pull
18:41 preaction to what end?
18:41 crayon joined #git
18:41 preaction do you want to know if it will succeed?
18:41 DolphinDream that woudl be a good use
18:41 preaction what are you trying to achieve by dry-running the pull?
18:42 DolphinDream ah.. immortality
18:43 zacts joined #git
18:43 alansaul joined #git
18:43 preaction alright, well, once you figure that out, then tell us and we'll try to help
18:43 preaction otherwise, the git pull manpage will tell you whether there is a dry-run option or not
18:43 DolphinDream yeah. there isnt
18:44 preaction another satisfied customer
18:44 * DolphinDream reached Nirvana
18:45 osse fetching is fairly harmless, so do that often all the time
18:45 * preaction reaches Farvana
18:45 osse all day every day
18:45 guampa joined #git
18:45 osse then you can do git merge --no-commit @{u} to merge in what pull would have merged. regardless of whether it succeeds or not you can do reset --hard to NUKE IT ALL TO HELL
18:46 osse no need for dry run when it's so easy to mop up a wet run after the fact, nahmean
18:46 preaction also if you don't want certain things to happen, you can restrict your merge so they do not happen
18:47 xaa joined #git
18:47 DolphinDream osse, i do scripting and user can provide an emulate/dryrun option.. so i want my script code that does git pull $extraFlags to emulate if the extra flags contains a --dry-run flag .. i guess if git pull does not have a --dry-run i'll have to conditionally run git pull or not based on the dryrun flag given to my script.
18:47 abyx joined #git
18:47 kbs DolphinDream: so, git push does a fetch followed by a merge. The merge can't run without the fetch, but thefetch has a dry-run - so maybe what you want is a git fetch --dry-run?
18:48 kbs s/push/pull
18:48 DolphinDream kbs, the fetch is fine to not be dry-run since its harmless.. but the following merge part of the pull i want to be dryrunned
18:49 osse what does a dry-run merge mean ?
18:49 bremner cowsay "here's yer pony"
18:50 osse there's the git merge-tree command
18:50 osse but i've never used it mahsleff
18:50 DolphinDream osse, my guess would be that it would emulate the merge :) tells you if the merge was successful or not without actually merging.
18:51 osse I don't think there's a way to properly do that without actually performing the merge
18:51 osse git checkout -b throwaway; if git merge foo; then echo it would have succeeded; git checkout -; git branch -D throwaway
18:51 Darren_ joined #git
18:52 osse i think the merge done by merge-tree isn't a full and proper attempt. not sure where I get that from though
18:52 DolphinDream how about git merge --no-commit --no-ff  ?
18:52 rj1 joined #git
18:52 osse i've already suggested that, so I just assumed you didn't like it
18:52 DolphinDream oh :) i missed it.. sowwy
18:52 Faylite_ joined #git
18:52 osse the --no-ff makes no difference, and git will still touch files
18:52 osse but other than that it should do the trick
18:53 DolphinDream kewl
18:53 DolphinDream gracias
18:53 yoan joined #git
18:53 osse de nada
18:53 eivarv joined #git
18:54 transhuman joined #git
18:56 fuchstronaut joined #git
18:56 transhuman given the git git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git , would i do the following to determine the kernel version applied to git-cherry-pick e94b1864d1ed5be54376404bb23d22ed0481feb ? to select it then git tag --contains?
18:57 roelmonnens joined #git
18:58 _ikke_ Sorry, I don't understand the question
18:58 ice9 joined #git
18:58 ice9 is it possible to move a local branch from repo to another?
18:59 transhuman ok I want to find out what exact kernel version the git listed above applies to is it 4.7 or 4.8 ? minor revision?
18:59 _ikke_ ice9: add the other repo as a remote and fetch from it
19:00 ice9 _ikke_, both rpos are local!
19:00 ice9 _ikke_, can you show and example please of the command?
19:00 transhuman sorry should say which kernel the commit applies to
19:00 _ikke_ git remote add other file://path/to/repo
19:01 ice9 ah
19:01 JamesBaxter joined #git
19:01 _ikke_ yeah, git tag --contains would work
19:01 _ikke_ if you cloned that repo
19:01 transhuman but the cherry-pick is the command that gets me there ? yes no?
19:02 _ikke_ No, cherry-pick does something different
19:02 _ikke_ it 'copies' a commit onto the current commit/branch
19:02 transhuman oh thats not what i want i just wan to select that git commit level
19:02 transhuman s/wan/want
19:03 _ikke_ select?
19:03 _ikke_ git checkout <commit>
19:03 transhuman ah same as svn checkout I see
19:03 transhuman thanks
19:05 _ikke_ Somewhat
19:05 catag87 joined #git
19:06 transhuman thanks _ikke
19:07 Guest93616 joined #git
19:07 ice9 if --depth 1 is used, it will only clone the latest from master, or also clone the latest from all branches of that repo?
19:08 _ikke_ ice9: only latest from the default branch (usually master)
19:08 settermjd joined #git
19:09 ice9 _ikke_, what if i wanna clone latest from all branches?
19:09 LeBlaaanc joined #git
19:11 _ikke_ No idea
19:11 _ikke_ I don't think that's directly supported
19:11 __name__ joined #git
19:13 vuoto joined #git
19:13 cyphase joined #git
19:14 Seveas --depth=1 --no-single-branch
19:14 Seveas as the manpage says :)
19:17 LeBlaaanc joined #git
19:17 ice9 ;)
19:17 mingrammer joined #git
19:18 dikiaap joined #git
19:22 johnny56 joined #git
19:22 byteflame joined #git
19:22 nettoweb joined #git
19:23 GavinMagnus joined #git
19:24 johnny56 joined #git
19:25 roelmonnens joined #git
19:25 eivarv joined #git
19:27 GavinMagnus left #git
19:28 roelmonnens joined #git
19:28 Spec-Chum joined #git
19:30 ahmedelgabri joined #git
19:31 roelmonnens joined #git
19:32 Sample joined #git
19:32 roelmonnens joined #git
19:33 Manuel15 joined #git
19:33 Manuel15 joined #git
19:34 nitin joined #git
19:35 cyphase joined #git
19:36 Guest93616 joined #git
19:36 Remram joined #git
19:36 Remram[m] joined #git
19:36 Remram[m] joined #git
19:36 fuchstronaut joined #git
19:38 oich joined #git
19:38 roelmonnens joined #git
19:39 Sample joined #git
19:40 stamina joined #git
19:40 m0viefreak joined #git
19:40 settermjd joined #git
19:44 nivag joined #git
19:48 zacts joined #git
19:51 eivarv joined #git
19:53 dmpeters63 joined #git
19:54 texinwien_ joined #git
19:55 jnewt trying to push a repo to my remote, and getting a bunch of error: refs/heads/x does not point to a valid object! (where x is several branches)
19:55 oich When rebasing from master onto other, I get a conflict with message blah was deleted in HEAD~9 and renamed in master. The change being integrated is moving one of the deleted files (renaming). Is there a way to have have "this type" of conflict automatically resolve, in the future?
19:56 vdv joined #git
19:56 jnewt status shows clean, log graph shows the branches on origin (i pulled prior to pushing the changes)
19:57 d5sx43 joined #git
19:57 robotroll joined #git
19:57 SporkWitch joined #git
19:59 fuchstronaut joined #git
19:59 johnmilton joined #git
20:00 settermjd joined #git
20:00 xall joined #git
20:02 oich jnewt i don't know the answer. but git ls-remote lists commit ids and git branch --no-abbrev -v shows commit ids. if any of those commits don't exist I would think they would not be pointing to a valid object
20:03 HoierM joined #git
20:03 SporkWitch joined #git
20:04 JexSys joined #git
20:05 johnmilton joined #git
20:05 saedelaere joined #git
20:08 thejoecarroll joined #git
20:10 settermjd joined #git
20:10 cyphase joined #git
20:11 marcm joined #git
20:11 clemf joined #git
20:17 clemf joined #git
20:17 morenoh149 I git amended too much. Is it possible to recover a previous version of a commit? an ammended commit.
20:18 osse morenoh149: git reflog HEAD
20:18 osse it's one of those
20:22 vassagus_ joined #git
20:22 nikivi joined #git
20:22 axisys joined #git
20:22 davidosomething joined #git
20:23 Guest5 joined #git
20:26 morenoh149 osse: sick
20:26 jagob joined #git
20:26 i7c Get well soon, osse.
20:26 cdg joined #git
20:27 eivarv joined #git
20:27 cdg joined #git
20:28 thiago joined #git
20:28 osse i7c: it's chronic :(
20:28 i7c aww, you poor thing.
20:29 NeonSynth joined #git
20:31 elosz joined #git
20:31 elosz hello. git push will push only my current branch or all branches to remote?
20:32 i7c Only current, and only if upstream is set for it.
20:32 elosz i7c: okay and what if I want to push all the branches?
20:33 i7c It has an --all option.
20:34 elosz okay
20:34 elosz ty
20:34 i7c You’re welcome. :)
20:34 osse elosz: just gotta ask. git version?
20:35 zacts joined #git
20:35 fuchstronaut joined #git
20:37 oich is there a tool that people use to post ascii art branches that look like those in the git man pages? like here: http://stackoverflow.com/questions/26402624/how-to-rebase-in-git-forcing-theirs-for-all-conflicts
20:37 zacts joined #git
20:37 codeman joined #git
20:38 thiago joined #git
20:39 cagomez joined #git
20:40 pasky joined #git
20:41 elosz osse: 2.8.1
20:41 osse good
20:41 osse then what i7c is true :p
20:41 i7c Means i7c didn’t fuck up.
20:41 ixian joined #git
20:41 i7c O:)
20:41 osse it changed from git 1 to 2
20:44 zeroed joined #git
20:45 i7c I think 1.9 already behaved like I described.
20:45 osse no, but it did have an annoying warning saying that 2.0 would
20:45 Pinkamena_D joined #git
20:45 osse and it was possible to configure it to behave like default 2.0
20:46 i7c Oh okay, I see.
20:48 elosz I've committed some files. Now I want to remove some files which i committed say with .exe extension. How do I do that?
20:48 elosz git rm filename?
20:48 osse yup
20:48 osse add --cached if you want to keep them on disk
20:49 mingrammer joined #git
20:49 Remram[m] joined #git
20:49 Remram[m] joined #git
20:49 afuentes joined #git
20:50 elosz okay osse
20:50 osse elosz: and then add '*.exe' to .gitignore to never having to do this again
20:51 jimi_ joined #git
20:52 Literphor joined #git
20:54 Noldorin joined #git
20:56 settermjd joined #git
20:57 AciD` left #git
20:57 EvilPenguin joined #git
21:05 vassagus joined #git
21:06 foist joined #git
21:06 cagomez joined #git
21:08 cagomez joined #git
21:09 CheckDavid joined #git
21:09 Faylite joined #git
21:10 prestorium joined #git
21:11 prestorium hi all. How do I get the commit from a given branch was created?
21:11 prestorium is that even possible?
21:11 pothepanda joined #git
21:11 osse prestorium: sort of maybe, do you know the branch it was created from?
21:11 fakenullie that
21:11 fakenullie what
21:11 prestorium osse, yes
21:11 osse prestorium: in that case  git merge-base motherbranch forkbranch
21:12 osse but even that is a bit blurry if there are merges between them and such
21:12 prestorium osse, yeah... :( I saw that... :(
21:12 prestorium thanks
21:13 elosz yep osse
21:13 osse prestorium: your best best is !lol and eyeballing it
21:13 gitinfo prestorium: A nifty view of branches, tags, and other refs: git log --oneline --graph --decorate --all
21:13 osse but why do you want to know? usually it's not that interesting
21:14 aidalgol joined #git
21:15 prestorium osse, someone made a *big* mess with merges
21:15 prestorium I messed with revert a bit, but my best bet is branch from a good tag and cherry-pick the commits
21:16 azerus joined #git
21:16 osse prestorium: take it from a battlescarred veteran. just leave it be. it's not worth the mental toll
21:17 NeverDie joined #git
21:17 khmarbaise joined #git
21:17 jeffreylevesque joined #git
21:17 prestorium osse, not an option... :(
21:17 prestorium it was a *big* mess
21:18 osse is it a big mess in the git log or a big mess in the current code base?
21:18 prestorium someone merged the sandbox branch into the master branch
21:18 prestorium both! :D
21:18 osse ah, ok then it's different
21:18 osse can you blame someone else?
21:19 prestorium osse, already did that :) Now I need to fix this someone else mess
21:19 prestorium "blame someone else": check.
21:20 osse so you're not in a position to delegate?
21:20 fakenullie maybe there's nobody to delegate to
21:20 osse prestorium: so assuming there are merges between foo and bar that confuses merge-base you can do this: git log foo..bar. this will list all commits in bar that are not in foo. so the earliest of these is the original first commit, and its parent is the original fork point
21:21 osse unless bar was merged *into* foo later also
21:21 osse then you're just fucked
21:21 osse use your eyes and brain studying  git log --oneline --graph --decorate foo bar
21:21 fakenullie git bisect for code state which is ok to you
21:22 unholycrab joined #git
21:22 ikmaak joined #git
21:22 matsaman joined #git
21:22 fakenullie prestorium: can't you spot commits which came from sandbox branch?
21:23 fakenullie then remove commits which parents include those
21:23 holgersson joined #git
21:23 matsaman so I'm reading in various places that things like git status --ignored (& similar) commands won't reliably get what people would expect, namely a list of all the files ignored by git--is this still the case?
21:23 chrispg__ joined #git
21:25 i7c git ls-files -o -i should show all ignored files that aren’t tracked.
21:25 i7c mhm or not
21:25 roelmonnens joined #git
21:26 prestorium fakenullie, trying... But it looks like a month-old problem
21:26 prestorium still trying to understand what happened
21:26 matsaman i7c: yes I've read that similar commands like that also don't quite work, at least not on their own
21:26 Faylite_ joined #git
21:27 matsaman specifically looking at https://semitwist.com/articles/article/view/do-not-use-git-status-ignored but it's from 2013, so I don't want to waste my time if it's outdated
21:27 fakenullie prestorium: doesn't merge commit message contain name of branch merged?
21:27 matsaman guess I should look for a tracking bug
21:27 i7c git ls-files -i -o --exclude-standard
21:28 fakenullie someone could rebase sandbox on top of master
21:28 Literphor joined #git
21:28 p4trix joined #git
21:28 PabloR joined #git
21:30 matsaman i7c: you reckon that addresses any shortcomings perceived by the blog poster I linked to?
21:31 PabloR left #git
21:33 i7c mhm
21:34 i7c The above command works for me but I'm too lazy to check the corner cases the author mentions. :)
21:34 matsaman k, thanks
21:34 jason237 joined #git
21:35 matsaman hrmm
21:35 matsaman now if I could just figure out why my irssi isn't logging =P
21:36 i7c Because it wants to be replaced by weechat.
21:36 matsaman nawp =)
21:36 matsaman but I like the pluck of that response regardless
21:36 i7c :D
21:37 matsaman weechat's on my radar, I'm not convinced, yet
21:37 matsaman =) later on
21:38 rwb joined #git
21:38 eijk joined #git
21:38 chrisgeorge joined #git
21:41 hjudt joined #git
21:41 cd-rum joined #git
21:43 mellernoia joined #git
21:45 Literphor joined #git
21:45 * kbs pets his emacs. It has extensibility in its tagline, has to be the right answer for everything :)
21:45 sagerdearia joined #git
21:45 hjudt hi! is it unwise to manage tomcat configuration _and_ binaries in a git repository? with binaries of course i only mean the components belonging to the tomcat standard installation, not any work/temp/webapps files. i am aware that repository size will grow when updating tomcat.
21:46 RichiH hjudt: why would you do that?
21:46 eijk_ joined #git
21:46 cd-rum joined #git
21:46 hjudt RichiH: because it allows me to clone from one repository but switch to different configurations using branches.
21:46 chrisgeorge joined #git
21:47 Seveas sounds bonkers to me
21:47 hjudt why?
21:48 Seveas git != config management. Use something like pupept to manage your installs and configs (and then git to store the puppet recipes)
21:48 chrisgeorge joined #git
21:49 rtyler joined #git
21:49 RichiH pretty much what Seveas said
21:49 hjudt i can't use puppet. constraints. i could use ansible, but i'd rather not manage that configuration but let it manage by other teams.
21:49 RichiH well, s/puppet/ansible/ because we're not unwashed savages ;)
21:49 osse guys
21:49 dsantiago joined #git
21:49 osse y u no salt
21:49 Seveas ansible is for savages though
21:49 rtyler I'm trying to execute `git clone` from within an ephemeral docker container and it's giving me 'fatal: You don't exist, go away'; I'm hoping that I can override whatever ident business git-clone(1) is looking for via environment variables
21:49 RichiH osse: cause Corey uses salt
21:50 Seveas rtyler: it'll be ssh saying that
21:50 RichiH osse: no can do
21:50 hjudt but you did not explain why it is "bonkers"
21:50 Seveas rtyler: and no, that's not overridable.
21:50 rtyler Seveas: the clone is over https
21:50 osse RichiH: is that the guy with the Docker rants?
21:50 RichiH hjudt: years of shared experience tell us that you are asking the wrong questions to arrive at this answer
21:50 RichiH if you really want to do it that way, it will work, though
21:50 Seveas rtyler: hmm. Rats. Then -ENOCLUE
21:51 rtyler heh
21:51 RichiH osse: he did rant on linux.com's rss feed lately, yes
21:51 Seveas osse: I use salt for remote execution, not for config management. Mostly because we were already using puppet.
21:51 spudowiar joined #git
21:51 * RichiH made proper amounts of fun of him for saying nothing but buzzwords
21:51 hjudt RichiH: actually i am looking for experience ;-) i know it is a bit unorthodox, but so are my circumstances. i am asking here because i am trying to find out whether i am overlooking something.
21:52 Seveas RichiH: didn't you docker your app in the cloud with agile?
21:52 RichiH hjudt: you can use vcsh to split out config and binaries into two repos
21:52 * Seveas hides
21:52 RichiH Seveas: no :)
21:52 osse RichiH: I must admit I chuckled IRL when I said "This is a so-and-so diagram" and without skipping a beat "This is why I drink so frequently"
21:52 RichiH :)
21:53 RichiH thing is, i do get the container approach, etc pp
21:53 RichiH if you do it _right_, it's awesome
21:53 RichiH problem is the doing it right bit
21:53 kbs hjudt: one thought just in general - it strikes me that (assuming one of the goals is to be able to rollback a production deployment when things catch fire) do you also plan to configure dependent binaries (e.g. jdk, possibly other os-specific items?)
21:53 kbs s/configure/archive
21:53 RichiH microservice all the things for your greenfield and/or brownfield and armies of brogrammers
21:54 Dougie187 left #git
21:54 * Seveas deploys a cluebar microservice to slap RichiH
21:54 Seveas cluebat* though cluebar does sound more fun.
21:54 hjudt kbs: no. only the tomcat. i can ignore everything else because this will be set up/updated by yet another team.
21:55 osse RichiH: let's say (purely hypothetically) that I develop a desktop application for Windoze. What, if anything, can I use these things for?
21:55 Seveas hjudt: I'd say your problem isn't git or tomcat, but rubbish cooperation between teams. No amount of tech will solve that.
21:55 RichiH Seveas: i agree. let's stick with cluebar
21:55 hjudt Seveas: i totally agree with you, but i am a prisoner of this dilemma.
21:55 * Seveas pours RichiH a Caol Ila Moch. Have some clue, sir.
21:55 RichiH osse: strawman; microservices kinda imply hosted services
21:56 RichiH Seveas: i'm a rum person
21:56 RichiH thanks, though
21:56 hjudt kbs: rollback is one advantage, but not the main reason. it is also logging provided by git history.
21:56 Seveas RichiH: hmm, I don't know my rum as well as my scotch. Got any tips?
21:56 osse RichiH: no strawman. genuine question since I'm clueless.
21:56 RichiH Seveas: mixed or pure?
21:57 Seveas pure
21:57 RichiH osse: for an standalone application: nothing really
21:57 RichiH unless you need ALL the threads
21:57 jason237 joined #git
21:58 hjudt i was also investigating into using docker as everyone is fond of it nowadays, but it too does not solve all these problems. i will probably use it in addition though.
21:58 RichiH osse: it's more about having one holy cow of an hosted machine, or a few rather important ones you can't touch, or a myriad of microservices which are designed to blink in and out of existence
21:58 osse I see.
21:58 osse Thanks
21:58 RichiH notsureifserious.jpg
21:58 osse I am very serious.
21:59 LiohAu joined #git
21:59 osse In my defense, I didn't emply me!
21:59 RichiH k. all else being equal, the latter allows for more resilient services as constant failure, spin-ups etc are part of the design
21:59 osse employ.
21:59 RichiH this does not translate well for some workloads
21:59 RichiH monitoring being one of them
22:00 RichiH even google seems to do huge services with monarch
22:00 Seveas microservices also allow for more elusive failures
22:00 hjudt kbs: if i answered your previous question with yes, you would have recommended to use docker, right?
22:00 RichiH Seveas: of course
22:01 osse In my short career this is all I've worked on: 1) an embedded platform, though a powerful one, doing this and that to LEDs and FPGAs and network interfaces and whatnot.  2) A standard desktop application that shows graphs based on data taken from a database. No http, no REST, no nuthin'.
22:01 RichiH Seveas: monitoring hopefully catches those
22:01 Seveas RichiH: monitoring is always done wrong, as is orchestration. Mostly because it's pretty damn hard to not do it wrong.
22:02 Seveas (source: expensive experience...)
22:02 sagerdearia joined #git
22:02 osse I have heard of using VMs for development to ensure all devs have similar environments and compilers and all that. But in those cases the buzzword I hear is Vagrant and I believe that is something completely different
22:02 osse But I'm not even sure of that :P
22:03 Seveas vagrant is the bastard offspring of config management and vm's
22:03 Seveas it's horrid.
22:04 osse It can't be worse than debugging something for hours and then realize the only reason it worked on my computer was because of some crappy Oracle thing I installed for some reason I can't remember now, and everyone else have to add the value of $ORACLE_HOME to $PATH or copy four DLLs to the build directory
22:04 RichiH Seveas: dunno
22:05 RichiH i love monitoring, i hate monitoring tools
22:05 RichiH historically speaking, zabbix was least shite
22:05 RichiH but now....
22:05 hjudt RichiH: thanks for the hint about vcsh, i will look into it
22:05 RichiH prometheus
22:05 Spec-Chum joined #git
22:05 RichiH hjudt: no worries
22:05 Seveas hjudt: if^Wwhen vcsh does weird things, just yell at RichiH. He wrote it.
22:07 kbs hjudt: sorry - distracted here. But my 0.02 fwiw - if the problem is framed as (basically) having an audit trail of binaries + configurations - sticking it in git seems no better or worse than anything else; each big binary will effectively turn into a big object under .git with any/all concommitant problems. I know some people use git-lfs, but have no personal experience with it
22:07 RichiH Seveas: got a bug report?
22:07 Seveas RichiH: the old one with vcsh dotfiles clone failing because it needs to overwrite my .bashrc
22:08 rkazak joined #git
22:09 kbs hjudt: but yes, I was initially wondering more around understanding the problem a bit better - docker might make sense ("dependending" :) but sounds like that's not the real problem anyway
22:09 RichiH Seveas: ah, hmm
22:10 RichiH i do have a plan in mind to fix that, actually
22:10 RichiH life keeps being rather busy
22:10 RichiH but i should really dedicate a quality afternoon
22:12 durham_ joined #git
22:12 EvilPenguin joined #git
22:13 Seveas I work around it with vcsh clone $url dotfiles && vcsh dotfiles checkout -f master --
22:13 hjudt kbs: yes, in my case, docker would probably only handle binaries a bit better, while requiring new infrastructure (hub etc.). but then i will not be updating to every new micro version of tomcat so often, and thus the repo should not grow that fast. if it does, i could still dump or rebuild it. at least, so have been my thoughts.
22:13 Seveas but that's not very nice as it kills all existing files
22:14 hjudt kbs: btw, i thought git-lfs was for _really_ large files like pictures, maybe music, movies etc.
22:15 hjudt the good thing about the git repository would be that everything is also easily accessible via web interface (gitlab).
22:17 hjudt RichiH, Seveas, kbs: thanks for your help so far.
22:18 cagmz joined #git
22:19 RichiH Seveas: you would prefer what? not checking out? seeing if there's a diff and only then checking out? simply ignoring existing files? staging changes for commit?
22:20 weox joined #git
22:20 Seveas RichiH: vcsh clone --backup-to=~/.local/dotfiles-archive $url dotfiles. Or auto-backup to ~/.config/vcsh/repos.d/dotfiles/backup
22:21 hjudt left #git
22:21 RichiH Seveas: hmm, with --backup, that makes sense
22:21 RichiH auto-backup could violate principle for least surprise
22:22 Seveas agreed
22:22 RichiH Seveas: do you want to toss that into an issue?
22:22 Seveas I'll add it to #190
22:22 durham joined #git
22:22 Seveas actually, it's already there
22:23 jason237 joined #git
22:23 Seveas https://github.com/RichiH/vcsh/issues/190#issuecomment-227611431 had similar ideas
22:24 sagerdearia joined #git
22:24 osse hah, I was reading that and suddenly your new comment appeared
22:24 osse I didn't even F5 !
22:24 shgysk8zer0 joined #git
22:24 Seveas long live websuckets
22:25 RichiH osse: github has had that for a few months
22:25 RichiH it's funky
22:25 * osse rips F5 key off keyboard
22:25 osse Oh shit, I use that to buy nades
22:25 * Seveas hands osse a glass of lemonade
22:26 Gianormagantrous joined #git
22:26 osse thanks
22:29 Rodya_ joined #git
22:31 stux joined #git
22:32 codebam I have 2 branches, master, and my new_feature branch. I added a new feature in the new feature branch, but I amended the commit message in master to fix a typo in the commit message. How can I merge the two branches together so they fast forward without a merge commit?
22:32 osse codebam: rebase new_feature on master
22:32 Seveas rebase the feature branch on top of master
22:33 codebam is there a branch I have to be on when I do that? I was trying to rebase but ended up destroying data and having to use reflog
22:34 osse codebam: on new_feature
22:34 Seveas git checkout feature; git rebase master
22:34 paws_ joined #git
22:35 osse codebam: or you can type 'git rebase master new_feature' to make it "branch agnostic"
22:35 codebam oh okay, I did it backwards
22:35 osse that should've been a noop I think
22:35 osse assuming changing the commit message was all you did
22:35 osse well not a noop but that master = new_feature afterwards
22:36 p4trix I have 2 files, config.php and Config.php. If I put config.php in my .gitignore file, it ignores both :|. Is that expected behavious?
22:36 codebam I think it did the opposite
22:36 codebam actually no you're right
22:36 p4trix They are not even in the same directory... A bit strange the situation.
22:37 osse p4trix: gitignore rules without / in them apply everywhere. Other than that... are you on windows or osx with hfs+ or whatsitcalled?
22:37 osse aka. a case-insensitive file system?
22:37 Dougie187 joined #git
22:37 Seveas or just core.ignorecase set
22:38 p4trix im on a linux virtual machine... But I might just put / in there then.
22:38 p4trix Thank you :)
22:38 codebam when I did rebase master new_feature, my new_feature branch changed to be the same as master
22:39 durham joined #git
22:39 osse what about this new feature you talked about? was that not a commit of its own with new code?
22:39 codebam its just one commit
22:39 diogenese joined #git
22:40 osse maybe this is because you did the rebase the wrong way around first. or did you undo it?
22:40 jnewt_ joined #git
22:40 codebam I undid it by checking out reflog, checkout -b master2, branch -D master, branch -m master2 master
22:42 arooni when git shows changes insertions/deletions; is that just done to scale ; or does every +/- represent a change
22:42 ahr3n joined #git
22:42 codebam each + and - is a line I think
22:42 Seveas arooni: you mean in git diff --stat?
22:43 kadoban arooni: What does "done to scale" mean?
22:43 arooni i.e. each - or + represent some number of lines changed
22:43 Seveas arooni: with which command?
22:43 arooni say a git pull
22:44 osse codebam: I don't really follow. But ok then: now master and new_feature are identical right? is your feature commit the newest one? in that case you can git branch -f master @~1 (or if you have master checked out: git reset --hard @~1)
22:44 osse and you'll get the result you want
22:45 codebam well I reverted it back again after it destroyed my changes. the feature commit is the newest one, okay I'll try that thanks!
22:45 madewokherd joined #git
22:46 osse ok, just check that the second newest one is the one you expect also (namely where master should be)
22:46 codebam but the whole point is that I want to keep the last amend in master
22:46 codebam it gets erased by that
22:47 osse ok so the second newest commit is NOT the amended one, but the original one?
22:47 Bizkit joined #git
22:47 codebam in master it is
22:47 osse but you said they were equal now!!1
22:47 osse so master is back where it's supposed to be? then  git rebase master new_feature
22:47 codebam in new_feature I have the old commit before the amend in master because I did checkout -b, and the top commit is the feature commit
22:48 cyan__ joined #git
22:48 HappyLoa- joined #git
22:49 Seveas ah, so the amended commit isn't the feature commit
22:49 cjbrambo joined #git
22:50 Seveas in that case, just git checkout master and git cherry-pick new_feature  and git branch -D new_feature
22:50 xall joined #git
22:51 venmx joined #git
22:51 codebam I'm getting this after cherry picking https://ptpb.pw/Zsnc
22:51 codebam I did git cherry-pick <sha sum>
22:52 osse then master already has the feature commit, or at least already has the changes
22:52 osse The branch is a lie!
22:52 osse or cake
22:52 codebam lol
22:52 Seveas are you cherry-picking the feature commit?
22:52 osse pastebin the first couple of lines of !lol
22:52 gitinfo A nifty view of branches, tags, and other refs: git log --oneline --graph --decorate --all
22:53 codebam you're right, the code is there
22:53 osse maybe replace --all with --branches
22:53 codebam oh I think I amended it by accident
22:54 Seveas you should probably stay away from the --amend :)
22:56 durham joined #git
22:56 d0nn1e joined #git
22:57 Literphor joined #git
22:58 cjbrambo joined #git
22:59 codebam yeah lol, probably right. well I got it fixed now :)
23:00 EvilPeng1 joined #git
23:01 abyx joined #git
23:03 afuentes joined #git
23:03 btree joined #git
23:04 clemf joined #git
23:06 Ph_ joined #git
23:07 mostlybadfly joined #git
23:11 xall joined #git
23:12 a3Dman joined #git
23:12 b1tchcakes joined #git
23:14 b1tchcakes joined #git
23:14 durham joined #git
23:15 Ph_ joined #git
23:18 d0nn1e joined #git
23:18 Thorn__ joined #git
23:19 durham joined #git
23:20 nettoweb joined #git
23:27 settermjd joined #git
23:28 _Cyclone_ joined #git
23:32 Derperperd joined #git
23:33 jimi_ joined #git
23:34 grant-h joined #git
23:37 grant-h I have a file in my repo that I have not modified, but git thinks that I've deleted each line of this file and then written them all back exactly as they were
23:37 grant-h (that's what git diff shows)
23:38 grant-h even when I do a git checkout on the file this status remains
23:38 bremner sounds like problems with line endings
23:38 grant-h any ideas on what might be causing this?
23:38 nunchuck joined #git
23:39 grant-h bremner: that would make sense, I've been dealing with that issue
23:40 RustyShackleford joined #git
23:41 grant-h bremner: I have core.autocrlf set to input now
23:41 cjbrambo joined #git
23:41 grant-h but I've been experimenting with the other options on that
23:41 grant-h any ideas on how to resolve this?  should I just commit what git believes to be changes?
23:42 bremner !line-endings
23:42 gitinfo [!crlf] to fix problems with line endings on different platforms, check out http://line-endings.gitrecipes.de/
23:42 bremner no personal experience, but see ^
23:43 grant-h thank you
23:43 theskillwithin joined #git
23:44 dsdeiz joined #git
23:44 dsdeiz joined #git
23:47 kulelu88 may I ask a travis-CI question here? the #travis channel is inactive
23:50 _sjs joined #git
23:51 Neico joined #git

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