Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2017-01-31

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

All times shown according to UTC.

Time Nick Message
00:00 Eugene streetwitch - sorry, without a transcript and a better idea of what you need to do(!xy) I can't help much. Also I'm wandering away for a meeting shortly, but I'm sure if you're patient somebody can help you out
00:00 gitinfo streetwitch: 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...
00:00 madewokherd joined #git
00:01 anuxivm left #git
00:02 streetwitch gitinfo, it is telling me I need to stash my changes now
00:02 streetwitch how do I do that?
00:03 streetwitch gitinfo, I need a friggen time machine
00:03 kyan gitinfo's a bot ;)
00:03 Hink joined #git
00:05 reznord left #git
00:05 diogenese joined #git
00:06 rucas__ joined #git
00:07 Vampire0 streetwitch, man git stash
00:07 gitinfo streetwitch: the git-stash manpage is available at http://jk.gs/git-stash.html
00:08 ecuanaso joined #git
00:08 Cabanossi joined #git
00:11 WaReZ joined #git
00:14 cdg joined #git
00:14 mehola joined #git
00:15 Gsham joined #git
00:16 matoro joined #git
00:17 kyan How can I escape newline/cr for .gitattributes?
00:18 dirtyroshi joined #git
00:21 WayToDoor joined #git
00:21 justanotheruser joined #git
00:24 peterpp joined #git
00:25 micah joined #git
00:25 shgysk8zer0_ joined #git
00:26 yohnnyjoe joined #git
00:27 matsaman joined #git
00:27 ShalokShalom joined #git
00:27 valar joined #git
00:27 rucas__ joined #git
00:28 njbair joined #git
00:28 wgrant joined #git
00:30 atomi joined #git
00:31 rucas__ joined #git
00:31 ISmithers joined #git
00:32 bcardarella joined #git
00:32 belak joined #git
00:32 boxrick1 joined #git
00:33 spriz joined #git
00:34 aep joined #git
00:34 jantje joined #git
00:34 dsdeiz joined #git
00:36 askb joined #git
00:38 ecuanaso joined #git
00:43 solenodic joined #git
00:43 hwrdprkns joined #git
00:44 LeBlaaanc joined #git
00:44 netj joined #git
00:44 micah left #git
00:46 Gsham joined #git
00:47 israelzuniga joined #git
00:48 jccn joined #git
00:49 telephone joined #git
00:49 telephone joined #git
00:50 spacelord_ joined #git
00:50 Ryanar joined #git
00:52 MarioBranco_2 joined #git
00:54 safe joined #git
00:56 kadoban kyan: Can you explain more the goal, exactly?
00:56 cyphase joined #git
00:58 kyan I want to be able to run "bash-gitattribute-script fileName" and have it add the file to the repository's .gitattributes file
00:58 JanC_ joined #git
00:58 kyan but I don't know how to properly escape the filename for that context
01:00 solenodic joined #git
01:03 ash_workz joined #git
01:07 qpdb joined #git
01:10 Mahpoul22 joined #git
01:11 cdg joined #git
01:13 ozmage joined #git
01:14 mizu_no_oto joined #git
01:16 salamanderrake joined #git
01:24 ecuanaso joined #git
01:25 rucas__ joined #git
01:26 ozmage_ joined #git
01:27 gugah joined #git
01:28 safe joined #git
01:31 rucas__ joined #git
01:31 krsr joined #git
01:34 WayToDoor joined #git
01:39 Bluebell_ joined #git
01:40 solenodic joined #git
01:40 pur3eval joined #git
01:49 Ken__ joined #git
01:50 hwrdprkns joined #git
01:50 stickperson joined #git
01:55 cluelessperson joined #git
01:59 Darren_ joined #git
02:01 solenodic joined #git
02:01 jlebar joined #git
02:02 jlebar joined #git
02:02 jlebar joined #git
02:05 justan0theruser joined #git
02:07 overlord_tm joined #git
02:13 thiago joined #git
02:15 ecuanaso joined #git
02:15 pks joined #git
02:15 mehola joined #git
02:17 DolphinDream joined #git
02:19 om_henners joined #git
02:22 om_henners joined #git
02:23 hwrdprkns joined #git
02:25 stickper1 joined #git
02:27 finalbeta joined #git
02:35 adymitruk is it possible to exclude git stash entries from gitk
02:35 adymitruk ?
02:37 Cabanossi joined #git
02:38 solenodic joined #git
02:40 d^sh joined #git
02:42 _28_ria joined #git
02:44 aavrug joined #git
02:46 adino joined #git
02:46 Vampire0_ joined #git
02:46 Darren_ joined #git
02:47 dreiss joined #git
02:49 WayToDoor joined #git
02:51 Nilesh_ joined #git
02:52 Dougie187 joined #git
02:53 adino joined #git
02:53 lucasem adymitruk: (it's likely that not many of us use gitk)
02:54 Limes joined #git
02:54 matsaman ...
02:54 dermoth_ joined #git
02:54 Orphis joined #git
02:54 stickper1 joined #git
02:56 _28_ria joined #git
02:57 mischat joined #git
02:59 gugah joined #git
03:07 Abbott joined #git
03:07 Abbott so whenever I `git push` to github, it asks for my username every time. Is there a way to set the repository to username@repository?
03:08 Cabanossi joined #git
03:10 lucasem Abbott: just use a personal access token or add your ssh key to github
03:13 govg joined #git
03:14 nckpz joined #git
03:14 fstd_ joined #git
03:15 Atrumx joined #git
03:17 SwiftMatt joined #git
03:17 arescorpio joined #git
03:17 Vortex34 joined #git
03:18 Radar joined #git
03:19 peeps[lappy] joined #git
03:20 andrew710 joined #git
03:20 Radar I have a commit way back on the master branch (the commit is 181e14) which contains a file at the path ./spec/app/employees/file_upload/validation_spec.rb. The latest master commit doesn't contain this file, and nor does " git log -- spec/app/employees/file_upload/validation_spec.rb" show any commit which has removed that file. How can I find out how the file was removed?
03:20 adino joined #git
03:21 valar joined #git
03:23 pur3eval joined #git
03:25 shgysk8zer0_ joined #git
03:25 rwp Radar, Hmm... I expect that to work to find that file. Have you tried "git whatchanged" and then simply searching for the file name by brute force?
03:25 batwork joined #git
03:27 Radar rwp: Me too! I've just used "git whatchanged" and the first commit there to list the file I'm looking for is that 181e14 commit.
03:27 tmcmahon joined #git
03:28 Bluebell_ left #git
03:28 rwp That is the first commit or the earliest commit? What is the latest commit?
03:28 Radar That is the first commit where that file appears. The latest commit is 46e952.
03:30 om_henners joined #git
03:30 rwp You might diff from that commit to HEAD.?? Not sure of your setup. But I need to run off! Sorry! Good luck!
03:30 Radar np, thanks for trying :)
03:33 howdoi joined #git
03:34 nopacienc3 joined #git
03:35 nopacienc3 Hi!
03:35 nopacienc3 how can i check the current point of a branch
03:35 nopacienc3 going to do a git pull, and i think i need the current commit to roll back if necessary ?
03:36 duderonomy joined #git
03:36 bernardio joined #git
03:36 nopacienc3 git rev-parse HEAD ?
03:36 Radar nopacienc3: yes
03:38 pks joined #git
03:39 LeBlaaanc joined #git
03:39 dima_ joined #git
03:41 sangy joined #git
03:43 omniwolf joined #git
03:44 omniwolf Hi, question about git pull
03:44 lucasem ask away
03:44 solenodic joined #git
03:45 svm_invictvs joined #git
03:45 omniwolf if i'm pulling a branch called "testing", and my local testing branch is different to the remote one, it does a merge, right?
03:45 lucasem is you local testing branch checking the remote testing branch?
03:45 omniwolf what does checking mean?
03:45 dreiss joined #git
03:45 omniwolf sorry, i'm pretty green with this stuff :)
03:46 phroa yes, pull is basically fetch and merge
03:46 lucasem tracking*
03:46 omniwolf yeah, they're the same branch. ok - the whole scenario.  i have a testing and beta branch.  i do a branch copy thing where i copy the contents of beta over the top of testing.  so now remote testing and remote beta are identical
03:47 omniwolf then on another computer, another budding developer has beta and testing branches locally
03:47 omniwolf but, they have the "old" testing
03:47 omniwolf if they simply do a git pull on the testing branch, it'll try and merge the old testing (local) and the new testing (remote), right?
03:48 a_thakur joined #git
03:48 cdg joined #git
03:49 Anja joined #git
03:49 a_thakur joined #git
03:54 chachasmooth joined #git
03:55 inflames joined #git
03:55 pur3eval joined #git
03:56 ozmage joined #git
03:57 jost_ joined #git
03:57 [huypn12] joined #git
03:59 mischat joined #git
03:59 sunri5e joined #git
03:59 hexagoxel joined #git
03:59 omniwolf bump
04:02 Cabanoss- joined #git
04:03 solenodic joined #git
04:04 ozmage joined #git
04:05 WayToDoor joined #git
04:06 [huypn12] joined #git
04:07 lucasem omniwolf: it would fast-forward. Not a 'merge' per-se
04:07 ozmage_ joined #git
04:07 Radar left #git
04:07 lucasem (provided the other developer made no changes to it)
04:07 Goplat joined #git
04:08 tristanp joined #git
04:10 omniwolf lucasem: ok.  what command would i use if the other developer just wanted to completely overwrite his local copy of testing with the remote version?
04:11 redhedded1 joined #git
04:11 lucasem omniwolf: he could delete his local branch `git branch -D testing`, then fetch from the remote, the `git checkout -b testing`
04:12 lucasem (that last command should automatically checkout from what was fetched from the remote)
04:12 omniwolf ok, does the checkout command do a pull as well in this case?  or will he/she need to do a git pull then too?
04:12 lucasem checkout doesn't pull or fetch
04:13 lucasem When i say "fetch from the remote", a `git pull` or a `git fetch` would work
04:13 lucasem (because `git pull` just does more than `git fetch`. It would likely change the beta branch)
04:14 omniwolf beta?? sorry what's beta got to do with this?
04:14 lucasem (I say "likely" because it depends on whether its tracking and whether there are changes on the beta branch)
04:14 lucasem using `git pull` could update every branch that is tracking with the remote.
04:14 lucasem using `git fetch` is quite harmless, though :)
04:14 hwrdprkns joined #git
04:15 omniwolf oh, i assumed the preceding "git checkout -b testing" meant that all further commands were performed on the testing branch?  or is that wrong
04:15 xall_ joined #git
04:15 ozmage joined #git
04:15 lucasem those previous commands don't depend on what branch you currently have checked out
04:15 lucasem (although, you can't actually delete a branch while you've checked it out)
04:15 xissburg joined #git
04:16 omniwolf ok.  i guess beta doesn't matter at this point.  if it's changed that's fine
04:16 omniwolf so "git branch -D testing" -> "git checkout -b testing" -> "git pull" should be all I need to make my local copy of testing the same as the server copy (and potentially pull other branches as well)
04:17 lucasem no, it'd be more like `git branch -D testing && git fetch && git checkout -b testing`
04:17 lucasem do you understand?
04:18 omniwolf mm i'm just gonna try and understand fetch 1 sec
04:18 lucasem the three commands are basically "delete branch, observe remote changes, create branch"
04:19 lucasem then the branch would be created using those changes from remote
04:19 phroa I feel like 'git reset --hard remote/branch' is a faster method
04:19 lucasem phroa:  after a fetch, though
04:20 omniwolf i see
04:20 phroa mm
04:20 [huypn12] joined #git
04:21 solenodic joined #git
04:21 omniwolf to make testing a copy of beta, i currently do "git checkout testing" -> "git reset --hard beta" -> "git push --force"
04:22 lucasem omniwolf: that would work
04:22 omniwolf then on a different computer to make the local testing branch the same as the remote testing branch, i can do "git fetch" -> "git checkout testing" -> "git reset --hard origin/testing" ??  is origin/testing the correct syntax?
04:22 lucasem omniwolf: though personally, I'd just my testing branch, recreate it while on beta, then push -f
04:22 aw1 joined #git
04:22 lucasem omniwolf: yep, that'd work!
04:23 omniwolf ok!
04:23 * lucasem is not fond of using `git reset --hard`
04:23 omniwolf what's the benefit of your way over reset --hard ?
04:23 ecuanaso joined #git
04:23 lucasem they accomplish the same thing in this case
04:24 lucasem reset --hard is just quite rash, and generally not good practice. Usually if you use it it's because you're doing something wrong
04:26 lucasem like in this case, if you keep testing to be whatever beta is, a) why even have the branch; b) why would you need to force push; c) why would you need to hard reset; ...
04:26 adino joined #git
04:27 snowkidind joined #git
04:27 lucasem `git reset origin/testing` should work if you haven't been doing anything very bad
04:30 PioneerAxon joined #git
04:30 lucasem actually, not that I think about, a merge would work perfectly fine
04:32 omniwolf hmm
04:32 omniwolf so it's probably not ideal coding practice, but this is how the lead dev has decided we'll do things
04:33 lucasem coolio
04:33 omniwolf there's a testing branch and a beta branch.  I work on testing, he works on beta.  every now and then we'll cherry pick commits from one branch to the other, and vice versa, so they stay in sync sorta
04:33 omniwolf and occasionally we will make testing identical to beta, eg overwrite testing with beta
04:33 omniwolf so we're 100% aligned
04:34 robotroll joined #git
04:34 omniwolf my above --hrad reset and --force push commands are to make the remote testing a copy of the remote beta
04:35 omniwolf but then i need to make sure that the local testing on his computer is the same the remote testing
04:35 lucasem omniwolf: normally that system would just be a single remote branch e.g. `master`, and each of you would have local branches where you make changes
04:36 lucasem (and you'd push/pull, and likely resolve conflicts as they arise)
04:36 omniwolf yeah, i tried to talk him into doing it that way, but he has some innate distrust of git, and wants to do it with separate branches.  *shrug*
04:37 ayogi joined #git
04:41 omniwolf ok, i appreciate your help, thanks!
04:42 lucasem yep! I recommend looking into rebasing :)
04:45 omniwolf ah.. maybe we could rebase instead of cherry picking
04:46 Emperor_Earth joined #git
04:47 rajkumar joined #git
04:47 Emperor_Earth joined #git
04:51 ayogi joined #git
04:51 glados1 joined #git
04:53 aartist joined #git
04:53 matoro joined #git
04:55 _28_ria joined #git
04:56 omniwolf ok, say i want to start with the beta branch, create a new branch for me to work on, then rebase my changes back on beta at some point in the future
04:57 msonntag joined #git
04:58 omniwolf git checkout beta && git branch beta-omniwolf && git checkout beta-omniwolf
04:58 omniwolf then i do coding and make some local commits
04:59 lucasem omniwolf: usually you have your own 'myfeature' branch, forked from beta at some point. When you're ready to commit, you'd pull for the beta branch, then rebase myfeature on top of the new beta branch. It'll probably work just fine, but occasionally there'd be a conflict.
04:59 omniwolf if i then do git checkout beta-omniwolf && git push
04:59 omniwolf ok, does the "git branch beta-omniwolf" command, is that forking ?
04:59 Cabanoss- joined #git
04:59 rivarun joined #git
05:00 lucasem just checkout. Like, to get started, `git pull && git checkout beta && git checkout -b
05:00 mischat joined #git
05:00 lucasem sry, that last part should be `git checkout -b mybranch`
05:01 valar joined #git
05:01 omniwolf ok.  that'll create a fork branch of beta called mybranch ?
05:01 lucasem yep!
05:01 omniwolf ok.  then i do coding, make commits, and do a git push
05:01 omniwolf that will create a new remote branch called mybranch with my commits?
05:03 lucasem it wouldn't create a remote branch
05:04 omniwolf oh ok, is that not a good idea to create a mybranch remote branch?
05:04 solenodic joined #git
05:04 mehola joined #git
05:05 lucasem you are correct. In this case, keep it local unless you want other people to work off of it
05:05 omniwolf i see.  i guess i like having it remote in case my computer dies
05:06 lucasem omniwolf: that could be the case, yeah
05:06 omniwolf if i wanted it push to create a new remote mybranch branch, how would i do that?
05:06 xall_ joined #git
05:07 lucasem git push origin mybranch
05:07 lucasem add the -u flag if you want to enable tracking
05:07 pur3eval joined #git
05:09 menip joined #git
05:09 omniwolf ok, tracking seems like a good idea
05:10 omniwolf ok, let's say i do the above (fork a branch "mybranch", make some commits, push it remotely).  then the other dev has a look at what i've done, decides he likes it, and wants it integrated into the beta branch
05:10 omniwolf i can then do: git checkout mybranch && git rebase beta
05:10 dtype also maybe consider a naming scheme on remote that is easy to follow personal branches? Not sure if it is best practice to maybe 'git push origin mybranch:myuser/mybranch', but others probably have more practical experience with what is normal.
05:11 omniwolf dtype: ok sounds reasonable!  there's only 2 people on this project so it shouldn't be too much of an issue :)
05:11 lindenle joined #git
05:12 dtype omniwolf: also don't take my word for it. I don't have lots of experience except on a couple of narrow things. Someone else would verify that this is a more standard convention
05:12 omniwolf no probs
05:12 dtype 2 people or not, always good to come up with sane conventions. :)
05:12 lucasem dtype: you're right. The best function is username/feature-or-bug-short-name
05:12 lucasem in practice, you see that often when pushing branches to the remote
05:13 omniwolf the tech lead of my project is i think 16 years old?  i'm really struggling with him.. teaching him good practice etc.  at the moment i'm focusing on coding practice, i think git can come later ;)
05:13 omniwolf so, i have done the above rebase command
05:14 omniwolf let's assume there are conflicts' as beta has had commits done to it as well
05:14 omniwolf is there where i will be told about the conflicts?
05:14 omniwolf when he does: git checkout beta && git merge mybranch
05:14 omniwolf ?
05:15 omniwolf sorry, add "or" in front of "when"
05:16 WayToDoor joined #git
05:16 lucasem the merge command would inform if there are conflicts. Though that's not how I would suggest you use your branch..
05:16 omniwolf ok, how would you do it?
05:17 lucasem yourself, ideally, you'd commit your changes locally to mybranch, fetch the latest beta branch, then `git rebase mybranch beta && git push origin mybranch:beta`
05:18 lucasem then any conflicts will come out of the `rebase` command that _you_ would deal with (and nobody else)
05:18 Lunatrius` joined #git
05:19 lucasem (and beta would be updated with your changes)
05:19 omniwolf ok, so the rebase command will surface any conflicts?
05:19 johnny56 joined #git
05:19 m4sk1n joined #git
05:21 dtype yes
05:21 dtype alternative to merge
05:21 lucasem actually you'd do `git rebase beta`, NOT ~~`git rebase mybranch beta`~~
05:21 lucasem but yes
05:22 omniwolf hmm ok, in the rebase thing i'm reading, it seems you do a rebase command, then you have to do a merge to "fast forward" the thingo
05:22 lucasem omniwolf: maybe what you're reading is talking about a different circumstance...
05:23 omniwolf https://git-scm.com/book/en/v2/Git-Branching-Rebasing
05:23 omniwolf under fig 37
05:23 dtype actually that whole git-scm book is a good end to end read. :)
05:24 lucasem omniwolf: yeah, you could either merge from the master branch then push, or `git push origin mybranch:master`
05:24 omniwolf ok.
05:24 red82 joined #git
05:25 lucasem the only difference is, you'd want to re-pull to update your master branch
05:25 lucasem after everything
05:25 omniwolf ok, i think your way could work
05:26 omniwolf i'll still need to my "mybranch" remote so the lead dev can see what i'm doing and decide if he's happy with my changes or not
05:27 omniwolf but then i can do the rebase and fix conflicts
05:27 lucasem yep!
05:27 omniwolf once i've done my rebase, then i assume i would probably just delete the local and remote mybranch branch, and then create a new fork when i'm ready to do some more diverging ?
05:28 lucasem omniwolf: well, you could keep working off of the same branch rather than re-creating it. Though normally, in practice, you'd make a new branch because it'd have a different name
05:28 ckruczek joined #git
05:28 lucasem like feature-1 is done with, move on to feature-2
05:29 omniwolf ok, i guess after rebasing, mybranch and beta will be the same
05:29 omniwolf oh hang on, no it might not be - beta might have changes that mybranch doesn't have if the other guy makes commits after i do the fork
05:29 lucasem omniwolf: only after _pushing_
05:29 omniwolf so i think re-forking is best
05:30 lucasem well his commits would ideally also update the beta branch
05:30 lucasem hence why you'd pull before rebasing
05:34 dtype abf? always be fetching?
05:34 lucasem yes. hopefully the other person is actually doing something to the remote branch, on occasion.
05:35 omniwolf hmm
05:35 omniwolf yeah agreed, but the mybranch fork wouldn't have his updates, which is why i'd be best off deleting the mybranch branch after rebasing
05:35 omniwolf right?
05:36 lucasem after pulling and rebasing, mybranch would have everyone's commits
05:36 lucasem it'd be the latest and greatest, hence why you'd `git push origin mybranch:beta`
05:37 pks joined #git
05:38 omniwolf ah i see
05:38 omniwolf i'm rebasing beta onto mybranch, not the other way around
05:38 lucasem it's the other way around
05:38 bocaneri joined #git
05:38 omniwolf so "git checkout mybranch" -> "git rebase beta"
05:39 omniwolf this will rebase mybranch onto beta
05:39 lucasem yes
05:39 omniwolf but you're saying that mybranch will also get all the latest beta commits by doing this
05:39 lucasem if you pulled before rebasing, yes
05:39 omniwolf ok, so rebasing changes both source and destination
05:39 omniwolf good to know
05:40 thiago you can do that in one go: git rebase beta mybranch
05:40 lucasem it doesn't change the destination
05:40 omniwolf ok i'm confused.  beta is the destination?
05:40 dtype beta doesn't change until you push mybranch onto it
05:41 Puffball joined #git
05:41 omniwolf i thought the rebase command would essentially replay the mybranch commits onto the beta branch
05:41 thiago it does
05:41 lucasem the result uses the mybranch ref though, not the beta ref.
05:41 valar joined #git
05:41 thiago to be clear: it modifies the mybranch branch, not beta
05:41 omniwolf oh wow
05:42 dtype then a push mybranch:beta, moves the beta pointer forward to mybranch
05:42 dtype keep in mind "branches" are just pointers
05:42 lucasem (the remote beta pointer)
05:42 lucasem (when you push mybranch:beta)
05:42 omniwolf so the rebase command will take beta in it's current form (assuming i have pulled the latest beta), then replay the mybranch commits on top of that, and put the results into the mybranch branch?
05:43 lucasem basically, yes
05:43 dtype does it from the common fork point, but yes
05:43 dtype nm what I just said
05:43 omniwolf yes this confuses me.  the common fork point.  this will be before any new commits to beta
05:43 dtype delete delete
05:43 omniwolf hehe ok
05:44 omniwolf so, then at this point it will alert me to any conflicts, at which point i can resolve them
05:44 lucasem yeah, they rebase will complain if there are conflicts.
05:44 lucasem the*
05:44 thiago it's called the merge base
05:44 thiago when you ask to rebase mybranch on beta, it needs to find the first commit in mybranch that is not in beta
05:44 thiago that's the merge-base
05:45 omniwolf then, i do "git push mybranch:beta", and it will push my fancy merged conflict free code to the beta branch on the remote server ?
05:45 lucasem yes!
05:45 dtype that's where the commits are played back from, and what I was trying to say with the fork point and said wrongly
05:45 omniwolf yep ok, but both new beta commits after the fork, and new mybranch commits after the fork are played back over the common fork point
05:46 thiago branch point and merge base are the same thing from Git's point of view
05:46 thiago from a user's point of view, they tend to think of "branch point" as the point in which they created the branch
05:46 thiago that is not kept in the history, though
05:46 thiago and every time you rebase or merge, it's like you deleted and rebranched
05:46 omniwolf mm
05:47 omniwolf i think i get it now.  this seems workable
05:47 dtype rebase is more common I think when you're trying to keep your code up with a common moving branch
05:47 dtype merge is how you take someone else's change and incorporate into yours
05:47 dtype at least conventionally
05:47 omniwolf ok, i think that's my scenario (rebase)
05:48 omniwolf say i make 3 commits to mybranch after the fork.  then the other dev looks at it, and says he doesn't want one of the commits, but the other 2 are OK
05:48 dtype there are some other mechanical implications of the differences
05:48 omniwolf can i achieve this with rebase?
05:48 lucasem generally speaking: merge is great for maintainers, and rebase is great for grunt developers.
05:48 lucasem torvalds basically said that in one of his emails
05:49 dtype there are some no-nos with rebasing, like rebasing public branches
05:49 dtype screws people up if you rebase a branch someone else is working from
05:49 lucasem yeah, hence grunt developers. They write their own code locally
05:49 dtype but that's not your workflow I think
05:49 thiago omniwolf: rebase -i can do it, yes
05:51 omniwolf ok will look into it
05:52 lucasem found the torvalds email: http://yarchive.net/comp/linux/git_rebase.html
05:54 freimatz joined #git
05:56 tirej joined #git
05:56 xall_ joined #git
05:58 _28_ria joined #git
06:00 blunaxela joined #git
06:00 solenodic joined #git
06:02 mischat joined #git
06:03 freimatz joined #git
06:06 brent__ joined #git
06:06 brent__ joined #git
06:08 Bizkit joined #git
06:08 pur3eval joined #git
06:10 Guest82 joined #git
06:11 Motoko joined #git
06:16 WayToDoor joined #git
06:18 rscata joined #git
06:21 cstrahan joined #git
06:28 dec0n joined #git
06:30 [0xAA] joined #git
06:35 freimatz3 joined #git
06:40 sdothum joined #git
06:40 Raging_Hog joined #git
06:40 overlord_tm joined #git
06:42 ngm joined #git
06:47 shgysk8zer0_ joined #git
06:48 ignacio G'night
06:48 sbulage joined #git
06:50 zeroed joined #git
06:50 zeroed joined #git
06:50 acetakwas joined #git
06:51 dreiss joined #git
06:54 chele joined #git
06:54 ThomasLocke joined #git
06:54 ThomasLocke joined #git
06:55 algun joined #git
07:03 omniwolf joined #git
07:03 qt-x joined #git
07:03 mischat joined #git
07:04 pks joined #git
07:08 seni joined #git
07:11 djb-irc joined #git
07:15 gettup joined #git
07:16 seni joined #git
07:16 Murii joined #git
07:18 zeroed joined #git
07:22 lmat joined #git
07:22 Guest6698 left #git
07:22 freimatz joined #git
07:22 ariver joined #git
07:24 matoro joined #git
07:26 noc_ joined #git
07:26 noc_ left #git
07:26 durre joined #git
07:26 ayogi joined #git
07:29 djb-irc joined #git
07:29 Lunatrius` joined #git
07:30 peterpp joined #git
07:30 Puffball joined #git
07:34 masuberu joined #git
07:38 sarri joined #git
07:38 sarri joined #git
07:40 nostrora joined #git
07:41 lb1c joined #git
07:41 ihradek joined #git
07:41 bill99 joined #git
07:45 vuoto joined #git
07:45 lucasem joined #git
07:47 JeroenT joined #git
07:47 bill99 joined #git
07:48 jknetl joined #git
07:49 zefferno joined #git
07:49 peterpp joined #git
07:50 jagob joined #git
07:53 jsolano joined #git
07:54 netj joined #git
07:56 Singmyr joined #git
08:00 aleb joined #git
08:02 Snugglebash joined #git
08:04 ronskisika joined #git
08:04 courrier joined #git
08:04 spacelord_ joined #git
08:04 mischat joined #git
08:07 aleb Hi, I'm writing a git tool which uses "git notes". I have it now displaying the info associated with the commits nicely formatted. I'd like to add a flag to enable an output which can be easily parsed by a script. How should I call the command line flag which enables this behavior?
08:10 TomyWork joined #git
08:11 aleb Hm, I just found git-status --porcelain "Give the output in an easy-to-parse format for scripts" - Why is it called "porcelain" when it enables the behaviour for plumbing?
08:12 _ikke_ It's meant to be consumed by porcelain
08:15 PHPanos joined #git
08:17 pihpah joined #git
08:19 mehola joined #git
08:19 mar77i joined #git
08:21 pR0Ps joined #git
08:23 overlord_tm joined #git
08:23 GodGinrai joined #git
08:26 Snugglebash joined #git
08:26 jknetl joined #git
08:27 rivarun joined #git
08:30 elect joined #git
08:30 theoceaniscool joined #git
08:32 [0xAA] joined #git
08:37 sea-gull joined #git
08:38 dsdeiz_ joined #git
08:38 King_Hual joined #git
08:41 rivarun joined #git
08:43 Balliad joined #git
08:43 jstein_ joined #git
08:44 Mogget joined #git
08:46 jstein Hi, I want to configure  a global mail and overwrite this for some repositories with a local one. I set [user] email=global@a.b.c in ~/.gitconfig and [user] email=local@a.b.c in .git/config
08:47 jstein but if I run git config --list | grep email
08:48 jstein I get a list with both. First the global mail, second the local one. Finally Github can not assign my contribution to the right mail.
08:48 jstein can't I overwrite the global setting with a local one?
08:53 nettoweb joined #git
08:53 Alienpruts joined #git
08:54 Mogget joined #git
08:54 lb1c38 jstein you should be able to set the global with "git config --system user.email 'foo@bar'" and the local with "git config --local user.email 'bar@baz.com'" in the repo you want to set it
08:56 lb1c38 jstein note that this obviously only applies to future commits. since the email is part of the commit metadata you cannot change the email for past commits unless you rewrite the history
08:56 navidr joined #git
08:57 jfxonq joined #git
08:58 elect_ joined #git
08:59 ecuanaso joined #git
09:01 ash_workz joined #git
09:02 redeemed joined #git
09:02 pur3eval joined #git
09:02 ocbtec joined #git
09:03 mikecmpbll joined #git
09:04 afuentes joined #git
09:04 dersand joined #git
09:05 mkoskar joined #git
09:07 xall joined #git
09:09 _ikke_ s/--system/--global
09:09 _ikke_ --system is in /etc
09:10 lb1c38 args, my bad
09:10 lb1c38 _ikke_ is right
09:11 * lb1c38 need to fill up the strategic coffee reserve
09:11 _ikke_ I just did :P
09:11 lb1c38 Hm seems like freenode disconnected me again... grml have to identify again :/
09:14 zeroed joined #git
09:14 kurkale6ka joined #git
09:14 achadwick joined #git
09:17 JeroenT joined #git
09:18 dec0n joined #git
09:19 jstein strange... I run git version 2.10.2 I can not believe there is such a stupid bug. But I can not see a mistake on my config too...
09:20 clmsy joined #git
09:20 jstein cat ~/.gitconfig | grep email returns one global email and
09:20 mda1 joined #git
09:21 yanik joined #git
09:22 AciD` left #git
09:22 Darcidride joined #git
09:22 silverdust joined #git
09:22 silverdust why git commit -am over git commit -m ?
09:22 jstein cat .git/config | grep mail has exactly one mail too
09:22 peterpp joined #git
09:23 JeroenT joined #git
09:23 _ikke_ silverdust: Some people are lazy, and want to commit everyting at once
09:23 silverdust I use the former and someone asked why that over just -m and I could only say it seems like the better way. The man doesn't describe it enough to me
09:23 _ikke_ -a mean --all
09:23 _ikke_ which means commit every change to tracked files
09:23 jstein the variables are both [user] and then email=.... is that correct
09:24 _ikke_ jstein: yeah, should be good
09:24 silverdust Oh but even without the -a it does commit every added/modified/deleted file
09:24 _ikke_ silverdust: No, it should only commit staged changes
09:25 Darcidride joined #git
09:25 silverdust So with -a I may not need to do a git add .
09:25 _ikke_ git add . would also add new files (though not stage removed files)
09:26 silverdust Oh I see. That makes it so much clearer. Thanks
09:26 marcogmonteiro joined #git
09:31 Darcidride joined #git
09:32 mar77i joined #git
09:34 ferr1 joined #git
09:34 algun_ joined #git
09:35 Darcidride joined #git
09:35 dec0n joined #git
09:36 mastro joined #git
09:37 manuelschneid3r joined #git
09:39 Aleric joined #git
09:40 Darcidride joined #git
09:42 solenodic joined #git
09:43 Lunatrius` joined #git
09:44 lb1c38 silverdust maybe you want to try git add --update (-u) also? stages every change to tracked files (not adding new stuff) but also staging deletions
09:45 fragMental joined #git
09:45 Darcidride joined #git
09:47 mischat joined #git
09:49 jsolano joined #git
09:49 mischat joined #git
09:50 mischat joined #git
09:53 theoceaniscool joined #git
09:55 chll_ joined #git
09:58 aard_ joined #git
09:58 Dumblez0r joined #git
09:58 aswen joined #git
09:58 Puffball joined #git
10:00 gajus joined #git
10:01 pusher joined #git
10:03 pusher hello, I have a bunch of local "merge remote branch" commits, every time I git pull it asks me to merge the remote branch without doing a fast forward, is there a way to fix this behaviour?
10:03 cdown joined #git
10:04 tobiasvl pusher: fix the behavior how? can the branches be fast-forwarded at all? or is the local branch ahead of the remote?
10:04 tobiasvl pusher: are you looking for !rebase?
10:04 gitinfo pusher: 'git rebase' takes away your local commits, updates your branch with new stuff from <upstream> (argument), then re-applies your local commits on top. This makes it look like your commits were created "after" the new stuff, and it can look cleaner than doing a !merge. Beware of !rewriting_public_history, though. Not to be confused with !interactive_rebase.
10:04 pusher tobiasvl: I think that given the local merge commits git thinks it is ahead of time, given the checked out repository state it could be fast forwarded
10:05 theoceaniscool joined #git
10:05 pusher tobiasvl: probably this issue emerged when I merged a different branch and not the common tracked master from which I'm pulling usually
10:05 tobiasvl well, given the local merge commits it IS ahead. git usually isn't confused about the state of the repo.
10:06 pusher yeah you are right, I don't know how to proceed without rebasing
10:06 tvw joined #git
10:06 pusher this is a deploy repo anyway that could have been messed up between merges
10:06 Darcidride joined #git
10:07 tobiasvl what's the problem though? why can't you rebase? because you can't rewrite history?
10:08 ronny joined #git
10:08 pusher tobiasvl: I would like to avoid any harmful operation given this is a production deployment, I'm not sure what rebasing would touch in this case
10:09 BlessJah whoah, why would you have commits in deploy repo?
10:09 pusher eh
10:09 Serpent7776 joined #git
10:09 pusher the deploy is a git pull
10:09 tobiasvl pusher: that's not recommended. !deploy
10:09 gitinfo pusher: Git is not a deployment tool, but you can build one around it (in simple environments) or use it as an object store(for complex ones). Here are some options/ideas to get you started: http://gitolite.com/deploy.html
10:10 pusher tobiasvl: thanks, I'm aware of the tradeoffs
10:10 Puffball joined #git
10:12 jast I'm guessing you never pushed your local merges back to upstream?
10:12 tobiasvl pusher: yeah, you see one of the tradeoffs here. local changes make for merges that might have to be manually resolved and that might break production. better to use "fetch + reset --hard $REMOTE_BRANCH" than pull. I would advise you to now somehow save the local changes to the remote, reset --hard (although not before you know the remote is fine) in production to the remote branch, and stop pulling in prod
10:13 tobiasvl forever
10:13 tobiasvl that "forever" seemed more ominous when irssi broke it to a separate line, but the emphasis is warranted ;)
10:13 cebor joined #git
10:14 digidog joined #git
10:14 pusher jast: tha'ts right, a colleague merged her own repo and now I'm dealing with this (no I'm not lying! :) )
10:14 BlessJah one shall never have to merge anything in deploy repo, because one must not to do local changes in deploy repo
10:14 jast okay, in that case there's simply no way to ever fast-forward again
10:14 pusher tobiasvl: ok thanks for the advices
10:15 pusher the thing is that probably those changes was also merged in dev and then master by me from her repo
10:15 jast fast-forward is defined as "there are changes on only the remote side", which is obviously false when you have any local-only commits
10:15 BlessJah pusher: just add the changes you've locally commited using your development checkout, push it to master and use reset
10:15 pusher the reality got lost in time
10:15 Bizkit joined #git
10:16 jast if all the same things got merged to the upstream branch, one more non-ff merge should "fix" this
10:16 jast if there's just a single commit of difference, though, it won't
10:17 Snugglebash joined #git
10:17 nickabbey joined #git
10:17 pusher I see, would cloning and restoring the .git directory on production solve the issue? given I'm sure the checked out files are the same on my local master?
10:17 lindenle joined #git
10:18 jast it should, yes. or you could do a hard reset to a commit that you believe is equivalent to what you have in production right now.
10:18 tobiasvl why not just reset in that case?
10:18 digidog joined #git
10:18 jast of course, "believing" is risky in a production env ;)
10:18 tobiasvl yeah what jast said
10:18 moritz pusher: please don't use git for deployments; read https://leanpub.com/deploy for why not, and how to do it better
10:18 solenodic joined #git
10:18 jast I'd at least make sure the two versions have the same tree ID
10:18 pusher moritz: thanks that was already given to me
10:18 moritz pusher: maybe you should start listening/reading then :-)
10:19 jast it's actually a different link
10:19 pusher moritz: I'm all ears, but I cannot move from the current position
10:19 tobiasvl jast: yeah, moritz posted a link to his self-published book
10:19 dumacdev joined #git
10:19 pusher uh
10:19 pusher sorry
10:20 tobiasvl pusher: well, you can use this opportunity to fix the environment :)
10:20 mehola joined #git
10:20 jast side note, I don't tend to buy books just because some random person mentions them to me ;P
10:20 pusher tobiasvl: trying
10:20 cdown joined #git
10:22 b100s joined #git
10:22 b100s hi2all
10:22 b100s how can i merge after revert merge ? : )
10:22 cdown_ joined #git
10:22 _ikke_ revert that revert again
10:23 b100s my problem is next: i'm on the branchA and doing 'merge banchB' got 'already up to date'; but diff says me a lot
10:24 b100s _ikke_, is it my case? i slightly remember that in past i do revert of merge and in can be cause
10:24 _ikke_ b100s: according to git, the branch is already merged (even though, the changes in the branch are reverted)
10:24 GodGinrai joined #git
10:24 b100s s/in/it/
10:27 andrew710 joined #git
10:28 b100s _ikke_, can be another solution, if i'm not wrong: merge brnahcA into branchB and after that just move pointer of A to B : )
10:29 b100s _ikke_, after revert 'revert "merge.." ' problem saves
10:30 alzagros joined #git
10:30 _ikke_ ?
10:31 b100s _ikke_, ?
10:31 _ikke_ ..
10:31 b100s : )
10:32 _ikke_ It's not clear to me what you mean
10:32 Hudu joined #git
10:32 durham joined #git
10:33 tobiasvl b100s: _ikke_'s first suggestion is the correct one, see !faulty_merge for more information on why
10:33 gitinfo b100s: If you have published a merge commit that turned out to be faulty and you'd like to get rid of it, you're probably not going to have a good time. Here's the full story: https://raw.github.com/git/git/master/Documentation/howto/revert-a-faulty-merge.txt
10:35 b100s tobiasvl, what if i change ponter of branchA to branchB ?
10:35 b100s because branchA is master and didnt change till revert of wrong merge was done
10:36 b100s s/till/since/
10:36 tobiasvl b100s: you can change the branch pointer if you want, but that will !rewrite history
10:36 gitinfo b100s: Rewriting public history is not recommended. Everyone who has pulled the old history will have to do work (and you'll have to tell them to), so it's infinitely better to just move on. If you must, you can use `git push --force-with-lease <remote> <branch>` to force (and the remote may reject that, anyway). See http://goo.gl/waqum
10:36 andrew710 joined #git
10:38 _ikke_ You can do a double merge. Merge the main branch in the reverted branch, and then fast-forward merge the reverter branch into the main branch
10:38 jstein Do you see the global and the local mail too, if you run git config --list | grep email
10:39 vktec joined #git
10:39 _ikke_ jstein: I do
10:39 _ikke_ Oh wait
10:39 _ikke_ hold on
10:40 Tobbi joined #git
10:40 vktec Is it possible to push repositories using an SSH jump host? I often use ssh -J foo bar to access servers behind firewalls, but I'm not sure how I'd do that using git
10:40 jstein perhaps I try to find a mistake while its a feature...
10:40 _ikke_ Yup
10:40 jnavila joined #git
10:40 _ikke_ git config -l | grep user.email | wc -l
10:41 _ikke_ -> 2
10:41 jstein _ikke_: OK, we both get global and local mail in the list. Then its the question, how does github know which to use?
10:41 _ikke_ note that git config is just a generic configuration editor tool
10:41 jstein ...or which is connected to my commit?
10:41 _ikke_ it does not know anyhting about the keys or values
10:42 _ikke_ git commit should pick the most specific one
10:42 _ikke_ github looks at the author and comitter fields in the commit
10:42 jstein how can I see the mail used for a commit?
10:42 _ikke_ git show <commit>
10:42 _ikke_ git show --format=fuller
10:44 jstein Ha! It lists my local mail. So why uses github my global mail?... Getting closer...
10:44 jast in case it's not clear, changing your name/e-mail address doesn't affect commits you made before that
10:45 jast so if you added a local e-mail address after committing ten times, those ten commit will still have your old address
10:45 jast i.e. the global one in this case
10:47 vktec Alternatively, can I specify an identityfile to use when pushing so I can use a port forward?
10:49 jast not directly
10:49 jast the closest is core.sshcommand
10:49 YuGiOhJCJ joined #git
10:52 jstein I changed the local email before my fetch, edit, commit, push
10:58 jast jstein: where exactly is github showing the wrong e-mail address?
10:58 jstein in my PR.
10:59 jast I believe PRs are created by Github itself, i.e. they use your Github profile data, not your local config
10:59 MineCoins joined #git
10:59 jstein my PR is assigned to an mailaddress, which is my primary github mail address and my global email in git global
11:00 marcogmonteiro joined #git
11:01 ShalokShalom joined #git
11:01 tobiasvl I believe github will always use the primary address for PRs, yes
11:02 finalbeta joined #git
11:02 jstein ok then I can search forever... Thank you
11:02 jstein I would call this a bug in github
11:02 mischat_ joined #git
11:02 tobiasvl what should github use instead? it doesn't have access to your local git settings
11:03 jstein isn't the email stored in the commit?
11:03 JeroenT joined #git
11:03 jast how are they supposed to know which commit to use to take your address from? for all they know you want to be using a different address in your PR
11:03 tobiasvl the PR isn't a commit
11:04 durham joined #git
11:04 jast it becomes one if you use the web interface to merge the PR
11:04 jast which is why you shouldn't use the web interface to merge things :}
11:05 jstein here it is... I pushed in my personal fork and made a PR to the main project
11:05 jstein so github should know
11:05 jstein however I think I will have to live with that
11:05 tobiasvl an email address is not connected to a push (or a branch for that matter), just to the commits that are pushed
11:06 jast you mean github should assume :) and other people will wish for github to assume differently. the only way to please everyone is to make it configurable.
11:06 jast and I guess it's not important enough to them to actually implement that
11:08 mar77i_ joined #git
11:14 Muzer joined #git
11:16 Snugglebash joined #git
11:16 rnsanchez joined #git
11:19 [0xAA] joined #git
11:20 matsaman joined #git
11:26 ShalokShalom_ joined #git
11:28 EY joined #git
11:31 finalbeta joined #git
11:34 DieguezZ joined #git
11:37 cdown joined #git
11:40 JeroenT joined #git
11:41 Hudu joined #git
11:43 jrm joined #git
11:47 ShapeShifter499 joined #git
11:48 nowhereman joined #git
11:48 kappter_ joined #git
11:49 diogenese joined #git
11:49 solenodic joined #git
11:49 d-fence joined #git
11:51 Puffball joined #git
11:51 Tobbi joined #git
11:52 ToBeCloud joined #git
11:53 durham joined #git
11:54 ertesx joined #git
12:00 dsdeiz joined #git
12:00 dsdeiz joined #git
12:00 bobdobbs joined #git
12:02 bobdobbs hi. I did some work on a branch. Then I commited that work. Then I switched to master. When I attempted to merge the branch into master I got this message: https://hastebin.com/eremagadig.sql
12:02 bobdobbs How can I just force a merge?
12:03 jast uh, does your commit show up in 'git log yourbranch'?
12:03 jast because the state you're in should have prevented committing, too
12:04 jast if you don't care about any details, you can throw away all uncommitted changes: git reset --hard
12:04 bobdobbs the message I create when I did my last commit shows at the top of the log, yes
12:04 bobdobbs thanks
12:05 bobdobbs oh crap. all the actual changes I did have gone though
12:05 Snugglebash joined #git
12:05 bobdobbs that's really odd. I committed them
12:06 jast you can use 'git show <commit>' to see what changes actually ended up in the commit
12:06 fuzzybear3965 joined #git
12:06 bobdobbs huh. looks like the changes I made weren't included into the last commit
12:07 bobdobbs that is weird
12:09 bobdobbs oh. the changes are still in the branch. but I can't merge the branch without a conflict being generated
12:10 bobdobbs so... how do I just pull all the changes from the branch to master?
12:10 adamru joined #git
12:10 tobiasvl pull implies a merge. a merge can have conflicts
12:10 b100s how can i get parent commit of commit?
12:10 tobiasvl bobdobbs: what do you really want to do here?
12:10 tobiasvl b100s: commit^
12:10 PHPanos joined #git
12:11 bobdobbs tobiasvl: I want to make it so that the changes I've made in the branch are moved to master, overwriting any differences
12:12 b100s tobiasvl, that says 'Stores the current contents of the index in a new commit along with a log message from the user describing the changes.'
12:13 tobiasvl b100s: I mean that to find the parent commit of the commit called COMMIT, you say COMMIT^ – for example, HEAD^ is the parent of HEAD
12:13 osse b100s: where did you read that?
12:13 leeN joined #git
12:13 b100s osse, git help commit
12:13 b100s tobiasvl, still cant get you
12:13 osse b100s: ok, then , what is "that" ?
12:13 tobiasvl bobdobbs: so you want to make master point to branch?
12:14 tobiasvl b100s: say that you have a commit called 01afb5 and you want to get that commit's parent. its parent will be 01afb5^
12:14 bobdobbs tobiasvl: I'm not entirely sure what that means, to be honest. I know that I want the work that I've been working on in the branch to become the work in master.
12:15 b100s osse, 'that' is what i got from tobiasvl message and what just said for you 'git help commit'
12:15 kurkale6ka joined #git
12:15 osse :S
12:15 tobiasvl bobdobbs: well, I don't know what that means either. what is "the work"? is it a singular commit? what do you want to overwrite?
12:16 tobiasvl b100s: to find the parent of a commit, you tack ^ at the end of the ref. I just called that ref "commit", which was maybe ill-advised
12:16 b100s tobiasvl, thanks :) 'git show deadbeaf^' shows me that commit
12:16 tobiasvl there you go
12:17 bobdobbs tobiasvl: I've been generating text files in the branch. I've made a few changes across a few text files. Master has exactly the same text files, but older versions.
12:17 tobiasvl bobdobbs: so you just want to merge the branch into master?
12:17 bobdobbs So now I want the the text files in master to reflect the changes I've created in the branch
12:17 bobdobbs yes
12:18 tobiasvl git checkout master; git merge the_branch
12:18 bobdobbs right...
12:18 tobiasvl but you said "overwriting any differences"… what do you mean by that?
12:18 b100s tobiasvl, but 'git log -p deadbeaf^' doesnt show me second commit as deadbeaf as i expect
12:19 bobdobbs tobiasvl: So in the branch Ive done 'git commit -a', and created a message and then committed. Then I did 'git checkout master'. Then I did 'git merge branch'...
12:19 bobdobbs tobiasvl: at that point I got the error message.
12:19 bobdobbs tobiasvl: by "overwriting the differences", I mean I want the old version of the files in Master to be overwritten by the newer versions that I've created in the branch.
12:21 Alienpruts joined #git
12:21 matsaman joined #git
12:21 tobiasvl OK, so no overwriting then. you just want master, when checked out, to display the contents of the files from the commit that is now in master. that is how git works and doesn't involve any overwriting (beyond the fact that git "overwrites" the working directory with HEAD at all times if there are no local conflicts)
12:21 tobiasvl but what error message? this one? https://hastebin.com/eremagadig.sql
12:22 bobdobbs yep. that's the message
12:22 tobiasvl you got that error after checking out a branch freshly and merging? that doesn't sound right
12:22 tobiasvl what does git status say?
12:22 bobdobbs output of 'git status': https://hastebin.com/ageqolamej.vbs
12:22 vktec left #git
12:23 mischat joined #git
12:24 bobdobbs I'm not sure how most of those files came to existence. I'm guessing files with BACKUP,  LOCAL, REMOTE and BASE came into existence. Maybe generated by git?
12:24 tobiasvl so I guess you did "git merge the_branch", got a message about conflicts, and then just tried to run "git merge the_branch" one more time and that's when you got the error message in https://hastebin.com/eremagadig.sql
12:24 bobdobbs at this point I can't remember
12:24 tobiasvl that's the only scenario I can think of here, that you ran "merge" twice
12:24 tobiasvl your terminal doesn't have history?
12:24 tobiasvl bobdobbs: anyway, what you need to do now is just to resolve the conflicts. !eek
12:24 gitinfo bobdobbs: [!eekaconflict] Merge conflicts are a natural part of collaboration. When facing one, *don't panic*. Read "How to resolve conflicts" in man git-merge and http://git-scm.com/book/ch3-2.html#Basic-Merge-Conflicts then carefully go through the conflicts. Picking one side verbatim is not always the right choice! A nice video explaining merge conflicts: https://www.youtube.com/watch?v=zz7NuSCH6II
12:24 bobdobbs it does. but it's kinda messy
12:24 GodGinrai joined #git
12:25 tobiasvl you just want to pick the branch's version for the css/css.map/scss files? was that what you meant earlier with "overwriting"?
12:25 des_consolado joined #git
12:25 bobdobbs I don't think I want a way to pick through all the changes. I just want to transfer all the changes wholesale into master
12:26 bobdobbs The branch is 'cannonical'. It contains  the desired state of the project.
12:27 bobdobbs and I really don't want to pick through  css sourcemaps.
12:27 tobiasvl bobdobbs: git checkout --theirs css/julieanne.css css/maps/julieanne.css.map src/sass/julieanne.scss
12:27 tobiasvl that tells git to check out "their" (ie. the branch's) version of those files
12:28 bobdobbs hmmmm
12:28 tobiasvl then ad the files and commit
12:28 tobiasvl add
12:28 bobdobbs is there a way to do that operation across all files at once?
12:29 ayogi joined #git
12:31 tobiasvl aren't those the only three with conflicts?
12:31 bobdobbs oh
12:31 bobdobbs yeah
12:31 tobiasvl I guess you can do something like this to find all files in conflicted state: grep -lr '<<<<<<<' * | xargs git checkout --theirs
12:32 osse or git ls-files --unmerged
12:32 osse but that's just me
12:34 bobdobbs conclusion. git is hard and my mental model of it is broken.
12:34 bobdobbs think I'm gonna hit up a book on it in the near future.
12:35 osse if you stop committing generated files you have less of a reason to \o/
12:35 bobdobbs truedat
12:36 mizu_no_oto joined #git
12:36 tobiasvl bobdobbs: !book
12:36 gitinfo bobdobbs: There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable
12:36 bobdobbs thanks
12:36 jagob joined #git
12:37 arnoldoree joined #git
12:37 djb-irc joined #git
12:37 tobiasvl bobdobbs: I think one small thing that can be corrected easily is to stop thinking about files, but commits. what git does with files is actually secondary. don't consider a file as being "overwritten" if you check out a different version of it from another commit/branch
12:37 tobiasvl bobdobbs: instead, consider that git tracks commits, which are sets of changes
12:38 bobdobbs I See
12:38 overyander joined #git
12:38 durham joined #git
12:39 rorro joined #git
12:40 tobiasvl (OK, git doesn't "track" commits, that was a weird word for me to use)
12:40 tobiasvl but yeah, read a book or tutorial on git
12:40 * bobdobbs nods
12:42 des_consolado left #git
12:44 Tobbi joined #git
12:48 Puffball joined #git
12:55 inflames joined #git
12:56 solenodic joined #git
12:57 troulouliou_div2 joined #git
12:58 bschindler joined #git
12:58 jsolano ~
12:59 Renter bobdobbs: the git lifestyle is honestly really good, commit often, commit small things, fetch/pull all the time
12:59 bschindler Hi - lets assume I have a branch and the latest commit is tagged (either lightweight or not). Now, I remove the branch (without merging it). Does the commit disappear at some point or does it count as referenced by git because it has a tag?
12:59 Renter for bigger wholes, make a branch
13:00 synthroid joined #git
13:00 qqx bschindler: Tags (both sorts) are references.
13:01 kappter_ joined #git
13:02 enckse joined #git
13:02 netj joined #git
13:02 bschindler qqx: thx
13:03 seni joined #git
13:03 DolphinDream joined #git
13:03 storrgie joined #git
13:03 nowhereman joined #git
13:05 kettlewell joined #git
13:07 noecc joined #git
13:08 Puffball joined #git
13:10 dcpc007 joined #git
13:11 Snugglebash joined #git
13:14 JeroenT joined #git
13:18 HardlySeen joined #git
13:18 cdg joined #git
13:20 digidog joined #git
13:20 frostythesnowman joined #git
13:20 cdg joined #git
13:22 raspado joined #git
13:23 Raging_Hog joined #git
13:24 digidog joined #git
13:26 acetakwas joined #git
13:28 sgfgdf hello, guys! why git revert cannot always perform the opposite operation if some more recent commits interfe with its modifications?
13:28 canton7 sgfgdf, let's say I made a commit which changed the line 'old' to 'new'. Then I made another commit which removed that line altogether. What should reverting the first commit do?
13:31 sgfgdf canton7: well, this scenario makes sense. my idea was if i can rely on revert to exclude a feature from already merged branch. and in this case probably not.
13:32 canton7 no. Any attempt to automatically include/exclude features is probably going to be open to failure (unless you, say, have feature gates built into your code). If all of your features modify different bits of code, as if often the case on a large mature project, you'll probably be OK
13:32 mischat_ joined #git
13:33 fuzzybear3965 joined #git
13:33 DolphinDream joined #git
13:34 Spherical joined #git
13:34 raspado joined #git
13:35 amagawdd joined #git
13:35 sgfgdf canton7: is there a good workflow which address this issue (handle removing something already merged) in a clean way?
13:35 solenodic joined #git
13:36 canton7 !faulty_merge
13:36 gitinfo If you have published a merge commit that turned out to be faulty and you'd like to get rid of it, you're probably not going to have a good time. Here's the full story: https://raw.github.com/git/git/master/Documentation/howto/revert-a-faulty-merge.txt
13:36 canton7 (reverting a merge commit is fine: reverting that revert gets painful)
13:36 Tobbi joined #git
13:36 styler2go joined #git
13:37 styler2go Hey everyone. I have a project with around 29.000 unstaged / changed files. Is there some easy way to get some overview etc.?
13:37 grawity maybe `git gui`
13:38 GodGinrai joined #git
13:39 CheckDavid joined #git
13:40 styler2go grawity, yes but which one.. just tried "GitKraken".. looks fancy but after all it's too fancy to give me an overview
13:40 Singmyr joined #git
13:41 re1 joined #git
13:41 ash_workz joined #git
13:41 rwb joined #git
13:44 nivag joined #git
13:45 Snugglebash joined #git
13:47 tinanoo1 joined #git
13:47 MattMake_ joined #git
13:49 styler2go Can i somehow purge files from a git repo in any commit etc?
13:50 jfr joined #git
13:50 whitby styler2go: https://help.github.com/articles/removing-sensitive-data-from-a-repository/
13:50 whitby I guess
13:52 nettoweb joined #git
13:52 gingitsune joined #git
13:52 gingitsune I have a some repos that are all essentially the same thing but just different versions
13:53 dave0x6d joined #git
13:53 gingitsune We want to consolidate all of the code under a single repo with versioned branches
13:54 gingitsune What I'm not sure wether it is possible to have a new repo and have a clone of the new repo that would act as a stub ofsorts
13:54 gingitsune so we don't have to relink all of our old deployments at once and we could do it at our leasure
13:54 sbulage joined #git
13:55 safe joined #git
13:55 gingitsune A simpified example we have the dir /git/v8/code,/git/v9/code
13:56 gingitsune We move thos to /git/code \w branches v8,v9
13:57 gingitsune yet we still have a /git/v8 master branch that points to /git/code
13:57 gingitsune a simlink of sorts
14:00 theoceaniscool joined #git
14:05 qqx gingitsune: Are those bare repos or with worktrees? And do you need to be able to push to (or fetch into) the old ones?
14:09 rorro joined #git
14:09 sgfgdf canton7: but as you said, it is not 100% applicable, because there is a chance that the revert can fail.
14:11 Gsham joined #git
14:11 gingitsune qqx: currently the old ones are nothing. Asking for best solution. One ought to be able to pull from the old repos and push to the new
14:12 Charliechin joined #git
14:12 endiruna joined #git
14:15 qqx gingitsune: There are a couple of ways that I'd consider to do that. First, just create separate v8 and v9 repositories, and have hooks in the main repo to update them.
14:16 qqx The other would be to setup v8 and v9 as directories with symlinks into the main repo to simulate the structure.
14:16 qqx The latter would be somewhat simpler to setup, but could cause large problems if somebody tries to update through those.
14:17 mSSM joined #git
14:18 sarri joined #git
14:18 sarri joined #git
14:19 manuelschneid3r joined #git
14:19 thebishop joined #git
14:20 umbritool joined #git
14:21 lindenle joined #git
14:21 NaStYdoG left #git
14:23 dviola joined #git
14:23 mehola joined #git
14:24 cdown joined #git
14:26 jimi_ joined #git
14:27 Murii joined #git
14:27 sammyo joined #git
14:27 Hudu joined #git
14:30 jantje can commit notes be added/modified at any time?
14:31 ShekharReddy joined #git
14:31 Random832 jantje, changing the commit message will change the hash id of the commit
14:31 Random832 and all commits after it
14:32 _ikke_ Random832: He's talking about notes
14:32 _ikke_ he / she
14:32 Derperperd joined #git
14:32 Random832 what are notes? i'd never heard of that as a thing separate from the commit message
14:32 jantje is the commit object tied to the notes ref? Jikes
14:32 _ikke_ No, it's not
14:33 _ikke_ A note references a commit, not the other way around
14:33 _ikke_ that's why you can add notes afterwards
14:33 jantje Ok. Phew, just wondering what will happen to the notes refs on clients
14:34 _ikke_ I'm not too familiar with the notes structure, but I suspect it would work similar like normal commit rewrites
14:34 _ikke_ So you can append
14:34 _ikke_ Random832: man git notes
14:34 gitinfo Random832: the git-notes manpage is available at http://jk.gs/git-notes.html
14:35 _ikke_ jantje: "Every notes change creates a new commit at the specified notes ref."
14:35 _ikke_ So even changing is append-only
14:35 jantje _ikke_: ok, so there will be no 'forced update' of the notes ref, good.
14:36 Tobbi joined #git
14:36 griffindy joined #git
14:38 xall joined #git
14:40 clmsy joined #git
14:40 Atemu joined #git
14:41 matsaman joined #git
14:42 jaafar joined #git
14:43 nowhereman joined #git
14:44 gingitsune qqx: A requirement is that the old repos have only one branch master associated with a versioned branches of the new repo
14:44 nickabbey joined #git
14:44 dhollinger joined #git
14:45 mischat joined #git
14:45 vj4 joined #git
14:46 rafalcpp joined #git
14:47 gingitsune qqx: The hooks approach sounds like it could work though
14:48 qqx gingitsune: Either of the methods that I mentioned could fill that requirement.
14:48 hashpuppy joined #git
14:49 ikelso joined #git
14:49 qqx Actually, the symlink one *could* run into problems if the main repo would start using packed refs for one of the versioned branches.
14:50 qqx But other than that, the versioned repos could have a symlink for the master ref to the versioned ref in the main repo.
14:50 qqx But, using a hook to update separate refs should be much safer.
14:50 Es0teric joined #git
14:53 neilthereildeil joined #git
14:54 neilthereildeil i try to git checkout -f master and run git status, but it still hows a lot of modified files and says "# On branch master  Your branch and 'origin/master' have diverged, and have 2 and 2 different commits each, respectively."
14:54 neilthereildeil any ideas how to get my tree in an unmodified clean state?
14:54 neilthereildeil i wanna throw away all the modified changes
14:54 solenodic joined #git
14:55 treia joined #git
14:56 dendazen joined #git
14:56 tobiasvl neilthereildeil: there are no modified changes
14:56 neilthereildeil even when i try to checkout a specific file thats listed as modified, git status still shows that file as modified
14:56 tobiasvl neilthereildeil: that message means that master and origin/master are different, not that there are local changes
14:57 neilthereildeil i have a whole list of files in red that are modified
14:57 AaronMT joined #git
14:58 tobiasvl neilthereildeil: and you want to throw it all away? git checkout -- .
14:58 umbritool left #git
14:58 tobiasvl I somehow missed that part of your original question, soirry
14:58 neilthereildeil nope, status still shows the same modified files
14:58 tobiasvl git clean -df
14:59 neilthereildeil nope
14:59 Tobbi joined #git
14:59 neilthereildeil still same modified files
14:59 tobiasvl !repro
14:59 gitinfo [!transcript] Please paste (using https://gist.github.com/ or similar) a transcript ( https://git.io/viMGr ) of your terminal session so we can see exactly what you see
14:59 endiruna joined #git
15:00 tobiasvl you can always go nuclear: git reset --hard
15:01 b100s joined #git
15:01 red82 joined #git
15:02 jeffreylevesque joined #git
15:03 dumacdev joined #git
15:03 yehowyada joined #git
15:05 eduardas_m joined #git
15:06 rivarun joined #git
15:07 ShapeShifter499 joined #git
15:07 ShapeShifter499 hi
15:08 ShapeShifter499 I'm able to "git apply" a patch, but the same patch will not "git apply -R"  any ideas why?
15:08 ShapeShifter499 https://github.com/dirty-hank/frankenclark/commit/4bd52955d5f2e6c2dd695f6eec753f00b903bdd5.patch   is the patch
15:09 marcogmonteiro joined #git
15:09 tcorneli joined #git
15:09 Dougie1871 joined #git
15:10 d10n-work joined #git
15:10 ecuanaso joined #git
15:11 ayogi joined #git
15:12 basiclaser joined #git
15:12 LeBlaaanc joined #git
15:13 ayogi joined #git
15:13 jsolano_ joined #git
15:14 Tims_Tech joined #git
15:14 Tims_Tech left #git
15:14 Tobbi joined #git
15:14 Derperperd joined #git
15:15 ayogi joined #git
15:19 JeroenT joined #git
15:19 mSSM joined #git
15:20 freimatz joined #git
15:21 JeroenT_ joined #git
15:21 borkr joined #git
15:24 shinnya joined #git
15:25 nettoweb joined #git
15:25 Eryn_1983_FL joined #git
15:25 cyphase joined #git
15:26 kpease joined #git
15:27 thebishop hi all.  when i do git diff, is it possible to only show the diffs under a particular path?
15:29 MineCoins joined #git
15:29 xall joined #git
15:29 jraccoon joined #git
15:30 Dougie1871 thebishop: I'm pretty sure you can do something like `git diff -- <path>`
15:31 seni joined #git
15:33 mikecmpbll joined #git
15:33 artisanIndia joined #git
15:35 al-damiri joined #git
15:35 kpease joined #git
15:35 mischat_ joined #git
15:36 artisanIndia hi I have a master branch and a social branch I made changes to social branch and then pushed and merged it into master branch now i wnat to un merge it
15:36 thebishop Dougie1871, cool thanks
15:36 moritz !unmerge
15:36 moritz !undo
15:36 gitinfo [!fixup] So you lost or broke something or need to otherwise find, fix, or delete commits? Look at http://sethrobertson.github.com/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history!
15:36 moritz hm, nope
15:36 moritz !brokenmerge
15:36 * moritz can't remember the right trigger
15:37 a_thakur joined #git
15:37 WayToDoor joined #git
15:38 Darren_ joined #git
15:38 pbrewczynski joined #git
15:38 osse !revert_merge
15:38 gitinfo [!faulty_merge] If you have published a merge commit that turned out to be faulty and you'd like to get rid of it, you're probably not going to have a good time. Here's the full story: https://raw.github.com/git/git/master/Documentation/howto/revert-a-faulty-merge.txt
15:38 osse artisanIndia:
15:39 artisanIndia yes osse
15:40 sborza joined #git
15:40 Es0teric joined #git
15:42 artisanIndia osse I am not good with git and I might break the repo even more doing wrong things
15:42 artisanIndia can you please still help me
15:43 m4sk1n joined #git
15:43 artisanIndia when I did git reset --hard <hash> it reverted but then I was not able to push because my master is ahead and git asks me to pull the code
15:45 Raging_Hog joined #git
15:46 e14 joined #git
15:47 seni joined #git
15:47 osse artisanIndia: You either have to do it a different way (git revert) or use push -f
15:49 crose joined #git
15:50 Tobbi joined #git
15:50 notbenh_ joined #git
15:52 notbenh_ I'm fairly sure this isn't possible to do via an alias but I just typed `git co ..` when I wanted to type `git co master` seems like a clever alias though. Just wondering if it's at all possible to some how config that?
15:53 notbenh_ trying to alias 'co ..'='checkout master' gave me a config error, not urgent but just a neat little thing that happened
15:54 valar joined #git
15:55 osse notbenh_: how about alias com ?
15:55 qqx notbenh_: If the value of an alias starts with ! the remainder is interpretted by the shell. You could have that check if the next argument is .. and then checkout master, otherwise do whatever.
15:56 notbenh_ osse not a bad suggestion, qqx ohh neat!
15:56 vuoto joined #git
15:57 Derperperd joined #git
15:57 darkmist left #git
15:58 osse notbenh_: btw, git already has - builtin as a shortcut for checking out the previous branch (whatever that may be)
15:58 osse sort of like 'cd -'
15:58 Charliechin joined #git
16:02 Derperperd joined #git
16:02 jraccoon joined #git
16:05 Raging_Hog joined #git
16:08 crose joined #git
16:09 synthroid joined #git
16:13 a_thakur joined #git
16:14 jabberwock Is this an appropriate channel to ask about the github API?
16:14 Snugglebash joined #git
16:14 barduck joined #git
16:15 gingitsune joined #git
16:15 dendazen joined #git
16:15 canton7 jabberwock, #github's probably better
16:15 jabberwock thanks
16:16 Noldorin joined #git
16:16 PHPanos joined #git
16:19 peeps[lappy] joined #git
16:19 neilthereildeil tobiasvl: i think the issue was that the modified files was due to permissions
16:19 jstimm joined #git
16:20 brent__ joined #git
16:21 e14 joined #git
16:21 gugah joined #git
16:24 mehola joined #git
16:25 adamru joined #git
16:28 nowhere_man joined #git
16:28 jccn joined #git
16:28 ecuanaso joined #git
16:29 Charliechin joined #git
16:30 ExeciN joined #git
16:32 Snugglebash joined #git
16:32 chardan joined #git
16:33 jstimm joined #git
16:34 mischat joined #git
16:35 nickabbey joined #git
16:36 govg joined #git
16:36 Snugglebash joined #git
16:37 cdown joined #git
16:39 emg joined #git
16:39 marcogmonteiro joined #git
16:39 rgrinberg joined #git
16:41 frostythesnowman joined #git
16:41 mikecmpbll joined #git
16:41 Keytap joined #git
16:41 TomyLobo joined #git
16:43 Keytap Morning gents.  Quick questions for you if you don't mind.  Is it a problem in any way to commit a file with literal escape characters in it, beyond the obvious issue of someone else's editor being unable to interpret it correctly?
16:43 Dougie1871 left #git
16:43 _ikke_ git should not have problems with it, though, it might identify it as a binary file and not do diffs and merges
16:43 Eugene git doesn't (really) care. The file might get flagged as binary.... that ^
16:44 xall joined #git
16:44 Keytap Whoopse, I mean control characters, but yeah, you got it.  And I figured that was the case.  Just wanted to be sure.
16:44 Keytap Appreciate the response.
16:45 nowhere_man joined #git
16:46 OMSQ joined #git
16:47 e14 joined #git
16:49 neilthereildeil left #git
16:50 matoro joined #git
16:52 LionsMane joined #git
16:52 jraccoon joined #git
16:55 jstimm joined #git
16:55 MillerBoss joined #git
16:56 xall joined #git
16:58 Sound joined #git
16:59 azerus joined #git
17:00 netj joined #git
17:01 frostythesnowman joined #git
17:02 gabi_ joined #git
17:04 shubhnik joined #git
17:04 azerus joined #git
17:05 thiago joined #git
17:07 tang^ joined #git
17:07 canuk joined #git
17:07 matsaman joined #git
17:07 diegoaguilar joined #git
17:08 nowhere_man joined #git
17:08 diegoaguilar I tried a git stash apply which ended up in a merge conflict (merging dev into my branch), how can I make that file look just like dev does and avoid conflicts
17:09 Eugene `git checkout <branch> -- <fle>`
17:09 Eugene (This will eat uncommited/unstashed changes)
17:09 diegoaguilar Eugene, can I do that while I should be fixing the merge conflict?
17:10 mischat joined #git
17:10 Eugene Yup, it'll automagically add it to the index for you, too
17:10 Eugene "pick one side" is a totally valid merge-resolution strategy
17:10 Ryanar joined #git
17:11 a_thakur joined #git
17:12 sathed joined #git
17:13 ispn joined #git
17:15 jstimm joined #git
17:16 eb0t joined #git
17:17 redhedded1 joined #git
17:18 g105b joined #git
17:21 tcorneli joined #git
17:23 matsaman joined #git
17:25 mischat_ joined #git
17:25 [0xAA] joined #git
17:26 Charliechin joined #git
17:26 tvw joined #git
17:27 Midas joined #git
17:28 gingitsune joined #git
17:29 shgysk8zer0_ joined #git
17:34 bill99 joined #git
17:34 cdown joined #git
17:35 madewokherd joined #git
17:38 spacelord_ joined #git
17:38 Tobbi joined #git
17:39 dreiss joined #git
17:41 nickabbey joined #git
17:41 rivarun joined #git
17:42 crater2150 joined #git
17:44 asifm joined #git
17:45 Doginal joined #git
17:46 robotroll joined #git
17:47 asifm joined #git
17:48 matsaman joined #git
17:48 Darren_ joined #git
17:49 Charliechin joined #git
17:49 spacelord_ joined #git
17:50 rorro joined #git
17:50 emg joined #git
17:50 lindenle joined #git
17:51 wget Hello everyone. I'm trying to add a submodule. I have made that a few times already. But adding a submodule is not something I often do, while I do work every day with projects that have submodules in them, without any issue.
17:51 wget The process I have is as this: git submodule add https://URL ./myFolder/subFolder; git submodule init; git submodule update
17:51 wget After these steps, do you know why ./myFolder/subFolder is still empty?
17:52 peepsalot joined #git
17:54 tyreld joined #git
17:57 eblip joined #git
17:57 svm_invictvs joined #git
17:59 xall joined #git
18:00 b100s left #git
18:00 nickabbey joined #git
18:02 Mikerhinos joined #git
18:03 crayon joined #git
18:05 ataldis joined #git
18:05 dreiss joined #git
18:06 ataldis left #git
18:06 freimatz joined #git
18:06 glowdemon1 joined #git
18:06 jraccoon joined #git
18:06 glowdemon1 I'm trying to move all files in a certain directory to a current directory with Git, but it gives me a "Bad source" error - "git mv www/* ."
18:09 hahuang61 joined #git
18:10 dumacdev joined #git
18:12 rivarun joined #git
18:13 marcogmonteiro joined #git
18:15 valar joined #git
18:17 kevwil joined #git
18:19 osse glowdemon1: try git mv 'www/*' .
18:19 glowdemon1 Just found out that windows shell didn't support the wildcard, or something
18:19 glowdemon1 Worked with Git bash
18:19 osse glowdemon1: the git cmd thing is evil
18:20 mathematic-alpha joined #git
18:20 osse not really, but it's pretty bad
18:20 glowdemon1 Do you mean git bash or git mv?
18:20 amagawdd joined #git
18:20 tang^ running git mv in git bash
18:20 [huypn12] joined #git
18:20 glowdemon1 Whats the bad part? :c
18:21 tang^ oh, osse said git in CMD.exe is not fun
18:21 tang^ powershell + posh-git is better
18:21 [huypn12] joined #git
18:21 glowdemon1 Ah
18:21 glowdemon1 Its not
18:21 glowdemon1 I mean, it is not fun
18:22 mathematic-alpha left #git
18:22 adino joined #git
18:22 glowdemon1 I generally use cmder, if that's the same :c
18:22 tang^ cmder with powershell + posh-git is what I was using a lot
18:23 tang^ cmder with git bash also works well
18:23 glowdemon1 Yup
18:23 brent___ joined #git
18:23 jstimm joined #git
18:24 ecuanaso joined #git
18:25 mehola joined #git
18:26 sea-gull joined #git
18:28 cgdub joined #git
18:29 kevwil_ joined #git
18:31 Oatmeal joined #git
18:32 svm_invictvs joined #git
18:33 cdg joined #git
18:34 ataldis joined #git
18:34 Gsham joined #git
18:34 ataldis left #git
18:36 aw1 joined #git
18:37 e14 joined #git
18:37 publio joined #git
18:37 Ryanar joined #git
18:39 smithbone joined #git
18:44 yehowyada joined #git
18:45 tvw joined #git
18:46 Singmyr joined #git
18:47 Es0teric joined #git
18:51 tvw joined #git
18:53 valar joined #git
18:54 Charliechin joined #git
18:57 wilbert joined #git
18:57 Charliec_ joined #git
18:59 jstimm joined #git
18:59 jccn joined #git
19:01 cdg_ joined #git
19:02 netj joined #git
19:04 Dumblez0r joined #git
19:04 jimi_ joined #git
19:07 Murii joined #git
19:08 sborza joined #git
19:08 Es0teric joined #git
19:09 yehowyada joined #git
19:11 Es0teric joined #git
19:12 jnewt joined #git
19:13 m0viefreak joined #git
19:13 anjen joined #git
19:17 e14 joined #git
19:18 gargola joined #git
19:19 Sasazuka joined #git
19:20 Endarked joined #git
19:20 jccn joined #git
19:25 jstimm joined #git
19:25 Gsham joined #git
19:25 Thesisus joined #git
19:27 treia joined #git
19:29 Hudu joined #git
19:29 grayjoc joined #git
19:29 tvw joined #git
19:33 Guest55 joined #git
19:34 valize joined #git
19:38 kulelu88 joined #git
19:39 adamru joined #git
19:41 nickabbey joined #git
19:47 Singmyr joined #git
19:49 jstimm joined #git
19:54 Dougie187 joined #git
19:54 ResidentBiscuit joined #git
19:54 jnavila joined #git
19:56 cdg joined #git
19:56 cdown joined #git
19:58 matoro joined #git
19:58 matsaman joined #git
20:00 Charliechin joined #git
20:02 cdg_ joined #git
20:03 zeroed joined #git
20:03 zeroed joined #git
20:03 spacelord_ joined #git
20:07 Ragnor joined #git
20:07 wilbert joined #git
20:08 noecc left #git
20:10 bill99 joined #git
20:11 matsaman joined #git
20:12 kevwil joined #git
20:12 doener joined #git
20:15 Guest24384 joined #git
20:16 ertes joined #git
20:17 finalbeta joined #git
20:17 senaps joined #git
20:18 senaps joined #git
20:18 DrSlony joined #git
20:19 duderonomy joined #git
20:19 boombatower joined #git
20:20 DrSlony Hello, I just did a "git revert" and git told me to --continue after fixing conflicts, but I don't see any message about what the conflicts are. Is there a command like "git mergetool" got sorting out merge conflicts, but for revert?
20:20 nettoweb joined #git
20:20 thierryp joined #git
20:20 DrSlony https://i.imgur.com/MPXBa7j.png
20:21 ResidentBiscuit `git status`
20:22 Balliad joined #git
20:22 DrSlony thanks ResidentBiscuit
20:22 isysd joined #git
20:23 ResidentBiscuit When in doubt, just check the status
20:23 brent__ joined #git
20:23 aw1 joined #git
20:24 Charliechin joined #git
20:24 jnavila joined #git
20:24 Charliechin joined #git
20:25 ecuanaso joined #git
20:26 nickabbey joined #git
20:27 aidalgol joined #git
20:27 duderono_ joined #git
20:30 re1 joined #git
20:34 jccn joined #git
20:37 Snugglebash joined #git
20:40 Max___ joined #git
20:40 Balliad joined #git
20:42 jstimm joined #git
20:43 askb joined #git
20:45 valar joined #git
20:46 lucasem joined #git
20:48 __builtin joined #git
20:50 yehowyada joined #git
20:50 nowhere_man joined #git
20:51 kettlewell joined #git
20:52 apotry joined #git
20:53 e14 joined #git
20:55 dan2k3k4 ok maybe this is more of a bash/pipe question but I basically want to do in one line: git diff "git rev-list --max-parents=0 HEAD | tail -n 1"..HEAD
20:55 dan2k3k4 as in compare first initial commit with current head... [for no reason, just for fun]
20:56 Guest55 joined #git
20:56 cdg joined #git
20:57 durham joined #git
20:57 pbrewczynski joined #git
20:58 jeffreylevesque joined #git
21:00 matsaman dan2k3k4: git diff $(git rev-list...)
21:00 matsaman more at #bash
21:01 Macaveli joined #git
21:02 Flaghacker joined #git
21:02 Flaghacker I just cloned a repository, but I don't see all of the branches when I run "git branch". Is there a way to pull all of the remote branches?
21:03 philipsd6 joined #git
21:03 m0viefreak git branch only shows local branches
21:03 m0viefreak use git branch -a
21:03 m0viefreak man git-branch
21:03 gitinfo the git-branch manpage is available at http://jk.gs/git-branch.html
21:05 _ikke_ or git branch -r to only show remote branches
21:05 hahuang61 joined #git
21:06 Flaghacker And why is not every remote branch also a local one? What's the difference in practice?
21:07 _ikke_ Flaghacker: git by default just creates on local branch when cloning
21:07 Ryanar joined #git
21:07 _ikke_ You can easily create the others (note that you already have all data)
21:08 Flaghacker Okay and one more thing, in git branch -a there is a line "remotes/origin/HEAD -> origin/master", what does that mean?
21:08 reynierpm left #git
21:08 cdg_ joined #git
21:10 _ikke_ Every repository has a symbolic ref called HEAD
21:10 dan2k3k4 thanks matsaman
21:10 _ikke_ When cloning you also get this symbolic ref from the remote. git uses it to determine the default branch to checkout when cloning
21:10 matsaman dan2k3k4: awww yeah
21:11 dsdeiz joined #git
21:11 dsdeiz joined #git
21:12 Snugglebash joined #git
21:12 durham joined #git
21:13 cdg joined #git
21:13 hashpuppy joined #git
21:13 Es0teric joined #git
21:14 wilbert joined #git
21:16 Balliad joined #git
21:16 Flaghacker _ikke_, So it's not something bad? I can just leave it as-is?
21:17 _ikke_ Flaghacker: You can just leave it
21:17 Flaghacker Okay thanks for all of the help, much appreciated!
21:17 durham joined #git
21:18 lucasem left #git
21:20 om_henners joined #git
21:22 mdk left #git
21:22 Limes joined #git
21:22 durham joined #git
21:24 clearcut22 joined #git
21:25 Guest85 joined #git
21:25 Flaghacker _ikke_, When I now try to merge two branches I get "not something we can merge", and it says I should put origin/ in front. Is there a way to really get a branch locally?
21:25 dtype if you git re-fetch a branch that you've not touched, but the origin has been updated, do you still have to marge to bring your local branch up to the remote one, if you have a tracking branch?
21:26 rivarun joined #git
21:26 dtype so a git --fetch all doesn't change my local branches at all (even if untouched), I'd need to pull/merge them to advance local pointers?
21:26 m0viefreak dtype: yes
21:26 clearcut22 left #git
21:26 dtype m0viefreak: thanks
21:27 m0viefreak doing a 'git status' on one of those branches, will tell you that you are "behind ... x commits"
21:27 dtype ok, makes sense. thanks.
21:28 Gsham joined #git
21:29 adino joined #git
21:29 Es0teric joined #git
21:30 shgysk8zer0_ joined #git
21:30 jagob joined #git
21:30 umpc joined #git
21:32 valar joined #git
21:33 durham joined #git
21:34 ecuanaso joined #git
21:35 Flaghacker Should I delete old development branches that are no longer useful? Eg. they were merged to master and they will never be used again.
21:37 OMGOMG joined #git
21:37 streetwitch joined #git
21:37 _ikke_ Flaghacker: That's you're own choice, but that's what I do
21:38 matsaman same
21:38 durham joined #git
21:38 streetwitch I tried a "git push test master"  and got 'everything up to date', but I don't see any changes.  Anyone have a minute to help someone with this?
21:39 Flaghacker And do you guys delete it locally only or also on the remote?
21:40 str joined #git
21:40 matsaman I delete locals with almost no hesitation
21:40 matsaman remote might be nice for posterity
21:40 gigq joined #git
21:40 matsaman since it's not going to be secret to you alone already
21:41 e14 joined #git
21:42 _ikke_ I usually record in history that what branch was merged, so I don't see any problem personally to even remove it remotely
21:44 matsaman you definitely don't want git branch output to just grow forever
21:45 _ikke_ right
21:45 cr34ton joined #git
21:46 polyrob_ joined #git
21:47 e14 joined #git
21:48 durham joined #git
21:50 adino joined #git
21:50 Electrometro joined #git
21:51 durham joined #git
21:52 glowdemon1 joined #git
21:52 glowdemon1 Hi. How would I use "git push heroku heroku:master" with git subtree? I tried "git subtree push --prefix www heroku heroku:master" but it gives me an error
21:53 solenodic joined #git
21:53 _ikke_ streetwitch: If git says it's uptodate, it probably is (for what you are pushing)
21:54 _ikke_ streetwitch: Where don't you see any changes
21:54 _ikke_ ?
21:55 mikecmpbll joined #git
21:56 durham joined #git
21:58 BlessJah belak: feature request: ssh git@bitbucket.org [--] [list|create [--public] <name>|destroy <name>]
21:59 cdown joined #git
22:00 sobczyk_ joined #git
22:00 jccn joined #git
22:00 Ragnor joined #git
22:00 emg joined #git
22:00 AKPWD joined #git
22:00 BlessJah belak: should I create ticket or sth, or will you forward the idea?
22:01 dermoth_ joined #git
22:01 cdg joined #git
22:02 Blkt joined #git
22:02 yehowyada joined #git
22:03 e14 joined #git
22:03 belak BlessJah: That's something we've considered... but not very seriously because of how few people would end up using it... you're welcome to file an issue, but I unfortunately doubt it would happen any time soon
22:04 hahuang61 joined #git
22:05 BlessJah is there public issue tracker?
22:05 LaptopBrain joined #git
22:05 SwiftMatt joined #git
22:05 jedahan joined #git
22:07 belak https://bitbucket.org/site/master
22:08 DrSlony Hello. We have two stable branches: "master" with gtk2 code and "gtk3" with gtk3 code. We are killig off gtk2. We want to adopt something like this http://nvie.com/posts/a-successful-git-branching-model/ and are wondering what to do about master. I'm thinking of branching "master" to a new branch called "gtk2" then deleting master. Would it be safe/wise to then create a new branch called "master" for the tags?
22:08 cdg joined #git
22:09 BlessJah heh, never seen that one oO
22:11 Endarked joined #git
22:12 belak DrSlony: a tag just points to a commit, so the branch that commit came from doesn't matter
22:12 belak You can rename master to gtk2 and it won't break tags
22:13 DrSlony my worry isnt about tags
22:14 DrSlony its about having a new branch called master which should have notihng to do with the old branch called master
22:14 DrSlony i dont know what side effects to expect
22:15 e14 joined #git
22:16 anuxivm joined #git
22:16 DrSlony there is this http://stackoverflow.com/a/3790682
22:16 anjen joined #git
22:16 invisbl joined #git
22:18 tristanp joined #git
22:19 moei joined #git
22:20 jstimm joined #git
22:21 fatalhalt joined #git
22:22 azerus joined #git
22:22 BlessJah belak: 13792
22:23 DrSlony i'll ask again tomorrow, good night
22:23 amrx joined #git
22:24 SwiftMatt joined #git
22:25 Charliechin joined #git
22:26 spacelord_ joined #git
22:26 mehola joined #git
22:28 SwiftMatt joined #git
22:31 e14 joined #git
22:32 netj joined #git
22:34 penanze joined #git
22:34 pR0Ps joined #git
22:36 ikelso joined #git
22:36 Gsham joined #git
22:40 matoro joined #git
22:41 Dougie187 left #git
22:41 jedahan joined #git
22:42 tang^ DrSlony: yes you can do what you described
22:42 tang^ folks with master already checked out may have to retrack their branches
22:43 e14 joined #git
22:43 jedahan joined #git
22:45 ab-rex joined #git
22:46 Charliechin joined #git
22:46 nowhere_man joined #git
22:47 Charliec_ joined #git
22:48 rwb joined #git
22:48 Atm0spher1c joined #git
22:49 BlaXpirit joined #git
22:49 sammyo joined #git
22:51 dumacdev joined #git
22:53 seni joined #git
22:55 cjohnson DrSlony: it's effectively the same as any other history rewrite
22:56 dirtyroshi joined #git
22:57 PHPanos joined #git
22:58 fission6 joined #git
22:58 fission6 i have a remote branch A and a local branch A, the remote is all wrong, how do i delete it and replace with my local
22:59 cjohnson force push
22:59 cjohnson anybody else who has A will have troubels when they pull though
23:00 fission6 i can't delete the remote branch
23:00 Karazhan joined #git
23:00 netj joined #git
23:00 fission6 basically the code in it is all wrong, my local is what it should be, and i don't want to pull it locally
23:01 fmeerkoetter joined #git
23:02 cjohnson right, force push
23:02 cjohnson it will overwrite the remote with whatever you have locally
23:04 thiago joined #git
23:04 solenodic joined #git
23:05 dave0x6d joined #git
23:06 adino joined #git
23:06 kyan joined #git
23:06 fission6 k
23:08 isysd joined #git
23:08 ShalokShalom_ joined #git
23:13 seni joined #git
23:14 WaReZ joined #git
23:15 amrx joined #git
23:17 Ryanar joined #git
23:20 jedahan joined #git
23:22 cdown joined #git
23:24 dreaddymck joined #git
23:25 p4trix joined #git
23:25 dendazen joined #git
23:26 Guest55 joined #git
23:29 dsdeiz joined #git
23:32 DolphinDream joined #git
23:32 sammyo joined #git
23:32 seni joined #git
23:32 crayon joined #git
23:34 Vinnie_win joined #git
23:36 solenodic joined #git
23:36 LeBlaaanc joined #git
23:37 krillr joined #git
23:39 dsdeiz_ joined #git
23:40 matoro joined #git
23:41 anuxivm left #git
23:42 prg3 joined #git
23:43 rivarun joined #git
23:45 lucasem joined #git
23:48 madewokherd joined #git
23:48 MrKira joined #git
23:49 mutex joined #git
23:52 diogenese joined #git
23:53 DieguezZ joined #git
23:53 redhedded1 joined #git
23:59 Kaisyu joined #git
23:59 spacelord_ joined #git
23:59 AlexMax joined #git
23:59 AlexMax Okay
23:59 AlexMax So I have two branches, A and B

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