Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2016-07-04

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

All times shown according to UTC.

Time Nick Message
00:02 boombatower joined #git
00:03 tristanp joined #git
00:04 netmink joined #git
00:06 torbjorn joined #git
00:06 tesuji_ joined #git
00:06 anushil051 joined #git
00:06 ffabi joined #git
00:07 ISmithers joined #git
00:07 ISmithers Morning all :>
00:07 fcrick afternoon
00:08 PixelPaul joined #git
00:09 zumba_addict joined #git
00:09 Douhan joined #git
00:11 ISmithers Has anyone here had to train other users on some git essentials to get them up and running? I'm curious as to a good format - they are coming from centralized version control systems.
00:11 PixelPaul left #git
00:12 PixelPaulaus joined #git
00:13 PixelPaul hi all
00:13 gitinfo PixelPaul: 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.
00:13 BrunoSaboia joined #git
00:13 PixelPaul i am hoping can someone can help me ( i am new to git and figuring it all out )
00:14 PixelPaul i have a gitlab setup, and i am using source tree on my local machine to connect to my repository on gitlab
00:14 PixelPaul but i want to setup a workspace on my VPS but not sure how, and if there is a web interface i can install to make it easier
00:14 Rodya_ joined #git
00:15 ISmithers PixelPaul Do you have SSH access to the VPS? If so you could connect and then use the command line to install Git and create a bare repository.
00:17 PixelPaul i do have SSH access.   so is that how it works. i setup a bare repository? and then that will create a workspace. and then when i am finishes working in it a push it to my gitlab?
00:18 ISmithers Someone may correct me, but I'm not sure the term 'workspace' exists in git. I think what you are after is creating a 'repository' and then adding that repository as a 'remote' for your local machine.
00:19 ISmithers So the idea behind a bare repository is that you won't ever work in it directly. They are perfect for remotes.
00:20 PixelPaul ohh i gotcha, they are more for pushing to a remote to test
00:20 PixelPaul but the thing is, i wan to work on the files on the VPS
00:20 PixelPaul and when done working on them push them to my development branch
00:22 PixelPaul i am just lost as to how to properly have a copy of my app folder on my VPS i can work on then be able to push back to my dev branch when done working on it
00:22 ISmithers So a setup I use looks something like:
00:22 ISmithers My local PC, has a repository called "myproject".
00:22 ISmithers My Bitbucket account has a repository called "myproject". So the same name.
00:22 ISmithers I add the Bitbucket repository to my local git repository as a 'remote'.
00:22 ISmithers Then I can do changes (including making new branches, swapping branches, merging and all the fun stuff) and when done, I push those changes to the remote in Bitbucket, so it is available to others.
00:23 Rodya_ joined #git
00:23 ISmithers (So you could get myproject from Bitbucket and clone it locally, then you would have my changes, and my branches).
00:23 PixelPaul that is exactly what i want to do, but instead of my local, i want to use my VPS and be able to define where the folder is located that i am working on as it will be in my nginx http folder
00:24 ISmithers So you don't need to work on the remote directly. You work locally - that's the main attractive feature of git, its distributed nature. Then you update the server as/when desired.
00:24 zacts joined #git
00:25 ISmithers ie What is the main reason you wish to work directly on the VPS? Why not just use the VPS as a remote repo hosted in the cloud for data redundancy/access purposes which is what most of us do I would hasten to add. :)
00:25 PixelPaul i have this VPS setup so i can work directly on it for testing purposes. as i need to test very often when i make changes otherwise it is to slow to make changes locally and keep pushing them up to test
00:26 PixelPaul i can't host the app locally on my machine to be able to test/run it each time i modify a file
00:26 Oog joined #git
00:27 ISmithers You can't? Howcome?
00:27 thinsoldier joined #git
00:28 jimi_ joined #git
00:29 PixelPaul the app connects to many different databases which all whitelist my VPS and i can't whitelist my home IP, the VPS has some tools running that my app uses which i can't install on my VPS.  the VPS is just a clone of my production setup which i can't really reproduce on my personal computer
00:30 PixelPaul i do, do some programming on my local computer when i don't need to do constant testing tho. which is great.
00:30 ISmithers Hmm OK. That makes it a little trickier (but really only from the workflow perspective). I think common practice is to have a local setup that you can test/make changes on your project.
00:31 ISmithers I guess with Git you could make a regular repository on the GPS and then connect via that URL in SourceTree. You may find it slow though.
00:31 israelzuniga joined #git
00:32 PixelPaul i am thinking more.... i treat the VPS as if it was my local computer.
00:32 fstd_ joined #git
00:32 PixelPaul as i am sftp'ed in to the folder when i am editing the files. which is fine for me. i don't need the speed of local.
00:33 PixelPaul i just am not sure of the best way to setup the VPS like i did with source tree on my computer as that is GUI, which the VPS is all SSH
00:34 ISmithers Well you would need to host the repository somewhere that you can access via HTTP.
00:35 PixelPaul i have it on Gitlab which i am using as the main place for it
00:36 ISmithers Like I have a NAS at home where my repositories are located on a drive that I can access via HTTPS should I need to.
00:37 ISmithers Yeah but you want it on the VPS, so you can do changes etc. It feels like a strange setup to me personally.
00:37 ISmithers Can you not make changes locally, and then deploy to the server for testing?
00:38 PixelPaul the problem with that is it is very slow... like i literally write a line of code. save it. test it.
00:40 dreiss joined #git
00:40 ISmithers I feel like you really need a local testing environment. Otherwise I see no real advantage - like why not just remote desktop into the VPS and work on it via RD the entire time?
00:42 ascarter joined #git
00:44 PixelPaul i have an idea... is this possible?   i work on it locally, and as soon as i save,add,remove a file or folder. any changes in a select branch that i have checkout.  it will auto push that change to the gitlab repository and the remote development server
00:44 ailaG joined #git
00:46 ailaG Hi. If I’ve committed code but not pushed it, will “git stash” return everything to the state after the last commit? Or further back?
00:46 ailaG I did some good things, then I committed them, then I did some crap I want to revert :)
00:47 ISmithers PixelPaul: I imagine you could set up some hooks or scripts to do that. But again I would ask what is the advantage of that, why not just work on the VPS directly with RD, and push changes to Gitlab from there. Why complicate it further by adding another machine in the picture?
00:49 PixelPaul i would prefer to work on the VPS directly. which is what i do now without Git setup.  i am just not confident in my SSH skills to do pushes and commits and git commands. so i was wanting a web interface. but i can't seem to find one that allows this kinds of stuff
00:49 ISmithers ailaG: I think what you want is to "git reset HEAD^" which will pop off the changes from the last commit, and then you can revert them from there via checkout.
00:50 ailaG ISmithers: Thanks… So that keeps my unpushed, local commit then?
00:51 ISmithers PixelPaul: I won't say it doesn't exist - I think I would just say its a very niche requirement and isn't the normal workflow. If it is really impossible to get a local setup done, then you are as you say constrained by that.
00:51 wodim how do I tell git-fsck to... well, fix stuff?
00:51 inflames joined #git
00:51 wodim I have lots of dangling commits, probalby because of rebasing. can I make them disappear? does it matter at all?
00:51 ISmithers ailaG: If you have commit#1 and then commit#2 on top of that, and you do the above command, it will revert commit#2 and unstage it's changes.
00:51 PixelPaul i'll muck around with it, thanks for your help
00:52 ISmithers PixelPaul: Yeah sorry I wasn't more helpful. I'm interested to know what you settle on.
00:52 ailaG ISmithers: I have commit #1 and then some uncommitted junk that I don’t need. So I need to get back to commit #1, which is only on my computer. No commit #2
00:53 ISmithers Are these files untracked ailaG?
00:54 musl joined #git
00:54 ailaG ISmithers: *confused* Do you mean if I’ve git added them or git committed their changes? They’re git added (meaning if I commit, the bad code will be added)
00:54 ailaG I need to learn git some more, this is helping
00:54 rgrinberg joined #git
00:55 ISmithers I mean are they new files ailaG or are they files that were already part of the git repository and you have modified them but don't want to retain those modifications?
00:55 ailaG ISmithers: not new
00:55 KeelOfSteel joined #git
00:56 oberststen joined #git
00:56 ailaG I’m working on an existing WordPress site. It had a couple of bugs. I had to trace through 32198492384 WP / theme files to find the culprit. I did that with var_dumps :) (don’t judge me!) So now I know where the bug is but I may have stray debug code anywhere in the system
00:57 ailaG Before that, I fixed a bug and committed the bugfix
00:57 oberststen Hey everyone, I want to create a hook, I have my --bare repository on my server (which is my origin), and a clone of that repository on another folder that is the nginx folder so it is visible for everyone, the hook must be on my --bare repository, right? I want it to auto "git pull origin master" on my site's folder
00:57 oberststen Would it be the post-update hook?
00:57 GandalfTheWizard joined #git
00:58 ISmithers Is commit#1 pushed to the remote server already (as the command I am about to suggest will reset you, but I want to make sure commit#1 won't be lost if something goes wrong).
00:59 ailaG OK, I hit undo numerous times and now git diff shows nothing. Ugly but I guess I’m OK… ISmithers  commit #1 is not pushed to the remote server
01:00 fuchstronaut joined #git
01:00 ailaG (I have the diff command open so I have its id, I wrote git diff <id> and it returned nothing)
01:00 ailaG Just want to reset to the latest (unpushed) commit
01:00 ISmithers OK so you can do "git checkout <filename>" and that will revert changes.
01:00 ISmithers Sort of manual for a lot.
01:01 dreiss joined #git
01:01 ISmithers Also "git checkout -p" which will prompt you per chunk.
01:01 ISmithers As well as "git add -i" and then select 3 to perform reverts.
01:02 thinsoldier joined #git
01:04 ailaG ISmithers: Thanks!
01:08 fscala joined #git
01:09 rubyonrailed joined #git
01:10 dpower joined #git
01:10 MattMaker joined #git
01:12 ochorocho__ joined #git
01:16 liquidjamm joined #git
01:17 ekinmur joined #git
01:18 rgrinberg joined #git
01:20 cqi joined #git
01:25 githogori joined #git
01:25 tristanp joined #git
01:30 jyjon joined #git
01:30 denimsoft joined #git
01:32 pks joined #git
01:33 PixelPaul is there a single command to add all changes to staging and commit,  such as all files i have added,edited, removed, renamed and also folders in the checkout
01:35 nettoweb joined #git
01:38 xaviergmail joined #git
01:38 githogori joined #git
01:39 fracting joined #git
01:41 thinsoldier joined #git
01:46 ensyde__ joined #git
01:46 liuzhen joined #git
01:46 harish joined #git
01:46 d4rklit3 joined #git
01:49 ilbot3 joined #git
01:49 Topic for #git is now Welcome to #git, the place for git help and contracts | Public logs at http://goo.gl/BuUi5o | Current stable version: 2.9.0 | First visit? Read: http://jk.gs/git | Getting "cannot send to channel"? /msg gitinfo .voice | Commit early, commit often. A tip for version controlling - not for relationships
01:49 arescorpio joined #git
01:50 fraid joined #git
01:52 liuzhen joined #git
01:54 iamthad joined #git
01:55 liuzhen joined #git
01:57 d4rklit3 joined #git
02:00 lindenle joined #git
02:00 liuzhen joined #git
02:00 ISmithers PixelPaul: "git add -A" will add and track every change/file"
02:01 ISmithers Then PixelPaul "git commit -m 'I made some changes, here is my commit message'" will do the commit. I don't know if they can be combined.
02:01 PixelPaul git add -A && git commit
02:01 PixelPaul :-)
02:05 liuzhen joined #git
02:05 Literphor joined #git
02:07 ISmithers Oh you can do that? That's cool :D
02:07 Vgr joined #git
02:08 dgarr joined #git
02:09 chigang_ joined #git
02:10 texinwien joined #git
02:11 chachasmooth joined #git
02:17 david007 joined #git
02:23 githogori joined #git
02:28 hexagoxel joined #git
02:30 jyjon joined #git
02:33 englishstudent joined #git
02:34 NeverDie joined #git
02:34 Sonicbit joined #git
02:35 NeverDie joined #git
02:39 liuzhen_ joined #git
02:42 durham joined #git
02:43 oich joined #git
02:43 Aboba joined #git
02:47 MTecknology left #git
02:49 Meow-J joined #git
02:50 rkazak joined #git
02:54 durham joined #git
02:57 rubyonrailed joined #git
02:57 romerocesar joined #git
03:00 fuchstronaut joined #git
03:02 rbern joined #git
03:03 boombatower joined #git
03:10 LeMike_ joined #git
03:13 ochorocho__ joined #git
03:17 nwf joined #git
03:19 dreiss joined #git
03:26 romerocesar joined #git
03:27 treaki_ joined #git
03:29 netj joined #git
03:30 NeXTSUN joined #git
03:35 chrisshattuck joined #git
03:37 chrisshattuck joined #git
03:39 sdothum joined #git
03:44 Rodya_ joined #git
03:52 robotroll joined #git
03:52 steve joined #git
03:53 oich joined #git
03:54 ome joined #git
03:56 Rodya_ joined #git
03:57 tesuji joined #git
03:57 sssilver joined #git
04:00 Goplat joined #git
04:00 sssilver joined #git
04:02 lindenle joined #git
04:03 Lynxium joined #git
04:03 AfroThundr joined #git
04:03 iateadonut joined #git
04:04 iateadonut i'm getting this error: pathspec did not match any file(s) known to git
04:04 iateadonut but the file shows up in 'git status'
04:07 moei joined #git
04:07 rscata joined #git
04:08 hk238 joined #git
04:08 rubyonrailed joined #git
04:08 iateadonut deleted:    laravel/app/Http/Controllers/Controller.php
04:08 iateadonut shows up as 'deleted'
04:08 chrisshattuck joined #git
04:09 chrisshattuck joined #git
04:09 dan2k3k4 joined #git
04:11 mwhooker joined #git
04:13 Neobenedict joined #git
04:13 Neobenedict joined #git
04:13 sssilver_ joined #git
04:13 ochorocho__ joined #git
04:15 romerocesar joined #git
04:17 chrisshattuck joined #git
04:17 sssilver joined #git
04:18 nilg joined #git
04:21 a_thakur joined #git
04:23 zumba_addict joined #git
04:29 mkopriva joined #git
04:30 ascarter joined #git
04:32 gelei007 joined #git
04:33 ferr joined #git
04:33 mwhooker joined #git
04:34 lordjancso joined #git
04:35 t2mkn joined #git
04:35 iateadonut i had to git reset head file then checkout
04:38 _ikke_ If it's deleted from the index, you cannot check it out from the index
04:38 _ikke_ if you wanted to checkout from the last commit, use HEAD
04:39 dreiss joined #git
04:39 jaguarma_ joined #git
04:41 jungsubk joined #git
04:42 a_thakur joined #git
04:43 stantonk joined #git
04:43 cyan__ joined #git
04:44 a_thakur_ joined #git
04:44 stantonk joined #git
04:46 freimatz joined #git
04:48 muthu joined #git
04:49 bgamari joined #git
04:51 subhojit777 joined #git
04:52 a_thakur joined #git
04:54 Cabanossi joined #git
04:54 freimatz joined #git
04:56 quetz joined #git
04:58 mwhooker joined #git
05:01 fuchstronaut joined #git
05:03 sssilver joined #git
05:05 sssilver joined #git
05:05 OnBrokenWings joined #git
05:09 romerocesar joined #git
05:14 a_thakur joined #git
05:14 zacts joined #git
05:14 ascarter joined #git
05:16 ochorocho__ joined #git
05:16 jyjon joined #git
05:18 mwhooker joined #git
05:18 sssilver joined #git
05:19 mwhooker joined #git
05:20 chrisshattuck joined #git
05:22 zumba_addict joined #git
05:23 chrisshattuck joined #git
05:23 Orbitrix joined #git
05:23 jyjon joined #git
05:24 tristanp joined #git
05:27 qt-x joined #git
05:28 VagabondAzulien joined #git
05:29 nilg joined #git
05:29 testild joined #git
05:30 oich joined #git
05:30 lindenle joined #git
05:30 circ-user-jjOUE joined #git
05:33 mkopriva joined #git
05:34 testild joined #git
05:36 glebihan joined #git
05:37 Shock- joined #git
05:38 a_thakur joined #git
05:42 kaquavivak joined #git
05:44 ThomasLocke_ joined #git
05:47 yuhlw_ joined #git
05:48 Repox joined #git
05:48 Alienpruts joined #git
05:49 jyjon joined #git
05:56 rubyonrailed joined #git
05:57 tvw joined #git
05:57 freimatz joined #git
05:57 phatypus joined #git
05:57 jaguarmagenta joined #git
06:00 liuzhen joined #git
06:00 xlegoman joined #git
06:00 githogori joined #git
06:00 sssilver joined #git
06:01 romerocesar joined #git
06:05 mwhooker joined #git
06:05 bluezone joined #git
06:07 sssilver joined #git
06:11 zeroed joined #git
06:11 zeroed joined #git
06:13 davidbz joined #git
06:13 phaleth joined #git
06:15 phaleth joined #git
06:16 safe joined #git
06:16 sssilver joined #git
06:20 jknetl joined #git
06:21 mkoskar joined #git
06:25 cdown joined #git
06:28 User458764 joined #git
06:29 shgysk8zer0_ joined #git
06:35 a_thakur joined #git
06:36 Sceorem joined #git
06:39 Alienpruts joined #git
06:39 dan2k3k4 joined #git
06:40 zumba_addict joined #git
06:40 cyphase joined #git
06:42 phatypus Hello.  I'm having an issue with multi-line git commit messages.  if you look at line 16 in this paste: http://pastebin.com/sNXKrCeZ , you can see my prompt is displaying in the middle of the multi-line commit message which it should not do.  I don't have a complex prompt.  Multi-line echo works as expected.  Anyone know what might cause this?
06:44 selckin and the problem is?
06:44 ISmithers Hang on, I don't think I followed the problem in the pastibin, the commit message looks OK - or do you mean that you WANT it over multiple lines, but it is going on as single line?
06:44 selckin should be multiline in git log
06:45 thiago is your complaint that the output from git commit shows it all in one line?
06:45 moritz phatypus: fwiw you should upgrade your bash
06:45 ISmithers I think it is. In which case, I think that is as designed isn't it? Do you need to add characters like \n or the bash equivalent?
06:45 phatypus hi all.  if you look at line 16, it reads "~/tmp/git $ Reviewer: xyz" when it should read "> Reviewer: xyz"
06:46 moritz phatypus: dunno if it changes anything here, but 3.2 is, like, *old*
06:46 sanketdg joined #git
06:46 thiago phatypus: that > comes from bash
06:46 thiago phatypus: it's not part of the commit message
06:46 thiago phatypus: so far, everything looks to be working as expected
06:46 phatypus hmm, then why does line 17 look correct
06:47 thiago everything looks correct
06:47 phatypus not quite
06:47 moritz also, for multi-line commit messages, opening an editor isn't such a bad idea :-)
06:47 thiago commit messages are split by empty lines. The first paragraph is the commit's subject.
06:47 akiva joined #git
06:47 phatypus if you look at the answer in this post, this is what i expect: http://stackoverflow.com/questions/29933349/how-can-i-make-git-commit-messages-divide-into-multiple-lines
06:47 dreiss joined #git
06:47 thiago you haven't shown anything that it isn'tsplit
06:47 thiago is it split in git show?
06:47 thiago or git log
06:48 phatypus line 2, 3, 4... do not have a prompt on it
06:49 thiago I still don't see anything wrong
06:49 thiago everything is working as expected
06:49 thiago answer my question
06:49 thiago as for the people complaining about bash's version.... well, Apple happened
06:50 moritz that doesn't make the complaint any less valid :-)
06:50 thiago no
06:50 thiago the user's choice of OS is forcing the use of old, unmaintained software
06:51 thiago the user should complain to the vendor
06:51 mkoskar_ joined #git
06:51 moritz so you're saying there's no way of installing a newer shell on a MacOS?
06:51 thiago not easily
06:51 thiago you'd have to get homebrew or macports
06:51 thiago not supported by Apple
06:52 moritz well, I'd expect most developers to use homebrew or macports anyway
06:52 moritz at least I haven't talked to any developers on MacOS that uses neither
06:52 Sasazuka joined #git
06:53 thiago phatypus: it doesn't take this long to run git log or git show
06:53 a_thakur joined #git
06:53 mkoskar_ joined #git
06:53 thiago phatypus: are you trying what I asked?
06:55 Balliad joined #git
06:57 oich joined #git
06:59 thiago ok, assuming you don't need help anymore
06:59 _ng joined #git
06:59 b100s joined #git
07:00 romerocesar joined #git
07:00 ISmithers You can see how these Live Chat! things on websites defined their user-experience.
07:00 wethu joined #git
07:01 sssilver joined #git
07:02 fuchstronaut joined #git
07:02 jyjon joined #git
07:02 wethu hey, I want to fetch a branch that was created using git-flow. however I do not want to use that, i just want to fetch and track the branch locally. I've tried `git co --track -b origin/feature/some-feature` however that, i think, just checked out a branch for me locally called origin/feature/some-feature. I think / in the branch name is causing problems, i tried quoting the branch name too
07:04 nnyk_ joined #git
07:04 _ikke_ / in branch names is not a problem
07:05 _ikke_ git branch -t feature/some-feature origin/feature-some-feature just creates the branch
07:06 ahmedelgabri joined #git
07:06 wethu _ikke_, Yeah you're right, just tried `git fetch origin feature/some-feature` and now that has some traction, still wouldn't mind working out the combined fetch and merge (pull)
07:07 _ikke_ just git fetch origin would already get the remote tracking branch
07:07 encod3 joined #git
07:08 _ikke_ Then you could just merge the remote tracking branch for example
07:08 _ikke_ or create a local branch from it
07:08 WayToDoor joined #git
07:09 demiro joined #git
07:10 [cust] joined #git
07:11 Hounddog joined #git
07:11 lindenle joined #git
07:12 demiro_ joined #git
07:13 TomyWork joined #git
07:14 zeroed joined #git
07:15 Rish joined #git
07:16 dreiss joined #git
07:17 sssilver joined #git
07:17 ochorocho__ joined #git
07:19 User458764 joined #git
07:21 texinwien joined #git
07:22 phatypus_ joined #git
07:23 Neobenedict_ joined #git
07:23 Neobenedict_ joined #git
07:23 LeMike joined #git
07:25 imjacobclark joined #git
07:26 rubyonrailed joined #git
07:26 stantonk joined #git
07:26 Rish joined #git
07:26 imjacobclark joined #git
07:29 garethdaine joined #git
07:30 Andrew_K joined #git
07:31 pvinis joined #git
07:31 denimsoft joined #git
07:32 pvinis hello. i have master and dev, and the only thing changing is a config file. so when i start a branch, should i start from dev, and merge back into dev, and then what? merge into master too, or merge dev into master?
07:32 sssilver joined #git
07:32 paakku left #git
07:32 irco joined #git
07:33 selckin or a 2nd repo for the config file
07:34 artemisart joined #git
07:35 sssilver joined #git
07:36 Softeisbieger joined #git
07:37 Rish joined #git
07:38 sssilver_ joined #git
07:39 Macaveli joined #git
07:39 phatypus_ thiago selckin moritz -- hi guys, really appreciate the interest earlier, sorry my attention was moved to family and i didn't respond sooner.  i've run the exact same set of commands on 2 different machines, my mac and a centos server.  check out the difference between line 35 on both this pastes: http://pastebin.com/KQEJNDFT http://pastebin.com/tUAb7ASb
07:40 ThomasLocke joined #git
07:41 phatypus_ the behavior on mac is not quite right (on the second time i commit to that repo)
07:41 selckin even if thats correct, why do you care about that being different? its pure cosmetic
07:42 pvinis selckin: hm.. and then pull it on compile?
07:42 selckin pvinis: i don't see how configuration warrents a branch, usually you don't commit it, maybe an exmaple config file
07:43 phatypus_ selckin your right, i should not care, but it's not quite right and it gets scratches my brain
07:43 phatypus_ -gets
07:43 Impaloo joined #git
07:44 selckin phatypus_: update bash then probably an issue there
07:44 pvinis selckin: config file as in we have feature toggles and shit it there
07:44 pvinis its like a build config file
07:45 fornax joined #git
07:45 Codfection joined #git
07:45 Steele joined #git
07:46 pvinis because it too weird trying to merge branch by branch when there ae 50 feature branches and also small fixes in the dev branch, so im changing things so we just merge everything back to master, and we just have a config file so it hides some stuff that are not ready for public release
07:47 selckin workflow advice like that is difficult, since so many  things factor into it
07:47 pvinis mm..
07:48 selckin i prefer to alwys develop on master, and backport to some stable/maintance branches
07:48 pvinis i will look online for advice with git and this kind of development
07:48 pvinis yea, its essentially the same for us
07:48 pvinis we develop and test in dev, and master is the public one
07:49 pvinis so we start things from dev branch, but when we have to merge back to master, things that are not ready might merge in
07:49 multi_io joined #git
07:49 selckin feature flags should be config in the build
07:50 selckin idk to many humans factors aswel
07:51 romerocesar joined #git
07:51 pvinis indeed
07:52 phatypus $ bash --version
07:52 phatypus GNU bash, version 4.3.46(1)-release (x86_64-apple-darwin15.5.0)
07:52 phatypus ... and the multi-line commit message on os x works as expected
07:52 phatypus :)
07:52 phatypus "brew cask install bash" to the rescue
07:53 moritz \o/
07:53 moritz turns out my advise on updating bash wasn't useless after all :-)
07:54 clmsy joined #git
07:56 Orbitrix joined #git
08:00 nilg` joined #git
08:01 a_thakur joined #git
08:02 mwhooker_ joined #git
08:02 phatypus moritz you are a valued member of this channel
08:05 sanketdg joined #git
08:08 blahdodo joined #git
08:10 mwhooker joined #git
08:11 Neobenedict joined #git
08:14 ntonjeta joined #git
08:15 demiro joined #git
08:23 notebox joined #git
08:23 texinwien_ joined #git
08:27 t0by joined #git
08:33 baraba joined #git
08:34 jbitdrop joined #git
08:39 netj joined #git
08:40 mwhooker joined #git
08:41 circ-user-Zs542 joined #git
08:43 jaguarmagenta joined #git
08:45 ahmedelgabri joined #git
08:45 Beams joined #git
08:46 romerocesar joined #git
08:46 rominronin joined #git
08:47 aguenokk joined #git
08:48 jaggz left #git
08:48 SimonNa joined #git
08:48 encod3 joined #git
08:50 tvw joined #git
08:51 keltvek joined #git
08:52 jatin30 joined #git
08:52 SimonNaa joined #git
08:53 b100s hi2all
08:53 chachasmooth is it a good idea to use git lfs / git annex for my ~/ ?
08:53 b100s i have branch with some changes since commitX. now i got some new branch witch based on commitX + some commits .. kind of master.
08:54 b100s so i have to commit some !files to new branch from previous one
08:54 gitinfo [!dotfiles] There are various issues associated with keeping dot-files in git (permissions, avoiding having a .git in ~/  or /). One attempt to manage this is http://dotfiles.github.com/ and there is a wiki about it here: http://vcs-home.branchable.com
08:54 rubyonrailed joined #git
08:55 b100s no. i want to take changes of files since commitX from branch1 and commit that commits to branch2.
08:55 b100s only for certain files
08:56 EvilDMP A colleague of mine ported forward some commits in https://github.com/divio/django-cms/pull/5479, but now when I do a comparison https://github.com/divio/django-cms/compare/develop...release/3.3.x, it still lists all the commits that were ported forward, as well as some new ones, so it's not clear from that what is actually very different between them.
08:56 EvilDMP Is there some way of tracking what needs to be ported forward?
08:56 EvilDMP i.e. what hasn't yet been ported?
08:56 Lynxium[] joined #git
08:57 EvilDMP or a way of marking a commit as having been ported forward, so Git/GitHub will know not to worry about it any longer?
09:00 Lynxium joined #git
09:00 ufk joined #git
09:01 ufk hi! :) my main git repository got totally corrupted.. i was able to fix it but not without deleting lots of commits and trees which makes it irrelevant. how can I create a new git repository from existing .git files ?
09:02 notebox joined #git
09:02 fuchstronaut joined #git
09:05 King_Hual joined #git
09:05 PHPanos joined #git
09:06 irqq joined #git
09:08 kurkale6ka joined #git
09:09 rasto joined #git
09:12 eijk joined #git
09:13 lindenle joined #git
09:17 SimonNa joined #git
09:17 cyphase joined #git
09:17 nivag joined #git
09:18 ochorocho__ joined #git
09:18 ploop joined #git
09:21 ahmed_elgabri joined #git
09:23 mischat joined #git
09:23 Seveas ufk: git clone /path/to/repo/.git
09:25 OnBrokenWings joined #git
09:27 stantonk joined #git
09:35 settermjd joined #git
09:36 harish_ joined #git
09:38 acetakwas joined #git
09:41 texinwien_ joined #git
09:42 fuchstronaut joined #git
09:45 jknetl joined #git
09:48 Oog joined #git
09:50 mkoskar_ joined #git
09:52 cdown joined #git
09:52 User458764 joined #git
09:55 PCatinean joined #git
09:55 thorsten` joined #git
10:00 thorsten` left #git
10:01 romerocesar joined #git
10:02 Beams_ joined #git
10:04 Beams__ joined #git
10:06 fuchstronaut is there a way to save a stash as a favorite or something? I have two files stashed that I have to apply on a regular basis (local config files that are not local, don't ask me). Now I have to type 'git stash apply stash@{n}' depending on how many stashes there are, so I  can't just create a shortcut cmd
10:07 canton7 create a branch/tag?
10:07 izabera would you mind explaining more about these local config files that are not local?
10:08 Chill_Surf joined #git
10:08 tobiasvl sounds like an !xy problem
10:08 gitinfo Woah, slow down for a bit. Are you sure that you need to jump through that particular hoop to achieve your goal? We suspect you don't, so why don't you back up a bit and tell us about the overall objective...
10:08 fuchstronaut .htaccess and a config file that is ment to be local but ended up in the repo,
10:09 fuchstronaut each developer needs to manipulate htaccess i.e. to set up paths for the local env
10:09 ntonjeta joined #git
10:09 fuchstronaut I know there are better way, i.e. really making the config file local, but it is what it is
10:10 fuchstronaut I actually won't die from typing out 'git stash apply stash@{n}', but I thought I might ask
10:10 Beams joined #git
10:10 j416 fuchstronaut: why can't you do what canton7 says?
10:10 j416 xy problem indeed, sounds like, but still
10:11 canton7 (!config also)
10:11 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
10:11 MrWoohoo joined #git
10:11 canton7 there are some half-way houses between all-local config and all-in-repo config
10:12 fuchstronaut by creating a branch I would have to merge this branch into my feature branch to have my config and later I would push it, wouldnt I?
10:12 fuchstronaut I think this is a case where you can seperate local config clearly, but it is not up to me
10:16 cyan__ joined #git
10:16 GandalfTheWizard joined #git
10:17 canton7 fuchstronaut, no, you can apply changes from a branch on top of your working copy, just like 'stash apply'
10:18 b100s has git ability to show diff in compatible patch format?
10:18 fuchstronaut canton7: can you point me to some resource about this? thank you!
10:19 tobiasvl b100s: doesn't git diff do this by default? perhaps with --no-prefix so you get rid of a/ and b/ prefixes in paths
10:19 b100s tobiasvl, thanks
10:19 canton7 fuchstronaut, e.g. cherry-pick --no-commit
10:20 tobiasvl b100s: the patches from git diff should be patch compatible by default too though, although you'd need to use patch -p1 to strip the prefixes
10:20 fuchstronaut ah, --no-commit makes sense
10:23 _ikke_ (-n short)
10:32 ewet joined #git
10:33 rubyonrailed joined #git
10:35 ntonjeta joined #git
10:36 Darcidride joined #git
10:38 ismithers joined #git
10:38 ismithers Evening kids :>
10:39 i7c Hi dad.
10:40 yqt joined #git
10:44 jaguarmagenta joined #git
10:45 sdothum joined #git
10:50 sanketdg joined #git
10:51 Lynxium joined #git
10:54 ismithers Long time no see, at least a few hours.
10:54 ismithers How is all in git-land?
11:01 ntonjeta joined #git
11:09 romerocesar joined #git
11:09 jatin30 joined #git
11:10 User458764 joined #git
11:11 jatin30 I created a new branch B1 and my HEAD was pointing to B1. Then I did git checkout -b B2 from this branch B1. when I did git log It did not include the commit made in B1. It just had the commits in the master. Why is that?
11:12 inflames joined #git
11:12 tobiasvl jatin30: !repro
11:12 gitinfo jatin30: [!transcript] Please paste (using https://gist.github.com/ or https://vomitb.in/ ) a transcript ( https://gist.github.com/2415442 ) of your terminal session so we can see exactly what you see
11:12 lordjancso joined #git
11:12 hlysig joined #git
11:13 EvilDMP when porting forward to another branch, is there a "best way" to do it? Assuming that all commits can come from the original branch, better to cherry-pick the range, do a rebase onto the new branch, merge the original into the new?
11:13 moritz jatin30: it should, which means most likely one of your assumptions is wrong
11:13 jatin30 https://bpaste.net/show/a77659f6096f
11:14 moritz EvilDMP: merging has the advantage that commits have the same hash in both branches
11:14 b100s guys, how can i put patches( diff between com1 and com2 of branch2 ) file by file to a HEAD of branch1 ?  i have to exclude some files
11:14 moritz EvilDMP: which means that if you merge those two branches in the end, they don't show up twice (as would happen if you rebased or cherry-picked)
11:14 jatin30 I am sorry
11:15 jatin30 I got the error
11:15 moritz b100s: git diff > somefile; edit somefile, take out the hunks; git apply
11:15 ojdo joined #git
11:15 jatin30 I checked out from master and created the new branch
11:15 lindenle joined #git
11:15 aguimaraes1986 joined #git
11:16 EvilDMP moritz: in the case of https://github.com/divio/django-cms/pull/5479, it looks like that was not what was done - would that explain why the comparison of https://github.com/divio/django-cms/compare/develop...release/3.3.x has the commits with different hashes?
11:16 moritz EvilDMP: yes
11:16 EvilDMP moritz: thanks
11:17 moritz EvilDMP: note that merging isn't always possible (like when  you want to exclude some commits)
11:17 b100s moritz, somehow it doesnt work by saying http://paste.org.ru/?9qsqr7
11:17 moritz b100s: that happens when the file that is being patched changed too much
11:18 moritz if you had done a merge instead, you'd have gotten a merge conflict here
11:18 moritz in both cases, the only thing you can do is manually resolve the issues
11:18 b100s moritz, got you, thanks
11:18 GavinMagnus joined #git
11:18 ochorocho__ joined #git
11:19 netj joined #git
11:23 mrkat joined #git
11:29 marcogmonteiro joined #git
11:29 stantonk joined #git
11:30 rnsanchez joined #git
11:33 kadoban joined #git
11:35 fscala joined #git
11:35 ximvan joined #git
11:36 johnmilton joined #git
11:38 iateadonut1 joined #git
11:40 ahmed_elgabri joined #git
11:40 Furai joined #git
11:41 Furai Hey, what is the best practice when importing hg repos to git?
11:42 CheckDavid joined #git
11:45 rudi_s Furai: git-remote-hg seems to work fine, just clone it and it gets converted.
11:45 HoierM joined #git
11:48 Ordentlig joined #git
11:50 Furai Ok, any other tools?
11:52 texinwien_ joined #git
11:54 johnmilton joined #git
11:55 settermjd joined #git
11:55 marcogmonteiro joined #git
11:56 Lynxium joined #git
11:56 wodim how do I tell git-fsck to... well, fix stuff? I have lots of dangling commits, probalby because of rebasing. can I make them disappear? does it matter at all?
11:57 Lynxium[] joined #git
11:58 _ikke_ dangling objects are innocent
11:58 _ikke_ rebasing, git adding without comitting can create those
11:58 _ikke_ will be removed eventually
11:59 _ikke_ wodim: Most problems git fsck detects, it cannot fix itself (if any)
12:00 texinwien joined #git
12:00 LarArT joined #git
12:00 cbreak git gc / git repack can "fix" such things to various degrees
12:00 cbreak as would a clone of the repository with a remote protocol
12:02 Sceorem joined #git
12:02 fracting joined #git
12:02 tomog999_ joined #git
12:03 Darcidride_ joined #git
12:06 rnsanchez joined #git
12:08 SoRoot joined #git
12:10 wodim thanks
12:12 j416 wodim: in case it was not clear; dangling commits are not a problem and it's normal for them to exist
12:13 wodim git gc failed, anyway
12:13 wodim left me with some error message re: repack
12:13 wodim but well, I'll just leave it there
12:13 j416 failed?
12:13 romerocesar joined #git
12:13 jaguarmagenta joined #git
12:14 wodim https://paste.pound-python.org/raw/IK3dj2nKX2YRPRUfsz0N/
12:15 nebg left #git
12:16 j416 huh. some permission error perhaps?
12:16 j416 disk space?
12:17 Lynxium joined #git
12:17 nilg joined #git
12:17 j416 more interestingly, why do you have a git repo inside .git/objects/pack/ ?
12:17 wodim not disk space
12:17 wodim I don't
12:18 j416 .git/objects/pack/.git/objects/pack
12:18 FuzzySockets joined #git
12:18 wodim `find -name .git' only yields one result
12:18 wodim I'm starting to think it might be because this is a ntfs partition on a virtualbox host mounted on linux
12:19 j416 huh.
12:19 j416 and you're running it from the linux machine?
12:19 wodim ya
12:20 wodim it's a vagrant virtual machine
12:20 kedare joined #git
12:20 wodim running on windows
12:20 wodim the .git folder is on ntfs on the host, mounted inside the linux vm
12:21 dimi1947 joined #git
12:22 espindola joined #git
12:22 j416 should probably run it where it actually is, not over net
12:22 wodim you know what, I also have a git client installed on windows, let's see what it says
12:22 digidog joined #git
12:23 catsup joined #git
12:23 wodim ah, it didn't error out this time!
12:23 wodim but it can't remove the dangling commit.
12:23 j416 those will stay
12:23 j416 for some time
12:23 wodim well, no matter, it was just my curiosity, I don't really care
12:23 wodim thanks
12:23 j416 it's a safety net
12:24 j416 they will be removed after a minimum of 14 days (I don't remember the exact times, but it's at least 14 days)
12:24 j416 so that you can get them back in case you screwed up
12:24 diegoviola joined #git
12:25 j416 so not to worry, they will go away eventually.
12:25 rafalcpp joined #git
12:25 j416 you don't need to care
12:25 Noldorin joined #git
12:25 xvesjj joined #git
12:28 ws2k3_ left #git
12:30 ws2k3 joined #git
12:31 piling_ joined #git
12:33 cdown joined #git
12:33 nettoweb joined #git
12:34 ismithers Hey guys, I have branch1 and branch2. branch2 contains two new files which branch1 doesn't have. When I checkout branch2, those files get added to my disk. When I checkout branch1, will they still be there or do they get removed automatically?
12:35 j416 ismithers: try it?
12:35 ismithers I get the feeling I see them there still, but this was earlier today at work and so now I'm trying to figure out what happened. Like checking out branch1 doesn't remove the untracked files.
12:35 ismithers I will in fact j416, how sensible. :p
12:35 j416 ismithers: they will be removed.
12:36 leehambley joined #git
12:37 optraz_ joined #git
12:38 ismithers Yeah they do get removed.
12:38 ismithers OK weird.
12:38 hosttor joined #git
12:38 lmatteis joined #git
12:39 lmatteis guys
12:39 ismithers Maybe Visual Studio is adding some files with different project versions or something.
12:39 lmatteis how do i ignore a file (for commits) but only for me
12:39 tobiasvl lmatteis: !ignore_tracked
12:39 gitinfo lmatteis: Git only applies ignore patterns to untracked files. You can't use ignore patterns to ignore changes to files that are already tracked by git. To remove files only from git, but keeping them on disk, use git rm --cached <file>. Still, see https://gist.github.com/1423106 for ways people have worked around the problem.
12:39 lmatteis i can change my .gitignore, but then that will be committed as well
12:39 selckin .git/info/exclude
12:40 tobiasvl hm, that should be in canton7's gist
12:44 cyan__ joined #git
12:46 shaggycat joined #git
12:46 bluepixe1 joined #git
12:46 DevAntoine joined #git
12:47 DevAntoine hi
12:47 BombStrike joined #git
12:48 DevAntoine I'd like to prevent people to directly pushing onto develop. Instead I want only merge commits to be pushed. I want to have PR and then merge these PRs to develop. I can do that using hooks? The pre-receive one?
12:49 ag4ve joined #git
12:49 _ikke_ yes
12:49 _ikke_ check and verify all the refs that are being pushed
12:50 oich joined #git
12:51 ismithers1 joined #git
12:51 DevAntoine _ikke_: sorry but I'm really new to this, how should I proceed? Check that the refs comes from another branch?
12:53 fracting joined #git
12:53 Neobenedict_ joined #git
12:53 Neobenedict_ joined #git
12:54 DevAntoine _ikke_: I guess sometinhg like this would work? https://github.com/jaspernbrouwer/git-flow-hooks/blob/master/modules/prevent-master-commits.sh
12:54 a_thakur joined #git
12:54 DevAntoine but what I don't understand is merge commits
12:54 DevAntoine it's commits too
12:55 DevAntoine so how do I make the difference between merges and "unit" commits?
12:56 cbreak number of parents
12:56 _ikke_ normal commits have one parent, merge commits 2 (or more)
12:57 cbreak you can look at the parents with various ways. git cat-file -p somecommit is one
12:57 cbreak it'll show the whole commit object in somewhat human readable form
12:57 _ikke_ git show --format="%P" should only have the parents
12:57 DevAntoine so considering the script linked above, how can someone merge something to master?
12:58 kus joined #git
12:59 a_thakur_ joined #git
12:59 Mekkis joined #git
12:59 cbreak _ikke_: or better git rev-parse somecommit^@
12:59 Corneliu joined #git
12:59 cbreak git checkout master && git merge foo
13:00 DevAntoine it seems complicated to setup a PR system to develop :/ I thought I'd be done in 1h
13:00 DevAntoine -_-'
13:00 bgamari- joined #git
13:01 cbreak just don't allow people that shouldn't to push to master (or what ever)
13:01 _ikke_ cbreak: what does ^@ mean? cannot find it in gitrevisions
13:01 _ikke_ Oh, just found it
13:01 cbreak all parents
13:01 _ikke_ Didn't know that one
13:02 cbreak man pages ftw
13:02 cbreak I looked it up just before :)
13:03 _ikke_ hehe
13:03 _ikke_ Odd that is only listed in the summary
13:05 DevAntoine "Pre-receive hooks do not fire when a pull-request is merged" ok now I understand. So basically I set up a pre-receive hook which prevents commits to a given branch. But when people we'll do PR and those PR will be merged the hook won't be triggered right?
13:06 moritz is this something github specific?
13:06 Lunatrius` joined #git
13:06 DevAntoine moritz: the pre-receive hook not being fired?
13:07 moritz DevAntoine: that, and the piece of documentation you quoted
13:07 DevAntoine https://confluence.atlassian.com/bitbucketserverkb/how-to-create-a-simple-hook-in-bitbucket-server-779171711.html
13:07 DevAntoine erf, I don't know if it's specific :/
13:07 djb-irc joined #git
13:08 mrkat joined #git
13:08 DevAntoine but considering this: https://github.com/jaspernbrouwer/git-flow-hooks/blob/master/modules/prevent-master-commits.sh I guess it's not Atlassian specific
13:09 WayToDoor joined #git
13:09 moritz ok, looks like bitbucket, not github or git
13:09 cdown joined #git
13:10 User458764 joined #git
13:10 PCatinean joined #git
13:11 DevAntoine erf, I'm completely lost
13:12 bgamari joined #git
13:14 ntonjeta joined #git
13:14 hlysig joined #git
13:15 bgamari joined #git
13:17 lindenle joined #git
13:17 a_thakur joined #git
13:18 moritz DevAntoine: what do you want to do?
13:19 Silenced joined #git
13:19 Silenced joined #git
13:19 DevAntoine moritz: I don't want people to make direct push to a given branch (develop), I want them instead to make PR and then push these merged PR to develop
13:19 ochorocho__ joined #git
13:20 moritz DevAntoine: does that mean somebody merges the PR locally, and then pushes the result?
13:20 moritz DevAntoine: if yes, I don't think the restriction mentioned above applies
13:20 DevAntoine moritz: yes, it could, but most likely it'll be done with the bitbucket's interface
13:20 moritz then it will likely apply, yes
13:21 moritz DevAntoine: but do you need a hook for that? for example in gitlab, one can "protect" a branch so that people can't just push to it
13:22 DevAntoine moritz: I can set the branch to limit pushes to an empty group but I guess merged PR will fail too
13:22 DevAntoine what I'd like is this: https://confluence.atlassian.com/bitbucketserver/using-branch-permissions-776639807.html "Prevent changes without a pull request" but it's not on bitbucket's interface
13:23 moritz DevAntoine: then maybe talk to the bitbucket community, or Atlassian?
13:23 noeden joined #git
13:23 moritz working around a feature that's supposed to exist doesn't seem to be a good way to spend your time
13:24 DevAntoine moritz: I thout I could easily do it using hooks, but I'm wrong
13:24 moritz ah, the docs seem to be for "Bitbucket Server", not "Bitbucket"
13:24 DevAntoine https://bitbucket.org/site/master/issues/12771/prevent-changes-without-a-pull-request
13:24 moritz maybe that's their "pro" version, or on-prmise
13:25 rnsanchez_ joined #git
13:25 DevAntoine so meanwhile my coworkers can still push bad code on develop and make it fails... That's bad :/
13:25 moritz DevAntoine: maybe you could also, like *talk* to them
13:25 moritz tell them to run their tests before they push
13:25 DevAntoine moritz: I do
13:25 moritz and have each who pushes a bad commit pay a cake or something
13:25 DevAntoine :D
13:26 DevAntoine or I'm the only one able to push to develop and then they'll have to make PR
13:26 DevAntoine but if I'm not there they can't push new code
13:27 bgamari joined #git
13:28 selckin the code review tools usually have this mostly out of the box, push to code review, triggers test build in jenkins report sback to code review, with button to merge if tests are ok and reviewed etc
13:28 DevAntoine selckin: it'd be nice but this workflow hasn't been set on my company
13:29 DevAntoine basically anyone can push to develop, when a push is done jenkins is triggered and then executes a build to deploy it on the dev server
13:29 DevAntoine the only safety we have is if the tests don't pass the build ain't deploy
13:30 selckin ppl will hate you if make yourself the gate keeper, if its software the blocks when not reviewed or the test build failed you're free
13:30 DevAntoine idc, I want to review their code
13:30 DevAntoine you don't see what I see each day
13:30 DevAntoine they have to level ip
13:30 DevAntoine up
13:31 stantonk joined #git
13:31 Lynxium[] joined #git
13:32 bgamari- joined #git
13:34 Softeisbieger joined #git
13:37 bgamari joined #git
13:38 Lynxium joined #git
13:41 notebox joined #git
13:43 benycze joined #git
13:43 benycze left #git
13:44 mdw joined #git
13:44 NwS joined #git
13:46 nfk joined #git
13:49 cdown joined #git
13:53 shinnya joined #git
13:54 ocbtec joined #git
13:54 diego2 joined #git
13:55 ejb joined #git
13:55 Lunatrius` joined #git
13:56 Chiko joined #git
13:56 cdown joined #git
13:57 bgamari- joined #git
13:58 Lynxium[] joined #git
13:58 Bizkit joined #git
13:58 m1dnight1 joined #git
13:59 bgamari joined #git
13:59 tabrez joined #git
14:00 nnyk_ joined #git
14:02 bgamari- joined #git
14:02 Gsham joined #git
14:02 sdothum joined #git
14:02 settermjd joined #git
14:05 Lynxium joined #git
14:05 D-Boy joined #git
14:06 cdown_ joined #git
14:10 bgamari joined #git
14:14 netcarver joined #git
14:16 khelkun joined #git
14:16 khelkun yz
14:17 Lynxium joined #git
14:19 _ng joined #git
14:19 Disar joined #git
14:19 a_thakur joined #git
14:20 settermjd joined #git
14:21 mischat_ joined #git
14:22 Disar Hi Q: At work we are using git to version our code+assets for games. Some games have multiple branches with different variations in assets/codes. I want to skimm down on size. As a test I just cloned a project ( about 11 gigs ). I made a new branch and only commited the head of the branch to it. Deleted the original branch. I assumed this would reduce the size of my repo but didn't.
14:22 cdown_ joined #git
14:22 Disar I deleted the .git file and created a clean repo and pushed all my content and it was reduced from 11 to 3 gigs.
14:22 dopesong joined #git
14:23 Disar So what exactly is containing all that data? Are all the other branches also affecting the size although I've only cloned one branch?
14:24 snowkidind joined #git
14:28 kaldoran joined #git
14:28 cdown joined #git
14:28 [cust] joined #git
14:30 Lynxium joined #git
14:30 Meteorhead say I have a commit 56efbcd is it possible to view the source of a file in the commit right before this one?
14:31 a_thakur_ joined #git
14:31 jast git show 56efbcd^:path/to/file
14:31 jast more info about the syntax of that last argument in man gitrevisions
14:31 gitinfo the gitrevisions manpage is available at http://jk.gs/gitrevisions.html
14:31 NeXTSUN joined #git
14:31 Meteorhead jast: hell yeah! thanks mate
14:32 cdown_ joined #git
14:34 fspy this gitinfo bot is smrt
14:35 gitinfo fspy: Yes, I am
14:35 fspy see wow
14:36 romerocesar joined #git
14:36 djb-irc joined #git
14:37 bgamari joined #git
14:39 cdown joined #git
14:40 Disar Anyone?
14:40 canton7 Disar, there are some hoops you need to jump through to reduce the size of a repo
14:41 canton7 there are some notes at the bottom of man git filter-branch
14:41 gitinfo the git-filter-branch manpage is available at http://jk.gs/git-filter-branch.html
14:41 canton7 but cloning the repo is an easy way
14:42 Silenced joined #git
14:42 Silenced joined #git
14:42 Disar Just came across those things but considering the amount of commits with several bulks of binary files it seems very tedious
14:42 Disar Perhaps I should just force a clean repo to remote
14:42 canton7 you read "Checklist for Shrinking a Repository"?
14:42 canton7 start with "Then there are two ways to get a smaller repository"
14:43 Disar If I were to cut off some of its history? Won't that help reduce size?
14:43 jaguarmagenta joined #git
14:43 Disar Thing is we don't really need to go back to the initial point. These games once done are in a good state
14:44 Disar Well at least I tried to cut the history by creating a new branch and only commit the head of the previous one, but that didn't help
14:44 bgamari joined #git
14:44 Disar ill look into that git-filter-branch page
14:45 canton7 yes it will reduce the size, but that won't reduce the size of .git straight away
14:45 Disar when will it then?
14:45 canton7 git keeps old stuff around for a while as a safety net
14:45 Disar ah
14:45 canton7 the bit I linked you to explains how to convince git to get rid of them
14:45 Disar right, saved me a few times when i went too many commits back
14:46 Disar Eh I think for most projects I think I can just delete the .git folder and initialize a new repo commit and force push to the server.
14:46 aqil joined #git
14:46 Disar and a backup on some external drive we have
14:47 bgamari- joined #git
14:47 Disar alright ill do some reading then and make a decision. Thanks for the help
14:47 Disar ill probably break compatibility with other branches I assume.[
14:47 Spacew00t joined #git
14:50 _ng joined #git
14:51 Lewix guys
14:51 Lewix im trying to track a branch but i get the following
14:51 Lewix https://www.irccloud.com/pastebin/c2GopHE2/
14:51 Lewix why is it ambigous?
14:51 Lewix https://www.irccloud.com/pastebin/vlxypsJV/
14:51 nettoweb joined #git
14:51 raijin joined #git
14:51 Lewix https://www.irccloud.com/pastebin/1ObeAuJE/
14:52 bgamari joined #git
14:52 happy-dude joined #git
14:52 ekinmur joined #git
14:54 osse Lewix: lines 7 and 21 of your last paste
14:54 osse you have a local branch with "origin/" in the name.
14:54 bruce_lee joined #git
14:54 User458764 joined #git
14:55 t0by joined #git
14:55 Lewix osse: what's wrong with that?
14:55 dan2k3k4 joined #git
14:55 acetakwas joined #git
14:55 SoRoot joined #git
14:55 osse Lewix: technically speaking there is nothing wrong with that
14:55 osse but it's what creates the ambiguity
14:56 Lewix so if i delete the local branch technically it should work?
14:56 osse yes, or you use the full names
14:56 cdown_ joined #git
14:56 osse usually when people have local branches with "origin/" in the name it's due to a typo in a git-checkout or git-branch command
14:56 Lewix how can i make the local branch track that branch
14:57 Lewix osse: the remote repo has origin in the name
14:57 osse yes that's common
14:58 osse use the full names
14:58 Lewix im gonna go
14:58 Lewix osse: so the remore branch is origin/feature/cpmt-foo
14:58 osse the first paste suggest you want to create a branch with the name "feature/blah", but you already have one named "origin/feature/blah".
14:58 Lewix and i have origin/feature/cpmt-foo
14:58 bgamari- joined #git
14:59 Lewix so im gonna do
14:59 Lewix git branch feature/cpmt-foo
14:59 Lewix git branch -d origin/feature/cpmt-foo
15:07 roknir_ joined #git
15:09 Lewix then i have to somehow track the remote branch to feature/cpmt-foo
15:09 osse your original command is find
15:09 bgamari joined #git
15:09 osse fine
15:09 osse it will work if you remove the branch you probably don't want in the first place
15:09 Lewix $ git branch backup-compt-38-create-preview-page
15:09 Lewix warning: refname 'origin/feature/CPMT-38-create-preview-page' is ambiguous.
15:09 Lewix fatal: Ambiguous object name: 'origin/feature/CPMT-38-create-preview-page'.
15:09 Lewix the problem i have is i have saved changes that i need
15:09 Lewix an i can create a new branch
15:09 thoraxe joined #git
15:09 Lewix maybe i should reset the head --soft stash and commit again
15:09 eijk_ joined #git
15:09 Vgr_ joined #git
15:09 thoraxe so i am trying to figure out why certain folders' contents are not being added to my repo
15:09 irqq_ joined #git
15:09 thoraxe in a repo I am cloning another repo and then deleting all the .git stuff in there
15:09 cerkauskas joined #git
15:09 ploop joined #git
15:09 thoraxe but for some reason "git status" isn't showing any of that stuff, and there's nothing mentioning those folders in .gitignore
15:09 osse thoraxe: git check-ignore -v path/to/stuff
15:09 thoraxe "git check-ignore -v ." shows nothing
15:09 jchonig_ joined #git
15:09 osse !repro
15:09 gitinfo [!transcript] Please paste (using https://gist.github.com/ or https://vomitb.in/ ) a transcript ( https://gist.github.com/2415442 ) of your terminal session so we can see exactly what you see
15:09 thoraxe how do I show nothing?
15:09 osse maybe git still thinks it's a repo
15:09 thoraxe there is no output.
15:09 polyzen
15:09 osse cd ..; git check-ignore -v path
15:09 thoraxe git check-ignore -v vim-ansible-yaml -> no output
15:09 mmmveggies joined #git
15:09 dan2k3k4 joined #git
15:09 brandonw_ joined #git
15:09 adante_ joined #git
15:09 a_thakur joined #git
15:09 osse cd vim-ansible-yaml; git rev-parse --show-toplevel
15:09 thoraxe osse: /home/thoraxe
15:09 jlewis_ joined #git
15:09 thoraxe pwd -> /home/thoraxe/.vim/bundle/vim-ansible-yaml
15:09 evil-t0by joined #git
15:09 izabera1 joined #git
15:09 retropikzel joined #git
15:09 thoraxe cat ~/.gitignore | grep -E "bundle|ansible|vim" -> nothing that would ignore this folder
15:09 monte joined #git
15:09 darlinger joined #git
15:09 WebDawg joined #git
15:09 osse is .vim/bundle part of the repo?
15:09 diogenese joined #git
15:09 thoraxe yes
15:09 NeverDie joined #git
15:09 mmmveggies joined #git
15:09 thoraxe because other files/folders in here are properly tracked
15:09 roknir_ joined #git
15:09 osse are you sure you haven't added them already?
15:09 thoraxe there are a few random folders that for some reason are not being tracked
15:10 thoraxe can you rephrase the question?
15:10 icedp joined #git
15:10 yazdmich joined #git
15:10 ejb joined #git
15:10 Gitzilla joined #git
15:10 Sigma joined #git
15:11 osse I am asking whether you have already done this
15:11 thoraxe osse: if i added them, then when I re-clone the repo I would have them
15:11 sharpy joined #git
15:11 thoraxe but I have just cloned the repo and those folders are empty.
15:11 sssilver joined #git
15:11 thoraxe so the folders are being tracked, but no files inside
15:11 thoraxe this repo is public
15:11 thoraxe https://github.com/thoraxe/homedir
15:11 KevinMGranger joined #git
15:11 osse that's impossible. git doesn't track empty dirs. i am starting to think they are submodules
15:11 thoraxe how do i list submodules
15:11 thoraxe wait
15:11 thoraxe you're misunderstanding
15:12 thoraxe the dirs have files on the system i am inspecting
15:12 Shentino joined #git
15:12 thoraxe but you can see that the dirs are empty in the repo on github
15:12 thoraxe they look like they are showing up as submodules
15:12 thoraxe but i have no submodules actually configured (to my knowledge)
15:13 osse they are submodules
15:13 thoraxe ok, i didn't want them to be submodules
15:13 evil-t0by joined #git
15:14 thoraxe so how do I unsubmodule them
15:14 thoraxe haha
15:14 djb-irc joined #git
15:14 sssilver joined #git
15:14 osse git rm vim-asciidoc~ asciidoc-vim Dockerfile vim-ansible-yaml
15:15 osse maybe you did git add on them before removing the .git stuff
15:15 osse that makes them submodules I think
15:15 thoraxe probably.
15:16 thoraxe it gives me an error about submodules and asks me if i want to use -rf
15:16 thoraxe i suppose i don't care i can always re-clone this stuff
15:16 thoraxe the following submodules (or one of their nested submodules) use a .git directory:
15:17 thoraxe (use 'rm -rf' if you really want to remove it including all of its history)
15:17 osse nuke it all to hell
15:18 thoraxe ok that appaers to be working
15:18 thoraxe thanks. gotta go through this
15:18 lindenle joined #git
15:20 ochorocho__ joined #git
15:21 thoraxe osse: fixed! thanks,.
15:21 thoraxe osse++
15:22 Urbany joined #git
15:23 raininja joined #git
15:24 MattMaker joined #git
15:26 artemisart joined #git
15:29 quetz joined #git
15:29 madewokherd joined #git
15:30 AaronMT joined #git
15:30 AaronMT joined #git
15:30 israelzuniga joined #git
15:32 rkazak joined #git
15:33 stantonk joined #git
15:33 sssilver joined #git
15:34 efdetonator joined #git
15:38 dreiss joined #git
15:38 xlegoman joined #git
15:39 fuchstronaut joined #git
15:42 ochorocho__ joined #git
15:43 ramsub07 joined #git
15:44 spudowiar joined #git
15:45 Alienpruts joined #git
15:48 fscala joined #git
15:48 cyan__ joined #git
15:48 User458764 joined #git
15:50 djb-irc joined #git
15:51 ochorocho__ joined #git
15:56 cagmz joined #git
15:59 OnkelTem joined #git
15:59 javajoe joined #git
16:00 pkmule joined #git
16:01 ahmed_elgabri joined #git
16:03 FuzzySockets joined #git
16:03 OnkelTem Hi all. I'd like to make `git pull` w/o switching my current branch (commited locally). So I did `git fetch`. Now how can I make git to update specific branch which is not current?
16:04 bgamari- joined #git
16:05 shgysk8zer0_ joined #git
16:05 sssilver joined #git
16:06 ojacobson generally, you can't. Merges and rebases - the two common strategies for integrating branches - both require a work tree.
16:06 OnkelTem ok
16:06 nettoweb joined #git
16:06 OnkelTem Just never did it. And will never do LOL
16:07 InfoTest joined #git
16:09 israelzuniga joined #git
16:10 dopesong joined #git
16:10 FuzzySockets joined #git
16:11 mischat joined #git
16:11 settermjd joined #git
16:11 basiclaser joined #git
16:14 israelzuniga_ joined #git
16:14 Silenced joined #git
16:14 Silenced joined #git
16:16 madewokherd joined #git
16:17 tristanp joined #git
16:18 ovii joined #git
16:18 israelzuniga joined #git
16:18 chrisshattuck joined #git
16:19 sssilver joined #git
16:22 chrissha_ joined #git
16:25 cbRoy joined #git
16:26 bgamari joined #git
16:26 TomyLobo joined #git
16:27 dopesong joined #git
16:28 Gsham joined #git
16:29 bgamari- joined #git
16:31 bgamari- joined #git
16:31 bluwave joined #git
16:32 yqt joined #git
16:34 bgamari joined #git
16:34 aard_ joined #git
16:34 sssilver_ joined #git
16:35 javajoe How do I revert back to my previous commit so the old code can be restored
16:36 ramsub07 joined #git
16:36 cookie3 left #git
16:36 ramsub07 Hello, i've had a PR opened on github. Someone else opened a PR to my branch and after merging that PR and a rebase with master, commits have duplicated in the PR. How do i fix this ?
16:37 abdulmagnun joined #git
16:38 bgamari- joined #git
16:38 israelzuniga joined #git
16:38 jhass ramsub07: got a link to the PR?
16:38 abdulmagnun hey guys, i am wondering, i have the commits A B C D E. i want to make D the last commit again but not loose the E commit, so i want to end up with A B C D E D. any ideas??
16:38 HoierM joined #git
16:38 jhass !revert
16:38 gitinfo That's a rather ambiguous question... options: a) make a commit that "undoes" the effects of an earlier commit [man git-revert]; b) discard uncommitted changes in the working tree [git reset --hard]; c) undo committing [git reset --soft HEAD^]; d) restore staged versions of files [git checkout -p]; e) move the current branch to a different point(possibly losing commits)[git reset --hard $COMMIT]?
16:38 ramsub07 jhass: https://github.com/Theano/Theano/pull/4570/commits
16:38 jhass javajoe: ^
16:39 abdulmagnun jhass: thank you, i tried reset --hard but it simply removes the E commit
16:40 jhass oh that wasn't for you :)
16:40 jhass abdulmagnun: where do you imagine E to stay?
16:40 abdulmagnun in the middle somewhere
16:40 abdulmagnun just in case
16:40 jhass do you mean you want to have the source code in the same state as it was with D?
16:40 abdulmagnun i mean the way i wrote it would be great: A B C D E D
16:40 abdulmagnun yes
16:41 jhass simply git revert E
16:41 toogley joined #git
16:41 abdulmagnun but i think that removes the E commit, no?
16:41 jhass it makes a new commit with the reverse patch of E
16:41 abdulmagnun and names it what?
16:41 jhass it's a new commit. The commit message is whatever you want
16:42 abdulmagnun ah ok
16:42 abdulmagnun so its git revert -m "asd"?
16:42 cbRoy or git reset HEAD~1
16:42 jhass if you want to, there also is a default of Revert "message of E"
16:42 abdulmagnun jhass: ah yes, this would be sufficient
16:44 abdulmagnun jhass: ok i am getting stderr: 'usage: git revert [<options>] <commit-ish>...
16:44 cbRoy you probably left out the commit
16:44 abdulmagnun what other info does it need? i want to go one step backwards
16:44 jaguarmagenta joined #git
16:44 abdulmagnun cbRoy: hmm how can i indicate that i want the last commit to be reverted?
16:44 cbRoy do you want to keep the last commit, or commit over it?
16:45 abdulmagnun cbRoy: i want to end up with A B C D E D
16:45 jhass ramsub07: mmh, not sure what went wrong, I suppose your local master branch has some of the commits. I would recommend you get a remote for upstream, (git remote add upstream https://github.com/Theano/Theano.git; git fetch upstream); and git rebase -i upstream/master; and throw everything out you don't want
16:45 cbRoy you need to specify the commit you want to revert
16:45 abdulmagnun cbRoy: the very last one
16:45 OnkelTem Folks, how to properly ignore local edits of a file? What does exactly do adding a pattern to .git/info/exclude for example?
16:45 israelzuniga_ joined #git
16:46 abdulmagnun cbRoy: so git revert HEAD~1?
16:46 jhass ramsub07: and in the future to rebase use git fetch upstream; git rebase upstream/master; and to create new feature branches use git fetch upstream; git checkout -b foo upstream/master
16:46 abdulmagnun no
16:46 cbRoy abdulmagnun, that should work
16:46 ramsub07 jhass: i've upstream as my remote. and i've pulled the recent upstream master. if i do git rebase -i upstream/master, would that be enough ?
16:47 ramsub07 jhass: -i is for interactive, i guess ?
16:47 jhass abdulmagnun: just git revert HEAD
16:47 jhass ramsub07: yes
16:47 jhass ramsub07: it won't magically remove duplicates, you have to remove those lines by hand there
16:48 ramsub07 jhass: yeah, there are two of them, suppose if i remove one, i won't end up ruining the code right ?
16:48 bgamari joined #git
16:48 jhass IME not
16:48 jhass in any case you can always get back with git reset --hard 7eabb8d870aa7eb92f8bf3cd6e2c5f56965d30dd
16:49 Bizkit joined #git
16:49 israelzuniga_ joined #git
16:49 sssilver joined #git
16:50 bgamari- joined #git
16:50 OnkelTem If I add an existing file to .git/info/exclude, then what will happen after a pull which modifies the excluded file?
16:51 grawity if it's untracked – the merge will stop with an error
16:51 SteffanW joined #git
16:51 grawity if it's already tracked by git – nothing; adding to .gitignore doesn't magically make it untracked.
16:52 OnkelTem so it will be overwritten silently by the merge?
16:52 grawity no, it'll be merged as usual
16:53 grawity if you untrack it on your side, then git will refuse to touch it, regardless of whether it's ignored or not (that merely hides it from `git status` anyway)
16:53 JanC_ joined #git
16:53 thiago joined #git
16:53 thiago joined #git
16:54 israelzuniga joined #git
16:57 sangy joined #git
16:57 tristanp_ joined #git
16:57 israelzuniga__ joined #git
16:57 chrisshattuck joined #git
17:00 espindola joined #git
17:02 freimatz joined #git
17:04 xaviergmail joined #git
17:05 israelzuniga_ joined #git
17:05 sssilver joined #git
17:06 tristanp joined #git
17:07 shinnya joined #git
17:08 ramsub07 i've a branch named "A" remotely(on my PC, haven't pushed). can i push that branch to an existing branch?
17:08 ailaG joined #git
17:10 tristanp_ joined #git
17:10 freimatz joined #git
17:13 InfoTest joined #git
17:13 espindola joined #git
17:14 texinwien joined #git
17:17 acetakwas I keep getting a fatal: unable to access 'http://github.com/XXX.git/': Could not resolve host: github.com
17:17 acetakwas What can I do?
17:17 ojacobson Fix your resolver
17:17 acetakwas This only happens on my current wifi network
17:17 ojacobson if your machine can't resolve github.com then git won't be able to either
17:17 acetakwas It doesn't happen on other networks
17:17 acetakwas How do I do that please?
17:17 ojacobson Then either that network's DNS setup is screwy or your local machine's DNS setup is screwy
17:18 bgamari joined #git
17:18 acetakwas It doesn't happen on my phone's hotspot connection either.
17:19 ojacobson 99% this isn't a git problem - can you ping github.com from that machine on that network?
17:20 bgamari- joined #git
17:20 lindenle joined #git
17:20 ojacobson 99% this isn't a git problem - can you ping github.com from that machine on that network?
17:21 jhass ramsub07: git push foo a:b pushes the local branch a to the remote branch b at the remote foo
17:21 ojacobson (had a network burp, sorry)
17:21 acetakwas PING github.com (192.30.253.113) 56(84) bytes of data.
17:21 acetakwas 64 bytes from 192.30.253.113: icmp_seq=1 ttl=50 time=401 ms
17:22 ojacobson neat
17:22 bgamari| joined #git
17:22 jhass what about ssh?
17:22 jhass ssh git@github.com
17:22 robotroll joined #git
17:22 ojacobson that's the 1% case - can you run (and pastebin the results of) GIT_CURL_VERBOSE=1 git clone https://github.com/foo.git
17:23 acetakwas ssh: Could not resolve hostname github.com: Temporary failure in name resolution
17:23 jhass sounds like your ping was just lucky to me
17:23 acetakwas ping keeps getting the same result
17:24 ojacobson I'm thinking ip6 vs ip4 issues
17:24 ojacobson try ping6 github.com as well
17:24 ojacobson (ping(1) only tries ip4, by design)
17:24 Vbolea joined #git
17:24 jhass github doesn't have AAAA records?
17:24 ojacobson ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 4328
17:24 ojacobson ;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 0
17:24 ojacobson it does not.
17:24 acetakwas ping6 github.com
17:24 acetakwas unknown host
17:25 jhass so ping6 can't possibly succeed
17:25 ojacobson Well, that's consistent, at least.
17:25 ojacobson Yeah. My theory right now is that something in acetakwas' network stack is preferring ip6 and failing, and for some reason not trying ip4
17:25 ojacobson (still not a git issue, but at least it'd explain what's going on)
17:25 ojacobson can you run the diagnostic git command, please
17:25 jhass acetakwas: curl -v http://github.com ?
17:25 acetakwas I get the same error over anything git
17:25 ojacobson or that
17:25 acetakwas Opnshift, Github, Bitbucker
17:26 acetakwas s/Bitbucker/Bitbucket
17:26 jhass what's your OS anyway?
17:26 ojacobson if it's got ping6 it's almost certainly either mac os or linux
17:26 ojacobson if ip6 is broken, it's probably not mac os
17:26 acetakwas * Rebuilt URL to: http://github.com/
17:26 acetakwas * Hostname was NOT found in DNS cache
17:26 acetakwas Ubuntu
17:26 jhass please use a pastebin
17:27 acetakwas curl: (6) Could not resolve host: github.com
17:27 acetakwas Sorry
17:27 acetakwas These are really short pastes
17:27 acetakwas that's why I'm pasting here.
17:27 jhass oh that's all?
17:27 jhass so at least curl can't consistently resolve
17:27 acetakwas Yes.
17:28 jhass while curl does have a tendency to mask other errors with that error message, given ssh couldn't resolve either it probably indeed is a resolver issue though
17:28 jhass do you have dig? does a dig github.com return an A record?
17:28 irco joined #git
17:29 acetakwas I guess. Any workaround?
17:29 martin_g joined #git
17:29 ojacobson Fix whatever bit of network misconfiguration is causing your machine to ignore ip4, I guess. Ask #ubuntu?
17:29 jhass I don't think ignoring v4 is the issue
17:29 jhass but we can easily find out
17:29 ojacobson sure :)
17:29 jhass what's curl -4 -v http://github.com giving?
17:29 bgamari joined #git
17:31 acetakwas `$ dig github.com`
17:31 acetakwas http://pastie.org/10898724
17:31 codeman joined #git
17:31 jhass 127.0.1.1#53
17:31 jhass so you run a local unbound or dnsmasq or something?
17:31 bgamari- joined #git
17:32 ag4ve joined #git
17:32 acetakwas `$ curl -4 -v http://github.com`
17:32 acetakwas http://pastie.org/10898726
17:33 acetakwas jhass::  I don't know what that is.
17:33 sdothum joined #git
17:33 fuchstronaut joined #git
17:33 jhass hrm so it's v4 vs v6 after all?
17:34 jhass 127.0.1.1#53 being your DNS server without you knowing anything about it is weird too
17:34 _ikke_ acetakwas: dnsmasq is a local caching dns server
17:34 jhass what's sudo ss -tlnp | grep 53 giving?
17:34 jhass er ulnp
17:35 sssilver joined #git
17:35 st0opkid joined #git
17:35 stantonk joined #git
17:37 acetakwas http://pastie.org/10898731
17:37 acetakwas jhass::  ^
17:38 jhass that's with sudo?
17:38 jhass or as root
17:39 _W_ joined #git
17:39 acetakwas http://pastie.org/10898733
17:42 jhass so you got a dnsmasq running for some reason
17:43 jhass is that just something ubuntu does these days? :/
17:43 nwf left #git
17:44 Bizkit joined #git
17:44 tang^ probably
17:46 jhass acetakwas: well given we've shown the issue in lower level tools, this has indeed become a question for #ubuntu I think, ask them why curl -4 github.com works while curl github.com doesn't
17:46 jhass (also given you never heard about dnsmasq but have one running)
17:47 espindola joined #git
17:48 Internet13 joined #git
17:50 sssilver joined #git
17:51 acetakwas jhass::  Okay. Thank you
17:52 ramsub07 hello, i'm doing an interactive rebase. I want to remove a commit and all the ones before that. How do i proceed ?
17:53 ojacobson delete those lines
17:53 ramsub07 ojacobson: squash wouldn't do ?
17:53 ojacobson squash/fixup will retain the effects of those commits, but combine the commits. removing them will also remove the effects.
17:53 ojacobson so it'd depend on which outcome you want.
17:54 ramsub07 ojacobson: actually the commits have duplicated
17:54 ramsub07 ojacobson:  https://github.com/Theano/Theano/pull/4570
17:54 ramsub07 this is my case. Now, what do i do ?
17:56 ahmed_elgabri joined #git
17:56 ramsub07 ojacobson: deleting those commits won't affect the changes, right ?
17:56 Gsham joined #git
17:58 User458764 joined #git
18:02 cerkauskas joined #git
18:03 t0by joined #git
18:03 ascarter joined #git
18:05 bgamari joined #git
18:10 Dyonisos joined #git
18:12 xaviergmail joined #git
18:12 bgamari joined #git
18:15 ascarter joined #git
18:19 Ordentlig joined #git
18:20 dreiss joined #git
18:20 bgamari joined #git
18:21 jost__ joined #git
18:22 ferr joined #git
18:23 bgamari- joined #git
18:24 nilg joined #git
18:24 kfunk joined #git
18:24 King_Hual joined #git
18:24 King_Hual joined #git
18:25 zumba_addict joined #git
18:27 MrJones joined #git
18:27 MrJones does anyone know what to do when git-lfs fails to track the files properly?
18:28 MrJones "Git LFS objects to be committed" is an empty list after doing "git add ." in the repository root, despite there being tons of .jpg files in there and despite me tracking .jpg ...
18:29 MrJones well I just nuked my .git folder which solved the problem... sigh
18:29 MrJones I wish git-lfs was more stable
18:30 ascarter joined #git
18:30 mablae joined #git
18:31 hosttor joined #git
18:34 bgamari joined #git
18:38 nitin joined #git
18:40 irqq joined #git
18:41 fuchstronaut joined #git
18:43 bgamari joined #git
18:45 jaguarmagenta joined #git
18:46 bgamari- joined #git
18:47 toogley joined #git
18:48 Discovery joined #git
18:49 bgamari joined #git
18:49 cookie3 joined #git
18:50 Ordentlig joined #git
18:50 cookie3 joined #git
18:51 IlNominePatre joined #git
18:53 ascarter joined #git
18:54 bgamari- joined #git
18:55 nilg joined #git
18:56 efdetonator joined #git
18:58 ajf- joined #git
18:59 nettoweb joined #git
18:59 ajf- git clone --depth=1 repo.git does not bring branch information ?
18:59 ajf- after doing that I'd like to change to dev but it is not there
19:00 romerocesar joined #git
19:01 settermjd joined #git
19:02 vuoto joined #git
19:03 abdulmagnun joined #git
19:03 afuentes joined #git
19:03 jhass ajf-: do git clone -b dev --depth=1 repo.git instead
19:04 ajf- great thank you
19:04 abdulmagnun i have the commits A B C D E F G, with G being the latest and i want to checkout B so that the situation of all the files and folders becomes like it was at B, and then commit so i have the file situations A B C D E F G B saved. how can i do this?
19:05 jon-mac joined #git
19:05 Literphor joined #git
19:06 _nwkj86 joined #git
19:06 ojacobson git rm -r :/ && git checkout B -- .
19:06 ojacobson (all of the punctuation there is important.)
19:06 abdulmagnun ojacobson: me?
19:06 ojacobson yes
19:06 abdulmagnun oh
19:07 abdulmagnun what does rm -r :/ do?
19:07 ojacobson removes every tracked, unmodified file from the work tree and the staging area, in preparation for checkout to replace them with the files from B
19:07 Impaloo joined #git
19:07 abdulmagnun ojacobson: its quite a bit operation though, no?
19:08 abdulmagnun and then i check it out and then what?
19:08 mafrasi2_ joined #git
19:08 _ikke_ commit
19:08 ojacobson Then you make any changes you need, and commit
19:08 retropikzel joined #git
19:11 settermjd joined #git
19:13 abdulmagnun but i dont want to do any changes
19:14 _ikke_ Then just commit
19:15 satifant joined #git
19:15 yqt joined #git
19:18 DJ_Spies joined #git
19:18 m0viefreak joined #git
19:18 Ivion joined #git
19:20 Literphor joined #git
19:20 ramsub08 joined #git
19:23 dreiss joined #git
19:24 digidog joined #git
19:24 ekinmur joined #git
19:28 ejb joined #git
19:28 shinnya joined #git
19:28 Pulp joined #git
19:30 ynanm joined #git
19:31 stonerfish joined #git
19:33 chitopunk joined #git
19:33 ahmed_elgabri joined #git
19:34 blackwind_123 joined #git
19:36 dopesong joined #git
19:36 Grum joined #git
19:37 stantonk joined #git
19:42 romerocesar joined #git
19:43 dopesong joined #git
19:43 ramsub07 joined #git
19:47 mafrasi2_ Hey, I merged branch1 into master (without committing yet) and now decided to rather do the reverse (ie. merge master into branch1)
19:47 mafrasi2_ Is that possible without losing my already solved conflicts?
19:49 m0viefreak mafrasi2_: i suppose you can backup the files, reset, merge again, replace files with backup, add -u
19:50 ojacobson mafrasi2_: for the future, you may want to enable rerere
19:50 ojacobson for now, ^what m0viefreak said^ - back up the work tree, abort the merge, redo the merge, and replace the work tree with your backup
19:51 m0viefreak there's probably some low-level commit-tree command that can commit with first and second parent reversed, but the backup solution is what i would do
19:51 jhass that only records on the finished commit though iirc
19:51 ojacobson there is, don't do that tho
19:51 ojacobson I've done it (i was creating an octopus merge that included conflict resolution - mostly a stunt, just to show it was possible) and it's a huge pain
19:53 chrisshattuck joined #git
19:53 mafrasi2_ Ok, thanks, I guess that's the easiest solution. I just hoped there would be a more "elegant" way.
19:54 jj995 joined #git
19:54 diogenese joined #git
19:55 nettoweb joined #git
19:55 jj995 How do I prevent git pull from doing automatic merges?  I already have "ff = only" in the [pull] section of my .gitconfig
19:56 jj995 stdout shows "Merge made by the 'recursive' strategy"
19:56 ag4ve could also create a patch file from just before the merge, checkout branch, merge ff, lay down the patch (I think)
19:57 Literphor joined #git
19:57 nettoweb joined #git
19:58 led_belly joined #git
19:59 m0viefreak or: commit; <remember sha>; reset to before; do the merge again; checkout <remembered_sha> -- .; add
20:00 jj995 this is very confusing.  I want git pull to fast forward with no merge commits if that is possible, otherwise abort.  isn't that exactly what ff=only is supposed to do?
20:02 m0viefreak the last two messages weren't directed at you, jj995
20:02 Goplat joined #git
20:02 m0viefreak ff=only sounds like it should work according to the manpage
20:02 m0viefreak not sure what goes wrong
20:04 jj995 m0viefreak: I am using git 1.9.1 -- maybe I need to upgrade?  I can't find any doc about a minimum version for ff-only
20:04 ag4ve ah I like that method better...
20:05 mellernoia joined #git
20:06 ag4ve (and sorry, hard to hl people on a phone - I should get off my ass and resolve that, but holiday)
20:07 m0viefreak jj995: greping RelNotes says it was added in 2.0.0
20:07 jj995 m0viefreak: thanks!
20:07 m0viefreak there's also merge.ff
20:08 m0viefreak that was added way earlier
20:08 m0viefreak you could use that
20:08 m0viefreak but it will affect manual merges, too
20:14 nettoweb joined #git
20:14 ekinmur joined #git
20:16 xpk joined #git
20:17 ochorocho__ joined #git
20:18 kadoban joined #git
20:20 astrofog joined #git
20:22 bgamari joined #git
20:26 SwiftMatt joined #git
20:27 sctskw joined #git
20:28 sssilver joined #git
20:32 Literphor joined #git
20:32 sleepee joined #git
20:32 _Patrick_ joined #git
20:34 b100s joined #git
20:35 sssilver joined #git
20:36 ejb joined #git
20:39 romerocesar joined #git
20:40 ekinmur joined #git
20:42 fuchstronaut joined #git
20:44 ramsub08 joined #git
20:44 bgamari- joined #git
20:46 efdetonator joined #git
20:46 dopesong joined #git
20:48 rolha joined #git
20:48 jjido joined #git
20:49 jjido How can I change the comment on a commit?
20:50 nettoweb joined #git
20:50 jhass git commit --amend
20:50 jhass don't if it's already pushed
20:51 jjido jhass: how do I un-push then?
20:52 jjido maybe that is not possible
20:52 jhass in many cases you just shouldn't
20:52 jhass where did you push it to?
20:53 jjido To github, but there are no downloads or pulls afaict
20:54 Sucks joined #git
20:58 jjido is it git commit --amend -c hashtag?
20:59 _tgunr joined #git
20:59 Neobenedict joined #git
21:04 albel727 joined #git
21:04 ahmed_elgabri joined #git
21:07 King_Hual joined #git
21:11 Literphor joined #git
21:15 digidog joined #git
21:16 LaptopBrain joined #git
21:19 boombatower joined #git
21:20 eijk_ joined #git
21:22 Ejumn joined #git
21:22 mykespb joined #git
21:24 HoierM joined #git
21:28 tgunr joined #git
21:29 sssilver joined #git
21:32 endiruna joined #git
21:34 a3Dman joined #git
21:37 endiruna joined #git
21:39 stantonk joined #git
21:39 zumba_addict joined #git
21:40 romerocesar joined #git
21:41 Gsham joined #git
21:43 fuchstronaut joined #git
21:43 mischat joined #git
21:44 leeN joined #git
21:45 vbsales joined #git
21:48 ahr3n joined #git
21:49 Brando753-o_O_o joined #git
21:50 dsop joined #git
21:51 rkazak joined #git
21:52 JeDa joined #git
21:53 rudi_s joined #git
21:54 artemisart joined #git
21:57 GavinMagnus joined #git
21:59 JeremyM_ joined #git
22:00 GavinMagnus left #git
22:00 joki joined #git
22:02 spudowiar joined #git
22:04 Ordentlig joined #git
22:05 Vcfr joined #git
22:06 blackwind_123 joined #git
22:06 jkt joined #git
22:06 bittin joined #git
22:08 hahuang61 joined #git
22:11 GavinMagnus joined #git
22:13 Impaloo joined #git
22:16 ahmed_elgabri joined #git
22:18 des_consolado joined #git
22:18 des_consolado How to remove a directory from the git thing, but leave the directory on the local filesystem?
22:18 GavinMagnus joined #git
22:19 anuxivm joined #git
22:19 des_consolado (python env/ folder, need that to not be in the actual git, have added it accidentally D:)
22:19 jhass des_consolado: git rm --cached
22:20 ahr3n joined #git
22:24 azbarcea joined #git
22:25 WaReZ joined #git
22:25 azbarcea what is the git mailing list? I would like to propose/submit a patch
22:25 ojacobson !ml
22:25 ojacobson hmm.
22:25 ojacobson !list
22:25 gitinfo [!mailing_list] The mailing list can be reached via git@vger.kernel.org. You don't need to subscribe to the list, you will always be put in cc on reply. Read archives at http://j.mp/gitlist
22:25 des_consolado jhass: thanks dude - is it also possible to have that env/ dir changes removed from all previous commits?
22:26 azbarcea thanks!
22:26 jhass des_consolado: only be rewriting the (entire) history of our repo. Which will give people that already clone the repo a hard time
22:26 jhass !bfg
22:26 gitinfo A tool designed to remove large files, or passwords from history: https://rtyley.github.io/bfg-repo-cleaner/ (!rewrite applies)
22:26 GavinMagnus joined #git
22:27 jhass eh, of your repo of course
22:27 ailaG joined #git
22:28 azbarcea yupee ... submited my first patch on the mailing list: "[PATCH] add documentation to mw-to-git"
22:30 t-mart joined #git
22:31 azbarcea I don't know if I respect all the conventions ... how does the mentoring part works in git community?
22:33 CEnnis91 joined #git
22:40 elastix joined #git
22:41 Goplat joined #git
22:46 jaguarmagenta joined #git
22:50 efdetonator joined #git
22:51 kbs joined #git
22:53 Arban joined #git
22:54 Impaloo joined #git
22:57 Ordentlig joined #git
23:01 arescorpio joined #git
23:01 jeffreylevesque joined #git
23:02 oich joined #git
23:05 BSaboia joined #git
23:06 Literphor joined #git
23:07 jj995 left #git
23:08 oich I want to create a sub-module but only use a sub-directory of the submodule. It is structured like :src/main, where main is effectively the root of the project. Is there a way to do this other than creating a branch of the submodule that has the path moved?
23:09 ojacobson nope.
23:09 ojacobson You could fake something up using symlinks, but honestly, this sounds like a java project
23:09 ojacobson so use gradle/maven to deal with dependencies, not git
23:10 ntonjeta joined #git
23:12 Gsham joined #git
23:12 hahuang65 joined #git
23:15 johnny56_ joined #git
23:16 mehola joined #git
23:16 ahmed_elgabri joined #git
23:18 axiom_1 joined #git
23:19 dunpeal joined #git
23:19 GavinMagnus left #git
23:20 pur3eval joined #git
23:20 oich ojacobson I would like to track the revision history and change the files, so it's different from a library dependency in some ways. But maybe I can make a branch of the submodule project where revisions can be made and then treat src/main as the root of another project that is the submodule
23:21 dunpeal Hi. How do I rebase the last N commits on branch foo onto master?
23:21 pur3eval joined #git
23:21 osse dunpeal: git rebase --onto=master foo~8 foo
23:21 dunpeal Alternatively, how do I rebase all commits since commit abc123 on branch foo onto master.
23:22 dunpeal Ah, that should be:
23:22 dunpeal git rebase --onto=master abc123 foo
23:22 dunpeal right?
23:22 ffabi joined #git
23:22 osse yes
23:23 go1111111 joined #git
23:23 dunpeal thanks
23:24 johntramp joined #git
23:26 Gsham joined #git
23:28 Literphor joined #git
23:29 ffabi joined #git
23:31 rolha joined #git
23:31 mablae_ joined #git
23:35 CheckDavid joined #git
23:35 oich joined #git
23:36 ffabi joined #git
23:37 Sucks joined #git
23:41 stantonk joined #git
23:43 chrisshattuck joined #git
23:44 fuchstronaut joined #git
23:47 ISmithers joined #git
23:50 ISmithers Ello o/
23:51 azerus joined #git
23:52 User458764 joined #git
23:52 ejb joined #git
23:57 noctux joined #git
23:57 WakiMiko joined #git
23:58 EgS joined #git
23:58 yushyin joined #git
23:58 nioncode joined #git
23:59 fiete joined #git
23:59 i7c joined #git

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