Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2017-05-03

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

All times shown according to UTC.

Time Nick Message
00:00 bilal joined #git
00:03 m1dnight_ joined #git
00:04 m1dnight1 joined #git
00:09 lil__ joined #git
00:09 lil__ Hi there, is this an alright spot to ask general questions about git use? I'm guessing so...
00:10 Noodlewitt joined #git
00:12 thiago you might choke if it's airtight...
00:12 thiago I'd prefer a more breezy place, lots of air
00:13 lil__ I'm stuck in the middle of a revert haha. So I have a previous commit I want to return to see if a bug was still in existence, but I get a warning to fix conflicts before continuing. How do I see what needs to be fixed?
00:14 anuxivm left #git
00:14 lil__ I think my problem is that I didn't run checkout with the SHA before git revert SHA
00:14 thiago you have to conclude or abort that revert first
00:14 Noodlewitt joined #git
00:15 jast you're aware that all 'git revert' does is create a new commit that reverses the changes from the commit you specify?
00:15 jast but anyway, to start over, 'git revert --abort'
00:15 cagomez joined #git
00:16 lil__ thiago, yes I think that's what I want. In git agnostic terms I want briefly explore the code pre changes I made today. Is that a revert case?
00:16 bannakaffalatta joined #git
00:16 cagomez joined #git
00:16 jast no, you can just checkout an old commit and then checkout your branch when you're done
00:16 dysfigured if i want to run a build command when after a remote client pushes to my repo, which git hook would i use? post-receive?
00:16 lil__ thiago, ok, so abort revert, and do checkout instead
00:16 lil__ ill give it a whirl
00:16 Noodlewitt joined #git
00:16 rocx joined #git
00:16 jast dysfigured: yes, or post-update. all that's different between the two is the semantics.
00:17 dysfigured thanks
00:17 cagomez joined #git
00:17 Cabanossi joined #git
00:19 TheSimonator joined #git
00:19 hotpancakes joined #git
00:21 Noodlewitt joined #git
00:22 cdown_ joined #git
00:25 cdown_ joined #git
00:26 caper joined #git
00:27 Noodlewitt joined #git
00:27 tocky joined #git
00:28 cdown joined #git
00:29 lil__ thiago, so I have now figured out that the issue wasn't with the code, but with how debug messages were being parsed by a separate program. Upon trying to return to the most recent work, I get a message "local changes about to be overwritten" is reset the best bet for this?
00:30 lil__ jast, haha I just looked back and saw that half of the replies were yours, sorry for not noticing earlier
00:30 jast if you're sure you don't need your uncommitted changes, you can use git checkout with -f to destroy them. reset will do it, too, but you'll still have to checkout as a second step
00:30 fstd_ joined #git
00:30 jast don't worry, I'll live :)
00:30 telephone joined #git
00:30 telephone joined #git
00:31 lil__ jast, cool, thanks for helping out! I have a particular cautious nature with the -f just because I at one point somehow completely reset my local git repo, and lost almost a semester of work.
00:31 jast whoops
00:32 lil__ yea, impulsive use of --hard somewhere
00:32 jast for future reference, everything you've committed can be restored if you don't wait too long
00:32 jast !lost
00:32 gitinfo As long as you have `git commit`ed your changes (or even `git add`ed them), your changes will not be lost for over two weeks unless you work really hard at it.  There are two places where "lost" changes can be hiding.  They might be in the reflog (`git log -g`) or they might be in lost&found (`git fsck --unreachable`). Type "!dangling" and "!reflog" into IRC for more info.
00:33 cdown joined #git
00:33 lil__ jast, seriously? I should have come on here immediately...
00:33 jast yeah, too much to hope for, isn't it? but it's true.
00:34 Noodlewitt joined #git
00:34 lil__ jast, ok, back to the issue at hand, I ran git reset, and then the checkout command, but got the same message...
00:35 cdown joined #git
00:35 jast git reset trashes only the index, but uncommitted changes in the working tree (physical files) will still stick around
00:35 jast those are what's in the way
00:35 lil__ right
00:35 jast either checkout -f or reset --hard will remove them. you can 'git diff HEAD' if you want to review them to make sure you don't need them
00:35 jast it's a good habit to build, I guess... :)
00:35 lil__ ok
00:37 lil__ I'm getting info from stuff a ways back that I didn't add. However these files aren't the ones popping up in the "Unstaged local changes "list
00:38 jast I *think* when switching branches it lists only the files that are clashing with what checkout needs to do
00:38 jast do these files from "Unstaged local changes" show up in 'git status', too?
00:39 jast oh, and did you abort the revert earlier?
00:41 lil__ yea aborted. And yes they showed up in the status check too. I have just done a git add . and commit with a message about housekeeping, but upon checkout of the previous newest changes branch I get a "leaving housekeeping commit behind" related error
00:42 lil__ not err per say, but warning
00:42 brent__ joined #git
00:42 druonysus joined #git
00:42 jast yeah, that's to be expected. when you checkout an old commit directly, you're no longer fully on the branch. in Git terms, HEAD (the reference to the current branch) is "detached", pointing directly at the commit instead of a branch.
00:42 jast so, you committed into the void
00:43 jast you can get your commit back by cherry-picking it (the commit ID is in the output from 'git commit'), though you may have to resolve some conflicts to make it work
00:43 dermoth joined #git
00:43 tnecniv joined #git
00:45 Atrumx joined #git
00:46 zivester joined #git
00:47 lil__ jast, gotcha. Well I haven't lost any work on this one, but I definitely could have been keeping better track of my commits, and I think taking care of this bug could have gotten easier... Thanks for your help though!
00:48 jast you're welcome
00:50 Noodlewitt joined #git
00:56 Noodlewitt joined #git
00:56 cagedwisdom joined #git
00:58 serialoverflow joined #git
00:59 telephone joined #git
00:59 telephone joined #git
01:02 ojdo joined #git
01:05 xpika joined #git
01:07 xpika Is there any way to perform a merger from a branch without influencing any of the current branches ancestory?
01:07 ehmk joined #git
01:07 svm_invictvs joined #git
01:08 Vampire0 xpika, what do you mean by "not influencing the ancestory"?
01:09 xpika I tried running git merge --no-commit but when I finally ran a commit the git log said that the current branch was merged from the other branch
01:10 xpika Vampire0: ping
01:10 MrWoohoo joined #git
01:11 jast so basically you want it to look like a normal commit?
01:11 TbobbyZ joined #git
01:11 xpika jast: yes
01:11 jast merge --squash
01:12 xpika jast: crosses fingers
01:12 durham_ joined #git
01:12 d^sh joined #git
01:12 caper joined #git
01:14 xpika jast: thanks
01:14 dreiss joined #git
01:15 awc737 joined #git
01:15 xpika jast: that works
01:15 matt|home hello
01:16 xpika didn't know squash also squashed branch ancestory
01:16 PresidentBiscuit joined #git
01:16 howitdo joined #git
01:16 jast it basically glosses over the history. the result is identical except the resulting commit has only one parent instead of multiple
01:17 jast and that's the only difference
01:17 Remram joined #git
01:18 cqi joined #git
01:21 owlbynight joined #git
01:21 caper joined #git
01:22 NeXTSUN joined #git
01:24 taoqy joined #git
01:24 dandaman joined #git
01:25 Wulf4 joined #git
01:27 sea-gull_ joined #git
01:27 sea-gull joined #git
01:33 tnecniv joined #git
01:33 otiose joined #git
01:33 reverse_light joined #git
01:35 njbair joined #git
01:35 tnecniv joined #git
01:36 rchavik joined #git
01:36 Goplat joined #git
01:37 tnecniv joined #git
01:41 hotpancakes joined #git
01:41 Milos joined #git
01:42 tnecniv joined #git
01:47 sunri5e joined #git
01:47 cagomez joined #git
01:47 Lyqyd joined #git
01:47 duderonomy joined #git
01:47 bitmod joined #git
01:49 ilbot3 joined #git
01:49 Topic for #git is now Welcome to #git, the place for git help. Just git help, apparently. | Public logs at https://gitirc.eu/log | First visit? Read: https://gitirc.eu | Current stable version: 2.12.2 | Getting "cannot send to channel"? /msg gitinfo .voice | This is serious business. https://madeitwor.se/git-jokes
01:50 brent__ joined #git
01:51 tnecniv joined #git
01:53 mel00010 joined #git
01:54 dreiss joined #git
01:59 kpease joined #git
02:00 dermoth joined #git
02:02 Vortex35 joined #git
02:04 jameser joined #git
02:05 telephone joined #git
02:05 telephone joined #git
02:07 hotpancakes joined #git
02:09 User458764 joined #git
02:11 chachasmooth joined #git
02:13 Ryanar joined #git
02:14 cagomez joined #git
02:15 dsdeiz joined #git
02:15 dsdeiz joined #git
02:16 watabou joined #git
02:16 svobhristob joined #git
02:17 gugah joined #git
02:17 Cabanossi joined #git
02:19 matt|home git checkout switches between branches?
02:20 doppo joined #git
02:21 Hello71 checks out a commit.
02:21 Hello71 or a file.
02:21 matt|home ah.. im a bit confused then
02:21 Hello71 man git-checkout
02:21 gitinfo the git-checkout manpage is available at https://gitirc.eu/git-checkout.html
02:22 roelmonnens joined #git
02:22 matt|home from my understanding, when you create "branches" in git that just essentially allows you to keep track of multiple versions of files at the same time.. i thought you could simply switch between branches and work on files like that
02:26 peepsalot joined #git
02:30 cagomez joined #git
02:32 dandaman joined #git
02:32 deltab matt|home: yes; checkout will change HEAD to point to the selected branch or commit, and adjust the files so that they match that version
02:32 a_thakur joined #git
02:33 matt|home ah
02:33 matt|home so if i change a file in one branch it wont modify files in any other branches deltab ?
02:33 deltab correct
02:33 matt|home thanks deltab
02:36 deltab you can experiment in a separate repository: create a directory, cd to it, git init, create some files, git add them, create branches, switch between them, etc.
02:36 matt|home mkay
02:36 matt|home i just had one question deltab (then i'll leave you alone)
02:37 matt|home https://pastebin.com/pnvKRMjr <-- that's the output of git branch -a when im in master
02:38 matt|home but when i switch to gh-pages , the output shows * gh-pages at the top above master , is that just indicitave of that being where i am?
02:39 johnflux_ joined #git
02:39 johnflux_ joined #git
02:39 matt|home oh nevermind i think i see, i have to manually set it up to track branches remotely
02:39 matt|home okay. i thought if you tracked master you tracked all branches at once. my fault
02:43 h12o joined #git
02:44 a_thakur joined #git
02:44 hexagoxel joined #git
02:45 Hello71 git status
02:49 twb joined #git
02:51 twb For dumb historical reasons, my codebase is mostly in A.git, but a few files are in a submodule in B.git.  I want to merge B back into A.  What docs should I be reading>
02:51 Lyqyd joined #git
02:52 twb (every commit in B.git is an ancestor of B master, so I don't think have to care about multiple branches.)
02:52 re1 joined #git
02:53 twb B working directory has about 4 top-level things, so I guess I can't merge -s subtree without mucking about with a filter-branch first
02:55 lagothrix joined #git
02:57 ExeciN joined #git
03:00 dave0x6d joined #git
03:00 TheSimonator joined #git
03:03 sondr3 joined #git
03:06 Hello71 !subproj
03:06 gitinfo [!subprojects] So, you want to add git repositories inside of other git repositories? Well, you have four main options. First is to just do it, add the repo to the outer project's .gitignore, and treat them entirely separately. Best if they are entirely separate. Otherwise your best options are "!submodule", "!gitslave", and "!subtree". Try those commands in this channel, or in a PM to avoid flooding.
03:06 Hello71 there is always the stupid option where you just import one as files into the other
03:07 twb yeah that's plan B; just throw away the history
03:07 twb ATM I'm looking at https://www.kernel.org/pub/software/scm/git/docs/howto/using-merge-subtree.html
03:07 twb trying it out to see if the history afterwards (especially git blame) is actually useful
03:14 TbobbyZ joined #git
03:20 twb Yeah OK this is annoying
03:21 twb After merging per that URL above, blame & log (with -w -M -C / --follow) claim that the entire subtree is my fault
03:22 chandu20 joined #git
03:22 robertparkerz joined #git
03:24 Hello71 joined #git
03:26 awc737 joined #git
03:26 hotpancakes joined #git
03:28 Pseudocrat joined #git
03:28 Pseudocrat joined #git
03:29 fission6 joined #git
03:31 fission6 joined #git
03:32 Cabanossi joined #git
03:33 otiose joined #git
03:33 v3i0 joined #git
03:34 v3i0 left #git
03:35 twb Grr.  Moving everything in B.git into B/ and then doing an ordinary merge (per http://stackoverflow.com/questions/19385565/fix-broken-git-history-after-merging-other-repo-into-subdir) is also not making history DWIM
03:35 kegster joined #git
03:36 twb The B repo is mostly flat, maybe I should just flatten it into a single branch and filter-branch it
03:41 aavrug joined #git
03:42 aavrug joined #git
03:43 awc737 joined #git
03:43 wcpan joined #git
03:43 davidpitkin left #git
03:45 Pseudocrat_ joined #git
03:45 h12o joined #git
03:47 fission6 joined #git
03:49 brent__ joined #git
03:49 sword joined #git
03:53 Pseudocrat joined #git
03:55 twb OK I think I solved it
03:55 twb previously I had a submodule commited in A.git B/, and I was trying to import B.git as B/.
03:56 twb The presence of the submodule stuff was confusing blame/log --follow, I think
03:56 Pseudocrat joined #git
03:56 Pseudocrat joined #git
03:57 twb So instead I went into B.git and moved everything into B2/, then merged that into A.git.  Now in A.git, blame shows the B.git history for B2/foo.c as expected
03:57 awc737 joined #git
03:57 twb git log still needs --follow, but I can live with that
04:02 hotpancakes joined #git
04:02 Cabanossi joined #git
04:10 roelmonnens joined #git
04:11 BeepBeepImASheep joined #git
04:11 BeepBeepImASheep rebase or merge?
04:12 twb merge
04:12 twb Oh sorry, thought you were asking about my problem
04:12 BeepBeepImASheep lol okay
04:12 BeepBeepImASheep rebase or merge when there's a merge conflict?
04:12 twb BeepBeepImASheep: in general, avoid rebasing anything someone else has pulled (i.e. stuff you've pushed to a public repo)
04:13 twb if you haven't pushed yet, you have more flexibility; it depends more on personal preference and which is easier for that particular conflict
04:14 BeepBeepImASheep Ahhh, okay. Makes sense
04:15 BeepBeepImASheep btw when do merge conflicts occur?
04:15 FilipNortic joined #git
04:15 rety joined #git
04:16 twb BeepBeepImASheep: roughly, when two branches make commits close together in the same file
04:17 BeepBeepImASheep close together as in nearly the same time?
04:17 twb as in nearly the same line
04:17 BeepBeepImASheep oh
04:17 twb in other words if alice and bob each have their own branches and are working on separate modules in the codebase, there probably won't be merge conflicts
04:17 twb but if alice and bob are both editing the same functions in the same module, conflicts are likely
04:18 BeepBeepImASheep Ahh
04:18 druonysus joined #git
04:18 rety left #git
04:18 rety joined #git
04:19 Noodlewitt joined #git
04:19 BeepBeepImASheep So suppose I resolve the conflict by choosing my commits to push to the master but alice has done the same for her commits, how does it work out then?
04:20 johnflux joined #git
04:20 twb BeepBeepImASheep: in a typical setup, you can only push "fast forward" changes, so you must resolve all conflicts locally, before pushing
04:22 njbair joined #git
04:22 BeepBeepImASheep Ah, okay!
04:28 tubal joined #git
04:31 robotroll joined #git
04:36 hussam joined #git
04:38 BeepBeepImASheep moar git jokes
04:38 durham joined #git
04:39 durham joined #git
04:40 miczac joined #git
04:44 awc737 joined #git
04:44 pifon joined #git
04:47 MistahKurtz joined #git
04:53 freimatz joined #git
04:53 swalladge joined #git
04:55 awc737 joined #git
04:56 h12o joined #git
04:58 mel00010 joined #git
04:58 ayogi joined #git
04:59 awc737 joined #git
05:00 ExeciN joined #git
05:02 overlord_tm joined #git
05:06 a_thakur joined #git
05:09 hotpancakes joined #git
05:09 Bock joined #git
05:09 dec0n joined #git
05:09 pks joined #git
05:13 schleppel joined #git
05:17 h12o joined #git
05:17 TbobbyZ joined #git
05:18 qt-x joined #git
05:21 venmx joined #git
05:23 netj joined #git
05:26 svm_invictvs joined #git
05:27 h12o joined #git
05:31 imack joined #git
05:32 scoobert1on joined #git
05:34 otiose joined #git
05:36 ppang joined #git
05:36 jameser joined #git
05:36 MattMaker joined #git
05:39 caliculk joined #git
05:41 HurricaneHarry joined #git
05:43 dersand joined #git
05:44 Taikibansei joined #git
05:44 raspado joined #git
05:47 h12o joined #git
05:48 miracle joined #git
05:50 miracle how to extract branch from one repo and transfer in other repo copy via files (no direct network connect) ?
05:51 raspado joined #git
05:52 markus-k miracle: maybe "git subtree" can help you
05:53 kadoban That plus git bundle probably
05:54 kadoban Oh wait, a branch not a directory you said. Yeah, just git bundle then.
05:54 miracle ok. i take timeout for  documentation
05:56 markus-k oh yes, if it's not just a subdirectory, subtree likely won't be of big help
05:56 kadoban It's also possible just to make a copy of the source repo as files and just copy that to the machine you want, then you can fetch from it (pointing to the files)
05:56 kadoban That's less fancy, but it does work.
05:58 Raging_Hog joined #git
05:58 roelmonnens joined #git
06:00 aax joined #git
06:00 dermoth joined #git
06:01 onehrxn joined #git
06:02 Cabanossi joined #git
06:03 Burgerz joined #git
06:04 Burgerz joined #git
06:05 Silmarilion joined #git
06:05 Burgerz joined #git
06:05 sbulage joined #git
06:07 zefferno joined #git
06:07 h12o joined #git
06:08 Burgerz joined #git
06:08 onehrxn joined #git
06:08 fakenerd joined #git
06:10 bgamari joined #git
06:11 navidr joined #git
06:12 rscata joined #git
06:12 cads joined #git
06:13 zefferno joined #git
06:17 chele joined #git
06:18 thebishop joined #git
06:19 zeroed joined #git
06:19 ftoulemon joined #git
06:19 rominronin joined #git
06:22 JeroenT joined #git
06:24 kegster joined #git
06:24 roelmonnens joined #git
06:26 aidalgol joined #git
06:28 h12o joined #git
06:28 cdown joined #git
06:31 freimatz joined #git
06:32 jagob joined #git
06:33 elect_ joined #git
06:36 lindenle joined #git
06:38 miracle i do something wrong. `git checkout cnetxms ; git bundle create cnetxms.bundle cnetmxs` on machine A,  `git checkout -b miracle ; git pull ..\cnetxms.bundle cnetxms` on B and i have now only modified files on B, but not commit history ? what i doing wrong ?
06:39 telephone joined #git
06:39 telephone joined #git
06:39 iamashishkebab2 joined #git
06:44 acetakwas joined #git
06:45 User458764 joined #git
06:46 bilal joined #git
06:47 zefferno joined #git
06:48 manuelschneid3r joined #git
06:48 h12o joined #git
06:49 rokups joined #git
06:50 karols146 joined #git
06:52 Lunatrius` joined #git
06:52 JeroenT joined #git
06:52 AndreaACME joined #git
06:53 flaviodesousa joined #git
06:54 cdown_ joined #git
06:54 bilal joined #git
06:55 MineCoins joined #git
06:57 diogenese joined #git
06:58 dreiss joined #git
06:59 zeroed joined #git
06:59 zeroed joined #git
07:00 _ng joined #git
07:00 esenegoid joined #git
07:01 King_Hual joined #git
07:01 King_Hual joined #git
07:01 YuGiOhJCJ joined #git
07:05 King_Hual joined #git
07:05 JeroenT joined #git
07:08 h12o joined #git
07:11 diogenese joined #git
07:11 elect__ joined #git
07:12 JeroenT joined #git
07:13 Tobbi joined #git
07:14 hotpancakes joined #git
07:14 misosoup joined #git
07:14 _Vi joined #git
07:15 Es0teric joined #git
07:16 jost joined #git
07:19 h12o joined #git
07:20 dfas joined #git
07:20 TbobbyZ joined #git
07:21 duderonomy joined #git
07:22 skre4 joined #git
07:22 fakenerd joined #git
07:23 Murr joined #git
07:24 dcpc007 left #git
07:25 ogny joined #git
07:25 ogny joined #git
07:26 dvaske_ joined #git
07:27 Gurkenglas joined #git
07:28 miracle how i can pull branch from bundle and create branch copy in existing repo with master ahead  ?
07:29 User458764 joined #git
07:29 kadoban fetch from it, create a branch from that if you want
07:30 Levex joined #git
07:31 miracle kadoban: i fetch it. but dont know what doing after
07:31 miracle From ..\cnetxms.bundle    * branch            cnetxms    -> FETCH_HEAD
07:32 osse miracle: 'git branch newbranch FETCH_HEAD'
07:32 miracle but `git branch -a`  dont show this brach
07:33 djb-irc joined #git
07:34 theoceaniscool joined #git
07:34 pbandark joined #git
07:35 miracle osse: ok . i do it . branch created but not have not one commit
07:35 otiose joined #git
07:35 miracle not one commit from bundle*
07:36 osse the branch now named "newbranch" is identical to the cnetxms branch from the bundle
07:36 miracle i test my bundle with clone repo from this. it is contain history what i need
07:37 miracle osse: i exec `git branch cnetxms FETCH_HEAD` for creating branch with same name in repo . it is important ?
07:37 Levex joined #git
07:38 osse depends on what you consider important
07:39 h12o joined #git
07:39 JeroenT joined #git
07:39 miracle ok. i do `git checkout cnetxms`  for switching to cnetxms. i `git log` not contain my commits
07:41 ertes joined #git
07:41 osse miracle: run 'git rev-parse FETCH_HEAD' and 'git rev-parse cnetxms'. Are the numbers the same?
07:42 karols146 joined #git
07:42 ogny joined #git
07:46 venmx joined #git
07:47 bcc_ left #git
07:48 bcc joined #git
07:48 bcc left #git
07:48 ogny joined #git
07:52 Levex joined #git
07:52 Introoter joined #git
07:52 hotpancakes joined #git
07:55 bilal joined #git
07:56 fakenerd joined #git
08:00 zxd joined #git
08:01 zxd what does -1  mean in  git rev-list HEAD -1
08:01 User458764 joined #git
08:02 osse zxd: same as -n 1
08:02 kurkale6ka joined #git
08:02 sea-gull joined #git
08:02 sea-gull_ joined #git
08:05 bilal joined #git
08:05 Snugglebash joined #git
08:07 elsevero joined #git
08:07 D4R5C0D3 joined #git
08:08 hurricanehrndz joined #git
08:09 zxd what is -n ?
08:09 zxd ah limit number of outputs ?
08:09 grawity yes
08:09 zxd how does it decide what to output
08:09 zxd from the list
08:10 zxd what to output first
08:10 grawity it starts with the one you gave, and goes backwards
08:10 h12o joined #git
08:10 zxd ah
08:10 zxd cool
08:10 zxd thanks
08:11 mikecmpbll joined #git
08:12 marcogmonteiro joined #git
08:12 iamasamosa joined #git
08:15 ams__ joined #git
08:16 Velizar joined #git
08:16 Velizar joined #git
08:16 a_thakur joined #git
08:16 miracle osse: yes. numbers is equal
08:17 osse miracle: then either you misunderstand, or the bundle is incomplete. your local branch cnetxms is identical to the one in the bundle
08:17 Cabanossi joined #git
08:17 wandering_vagran joined #git
08:18 miracle osse: ok. should git log show my last commits if  cnetxms checkouted ?
08:18 osse yes
08:19 Levex joined #git
08:19 overlord_tm joined #git
08:20 zeroed joined #git
08:21 netj joined #git
08:21 miracle git ls-remote cnetxms.bundle
08:21 miracle 7b0fb47a835f62cbc365cd0602a6fa6c9c4dcea9        refs/heads/cnetxms
08:22 nickel_dime_look joined #git
08:22 miracle i can clone from my bundle to new repo. my bundle is ok ?
08:22 nickel_dime_look joined #git
08:22 osse it is "ok" in the sense that it works. but if it contains the right things I cannot know
08:22 m4sk1n joined #git
08:22 miracle and i see my commits in new repo
08:23 aidalgol joined #git
08:23 Murii joined #git
08:25 tymczenko joined #git
08:25 xaa joined #git
08:27 Raging_Hog joined #git
08:27 rudrab joined #git
08:28 uranellus joined #git
08:28 uranellus joined #git
08:28 a_thakur joined #git
08:28 cstratak joined #git
08:30 h12o joined #git
08:31 jagobdue joined #git
08:31 laumann joined #git
08:32 aidalgol joined #git
08:32 miracle bundle for branch contain commits from head of branch to begin of branch ? or all commits with commits of parent branch ?
08:33 Snugglebash joined #git
08:33 acetakwas joined #git
08:33 canton7 miracle, a branch doesn't "begin" until it reaches the first commit in the repository (ignoring orphan branches)
08:34 miracle canton7: i not understand = branch begin in first commit of repo or in something in middle it?
08:35 Raging_Hog joined #git
08:36 canton7 miracle, if you create a new branch based on master, then all of the commits in master are also part of that new branch
08:36 miracle ok, thanks
08:37 cstratak joined #git
08:37 miracle so my bundle for branch must contain all commits from first commit to head of branch
08:37 canton7 yes
08:37 jagob joined #git
08:39 DomeMasterYolo18 joined #git
08:43 dcpc007 joined #git
08:43 lindenle joined #git
08:45 iamashishkebab2 joined #git
08:46 miracle ok. i want delete cnetxms branch from repo (and try all steps again) but take error `git branch -d cnetxms`
08:46 miracle error: The branch 'cnetxms' is not fully merged.
08:46 miracle If you are sure you want to delete it, run 'git branch -D cnetxms'.
08:46 DomeMasterYolo18 joined #git
08:46 miracle git status inside branch cnetxms say me `nothing to commit, working tree clean`
08:47 miracle something going wrong with this branch
08:47 iamashishkebab2 joined #git
08:47 Cabanossi joined #git
08:47 iamasamosa joined #git
08:50 DomeMasterYolo18 joined #git
08:50 h12o joined #git
08:51 hotpancakes joined #git
08:51 flaviodesousa left #git
08:51 tvw joined #git
08:51 velizar_ joined #git
08:51 velizar_ joined #git
08:51 barbos joined #git
08:52 telephone joined #git
08:52 telephone joined #git
08:52 Mogget joined #git
08:53 osse If you are sure you want to delete it, run 'git branch -D cnetxms'.
08:53 Xenophon1 joined #git
08:54 Wulf Good Morning. How to rewrite author's email address in previous commit? Yes, rewriting history is okay here, no copies of the repo exist yet.
08:54 DomeMasterYolo19 joined #git
08:54 demize How previous?
08:54 demize As in the last one, or before that?
08:54 Wulf demize: before that
08:55 demize If before the last one, then do an interactive rebase to a commit before the one you want to change, then set the commit you want to change to edit
08:55 demize Then do a `git commit --amend --author="..."`
08:56 miracle osse: yeah. i delete it with -D .
08:56 miracle now i return to begin. my bundle seems is ok. https://pastebin.com/GF9e2q7u
08:57 demize Wulf: So if you want to change commit FOO, then `git rebase -i FOO^`, then change `pick` on the first line to `edit`
08:57 TbobbyZ joined #git
08:57 demize Then run that git-commit command, then `git rebase --continue
08:57 demize `
08:58 miracle how to fetch branch cnetmxs from bundle and create new branch in repo (probably with same name cnetxms)
08:58 Wulf demize: yep, worked. Thanks
08:58 Wulf demize: but how do I modify the first commit?
08:59 miracle *new branch which contain comits from bundle
09:00 demize Wulf: `git rebase --root`
09:00 Wulf thanks!
09:00 demize No problem o/
09:01 barbos joined #git
09:02 osse miracle: I think you did it right the first time. Is the number from ls-remote the same as rev-parse earlier?
09:02 lyxus joined #git
09:02 JeroenT_ joined #git
09:03 JeroenT_ joined #git
09:03 miracle osse: no. it is different
09:04 osse ok, that's good
09:04 osse do git fetch cnetxms.bundle refs/heads/cnetxms:refs/heads/cnetxms
09:05 a_thakur joined #git
09:05 miracle i have lunch time. some later plz :)
09:06 bilal_ joined #git
09:06 marcogmonteiro joined #git
09:07 bilal joined #git
09:08 hussam joined #git
09:09 cdown joined #git
09:10 User458764 joined #git
09:11 clmsy joined #git
09:11 bilal joined #git
09:13 jagobdue joined #git
09:14 aax joined #git
09:16 mrkake joined #git
09:16 sarri joined #git
09:16 sarri joined #git
09:17 auzty joined #git
09:19 tvw joined #git
09:20 barbos joined #git
09:21 h12o joined #git
09:23 mozzarella joined #git
09:23 chele joined #git
09:23 g5pw joined #git
09:24 rishi`` joined #git
09:24 arooni_team_b joined #git
09:24 _lvb__ joined #git
09:24 selckin1 joined #git
09:25 Kobaz joined #git
09:26 jagob joined #git
09:26 MatthewAllan93 joined #git
09:26 IRQ` joined #git
09:27 bilal joined #git
09:27 Industrial joined #git
09:27 Industrial hi. "git push --set-upstream origin feature/thingy" I'm getting an error that I should run this command
09:28 Industrial Why doesnt it just do that instead of tell me to do so?
09:28 Industrial I haven't in 10 years found a situation where I wouldn't want to track a branch I'm pushing to remote that doesn't exist there.
09:32 kexmex joined #git
09:34 rudi_s Industrial: Because you might have more than one remote and you can set only one upstream remote for a branch.
09:34 romank joined #git
09:36 otiose joined #git
09:36 fakenerd joined #git
09:41 h12o joined #git
09:42 ferr joined #git
09:43 JeroenT joined #git
09:43 barbos joined #git
09:47 Cabanossi joined #git
09:49 venmx joined #git
09:51 aard joined #git
09:52 h12o joined #git
09:55 chovy joined #git
09:55 chovy joined #git
09:57 seni joined #git
09:58 blahdodo joined #git
09:59 swift1 joined #git
09:59 fakenerd joined #git
10:01 fakenerd joined #git
10:08 permalac joined #git
10:09 clmsy joined #git
10:10 jagobdue joined #git
10:11 Gurkenglas joined #git
10:12 h12o joined #git
10:12 skre4 joined #git
10:12 xaa joined #git
10:16 Ajit_ joined #git
10:16 AstralStorm joined #git
10:16 jagob joined #git
10:19 chardan joined #git
10:19 zeroed joined #git
10:20 Snugglebash joined #git
10:23 diogenese joined #git
10:25 dsdeiz joined #git
10:25 dsdeiz joined #git
10:28 Silox| joined #git
10:28 _ng joined #git
10:29 velizar_ joined #git
10:30 miracle joined #git
10:32 tvw joined #git
10:32 h12o joined #git
10:33 Raging_Hog joined #git
10:34 MaximB joined #git
10:34 hotpancakes joined #git
10:34 Altiare joined #git
10:35 MaximB Hi, i did git init and it greated a .git dir . I have a dir named Programming where I keep all my scripts. how can I use it with git? to where do I need to copy the dir and how to make git use it?
10:35 _ikke_ MaximB: try git status
10:36 moritz MaximB: it sounds like you might want to read a !book as an introduction
10:36 gitinfo MaximB: 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
10:36 MaximB I want only the basics
10:36 moritz so stop reading after chapter 2 or 3
10:39 earnestly joined #git
10:39 MaximB ok, but basically if i do git add <dir> , it starts using it? does it copy the dir to the .git dir repo or use it from where it is?
10:39 zerow joined #git
10:39 karols146 joined #git
10:40 moritz it uses it where it is, but it also copies a compressed copy to its .git dir for internal use by git
10:40 MACscr joined #git
10:40 kbeflo joined #git
10:40 osse MaximB: the way you use git is normal to cd to Programming and run git init there.
10:41 JeroenT joined #git
10:41 buck1 joined #git
10:42 Karazhan joined #git
10:42 kdeal joined #git
10:44 chovy joined #git
10:44 chovy joined #git
10:48 marcogmonteiro joined #git
10:52 Tobbi joined #git
10:53 h12o joined #git
10:53 MaximB I did git comming inside Programming, did git add DirX inside Programming, git config, git commit .. all files were added nice, then I edit 2 files and did git commit -a -m "edit" and got this msg:  2 files changed, 2 deletions(-) . what it deleted? the old versions of the files?
10:53 dosmachko joined #git
10:53 visual morning gits
10:54 visual so im trying to do a bit of an unconventional thing
10:54 visual maybe you could give me some pointers if you tried it before
10:54 visual as of now - 2 working branches, master and {branch 1}
10:55 visual i want to switch to another branch {branch 2}, do some work, merge it with master, and then go back to {branch 1}
10:55 visual how would i go about merging the changes from {branch 2} with {branch 1} ?
10:55 visual on branch 1
10:55 osse MaximB: two lines in total
10:55 dosmachko joined #git
10:56 MaximB nice, thank you
10:57 jackrandom joined #git
10:58 AndreaACME Thanks osse   (from MaximB )
10:58 AndreaACME Thanks osse
10:59 visual >tfw
10:59 visual >no reply
10:59 osse visual: git checkout {branch 2}; git merge {branch 1}
11:00 dosmachko joined #git
11:00 TbobbyZ joined #git
11:00 apotry joined #git
11:00 jagobdue joined #git
11:01 visual well that was straightforward
11:01 visual thanks osse
11:01 kbeflo_ joined #git
11:02 Cabanossi joined #git
11:03 osse visual: what did you create branch 2 from?
11:03 MaximB left #git
11:03 osse if you merge like I said you get everything in branch 2 that's not already ibn branch 1. That might be some bits of master you don't want
11:04 kbeflo_ joined #git
11:04 visual umm, good question
11:04 visual i think i'll create it from master
11:05 Snugglebash joined #git
11:05 visual "might be some bits of master you dont want"
11:05 visual why?
11:05 visual how?
11:05 jagob joined #git
11:05 visual if master was updated after {branch 1} was created?
11:06 Tobbi joined #git
11:06 visual so some bits wil got to branch 2, got it
11:06 visual no, not a problem
11:06 visual on a related note, is there a way to see diff {branch 1} {master/branch2} ?
11:06 brent__ joined #git
11:06 apotry joined #git
11:08 dsdeiz joined #git
11:08 dsdeiz joined #git
11:08 sondr3 joined #git
11:11 ExeciN joined #git
11:12 raijin joined #git
11:13 h12o joined #git
11:14 miczac joined #git
11:16 nikow joined #git
11:19 nowhereman joined #git
11:19 visual SourceTree vs Gitlab?
11:22 cstratak joined #git
11:22 VampiricPadraig joined #git
11:22 VampiricPadraig joined #git
11:23 Tobbi joined #git
11:25 cbreak sure you can diff different things
11:25 cbreak visual: sourcetree and gitlab are unrelated
11:25 visual i know, asking about their /branch diffing/ tool
11:27 chll_ joined #git
11:27 Junior joined #git
11:33 h12o joined #git
11:37 otiose joined #git
11:37 bernd27 joined #git
11:38 jeffreylevesque joined #git
11:40 tocky joined #git
11:42 fission6 joined #git
11:42 jagobdue joined #git
11:42 fakenerd_ joined #git
11:43 hotpancakes joined #git
11:47 Cabanossi joined #git
11:47 manuelschneid3r joined #git
11:50 venmx joined #git
11:50 Junior joined #git
11:52 slidercrank joined #git
11:52 Vampire0 visual, !gui ;-)
11:52 gitinfo visual: Graphical user interfaces are not supported here. If you want to get support, it needs to be through the git CLI. Reasons: 1) Because very few people here use the graphical interface. 2) Because giving instructions for GUI's is difficult. 3) The command line gives you a history of what commands you have executed.
11:52 blackwind_123 joined #git
11:52 visual Vampire0: thank you
11:54 h12o joined #git
11:54 UrsoBranco joined #git
11:55 arecaceae joined #git
11:55 User458764 joined #git
11:55 Vampire0 visual, actually in GitLab, you go to "Repository", then on "Compare", then you can select what to compare
11:55 visual yep yep, found it
11:55 visual you can also do that in phpstorm if that's your thing
11:56 Vampire0 My thing is the command line client mostly :-)
11:56 gnixev joined #git
11:57 Tobbi joined #git
11:57 visual i hear you, but diff file1 file2 is one thing, diff branch1..branch2 is hard without a visualization tool
11:58 gableroux joined #git
12:01 karols146 joined #git
12:01 xaa joined #git
12:01 osse git difftool branch1..branch12
12:01 Levex joined #git
12:01 osse 2
12:02 gableroux joined #git
12:02 gableroux joined #git
12:02 osse also, I highly recommend getting comfy with the output of git diff branch1 branch2, because that sort of output is everywhere
12:04 caper joined #git
12:04 random123 joined #git
12:04 aax joined #git
12:06 moei joined #git
12:07 scottjl joined #git
12:07 watersoul joined #git
12:07 random123 Hi, I can't find documentation about --first-parent with git show and merge commit
12:07 random123 I have merge commit with parents ce3 and 7b3 (in that order)
12:07 random123 but git show -m --first parent show me actually second branch
12:08 random123 actually it's exactly what I've needed
12:08 random123 but I need to be sure
12:09 hotpancakes joined #git
12:09 j416 random123: afaik git show does not take either of -m or --first-parent. Do you mean git log
12:09 j416 ?
12:10 random123 j416: it did, and produce desired result
12:10 random123 -m are ignored
12:11 j416 random123: so what is your question?
12:11 random123 j416: but the question still in place why first-branch outputs second branch
12:11 j416 random123: I have no idea what you are doing
12:12 j416 random123: what is first-branch?
12:12 random123 s/branch/parent/
12:12 random123 first-parent outputs second-parent
12:12 j416 random123: what do you want?
12:12 osse random123: the diff between a merge commit and one of the parents is the entirety of the diff of the *other* branch
12:12 thiago joined #git
12:12 jrm joined #git
12:12 osse because what's changed between them is that another branch was merged in
12:13 random123 osse: now it makes sence
12:14 h12o joined #git
12:15 random123 osse: so "git show --first-parent foo" equals to "git diff foo..foo^1"
12:15 jost joined #git
12:15 cdown_ joined #git
12:15 osse equal to git diff foo^1..foo but yes
12:15 osse which btw is the same as git diff foo^1 foo
12:16 Murii joined #git
12:16 ZexaronS-2 joined #git
12:16 ski7777 joined #git
12:16 cagomez joined #git
12:16 random123 probably I should send pull request to git to fix this counter-intuitive argument
12:17 random123 osse: j416: thank you
12:17 jfr__ joined #git
12:17 dglambert joined #git
12:17 dglambert yo
12:17 j416 where's --first-parent to git-show documented? Doesn't seem git-diff-tree documents it either? What is it an argument to?
12:17 osse random123: it is really meant for git log, not git show
12:18 fakenerd joined #git
12:19 random123 j416: --first-parent doesn't documented but works, this is why I came here to understand how it works
12:19 j416 random123: probably better off not using it
12:19 j416 random123: use git show for single commits, use diff or log for other things
12:20 j416 in non-ancient versions of git though I believe 'git show' on a merge-commit without extra arguments will show what that merge commit changes apart from what the branch(es) it merges in change
12:20 random123 j416: my first intent was to list files changed in topic when topic merged to master
12:21 j416 random123: git diff --name-only master..topic
12:21 osse y u use double dot :((
12:21 j416 random123: sorry, git diff --name-only master...topic
12:21 j416 osse: because I type too fast for my fingers to follow me
12:21 j416 :(
12:22 j416 :::(
12:22 osse hmm I think git diff --name-only master^1 master is correct here
12:22 j416 if it has been merged
12:22 j416 yeh
12:22 osse yours may show files changed in master but not in topic
12:22 j416 so many questions, so little information
12:23 random123 j416: more interesting that I get refs from git hub fake branches created for pull request and I have only merge-commint-for-pr-1234 and last-commit-in-pr-1234
12:23 random123 so I have only merge and topic
12:23 ferr joined #git
12:23 random123 without knowing of merge-into branch
12:24 gableroux joined #git
12:24 rasto joined #git
12:25 random123 git diff merge-pr-1234^ merge-pr-1234
12:25 random123 works for me
12:25 j416 random123: the "merge-into" branch is the first parent of the merge.
12:25 paissad joined #git
12:25 j416 random123: that is, merge-pr-1234^
12:26 random123 I mean after merge
12:26 leeN joined #git
12:26 paissad hello guys, i would like to see the changes in the remote origin branch before i do a "git pull" or "git fetch"
12:26 j416 random123: after the merge it is the merge commit itself.
12:26 j416 paissad: why?
12:27 paissad j416: to see if i'm about to face some conflicts
12:27 random123 j416: correct https://help.github.com/articles/checking-out-pull-requests-locally/
12:27 j416 paissad: git fetch is safe to run, it won't modify your local branches (only your read-only remote-tracking branches)
12:27 paissad j416: or know what has changed to see if i really need to update now ... whatever
12:27 j416 random123: haven't read that, sorry
12:27 paissad i prefer to know what i'm "pulling"
12:27 seveneve1 joined #git
12:27 j416 paissad: git fetch, the compare against what you got
12:28 random123 paissad: git merge --no-commit topic origin/topic
12:28 random123 or diff
12:28 hotpancakes joined #git
12:28 paissad random123: so basically "git merge --no-commit master origin/master" ?
12:29 j416 random123: his point was that he wanted to compare them before merging.
12:30 j416 paissad: git fetch; then diff against your branch upstream (that's origin/whatever for example, or the synonym @{upstream})
12:30 Junior joined #git
12:30 j416 paissad: git fetch && git log --oneline --decorate ..@{upstream}
12:30 j416 try that
12:31 j416 not sure exactly what you want
12:31 paissad j416: ok, thanks
12:31 Chuguniy joined #git
12:31 roelmonn_ joined #git
12:34 plos joined #git
12:34 tarkus joined #git
12:34 h12o joined #git
12:37 YuGiOhJCJ joined #git
12:38 Silmarilion joined #git
12:39 rtc joined #git
12:39 Matsv joined #git
12:39 miracle joined #git
12:39 rtc Hi, how do I make git diff show the file timestamps, as plain diff does?
12:39 hotpancakes joined #git
12:40 seni joined #git
12:40 cbreak rtc: what for?
12:41 masuberu joined #git
12:41 rtc cbreak: I want to see when the last commit of the file was done, and the current timestamp of the file...
12:42 binary106 joined #git
12:42 cbreak git does not store timestamps of files
12:42 cbreak the closest you can get is to look at the commit date or author date
12:43 cbreak but diff of course doesn't look at that
12:43 rtc the filesystem has the current timestamp, and git should know the last commit of the file, no?
12:43 cbreak depends
12:43 cbreak if you do something like git log -p -- filename
12:43 cbreak then it knows
12:44 cbreak it'll show you the commit including date
12:44 cbreak if you use diff, then it does not know
12:44 cbreak because then you don't compare commits, but just blobs
12:44 cbreak or maybe trees
12:45 rtc doesn't it know the commit with which the revision was committed?
12:45 cbreak revision?
12:45 _Vi joined #git
12:46 cbreak any blob can be reached from any number of commits
12:46 j416 rtc: what do you want to know?
12:46 digidog joined #git
12:47 binary106 hi guys I need some advice. I'm an external consultant for a webapp, which is git controlled. My edits only consist of a subfolder. And only I edit that subfolder. I wish to push my changes to this subfolder at any time - and without affecting anyone elses work.. is this possible and if so, how should we go about it
12:47 j416 rtc: git log -- path/to/file
12:47 j416 rtc: does that show you what you need+
12:47 rwb joined #git
12:47 j416 ?
12:47 cads joined #git
12:47 j416 binary106: is it one repository?
12:47 binary106 yea 1 repo
12:47 j416 binary106: or several
12:47 binary106 (atm)
12:48 j416 binary106: then subdirectory or not does not matter. You're in the same playground as the others.
12:48 j416 binary106: you cannot expect to change files anywhere in the same repo without affecting others
12:48 Stumbler joined #git
12:49 binary106 problem is i need to release my code to the web server at different schedule to the main guys
12:49 rtc cbreak: git diff doesn't just compare blobs, it compares commits.... so it should know the date of the commit
12:49 j416 binary106: is it the same server?
12:49 cbreak not really
12:49 cbreak blobs or trees
12:49 binary106 i was thinking could i make the folder a remote upstream to my own repo? and only update that via ssh
12:49 binary106 yea same server
12:49 j416 binary106: your thinking is a bit off here
12:49 cbreak binary106: no
12:49 j416 binary106: if you change something, it affects them
12:50 rtc j416: it shows the info in some way, but I'd like to have it the same way as plain diff
12:50 j416 binary106: so, if you need to deploy more often, then have to be in on that as well, as their stuff might break
12:50 fission6 joined #git
12:51 masuberu joined #git
12:51 j416 binary106: and, if your changes are in fact not related to their changes (i.e. they have nothing to do with one another), consider splitting them into two repos
12:51 binary106 ok how about a .gitignore then, and I FTP into my special folder?
12:51 j416 rtc: sorry, I'm not very familiar with that format
12:51 j416 binary106: skip the technical aspect and think about how it affects the surrounding code
12:51 binary106 yea 2 repo's sound nice - but my folder is a subfolder of their main repo. I mean I'm inside their root, but my subfolder doesn't have to exist on their repo, just mine
12:51 Vampire0 binary106, or do not use Git as a !deploy tool ;-)
12:51 gitinfo binary106: 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
12:52 rtc j416: It's not a different format, it's what the diff utility outputs by default, which includes timestamps for files, which git diuff omits...
12:52 Vampire0 binary106, if your suggestion with .gitignore meant ignoring changes to tracked files, this will not work
12:52 Ryanar joined #git
12:53 j416 rtc: yeah, I almost never use regular diff so I'm not sure how that looks, sorry
12:53 Vampire0 rtc, `diff` diffs files which have timestamps. `git diff` does not compare files, but text blobs those do not have timestamps
12:53 Vampire0 j416, it just shows the file timestamp after the file name in the diff header
12:53 j416 my 'diff' does not show timestamps by default
12:54 j416 I see
12:54 Vampire0 j416, not? interesting
12:54 binary106 ok, what would you do in this situation? we used to simply FTP into the web server, and I would only edit my subfolder, and the other guys would edit anything else
12:54 rtc Vampire0: the git diff manual page certainly has "git diff [options] <blob> <blob>" but it also has "git diff [options] [<commit>] [--] [<path>...]"
12:54 Vampire0 j416, it is `diff -u`that shows them
12:55 j416 binary106: like I told you. Separate into two repos if there is no dependency between your things and theirs.
12:55 h12o joined #git
12:55 hotpancakes joined #git
12:55 j416 Vampire0: ah, indeed.
12:55 Vampire0 rtc, but it still diffs blobs. You just tell Git where in the tree object it finds the blobs
12:55 j416 rtc: why do you need this specific format?
12:55 robertparkerx joined #git
12:56 Noldorin joined #git
12:56 DevAntoine joined #git
12:56 DevAntoine hi
12:56 Vampire0 !hi
12:56 gitinfo [!welcome] Welcome to #git, a place full of helpful gits. If you have a question, feel free to just go ahead and ask—somebody should answer shortly. For more info on this channel, see https://gitirc.eu/ - Take backups (type !backup to learn how) before taking advice.
12:57 rtc Vampire0: When searching for the blogs it needs to look for the commit, so it should know the timestamp of that commit. I don't ssee any reason for why showing the timestamp should be impossible...
12:57 j416 rtc: the timestamp is not a property of the file, it is a property of the commit, hence, it is not shown per-file
12:58 Vampire0 rtc, but it would be the timestamp of the commit, not the timestamp of the file. The file might not be touched in that commit. It is a property of the commit, not the "file" or the blob
12:59 cbreak rtc: it compares blobs or trees
12:59 rtc j416: I need that format because it's just handy to see ...
12:59 cbreak I don't think there's a tool to compare commits
12:59 SantaClauze joined #git
13:00 Vampire0 cbreak, waht do you mean by "compare commits"?
13:00 j416 rtc: it's even more handy to see the commits
13:00 cbreak Vampire0: find the differences between commits
13:00 mohabaks joined #git
13:00 Snugglebash joined #git
13:00 cbreak the closest I can think of is git log -p
13:00 cbreak i.e. git log -p a..b or git log -p a...b
13:00 _ikke_ Well, doesn't that just do the same as git diff for that commit?
13:00 Vampire0 cbreak, you mean between the changes the two commits introduce?
13:01 cbreak Vampire0: no
13:01 rtc Vampire0, cbreak: okay, fair enough, but there is certainly the most recent commit before the one specified in which the file was last modified
13:01 h12o joined #git
13:01 cbreak Vampire0: that's easy to do by simply comparing the trees
13:01 cbreak rtc: diff doesn't care about commits
13:02 Vampire0 cbreak, so what did you mean? *confused* What would the output of such a tool if it exists show?
13:02 Cabanossi joined #git
13:02 cbreak Vampire0: the difference between two commits is obviously history
13:02 cbreak so it has to show parent and child commits that can be traversed to reach the other commit
13:02 cbreak as I said, git log can do it somewhat
13:02 Vampire0 rtc, only in linear history. If you have a merge it is not that easy
13:03 TbobbyZ joined #git
13:03 cbreak rtc: if you want to see dates, take a look at git gui blame
13:03 cbreak rtc: it'll show you dates for every line in a file
13:03 Vampire0 rtc, you can even have younger commits with author dates before their parent commits easily by rebasing
13:03 cbreak obviously this is done by looking at history
13:03 cbreak unlike git diff, which does not look at history
13:05 mohabaks joined #git
13:06 rtc ok I see. thanks
13:07 marcogmonteiro joined #git
13:08 ExoUNX_ joined #git
13:08 gugah joined #git
13:09 gugah joined #git
13:09 rtc left #git
13:10 theoceaniscool joined #git
13:11 mar77i hmm. I'd need to edit a past commit, but I wonder what happens with the commits that were made since in other branches. I wonder if I could just go and just merge into those other branches afterwards.
13:12 mohabaks joined #git
13:12 j416 mar77i: you'd have to rewrite the other branches as well for that to make sense
13:12 j416 mar77i: or, just bite the apple and don't modify history
13:12 Vampire0 mar77i, you should probably rewrite. Much like the usual problems you have with !rewrite
13:12 gitinfo mar77i: 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
13:12 j416 Vampire0: "you should probably rewrite"?
13:12 j416 or.. you know, avoid rewriting at all ^
13:12 j416 :D
13:12 Vampire0 rebase I meant
13:13 Tobbi joined #git
13:13 j416 (rebase is rewriting)
13:13 mar77i I could create a waqum...
13:13 j416 a waqum
13:13 Vampire0 j416, I know, that's why I added the rewrite trigger
13:13 peacememories joined #git
13:13 j416 where is the party parrot emoji
13:13 mar77i I dysoned it up earlier.
13:14 mar77i wacuumed it away.
13:14 j416 you vacuumed it up
13:14 j416 was your question answered above?
13:15 askb joined #git
13:15 AstralStorm joined #git
13:16 cdg joined #git
13:16 mar77i yes and no. I'm the only person on the project, but I have better things to do. I'll create an empty commit containing the commit id I intend to fix and let the referencing parsers do their magic.
13:17 kexmex joined #git
13:17 j416 !next
13:17 gitinfo Another satisfied customer. NEXT!
13:18 cbreak mar77i: just change what you want to change and make a new commit
13:18 kbeflo joined #git
13:18 cbreak mar77i: when committing, use --fixup
13:19 ChristopheT joined #git
13:19 cbreak then you can use rebase --autosquash later when you have time, and it'll automatically fix that commit
13:19 cbreak mar77i: see man git-commit
13:19 gitinfo mar77i: the git-commit manpage is available at https://gitirc.eu/git-commit.html
13:20 relipse joined #git
13:21 h12o joined #git
13:22 brent__ joined #git
13:22 ChristopheT left #git
13:23 tymczenko joined #git
13:24 snowkidind joined #git
13:24 skre4 joined #git
13:25 apotry joined #git
13:26 j416 cbreak: then take into account that he needs to correct the other branches
13:26 m4sk1n joined #git
13:26 j416 >.>
13:26 cbreak he'll just have to rebase those as well
13:26 cbreak happy times had by everyone!
13:26 bket joined #git
13:27 j416 > 15:16:40     mar77i | [...] I have better things to do. [...]
13:27 j416 ¯\_(ツ)_/¯ :)
13:27 mar77i cbreak: I was talking about editing the commit message.
13:27 cbreak that's what people say when they don't really have anything better to do :P
13:27 j416 but then, how anything could be better than rebasing beats me
13:28 j416 rebasing is so fun
13:28 mar77i let's discuss this for several hours then.
13:28 j416 mar77i: that's a change of history, just like any other
13:28 cbreak mar77i: for that you don't even need a fixup commit, just rebase -i, then amend
13:28 j416 mar77i: discussing ≠ rebasing :)
13:28 j416 cbreak: just rebase -i, the reword *
13:28 j416 then* /me cant spel either
13:29 netroxen joined #git
13:29 mar77i also I didn't notice I had a `git subtree pull --prefix` commit in between, which messes up the rebase royally.
13:30 j416 rebase has --preserve-merges
13:30 j416 but, use with care
13:30 cbreak you'll need to preserve merges for that
13:30 mar77i hmm. I could try that.
13:30 Vampire0 mar77i,  !justdo :-)
13:30 gitinfo mar77i: [!justdoit] If you go to the trouble of asking questions, please go to the trouble of taking the advice offered, Take backups (type "!backup" as a reply to learn more) if necessary -- or in fact, it's a good idea anyway. You can almost always censor log messages if you have to.  We are literally helpless unless you tell us what we need or do what we ask.  Please, Just Do It™.
13:30 Vampire0 oh, wrong trigger :-/
13:31 Vampire0 mar77i,  !justtry :-)
13:31 Vampire0 hm, what was it
13:31 j416 we'll be right back with another Nike™-sponsored FAQ entry
13:31 j416 stay tuned
13:31 Vampire0 !try
13:31 gitinfo [!tias] Try it and see™. You learn much more by experimentation than by asking without having even tried. If in doubt, make backups before you experiment (see !backup). http://gitolite.com/tias.html may help with git-specific TIAS.
13:31 Vampire0 ah, that was it
13:31 j416 !doordonot
13:31 j416 doh.
13:32 mar77i ...nope.
13:32 lewix joined #git
13:32 Dougie187 joined #git
13:32 mar77i rebase -pi also fails in my case, because of the --prefix, I guess.
13:35 dsdeiz joined #git
13:35 Vampire0 how about `rebase -3.14159`? :-D
13:35 AaronMT joined #git
13:36 kbeflo_ joined #git
13:36 _ikke_ :P
13:36 mar77i instead of the subtree pull with prefix being run, the whole tree appears to be deleted, because the rebase doesn't take the prefix into account
13:38 otiose joined #git
13:39 yanome if you use the WeeChat IRC client and haven't already upgrade, you should upgrade: https://weechat.org/news/95/20170422-Version-1.7.1/
13:41 h12o joined #git
13:42 j416 mar77i: !doesntwork
13:42 gitinfo mar77i: Sorry to hear it doesn't work, but how about some details? At this point we don't even know whether it did nothing, spit out an error message, or abducted your pet unicorn in a flying saucer. That makes it hard to troubleshoot – unicorn abduction requires a completely different solution than an error message, for instance. So, please be specific...
13:42 j416 ...I don't really like the wording of this entry. But oh well :/
13:43 peacememories joined #git
13:43 Vampire0 j416, I quite like it. :-D What would you prefer?
13:43 noidea joined #git
13:43 noidea joined #git
13:44 j416 mar77i: ah, perhaps it just recreates the merge "normally" without looking at what actually happened
13:44 laumann joined #git
13:44 j416 mar77i: git-replace might help.
13:44 j416 mar77i: there's --edit
13:45 j416 Vampire0: it's a bit offensive
13:45 Vampire0 Is it? I find it more funny and a bit sarcastic :-D
13:45 Noldorin joined #git
13:45 j416 sarcasm is often lost in text
13:46 _ikke_ Right, but do we really want to be sarcistic to people who seek help here?
13:46 j416 Vampire0: sounds a bit like "I'm sorry you just broke your neck, how about you buy me a coffee first and then we'll see if we can get your "injury" fixed, somehow. Maybe"
13:46 Vampire0 It seems jast thinks we should and noone changed it since 2013 :-D
13:47 Vampire0 j416, no, not really. It sounds like "I'm sorry you are injured, but tell me which limb is aching to be able to help you"
13:47 _ikke_ Or no one bother to change it
13:47 kadoban j416: Your comparison seems poor. Providing details is a basic part of asking a good question. Buying me a coffee isn't, and nobody asked for that.
13:47 peacememories joined #git
13:47 j416 kadoban: indeed poor
13:48 j416 it is the effect of subpar imagination
13:48 kadoban The wording is a little harsh, but I'm not sure I'd change it or not. It's amusingly appropriate fairly often in my opinion.
13:49 j416 kadoban: but you're not the one being helped, most of the time, so you get no say :P
13:49 kadoban ;)
13:50 _ikke_ kadoban: We try to be very helpful and welcoming here. I'm not sure this kind of response fits in that
13:50 MistahKurtz joined #git
13:50 kadoban Perhaps not. I'm not sure :-/
13:50 jagob joined #git
13:51 venmx joined #git
13:51 TingPing joined #git
13:51 kadoban I certainly wouldn't use it as a first response to someone using the words "doesn't work", but then I'm not sure what it's for otherwise.
13:51 _ikke_ We can also ask for details in a more constructive way
13:51 raijin joined #git
13:51 TingPing what would be the best method to *non-interactively* remove the root commit from a branch?
13:52 h12o joined #git
13:52 kadoban TingPing: I probably won't know either way, but what do you want to do with the changes from that commit? Keep them or nuke them?
13:52 cbreak TingPing: git filter-branch's parent filter?
13:52 TingPing kadoban, drop them
13:52 cbreak just remove the root commit from all commits as parent
13:52 otiose joined #git
13:53 kadoban Ya, that's a good idea
13:53 hotpanca_ joined #git
13:54 cbreak git filter-branch --parent-filter 'sed 's/-p 12345//' --all
13:54 cbreak or something like that
13:55 kurkale6ka joined #git
13:55 TingPing actually maybe its easier to mention the root commit is an empty one
13:55 cbreak doesn't matter
13:55 cbreak with that technique you can cut at any commit
13:56 cbreak obviously this will rewrite history
13:57 TingPing cbreak, that filter commant doesn't seem to work, i presume 12345 is the hash but it doesn't seem to be enough args
13:57 tonyt left #git
13:57 cbreak yeah
13:57 cbreak needs to be sed -e "..."
13:58 j416 also a missing '
13:58 cbreak probably using - as file
13:58 TingPing git just prints the usage out
13:58 cbreak you might also have to escape the ' and " properly
13:59 hotpancakes joined #git
14:00 MDA2 joined #git
14:00 TingPing ah you were missing a --
14:01 MistahKurtz joined #git
14:02 dandaman joined #git
14:03 cbreak -- is optional
14:03 cbreak ... sometimes
14:04 TingPing seemed to require it, works like a charm thanks
14:04 Mir4g3 joined #git
14:04 JeroenT joined #git
14:05 Starky joined #git
14:05 otiose joined #git
14:05 pbrewczynski_c joined #git
14:06 mohabaks joined #git
14:06 dandaman1 joined #git
14:07 zivester joined #git
14:08 gablerou_ joined #git
14:10 dbedrenko left #git
14:11 mar77i oh actually. `git replace --edit` together with `git filter-branch -- --all` actually did the trick
14:12 gableroux joined #git
14:13 j416 mar77i: yep
14:13 j416 mar77i: now don't forget to force-push all modified branches
14:13 j416 and tags
14:13 Es0teric joined #git
14:13 j416 (oh, filter-branch doesn't update tags by default, I think, so you'll need to handle that if you've tags)
14:14 gablerou_ joined #git
14:15 rebrec joined #git
14:15 cbreak there's tag-name-filter
14:17 fission6 joined #git
14:19 j416 there's also coffee
14:20 Vampire0 But coffee doesn't help, tag-name-filter does :-
14:20 Tobbi joined #git
14:20 Vampire0 s/:-/:-P/
14:22 h12o joined #git
14:22 Guest52085 joined #git
14:22 mar77i no tags in the way, luckily.
14:22 TingPing left #git
14:22 j416 o/
14:22 j416 Vampire0: helps with what
14:22 hanthings joined #git
14:23 rkazak joined #git
14:23 dandaman joined #git
14:23 Vampire0 j416, handling tags for filter-branch. If you add `--tag-name-filter cat`, the tags are replaced onto the rewritten commits
14:23 d10n-work joined #git
14:23 ResidentBiscuit joined #git
14:24 mar77i now that you mentioned it I double checked. the thing I replaced was the third commit past the most previous tag
14:24 j416 Vampire0: that's the missing link. Thanks :)
14:24 ResidentBiscuit joined #git
14:24 Vampire0 j416, yw :-)
14:25 mar77i git is kind of a whole new approach to something that reminds me of a file system.
14:25 Vampire0 mar77i, if no tag pointed to a rewritten commit, you should be fine
14:26 mar77i lucky me, they're all further down the history.
14:26 j416 you can easily check: git tag --contains theoldcommit
14:26 dandaman1 joined #git
14:27 mar77i or graphically ^f the gui for the tag's commit id. :)
14:27 mar77i "graphically" webbrowseing the interactive thingamajig
14:29 gableroux joined #git
14:29 j416 sounds cumbersome
14:30 bariscant joined #git
14:31 gablerou_ joined #git
14:31 nikow joined #git
14:33 nikow joined #git
14:33 LeBlaaanc joined #git
14:35 dandaman joined #git
14:36 dsdeiz joined #git
14:36 Satchmo1[m] joined #git
14:37 davis joined #git
14:37 davis hello
14:37 Industrial joined #git
14:38 thinkpad joined #git
14:38 wcstok joined #git
14:39 davis regarding git workflow, I thought this was common, "Adding a new feature? Create a branch which implmenents the new feature.  When its accepted, its promoted from feature branch to main. ditto for bug fix. new branch, when accepted, promote to master branch."
14:40 Vampire0 davis, that is a workflow you can follow if you like, what is your question?
14:40 ResidentBiscuit joined #git
14:41 ResidentBiscuit joined #git
14:41 uther left #git
14:42 h12o joined #git
14:43 a_thakur joined #git
14:43 Snugglebash joined #git
14:44 a_thakur joined #git
14:44 LionsMane joined #git
14:45 roelmonnens joined #git
14:45 basiclaser joined #git
14:46 AstralStorm joined #git
14:48 kexmex joined #git
14:48 TikityTik joined #git
14:49 davis Vampire0, its just that at this new job, that approach prompted the lead to say that my repot for this work was a mess. I thought it was a common method.
14:50 Vampire0 davis, there is no "common" method. There are many different !workflows you can follow
14:50 gitinfo davis: [!workflow] Finding the right workflow for you is critical for the success of any SCM project.  Git is very flexible with respect to workflow.  See http://sethrobertson.github.com/GitBestPractices/#workflow for a list of references about choosing branching and distributed workflows.
14:51 peacememories joined #git
14:51 rmesta joined #git
14:53 davis thanks, i'll take a look.
14:53 al-damiri joined #git
14:53 a_thakur joined #git
14:54 Vampire0 davis, yw
14:54 gitinfo davis: This channel tracks karma based on who has gotten lots of thanks for being helpful. If you want to help someone reach karmic nirvana, please mention their name when thanking them with "thank you", "thankyou", "thanks", "thx", "ty" or "cheers". Try ".karma <nick>" or ".topkarma" to show karma status of a person.
14:55 davis Vampire0, thanks
14:55 h12o joined #git
14:56 bket joined #git
14:57 apax joined #git
14:58 dka joined #git
14:59 skalpin when I 'git checkout <branch> -- *' it stages the files. Can I stop it from doing that? or do I just need to call 'git reset .' if I want to unstage the changes
15:00 thiago no, you can't
15:03 cstratak joined #git
15:04 kpease joined #git
15:06 TbobbyZ joined #git
15:06 cstratak joined #git
15:06 MistahKurtz joined #git
15:09 hashpuppy joined #git
15:09 ayogi joined #git
15:10 Snugglebash joined #git
15:10 lewix joined #git
15:10 dsdeiz joined #git
15:10 dsdeiz joined #git
15:11 rominronin joined #git
15:11 AstralStorm joined #git
15:12 ozcanesen joined #git
15:12 DomeMasterYolo19 joined #git
15:14 h12o joined #git
15:15 AndChat-484449 joined #git
15:15 elect joined #git
15:16 publio joined #git
15:16 dreiss joined #git
15:17 kurkale61 joined #git
15:21 hotpanca_ joined #git
15:21 DomeMasterYolo11 joined #git
15:22 miklcct joined #git
15:22 AstralStorm joined #git
15:22 kurkale6ka joined #git
15:23 kpease joined #git
15:23 ssmall joined #git
15:24 Silox| joined #git
15:24 cagomez joined #git
15:25 jagobdue joined #git
15:26 _ADN_ joined #git
15:26 ResidentBiscuit joined #git
15:28 rudrab joined #git
15:31 apotry joined #git
15:32 User458764 joined #git
15:33 HoierM joined #git
15:35 davidfetter_ge joined #git
15:35 rokups joined #git
15:36 markmarkmark joined #git
15:36 brent__ joined #git
15:37 hotpancakes joined #git
15:37 hotpanca_ joined #git
15:40 nixjdm joined #git
15:40 gableroux joined #git
15:40 lindenle joined #git
15:43 TbobbyZ joined #git
15:43 privacy joined #git
15:43 Gurkenglas joined #git
15:44 hotpancakes joined #git
15:45 hotpanca_ joined #git
15:45 dsdeiz joined #git
15:47 Cabanossi joined #git
15:49 hotpancakes joined #git
15:49 Murii joined #git
15:52 tlaxkit joined #git
15:54 gableroux joined #git
15:56 hotpancakes joined #git
15:56 King_Hual joined #git
15:56 peacememories joined #git
15:56 netj joined #git
15:57 ResidentBiscuit joined #git
15:58 Tobbi joined #git
15:59 svm_invictvs joined #git
16:00 thiago joined #git
16:01 raspado joined #git
16:02 blackwind_123 joined #git
16:03 dsdeiz joined #git
16:04 madewokherd joined #git
16:05 mel00010 joined #git
16:05 chubek_ joined #git
16:07 Vadi joined #git
16:08 s1scha joined #git
16:08 kiichi_ joined #git
16:09 Vadi I'm trying to bisect between latest development and a certain commit in time, and git bisect is just going off on a tangent and selecting random commits way back in history. What am I doing wrong? https://pastebin.com/uLyf6rY6
16:10 Vadi I say that current development is bad, 2d2d72ea87301a4959f5d433c3c586b6feb3f780 is good and was on 2017-04-12 and the first commit it selects is on 2017-04-02 with the next one last year entirely (?!)
16:10 User458764 joined #git
16:10 _ikke_ Vadi: complex history might make it necessary to test older commits
16:11 Vadi okay - I'll try and follow through with its suggestions then
16:12 rominronin joined #git
16:12 rominron_ joined #git
16:14 _ikke_ You might want to try git bisect visualize to see what it still needs to test
16:14 m4sk1n joined #git
16:14 Vadi It's on a commit 3 years ago ...
16:14 Vadi Which isn't helpful because another issue in such ancient commits is preventing me from testing
16:15 Vampire0 Vadi, could be if you have merge commits. If you cant test a commit, tell bisect to skip it
16:15 DomeMasterYolo11 joined #git
16:15 Vampire0 _ikke_, does that do a textual representation or something else?
16:16 _ikke_ Vampire0: By default it uses gitk, and when DISPLAY is not set, it uses git log
16:16 Vadi I'll try skipping, because bad all the way (due to the other issue) didn't help
16:16 _ikke_ Vampire0: (just read it in the man)
16:17 roelmonnens joined #git
16:17 Vampire0 I was sure it will be there, but it was faster to ask you :-D
16:17 Vampire0 lame, I know *g*
16:17 Vampire0 thanks _ikke_
16:18 _ikke_ Vampire0: That was just to indicate that I didn't know that before either
16:19 Vampire0 Vadi, bad because of another issue will confuse the algorithm. bad (or whatever stanza you chose for it) *must* only used for the concrete issues you are investigating. Not testable commits due to other problems must be skipped always
16:19 Vadi Right
16:19 Vampire0 _ikke_, ah, ok, I understood I should read it there, not you read it there. I hate irregular verbs where all times are identical
16:20 DelTree would not be with accented letters...
16:20 DomeMasterYolo11 joined #git
16:20 nitric joined #git
16:21 _ikke_ hehe
16:21 grimTriggers joined #git
16:22 Vampire0 sorry?
16:22 roelmonnens joined #git
16:22 DelTree bah try but... réad vs reàd...
16:22 DelTree bad*
16:23 DelTree the times are *not* identical, they don't sound the same...
16:24 Vampire0 ah, know what you mean, but as we are in a textual conversation, I meant the written form exclusively of course. context, you know? ;-)
16:25 svm_invictvs joined #git
16:26 durham joined #git
16:27 mkillebrew joined #git
16:27 LionsMane joined #git
16:28 Kerwin_chain joined #git
16:28 Nugget joined #git
16:29 tarkus joined #git
16:30 anishathalye joined #git
16:30 mkillebrew forked a repo, go to github.com/user/repo/pull/new/branch, gives a 500 error. did I fuck something up or is there some time require for internal rejiggering after forking and before pulling
16:31 dglambert joined #git
16:31 qbolt joined #git
16:31 _ikke_ mkillebrew: you mean before making a pull request?
16:31 Dougie187 Forking doesn't copy pull requests.
16:32 mkillebrew forked, pushed change, pull request on web interface gives me just a 500 server error
16:33 Dougie187 Did you create a new PR yet?
16:33 Dougie187 Is this a public repo?
16:33 mkillebrew attempting to create a pull gives a 500
16:34 Dougie187 Are you navigating to the PR page by typing in that URL? or are you clicking through github's interface?
16:34 mkillebrew either
16:35 Dougie187 Is this a public repo?
16:35 mkillebrew eh, works now
16:35 boombatower joined #git
16:35 mkillebrew I guess... rejiggering time?
16:35 Dougie187 Glad I could help! :P
16:35 Dougie187 I'll never know.
16:35 unbalancedparen joined #git
16:35 Dougie187 because I still don't know if it's a public repo. :P
16:36 mkillebrew it's not
16:38 sunri5e joined #git
16:41 ertes joined #git
16:41 onehrxn_ joined #git
16:42 roelmonnens joined #git
16:43 SteffanW joined #git
16:44 Introoter joined #git
16:44 bernd9000 joined #git
16:45 Pseudocrat joined #git
16:46 publio I've done a hard reset of master and am now trying to push the old version on top of the new one, but am getting a "current branch is behind" error. What's a fix for this? Will making a new commit help?
16:46 gopar joined #git
16:46 raijin joined #git
16:46 ToBeCloud joined #git
16:47 dreamseeker joined #git
16:47 Dougie187 publio: No. Typically the fix is to force push, but whether that's a good idea or not in your specific case is still a question.
16:49 publio Dougie187: Does it delete the future changes? I branched the new version to another branch. So, I should be able to go back, if needed.
16:49 Dougie187 Well, why'd you do the reset, if you're just going to push it to another branch?
16:50 Dougie187 Are other people working off of the commit you're resetting from?
16:50 publio Dougie187: No. I'm abandoning the new version. I'll continue with the old, but don't want to delete the new version entirely
16:51 Dougie187 Are you the only developer?
16:52 publio Dougie187: Yes. I usually deploy off the master tho
16:52 Dougie187 Well, if you're the only developer, you can probably force push safely.
16:52 ExoUNX joined #git
16:52 Dougie187 User's might be messed up too, but who knows.
16:53 publio Dougie187: Would I have had to revert the individual changes, if there were other devs?
16:53 Gustavo6046 joined #git
16:53 Dougie187 Hard resetting a commit away removes it, and messes up branches in the future. So if anyone ever tries to get a branch merged that was based on the commit you're removing from master, it'll try to make it's way back in.
16:53 venmx joined #git
16:53 Doginal left #git
16:53 Dougie187 You're keeping that commit on a different branch, so it won't be deleted, but you still might have to deal with that case in the future.
16:53 dreamseeker_ joined #git
16:53 Dougie187 (where someone wants to merge in that commit)
16:54 lacrymology joined #git
16:55 publio Dougie187: Yea, never had to go back the chain before, so all the nuances are difficult.. Thanks for the info
16:55 Dougie187 np
16:56 * davidfetter_ge ponders nondeterministic polynomial time complexity
16:58 regedit joined #git
16:59 rwb joined #git
16:59 holodoc joined #git
17:00 hotpanca_ joined #git
17:02 Cabanossi joined #git
17:04 dsdeiz joined #git
17:05 hotpancakes joined #git
17:06 theoceaniscool joined #git
17:06 ter0 joined #git
17:09 bitmod joined #git
17:09 jackrandom joined #git
17:11 m4sk1n joined #git
17:12 CussBot_ joined #git
17:13 cdg joined #git
17:15 bernd27 joined #git
17:15 Es0teric joined #git
17:16 Cassiopaya joined #git
17:17 lacrymology joined #git
17:17 mda1 joined #git
17:18 gableroux joined #git
17:19 unbalancedparen joined #git
17:19 anishathalye left #git
17:19 iamashishkebab2 joined #git
17:19 gableroux joined #git
17:20 tymczenko joined #git
17:20 MattMaker joined #git
17:21 MattMaker joined #git
17:21 kerwincsc joined #git
17:21 freimatz joined #git
17:23 gableroux joined #git
17:24 kerwincsc joined #git
17:25 romank joined #git
17:28 bitmod joined #git
17:28 User458764 joined #git
17:29 hobodave joined #git
17:29 tnecniv joined #git
17:30 hotpancakes joined #git
17:31 gablerou_ joined #git
17:31 thiago joined #git
17:32 dreamseeker joined #git
17:33 ssmall joined #git
17:34 peacememories joined #git
17:34 lewix joined #git
17:34 fission6 joined #git
17:34 gableroux joined #git
17:34 satifant joined #git
17:36 iamashishkebab2 joined #git
17:38 dave0x6d joined #git
17:38 dsdeiz joined #git
17:38 dsdeiz joined #git
17:41 matsaman joined #git
17:41 gableroux joined #git
17:42 matsaman I merged master into a feature branch, and some files have simply disappeared, and what's even stranger is git diff master fbranch doesn't even say they've disappeared
17:42 matsaman it doesn't acknowledge that they are even present in master (which they are), and not present in the feature branch (which they are [not present])
17:42 matsaman s/it doesn't/git diff doesn't/
17:42 apax joined #git
17:43 misosoup joined #git
17:43 raijin joined #git
17:44 rominronin joined #git
17:44 gableroux joined #git
17:44 texasmynsted I have a local branch foo-old that tracks a remote branch foo.  I want to remove the remote branch foo or rename it to foo-old.  How would I do that without harming my local foo-old branch?
17:45 MDA2 joined #git
17:45 zturner joined #git
17:45 texasmynsted oh so to delete
17:45 texasmynsted I could do
17:45 texasmynsted git push origin :foo
17:45 iamasamosa joined #git
17:45 tyreld joined #git
17:45 texasmynsted and then foo-old would figure out that the remote branch does not exist right?
17:46 zturner how can i make git generate a diff of 1 file as if the entire file had just been created?  I want the "left" to be nothing, and the "right" to be the entire file
17:46 Altiare joined #git
17:47 Cabanossi joined #git
17:48 cdg_ joined #git
17:48 relipse joined #git
17:49 _ikke_ matsaman: does git ls-tree master <path/to/file> return something?
17:49 tanuki joined #git
17:49 _ikke_ zturner: git diff --no-index /dev/null <file>
17:49 matsaman _ikke_: yes
17:50 _ikke_ and git ls-tree HEAD <path/to/file>?
17:50 _ikke_ (assuming you are now in your feature branch)
17:51 zturner _ikke_: Thanks!
17:51 matsaman _ikke_: HEAD returns nothing (in feature branch)
17:52 _ikke_ matsaman: git log --diff-filter=D path/to/file
17:53 matsaman _ikke_: single commit from when it was added in the past
17:54 matsaman or, I guess when it was deleted?
17:54 pwnz0r joined #git
17:54 matsaman so that's the explanation?
17:54 vuoto joined #git
17:54 _ikke_ matsaman: It should only list commits that deletes the file
17:54 zturner _ikke_: What if I want multiple files included in the same diff?  listing them one after the other doesn't work
17:54 roelmonnens joined #git
17:55 olbat joined #git
17:55 _ikke_ zturner: just concatenate them
17:55 matsaman _ikke_: hrmmm, so maybe it was erroneously delete in the past...
17:55 _ikke_ matsaman: probably
17:56 matsaman I still don't get it, though
17:56 matsaman if I merge master, with a file that exists, into my feature branch
17:56 matsaman shouldn't it be added, and not sent to oblivion forever?
17:57 matsaman and further: shouldn't git diff master fbranch _show_ that the existence of that file differs between the branches (as it indeed does)?
17:57 _ikke_ matsaman: You are diffing the merge
17:58 matsaman 'git diff branch1 branch2' diffs... a merge?
17:58 texasmynsted did not work.  Hmm. If I want to just rename the remote branch
17:58 * matsaman boggles
17:58 matsaman will man up, bah AFK a few...
17:58 _ikke_ matsaman: you did merge master in, right?
17:58 matsaman yes I did
17:58 _ikke_ so the last commit is a merge?
17:59 matsaman but I thought git diff branch otherbranch did something more like 'diff foo bar', that is diff'ing everything
17:59 matsaman meh, sorry I have a surprise meeting
17:59 matsaman bbl =)
18:00 dsdeiz joined #git
18:02 texasmynsted I think this worked
18:02 texasmynsted git push origin origin/feature/foo:refs/heads/feature/foo-old
18:02 lacrymology I'm looking for a reasonable way to have `git merge` default to `--no-ff` WITHOUT this affecting `git pull`. Is this even possible?
18:02 gableroux joined #git
18:02 Eugene lacrymology - add an alias ;-)
18:03 bitmod joined #git
18:03 Eugene You can't re-define builtin git commands, but try something like: git config --global alias.blend merge --no-ff
18:04 lacrymology Eugene: thing is I wanted this to affect 3rd party plugins as well. I just discovered git-extras which has some very nice tools, but it's not very configurable. It's all scripts and it doesn't take many options
18:05 Eugene There are separate options for merge.ff and pull.ff
18:05 Eugene It even says that pull.ff is meant to override merge.ff
18:08 roelmonnens joined #git
18:08 dreiss joined #git
18:08 _ikke_ right, but you cannot unset it
18:09 _ikke_ you have to choose either false or only
18:10 gaucheph joined #git
18:11 robattila256 joined #git
18:12 elsevero joined #git
18:13 nick123 joined #git
18:13 mohabaks joined #git
18:13 tyoc213 joined #git
18:14 karols146 joined #git
18:14 tyoc213 Hi there, we have a remote SVN repo (from other company) but locally we use git... how can we manage this?
18:14 mkillebrew left #git
18:17 Cabanossi joined #git
18:19 _ikke_ tyoc213: git-svn is probably your best option then
18:19 _ikke_ !svn
18:19 gitinfo [!unlearn] Users of centralized VCSes will need to unlearn a number of habits in order to learn git. Many commands, such as "checkout" and "revert" bear little in common to those holding the same name in other systems.
18:19 _ikke_ sorry, that's not what I meant
18:19 tyoc213 np, searching for git-svn
18:22 ozcanesen joined #git
18:23 onehrxn joined #git
18:24 gableroux joined #git
18:26 johnsnow joined #git
18:26 sea-gull joined #git
18:26 sea-gull_ joined #git
18:26 bitmod joined #git
18:27 johnsnow i had detached head, and already have made some changes
18:27 johnsnow i then did "gi t stash"
18:28 johnsnow "git checkout master" and " git stash pop"
18:28 johnsnow and it is showing unmerged file
18:29 johnsnow and "unable to refresh index"
18:29 johnsnow how do i resolve that ?
18:29 duderonomy joined #git
18:32 cjohnson quick google gives me this: http://stackoverflow.com/questions/9739352/git-stash-pop-needs-merge-unable-to-refresh-index
18:32 cjohnson probably relevant
18:32 peterbecich joined #git
18:32 serialoverflow joined #git
18:33 NeverDie joined #git
18:34 bitmod joined #git
18:34 ejbs joined #git
18:34 Marumarsu joined #git
18:35 dsdeiz joined #git
18:35 mikecmpbll joined #git
18:35 ejbs Hi, is there a way of merging a branch so that I can keep the commits that was made on the other branch and not only have a generic "Merged with branch X" commit message?
18:35 apax joined #git
18:36 gableroux joined #git
18:36 dvaske joined #git
18:36 cjohnson that's what merging does
18:36 cjohnson the merge itself has a generic message but it brings over the other commits too
18:37 Qilibrun joined #git
18:37 gugah_ joined #git
18:38 jnewt joined #git
18:38 cjohnson ejbs: http://imgur.com/a/4DSit
18:38 cjohnson in this case it was a ff so no merge commit needed, but even in non-ff ones
18:40 tnecniv joined #git
18:40 tnecniv joined #git
18:42 serialoverflow joined #git
18:42 gableroux joined #git
18:44 JanC_ joined #git
18:44 Acerific joined #git
18:44 ejbs Sorry I completely forgot to reply cjohnson. Also yeah, that's the behavior I expecte
18:44 ejbs d
18:44 ejbs But it's not the one I get
18:45 navidr joined #git
18:46 ejbs I do git merge origin gsprints and git asks me to give a commit message, maybe that's where I missed something cuz I didn't read and I screwed it up
18:46 gableroux joined #git
18:47 ejbs Huh, and now it worked. Well, I dunno, maybe there's something implicit that I don't understand
18:47 Cabanossi joined #git
18:47 ejbs Thank you cjohnson!
18:47 ejbs left #git
18:47 tyoc joined #git
18:48 iamashishkebab2 joined #git
18:48 robotroll joined #git
18:51 gableroux joined #git
18:51 cjohnson the commit message is just for the merge commit itself
18:52 cjohnson it makes a new commit for non-ff merges
18:52 cjohnson detailing that you had to do some manual  work (even if git does it automatically for you)
18:54 vuoto joined #git
18:56 dfaught joined #git
18:56 peterbecich joined #git
18:57 vuoto joined #git
18:57 ivo_ joined #git
18:57 peterbec` joined #git
18:59 Aoiu joined #git
19:01 dsdeiz joined #git
19:01 dsdeiz joined #git
19:02 jackrandom joined #git
19:04 zealsham_ joined #git
19:05 zealsham_ how do i set up a private git repo for me and my team
19:06 gableroux joined #git
19:07 apax joined #git
19:07 gugah joined #git
19:08 DomeMasterYolo12 joined #git
19:08 Achylles joined #git
19:09 apax joined #git
19:09 Altiare joined #git
19:11 a3Dman joined #git
19:12 holodoc joined #git
19:12 altendky joined #git
19:12 yitzchak joined #git
19:12 sea-gull joined #git
19:14 sondr3 joined #git
19:14 tymczenko joined #git
19:15 rominronin joined #git
19:16 re1 joined #git
19:17 sea-gull_ joined #git
19:18 apax joined #git
19:19 _ikke_ No patience :-(
19:19 serialoverflow joined #git
19:20 barbos joined #git
19:22 caper joined #git
19:22 nonconvergent joined #git
19:23 nonconvergent https://stackoverflow.com/questions/18371741/git-branching-strategy-integated-with-testing-qa-process/18868914#18868914 Is he saying to rebase feature onto devel then push to remote? I don't understand what this guy means by push feature.
19:23 ertes joined #git
19:23 ResidentBiscuit joined #git
19:24 apax joined #git
19:24 raffo joined #git
19:24 netj joined #git
19:25 qbolt so I'm trying to get the raw files from a gist with multiple files
19:25 qbolt https://gist.github.com/jsuwo/9038610
19:25 qbolt that's the gist
19:25 venmx joined #git
19:26 qbolt if I do https://gist.githubusercontent.com/jsuwo/9038610/raw it only returns that last java class
19:26 qbolt Anyone know how I can get all of the files in the gist?
19:26 qbolt in raw form
19:26 sondr3 joined #git
19:26 _ikke_ nonconvergent: That's what he means
19:26 _ikke_ rebasing makes sure that branch has the latest changes from develop
19:28 _ikke_ qbolt: I only see how you can get each file individually raw
19:29 _ikke_ https://gist.githubusercontent.com/jsuwo/9038610/raw/1a13e9fb7ad325ada81bc386aabc69dc9c5b64df/TestBankAccount.java
19:29 qbolt So I'll have to parse the thing for those links?
19:30 _ikke_ qbolt: isnt't it easer to clone the repo and then have the files?
19:30 _ikke_ qbolt: And ask #github also
19:30 qbolt Well, I'm trying to find a way to view gists at work - gist is blocked on my network
19:31 qbolt So I'm sshing into a home server and trying to use curl to view them
19:31 _ikke_ ugh :-/
19:31 nonconvergent _ikke_: I think I'm still confused. Could you describe what the actual workflow would look like, as in the git commands?
19:32 _ikke_ nonconvergent: git rebase develop feature; git push (--force-with-lease) origin feature
19:33 _ikke_ qbolt: perhaps this is giving you an option: https://developer.github.com/v3/gists/
19:33 Sasazuka joined #git
19:33 _ikke_ qbolt: it should give you all the files with the link to the raw gist
19:33 qbolt oooh that looks promising
19:33 qbolt Thanks!
19:34 barbos joined #git
19:34 nonconvergent _ikke_: And git rebase develop feature will take all of the new commits on develop, stack them up underneath all the new commits on feature, and leave the changes on the feature?
19:34 _ikke_ nonconvergent: the other way around
19:35 nonconvergent _ikke_: I'm looking at this strategy because we have merge conflicts GALORE.
19:35 _ikke_ it takes the commits of feature, and reapplies them on top of develop
19:35 _ikke_ nonconvergent: Well, rebasing does not prevent conflicts
19:35 nonconvergent _ikke_: but that leaves develop possibly polluted with an untested feature's code?
19:36 nonconvergent _ikke_: Our problem isn't that conflicts happen, its that EVERYONE seems to have to resolve them as they keep their features up to date.
19:36 rwp nonconvergent, If you are having conflicts because there are two parallel paths of development on the same sections of files then there is no magic tool to avoid conflicts.
19:36 _ikke_ right
19:36 nixjdm joined #git
19:37 nonconvergent Then what's the benefit of rebase over merge?
19:37 _ikke_ nonconvergent: it depends on the workflow (who is merging)
19:37 rwp That is a complicated question to answer in short one line interaction.
19:38 _ikke_ if the developer is doing the merging theirselves, it doesn't help much (except perhaps only having to solve conflicts in steps, instead of all at the same time)
19:38 rwp But as far as conflicts go... I don't think there is any significant difference. The difference is elsewhere unrelated to conflicts.
19:38 bernd27 joined #git
19:39 rwp If you are doing simultaneous development on multiple branches in parallel then you are creating conflicts which will be seen when doing either a merge or a rebase.
19:39 rwp If you want to avoid conflicts then you will need to refactor your problem such that this will be avoided.
19:40 jackrandom joined #git
19:40 soerenb joined #git
19:40 soerenb !dots
19:40 gitinfo A..B = stuff that happened between A and B (if A and B are related; otherwise refer to "man gitrevisions"), A...B = (a) in history viewers: stuff that isn't in both A and B yet; (b) in "git diff": stuff that happened in B since the two diverged; (c) in "git checkout": the merge base of A and B. "master.." is the same as "master..HEAD" and "..master" is the same as "HEAD..master", and so forth.
19:43 druonysus joined #git
19:43 matsaman it used to be if I :q'd from vimdiff (via git mergetool), it'd ask me if the merge was successful, but now it does nothing and leaves crufty file.orig files
19:44 a3Dman joined #git
19:44 zealsham_ joined #git
19:44 _ikke_ those two are unrelated
19:45 bernd27 joined #git
19:45 _ikke_ matsaman: git config (--global) mergetool.keepBackup false
19:46 nonconvergent So we have problems with branches, testing, and merges. We work on feature branches, were merging to develop, fixing on develop, deciding to only ship a few things, then got into a revert/cherry-pick hell.
19:46 matsaman _ikke_: that's a changed default in more recent git versions?
19:46 matsaman like, how are people meant to deal with aborting mergetool stuff if not that way?
19:47 nonconvergent Meanwhile QA is testing on develop the whole time.
19:47 tgummerer joined #git
19:47 Cabanossi joined #git
19:49 nonconvergent _ikke_: OH
19:49 nonconvergent _ikke_: OHHHHHHHHHHH
19:50 Wcuju joined #git
19:50 MistahKurtz joined #git
19:50 matsaman I did update git from 2.9.2 to 2.12.2, but I didn't alter my configuration
19:51 nonconvergent _ikke_: I think I see what was confusing me. You "git rebase develop feature;" Your local copy of develop now has the feature on top. "git push (--force-with-lease) origin feature" You push your local develop to origin/feature (does force-with-lease clobber that branch if you were preserving changes there before rebasing?)
19:52 ozcanesen I recntly removed git submodule from .gitmodules file but when run git submodule it shows me `fatal: no submodule mapping found in .gitmodules for path 'lib/simplemde-markdown-editor'`
19:52 ozcanesen i also deleted that directory
19:52 _ikke_ matsaman: no, has been like that for quite some time
19:53 Sasazuka joined #git
19:53 ozcanesen but i think i should also do git rm :9
19:53 NeXTSUN joined #git
19:53 zeroed joined #git
19:54 KP580BM1 joined #git
19:55 _ikke_ ozcanesen: yes, git rm --cached <submodule_path>
19:56 jegaxd26 joined #git
19:56 cdown joined #git
19:57 nonconvergent _ikke_: Then QA finds a problem and deletes origin/feature, developer cuts a new feature from their local develop, fixes it, rebases develop feature again, pushes one more time. But what do you do when you're ready to ship a few feature branches? cut a release branch from develop, merge the features in, and push it out your CI pipeline?
19:58 apax joined #git
19:59 cdg joined #git
19:59 pietv joined #git
20:00 bitmod joined #git
20:01 ash_workz joined #git
20:01 _ikke_ Why the whole deleting / recreating branches?
20:02 nonconvergent Also, if devs are locally polluting their develop branches with history altering rebases, what if someone accidentally pushes to origin?
20:02 nonconvergent _ikke_: Because that's what the SO post I linked to said
20:02 dsdeiz joined #git
20:02 nonconvergent I'm trying to come up with a flow to ease some of our pain.
20:03 matsaman _ikke_: quite some time including before 2.9.2? Because I really haven't changed my config; mostly though I don't really understand how one is supposed to deal with merge conflicts with the keepbackups default (of making .orig files and not allowing you to re-run mergetool to re-address the conflict)
20:04 jnewt joined #git
20:04 matsaman used to be if I had to interrupt mergetool, I could quit vimdiff, it'd ask me if it was successful, I'd say no, and no I don't want to continue; then at any time I could re-run mergetool to finish
20:04 matsaman now I don't know what's meant to be done
20:05 _ikke_ matsaman: you mean with :cq?
20:05 matsaman I've just used :qa! for as long as I can remember
20:06 a3Dman joined #git
20:06 a3Dman joined #git
20:06 qqx There was a somewhat recent change to mergetool's vim driver to respect the exit code.
20:06 matsaman so I need to start using 'c'?
20:06 theoceaniscool joined #git
20:06 * matsaman tests
20:06 qqx matsaman: :cq
20:07 qqx If you want to indicate that the merge wasn't finished.
20:07 matsaman mmm, that's annoying, I don't use 'c' in Vim anywhere
20:07 qqx IIRC you can configure mergetool to not honor the exit code as it used to.
20:07 nixjdm joined #git
20:08 qqx mergetool.<tool>.trustExitCode
20:08 stpa5 joined #git
20:09 matsaman and it doesn't allow me to even switch from :qa! to :cqa!
20:09 _ikke_ but then you have to always confirm, right?
20:09 karols146 joined #git
20:09 _ikke_ matsaman: just :cq
20:09 matsaman think I'll just downgrade git for now, I don't have time for this kind of obnoxious thing today
20:09 _ikke_ matsaman: that quits vim completely
20:10 matsaman yeah I see that :cq works similarly to how :qa! did
20:10 matsaman I just don't use :cq ever in Vim, so it's foreign and annoying
20:11 _ikke_ (always convenion to have a repo lying around that I can easily create conflicts with :P)
20:12 matsaman compiler quit? <eyeroll>
20:14 matsaman like, why wouldn't it trust the exit code? <grumbles>
20:14 * matsaman hates finding problems when he's trying to get around some other problem
20:14 _ikke_ It will trust the exitcode
20:15 _ikke_ and with qa!, the exitcode is success
20:16 matsaman so... what was wrong with the old way?
20:16 _ikke_ matsaman: hmm, something might have changed
20:16 matsaman because it was already detecting if I saved it normally or force quit
20:16 matsaman and now it's just always assuming any quit is success
20:16 matsaman ...or something
20:17 qqx Before it was assuming that if you wrote the file that it was success.
20:17 _ikke_ matsaman: https://github.com/git/git/commit/b3dac9c07858e7b8b9c99a371c67129c2e7909e3
20:17 nonconvergent https://stackoverflow.com/questions/18371741/git-branching-strategy-integated-with-testing-qa-process/18868914#18868914 So is this a good strategy?
20:17 _ikke_ So what changed is that trustExitCode is now applied to built-in tools
20:17 matsaman qqx: and now it's doing what?
20:17 _ikke_ 2.11.1 changed that
20:18 qqx It now has a way to allow saving after handling part of the merge, but still indicate that the merge was a failure.
20:18 _ikke_ matsaman: it looks at the exit code of the editor
20:19 matsaman so one might say Vim is broken for not sending failure with a force quit?
20:19 rmesta joined #git
20:20 miczac joined #git
20:20 qqx No, force quit doesn't imply that there was a failure.
20:20 mostlybadfly joined #git
20:20 qqx That's specifically what :cq is for.
20:20 brent__ joined #git
20:21 qqx Also, as I said before. You can configure the current git mergetool to ignore the exit status, just as it used to.
20:21 qqx Which is also mentioned in the log message of the commit which _ikke_ referenced.
20:21 brent__ joined #git
20:22 _ikke_ force quit only tells vim to ignore unsaved changes
20:22 matsaman yeah but I hate that, I don't want to rely on highly configured dot files
20:22 matsaman okay thanks
20:22 matsaman in my mind I will file this under: Vim is stupid (even if usually less stupid than most things)
20:22 _ikke_ You don't need highly configured dotfiles
20:23 _ikke_ just learn that :cq makes vim return an error code, and you're set
20:23 matsaman I want to know how a tool is behaving by default, I don't want to need my configs
20:23 matsaman yes, that is learnt
20:23 onomatopiea joined #git
20:23 matsaman I can definitely see the argument for making git change to require :cq over :qa!, for example, to get a 1
20:24 _ikke_ :qa! does not return an error code
20:24 _ikke_ it returns success
20:25 Hello71 joined #git
20:25 _ikke_ The only thing that has changed is that git looks at the exit code, and if it returns sucess, then it's done
20:25 m0viefreak joined #git
20:26 _ikke_ So that if the conflict was resolved correctly already by the mergetool, it can signal that
20:26 DelTree what's the a for ?
20:26 _ikke_ all
20:26 DelTree oh...
20:26 DelTree right...
20:26 DelTree rarely edit several files with vim... ^_^
20:27 finalbeta joined #git
20:27 matsaman yeah that makes sense in general, for sure
20:28 _ikke_ And this is already in git for quite some time
20:28 matsaman that git mergetool should "just use the exit code", that is
20:28 matsaman that Vim's :q! doesn't send 1 I'm not so sure =P
20:28 matsaman but not this channel's problem, heh
20:29 _ikke_ matsaman: because vim has :cq :P
20:29 DelTree the opposite... :cq exists because :q! fails to fail...
20:30 _ikke_ I would not expect it to
20:30 _ikke_ :q! says to me: I changed my mind, don't store any changes
20:31 rnsanchez joined #git
20:31 qqx :cq and :q! are completely orthogonal. There are times that I want vim to exit with an error even if there aren't unwritten changes.
20:31 DelTree :q! says to me: I'm lost, get me out and leave my file alone...
20:32 mquin joined #git
20:32 _ikke_ qqx: right, I agree
20:32 DelTree does :cq exist in vi or is it some dark part of the m ?
20:32 _ikke_ for example, to abandon a commit in the editor
20:33 _ikke_ it's apparently vim only
20:33 _ikke_ when running vi, it does not accept it
20:33 DelTree ok... that's why it's bad...
20:34 relipse joined #git
20:34 DelTree new thing to make everyone messed up...
20:34 maroloccio joined #git
20:36 DelTree so... for the recap... the git thing to never do is...... mergetool, right ?
20:36 dsdeiz joined #git
20:36 dsdeiz joined #git
20:36 matsaman heheh
20:37 _ikke_ DelTree: It's not for nothing why it's called vi Improved :P
20:37 matsaman I'm not even sure what it does if you're not using vimdiff
20:38 DelTree _ikke_: having the arrows behave somewhat cleanly is an improvement... :cq is a sad headache-making complexification...
20:38 _ikke_ why?
20:38 _ikke_ you are not forced to use it
20:39 Gurkenglas joined #git
20:39 _ikke_ it's just setting an exit code when it exits, nothing complicated
20:39 matsaman how does vi handle exit codes, then?
20:39 matsaman doesn't?
20:39 DelTree seems in git mergetools I would be...
20:39 _ikke_ doesn't indeed
20:39 rmesta joined #git
20:39 matsaman mmm
20:39 matsaman so back to git for a moment =P
20:39 _ikke_ DelTree: You can always set trustExitCode to false
20:39 matsaman why isn't 'git merge master' putting a file that's in master into this branch?
20:40 DelTree I can set "git mergetool" to no-go...
20:41 mquin joined #git
20:44 matsaman it's an empty file, but there are other empty files that exist
20:44 matsaman and are present
20:45 apax joined #git
20:46 rominronin joined #git
20:46 DelTree _ikke_: how does git react if I kill -9 the poor vim ?
20:47 _ikke_ DelTree: try it
20:47 DelTree I won't
20:47 Hez joined #git
20:48 Es0teric joined #git
20:49 _ikke_ DelTree: Git asks to continue
20:50 DelTree if I ever get in the situation not knowing how git will react to my vim action... I'll kill git first, then vim...
20:50 matsaman my merge commit also shows no changes whatsoever in git log --stat/-p
20:52 Hez joined #git
20:52 jaziz joined #git
20:52 _ADN_ joined #git
20:53 aax joined #git
20:54 Hez joined #git
20:55 gopar joined #git
20:55 Bennux joined #git
20:57 matsaman why would git merge branch ever not copy a file into the new branch?
20:57 peterbecich joined #git
20:57 stemid joined #git
20:58 stemid left #git
20:58 skalpin it could be that the merge-base is a commit after the commit where the file was created
20:59 matsaman okay let's say it's something like that for a moment
20:59 matsaman what then is the command for 'just copy the files please'?
20:59 matsaman manual checkouts?
20:59 skalpin I had someone on my team who would merge --no-commit --no-ff, and remove a bunch of changes, then when we tried to merge things back to develop other peoples changes would disappear
20:59 skalpin checkout -- <filepath>
20:59 matsaman god
20:59 matsaman and what would be the command to generate all those filepaths/
20:59 skalpin git checkout <branchnam> -- *
20:59 matsaman s@/@?
21:00 skalpin that would pull all the changes into your working directory
21:00 oskarkv joined #git
21:00 matsaman but will that produce conflicts for me to manually merge, or just overwrite?
21:00 anuxivm joined #git
21:01 skalpin it will stage anything that's different and make it ready for commit
21:01 askb joined #git
21:01 skalpin so you can easily undo it with a 'git reset --hard' (I'm assuming your working directory is clean)
21:01 venmx joined #git
21:02 matsaman so why is there merge if there's checkout foo -- *?
21:03 dsdeiz joined #git
21:03 Dro joined #git
21:04 Dro in phpstorm "compare with branch" is not clickable.. anyone have an idea how to enable it?
21:04 skalpin matsaman: merge takes your history too
21:04 cdown joined #git
21:04 _ikke_ Dro: No idea
21:04 matsaman skalpin: 'too'? But it doesn't take the actual files
21:05 matsaman the history is nice, but the files are paramount
21:05 matsaman I need the files, I need to merge the file from one branch to another
21:05 MTecknology joined #git
21:05 skalpin lol true, it should take all your files, but a merge works by walking the tree which can get strange sometimes
21:06 * matsaman kills self =P
21:06 skalpin lol
21:06 skalpin did the checkout work for you?
21:06 matsaman let me try it
21:06 matsaman the command itself kind of makes me want to die
21:06 matsaman also, it's 100% not something I can tell my GUI-using coworkers how to utilize
21:07 matsaman like, 100000%
21:07 skalpin I get frustrated with the gui guys too
21:07 skalpin which one? sourcetree?
21:07 matsaman I definitely get frustrated with them, but I also think git should just merge the #$@#$%ing file
21:07 matsaman oh we have all kinds, I don't even keep track
21:07 matsaman PyCharm for one, though, definitely
21:07 Dro I tried to do like this http://stackoverflow.com/questions/34008599/how-to-compare-local-with-remote-git-repo-in-phpstorm
21:08 _ikke_ matsaman: Usually there is a reason git is not merging the file. For example the file being deleted in one branch
21:08 nixjdm joined #git
21:08 Dro Fetch working.. but "compare with branch" is not enabled
21:08 Dro :|
21:08 _ikke_ matsaman: if there is no such thingk it would be considered a bug
21:08 matsaman _ikke_: like, deleted in the past?
21:08 _ikke_ matsaman: deleted in the branch you are merging into
21:08 matsaman Dro: you using git svn?
21:09 Dro matsaman, no its git I guess
21:09 matsaman _ikke_: like delete a file in branch B, merge A into B where A has the file, and git will not add the file back to B? That's "correct" to git?
21:09 _ikke_ correct
21:09 skalpin yep
21:09 matsaman ...
21:09 matsaman but...
21:09 _ikke_ merging is not copy files from branch A to branch B
21:10 matsaman what is?
21:10 skalpin checkout
21:10 matsaman ...
21:10 _ikke_ git checkout master .
21:10 matsaman are you aware of any GUI that has a facility to checkout everything from one branch to another?
21:11 gopar joined #git
21:11 xco joined #git
21:11 _ikke_ matsaman: why are you branching if you want the same files in both branches?
21:12 xco hi. i have say 4 commits, with messages “one”, “two”, “three”, “four” with commit “four” being the latest. is there a way i can go to say commit “two” reset some files while keep the rest of the commits intact?
21:12 _ikke_ if you delete a file and commit that in a branch, the merge resolution is that that file is going to be deleted
21:12 _Vi joined #git
21:12 matsaman I don't want identical branches, I simply want extra and modified files from branch A to exist or be merged (combined manually) into branch B
21:12 _ikke_ if you modified the file in another branch, it would result in a conflict
21:12 apax joined #git
21:12 matsaman I'm not getting a conflict, it simply is not added to the branch
21:13 _ikke_ matsaman: if you *add* a file in A, then that would get merged in. but if A and B had a file, and you removed it from B, then the merge is not getting it back
21:13 matsaman xco: without making another commit (five)?
21:13 matsaman _ikke_: what if it was removed from both and then re-added to A?
21:13 acetakwas joined #git
21:14 xco matsaman: i can’t make another commit “five” doesn’t matter. most important is make some changes in previous commits and keep the rest
21:14 _ikke_ xco: you'd need to use rebase --interactive
21:14 matsaman xco: well, you could certainly make new branches based off of each past commit, then re-merge them all in order
21:15 Eugene matsaman - what you want is `git checkout <branch> -- path/to/file.txt`. I don't know what the equivalent operation would be in any git GUI off the top of my head; I would suspect that they don't really support it
21:15 xco matsaman: can’t i avoid extra branches?
21:15 matsaman but those'd be new commits, technically, I guess
21:15 matsaman Eugene: but won't that overwrite files that already exist in the currect branch, without prompting to merge them?
21:15 Eugene What I can suggest is to copy/paste the file to another (temporary) directory; switch branches in your GUI, then copy/paste them back in where needed
21:15 Eugene Yup, sure does.
21:15 justin2 joined #git
21:15 uther joined #git
21:15 uther left #git
21:15 Eugene If you want to merge differences between branches you should use merge
21:15 Dro matsaman, btw I used "git clone" to get the project
21:16 Eugene Or a third-party patch utility ;-)
21:16 matsaman Eugene: that is also my understanding
21:16 _ikke_ matsaman: git checkout is not merging files
21:16 zeroed joined #git
21:16 zeroed joined #git
21:16 matsaman _ikke_: that's my understanding as well, but people are suggesting it will help me
21:16 matsaman I'm not yet seeing it
21:16 a_thakur joined #git
21:16 frankpf joined #git
21:16 Dro matsaman, compare with branch appear when I used "pull" in other folder with the same project
21:16 _ikke_ matsaman: it's due to the way you are describing what you want
21:16 Dro wondering if the problem is that I used "git clone" instead of pull
21:17 matsaman Dro: as opposed to what, 'fetch' alone?
21:17 Cabanossi joined #git
21:17 matsaman fetch & pull do different things, for sure
21:17 _ikke_ Dro: You always need to use git clone the first time to get the repository
21:17 _ikke_ pull = fetch + merge
21:18 matsaman _ikke_: I don't see how you can say that without knowing what I want, so what was the solution again?
21:18 f0rks joined #git
21:18 Dro matsaman, yes but wondering why "compare with branch" don't appear a project, and appear for another project (the same git rep)
21:18 perlpilot joined #git
21:18 Dro what are the necessary steps to link a project with git correctly? except git clone and git remote add ?
21:18 frankpf is there any way to make interactive rebase preserve unchanged commits?
21:19 matsaman Dro: if you did a fetch on one and a pull on the other, then that's probably why
21:20 bernd27 joined #git
21:20 _ikke_ matsaman: A normal merge should aready do what you want
21:20 matsaman Dro: you'll have to read up on whether you want fetch/merge or pull in any given situation
21:20 matsaman Dro: but a fetch is usually followed by a merge at some point
21:20 _ikke_ if it doesn't, then I would look at the history to see why
21:20 matsaman _ikke_: I agree, it should, but it doesn't
21:20 _ikke_ Then verify your history
21:20 matsaman I don't really care why, though, if I can get a command to just make it do what I want
21:20 a3Dman joined #git
21:20 matsaman you don't know of any such command?
21:21 frankpf https://gist.github.com/anonymous/3e438223928aafdffe13df4b75fd03f9
21:21 matsaman short of a manual diff of both hierarchies
21:21 _ikke_ matsaman: No, there is no such command
21:21 matsaman okay thanks
21:21 * matsaman ponders
21:21 frankpf how can I tell git to not modify the root commit in this case?
21:21 Dro so: git branch give no output for a folder, and give * master for the other folder
21:21 Dro maybe thats why "compare with branch" don't work for the first folder
21:22 Dro but how can I correct this?
21:22 Hello71 joined #git
21:22 qqx frankpf: It should automatically leavel alone any initial commits which are being "pick"ed in the same order.
21:22 matsaman Dro: no output?
21:22 qqx s/leavel/leave/
21:22 Dro matsaman, yes
21:23 Dro "git branch" give no result at all for the first folder
21:23 matsaman Dro: that's pretty weird... you might have like a terrible alias or wrapper you're not aware of
21:23 frankpf qqx: if I use the same pick orders as in the gist, it changes the root commit hash
21:23 Dro matsaman, maybe because I copied files after "git clone".. to another directory
21:23 matsaman but chances are knowing why it's doing that will only satisfy your curiosity, and not actually help you move forward =)
21:23 matsaman Dro: what is it you actually want to accomplish?
21:24 Atm0spher1c joined #git
21:24 Dro matsaman, I want to make that folder correctly configured so that "git branch" show master
21:24 bilal joined #git
21:24 Dro and "compare with branch" will surely work than
21:27 matsaman sounds like what you want, in the end, is to make compare with branch work
21:27 qqx frankpf: Yeah, I can reproduce that. I'd say that's a bug in the handling of root commits.
21:27 matsaman unfortunately there is no non-magical reason why 'git branch' would produce literally no output
21:28 matsaman so you probably have significantly other things at play
21:28 matsaman Dro: is the copy that works just so you could test if the other copy was strange?
21:28 Dro Dro, what I have done exactly :
21:29 Dro git clone git_url  <== its created a folder x
21:29 Dro i copied x/* to my current folder and deleted x
21:29 cdown joined #git
21:29 Dro than I used git remote add origin git_url
21:30 saml_ joined #git
21:30 foreal joined #git
21:30 matsaman Dro: oh okay
21:30 Dro so maybe moving folder erased all the configuration
21:30 saml_ i have few files in feature branch.  how do I move some of those to master branch and remove from the feature?
21:30 matsaman Dro: you probably should've not used /*
21:30 iamashishkebab2 joined #git
21:30 saml_ nothing is pushed yet (master or feature). it's a new repo
21:31 Dro matsaman, so no solution for this case? I have to git clone again?
21:32 matsaman Dro: you can probably just copy the .git dir from one to the other
21:33 Dro matsaman, that what I was thinking about but.. maybe it will erase all local git history
21:34 Dro or config..
21:34 matsaman Dro: * is tricky, it's actually the shell that evaluates that
21:34 robattila256 joined #git
21:34 frankpf qqx: yeah, seems like a bug. It only happens with git rebase -i --root
21:35 matsaman usually it's more reliable to do something like: rsync -a ./path/to/dir/ ./new/path/
21:35 pandeiro joined #git
21:35 matsaman Dro: cp -nr ./path/to/.git ./new/path/ will probably work safely, though, if you want to do just that
21:35 Karazhan joined #git
21:35 saml_ git checkout feature -- file1
21:35 matsaman Dro: rsync -av --ignore-existing --dry-run ./path/to/dir/ ./new/path/ will probably also be useful (with & without --dry-run, if you like the output)
21:36 Dro matsaman, ok thank you, i'll try it!
21:36 matsaman Dro: that doesn't really explain 'git branch' producing no output, it should give you an error if there is no .git dir, but my guess is still that you failed to copy the .git dir
21:36 matsaman Dro: if you can afford to do it all again, I would just rsync -a ./path/to/original/ ./path/to/new/place/
21:37 matsaman rsync used in that fashion is very, very reliable
21:37 Dro matsaman, I just checked the .git/config of both folders.. its seems that branch configuration is missing in one of them
21:37 dsdeiz joined #git
21:38 Dro but maybe its not enough to copy the configuration from .git/config ..
21:38 Karazhan joined #git
21:38 matsaman cp -r ./original ./new would also have worked (vs what you did: ./original/* ./new/), but rsync is still better than cp for large copies like this
21:38 prg3 joined #git
21:39 Dro thats ok
21:39 Dro strange how git don't have a command to automatically fix missing configurations/files :P
21:40 matsaman well, that would fail about 99 out of 100 times, so =P
21:41 a3Dman joined #git
21:42 nick123 joined #git
21:43 rwb joined #git
21:46 gugah joined #git
21:47 Cabanossi joined #git
21:48 gnixev joined #git
21:49 texasmynsted I search the web a little but did not find a good answer.  I have an old branch that I need to revisit.  The main branch has changed dramatically.  I created a new branch off of the master head.  Now there are a number of old changes on the old branch that will conflict all over the place.
21:50 texasmynsted I was thinking about rebasing and squashing most of the commits I need from the old branch then trying to cherry-pick them in and resolve all the conflicts.
21:50 texasmynsted Is this a good idea or should I just try to merge the old branch into the new and hope for the best?
21:53 oprypin texasmynsted, no matter what thing you pick (merge, rebase, cherry pick),  you're gonna get about the same merge conflicts
21:54 Dougie187 left #git
21:54 oprypin cherry picking in my mind is just a generalization of rebase,  but a normal rebase seems to fit the case just fine.  and it's what i find the least confusing.
21:54 Karazhan joined #git
21:54 romank joined #git
21:56 preyalone joined #git
21:58 malide joined #git
21:58 barbos joined #git
21:58 texasmynsted Ok.  I guess I could interactive rebase and not bring in some changes that I "do not need" from the old branch
21:59 texasmynsted That might limit some of the breakage
22:00 dowao joined #git
22:00 texasmynsted thank ou
22:01 texasmynsted or you rather
22:02 gugah joined #git
22:04 ertes joined #git
22:05 YuGiOhJCJ joined #git
22:07 nixjdm joined #git
22:07 malide joined #git
22:08 cdown joined #git
22:09 diogenese joined #git
22:10 Sound joined #git
22:10 Sound joined #git
22:11 neilthereildeil joined #git
22:12 neilthereildeil theres a remote repo wit multiple branches. how can i clone the remote repo AND all its branches?
22:12 fscala joined #git
22:12 texasmynsted git clone
22:13 texasmynsted Are you missing branches?
22:13 texasmynsted git branch -av
22:13 texasmynsted from the newly cloned repo and see what you got
22:13 dave0x6d joined #git
22:13 neilthereildeil i see the branches in red under refs/remotes/origin/BRANCH_NAME
22:14 Levex joined #git
22:14 texasmynsted ok
22:14 texasmynsted Sounds like you have them
22:14 texasmynsted say you have a remote named foo
22:15 Tobbi joined #git
22:15 texasmynsted just check it out and you should get a local branch that tracks that remote.
22:15 neilthereildeil so i just checked out a branch, and it looks like it created the branch locally, because branch -av not shows that branchname without remotes/origin in front of the name
22:15 neilthereildeil is that what happneded?
22:16 texasmynsted I am sure there is a git command for this but...
22:16 texasmynsted if you cat .git/config
22:16 neilthereildeil ie i had remotes/origin/branchName. then i did git checkout branchName.
22:17 texasmynsted you should see your local branches and what they trrack
22:17 neilthereildeil now git branch -av shows branchName AND remotes/origin/branchName
22:17 neilthereildeil whats going on?
22:17 rominronin joined #git
22:17 texasmynsted There is a branch named branchName in the remote repo
22:18 texasmynsted and a local branch named branchName
22:18 texasmynsted The local branchName tracks the remote branchName
22:18 raijin joined #git
22:18 neilthereildeil why 2? i just want a copy of whatever exists remotely
22:18 texasmynsted you can push your local changes to the remote
22:19 texasmynsted right now you do have a copy of remote
22:19 neilthereildeil no, right now i have an extra branch with the same name
22:19 texasmynsted but as soon as you alter your local, or somebody alters the remote, the two will no longer be a match
22:19 aidalgol joined #git
22:19 texasmynsted cat .git/config
22:19 texasmynsted show me
22:21 cdown_ joined #git
22:21 TomyLobo2 joined #git
22:21 texasmynsted just paste the result to a pastebin or gist or something
22:21 neilthereildeil https://pastebin.com/QWsTnpYr
22:22 waltervargas[m] joined #git
22:22 texasmynsted what is the extra?
22:23 rwb1 joined #git
22:23 texasmynsted Looks like you have a local branch named Dash, and it tracks a remote branch named Dash
22:23 krainboltgreene_ joined #git
22:23 neilthereildeil git branch -av shows Dash and remotes/origin/Dash
22:23 altendky_ joined #git
22:23 preyalone_ joined #git
22:23 texasmynsted your remote, in this case is named "origin"
22:23 texasmynsted yes
22:23 kenansul- joined #git
22:24 ams___ joined #git
22:24 texasmynsted two branches
22:24 neilthereildeil how do you know its tracking?
22:24 texasmynsted 1. the branch named Dash that lives in the remote
22:24 al-damiri_ joined #git
22:24 ketas- joined #git
22:24 kexmex joined #git
22:24 texasmynsted 2. the local branch that lives in your local repo, also named Dash.
22:24 pifon_ joined #git
22:24 neilthereildeil left #git
22:25 catern hey #git, I have a repository A, and a folder containing the almost the same code with a few difference in a folder B (extracted from a tarball)
22:25 neilthereildeil joined #git
22:25 neilthereildeil where do u see the remote Dash branch?
22:25 catern how can I get the diff between them?
22:25 HurricaneHarry_ joined #git
22:25 preaction catern: diff -ru
22:25 DistantS1ar joined #git
22:26 TbobbyZ_ joined #git
22:26 cite_ joined #git
22:26 texasmynsted well you have tracking defined in the git config to show which remote it tracks and the name of the remote branch
22:26 QshelTier joined #git
22:26 mathu_ joined #git
22:26 immae_ joined #git
22:26 TSS3000RO joined #git
22:26 Chinorro_ joined #git
22:26 stfn1 joined #git
22:26 pulec_ joined #git
22:26 dikiaap1 joined #git
22:26 ablackack_ joined #git
22:26 neilthereildeil texasmynsted: where do u see that?
22:26 neilthereildeil which lines?
22:26 Hobbyboy|BNC joined #git
22:26 kba_ joined #git
22:26 texasmynsted 12
22:26 Riviera__ joined #git
22:26 texasmynsted [branch "Dash"]
22:27 neilthereildeil ok so i see a local branch called Dash
22:27 texasmynsted So you have a local branch name Dash
22:27 catern preaction: sorry, an additional condition: I want to ignore files ignored by .gitignore
22:27 mkoskar_ joined #git
22:27 texasmynsted it tracks a remote repo
22:27 canton7 joined #git
22:27 barteks2- joined #git
22:27 neilthereildeil now where do u see it tracking a remote branch called Dash?
22:27 AndreasLutro joined #git
22:27 mani2 joined #git
22:27 j416_ joined #git
22:27 roflincopter joined #git
22:27 Takumo joined #git
22:27 Takumo joined #git
22:27 cmn_ joined #git
22:27 texasmynsted the remote that has its own section in that file, in this case named for you as "origin"
22:28 Calinou joined #git
22:28 Levex joined #git
22:28 tgummerer joined #git
22:28 heftig1 joined #git
22:28 zifnab| joined #git
22:28 neilthereildeil ok. and because of #13, we know there must be a remote branch with the same name on line #12?
22:28 kyledeal joined #git
22:28 texasmynsted and the remote branch that your Dash tracks is also named Dash
22:28 lightstalker joined #git
22:29 Jonuz_ joined #git
22:29 ssmall joined #git
22:29 preaction catern: use git ls-files to list the files and then pipe to xargs diff to get a diff of them, is one option. you could init folder B as a repo, add it as a remote for repo A, fetch and then "git diff" will work on it, is another option
22:29 texasmynsted I have likely done a bad thing by describing this from the config file
22:29 mathu_ left #git
22:29 Sonderblade joined #git
22:29 Learath2 joined #git
22:29 texasmynsted line 13 shows which remote repo that your local Dash, tracks.
22:29 neilthereildeil texasmynsted: so a local branch is not created until i checkout, the name of the remote branch right?
22:30 texasmynsted You could have another remote, and a different local branch that points to a different remote
22:30 blackwind_123 joined #git
22:30 texasmynsted you can create a local branch that does not point to any remote.
22:30 qlkzy joined #git
22:30 Nothing4You joined #git
22:31 Dev0n joined #git
22:31 bilal joined #git
22:31 texasmynsted But it is convenient to create a local branch to track a remote and have them both named with the same identifier
22:31 neilthereildeil texasmynsted: ok thanks
22:31 levex_ joined #git
22:31 ahf joined #git
22:32 NeXTSUN joined #git
22:32 Imut joined #git
22:32 neilthereildeil left #git
22:32 demize joined #git
22:35 catern preaction: I see. but also: I get patch detection failed from trying to git apply the diff -ru
22:36 preaction since you're doing diff -r in the parent dir, you likely need patch -p1
22:36 pulec_ left #git
22:36 frankpf joined #git
22:36 catern also won't that delete my .git directory?
22:38 thiago joined #git
22:39 raijin joined #git
22:39 mquin joined #git
22:40 rmesta joined #git
22:41 rmesta joined #git
22:41 cagomez joined #git
22:42 Levex joined #git
22:42 rmesta joined #git
22:42 nick123 joined #git
22:43 rmesta joined #git
22:44 rmesta joined #git
22:46 zealsham_ joined #git
22:46 dxtr joined #git
22:46 mquin joined #git
22:48 gopar joined #git
22:49 madewokherd joined #git
22:49 infernix joined #git
22:49 TbobbyZ joined #git
22:50 ljc joined #git
22:52 _main_ joined #git
22:53 __main__ joined #git
22:53 Levex joined #git
22:54 jab416171 joined #git
22:54 zealsham_ joined #git
22:54 zealsham_ how do i set up a git server on ubuntu for me and my team
22:55 sondr3 joined #git
22:55 __main__ joined #git
22:55 ssmall joined #git
22:55 Hink joined #git
22:57 serialoverflow joined #git
22:58 Eao joined #git
23:03 deltab zealsham_: you can have a machine everyone can ssh into, or you can install extra software such as gogs
23:04 cseder joined #git
23:04 texasmynsted depends on what you and your team need
23:05 acln joined #git
23:05 zealsham_ deltab: whats gogs
23:05 texasmynsted some people use things like gitlab
23:05 zealsham_ texasmynsted:  i just need it to host a repo will could commit to
23:06 zealsham_ texasmynsted:  gitlab is also like github right ? they dont let me host my server , they kind of host it for mezzzz/
23:06 texasmynsted I am not promoting gitlab, but https://about.gitlab.com/installation/
23:07 nixjdm joined #git
23:07 romank joined #git
23:07 texasmynsted but what deltab said
23:07 zealsham_ i dont know what gogs is
23:07 texasmynsted If it were me I might install docker on some server someplace, and through up a "git server" docker image on the box
23:08 texasmynsted Did you try googling it?
23:08 zealsham_ yeah i tried, and the commandline options i saw was scary
23:09 texasmynsted https://gogs.io/
23:10 texasmynsted oh ok
23:10 zealsham_ any easier means
23:11 texasmynsted Sorry but you could either setup a box and grant people ssh access to it and setup git, or you could pay somebody to host it for you.
23:12 Hink joined #git
23:13 dandaman joined #git
23:13 HedgeMage zealsham_: if command line work scares you, why are you in charge of setting up a git server?
23:14 texasmynsted how many people?
23:14 HedgeMage zealsham_: That sounds like you should have someone host it for you... there are some pretty simple solutions out there, but nothing that doesn't require you to understand Sysadmin 101
23:14 texasmynsted You could use something like bitbucket if there is a small enough team.
23:15 theoceaniscool joined #git
23:15 texasmynsted if the team is too big for free private hosting via bitbucket then it would likely be worth the money to pay for the service for more people.
23:16 texasmynsted Heh, I suppose you could find somebody here to host it for you.
23:17 a_thakur joined #git
23:19 zealsham_ thing is
23:19 zealsham_ my job requires
23:19 zealsham_ i tried explaining how github has private repos and bitbucket
23:19 zealsham_ but my boss wants me to build a server for git that will reside at the workplace and we push code their
23:20 Vampire0 zealsham_, the easiest is a network share everyone has access to and you put a bare repo there, that's it if you don't need more like access control and so on
23:20 HedgeMage zealsham_: Well, that's easy enough to do, but how are we supposed to teach you how to run a Linux server so you have something to run it on?
23:20 Vampire0 zealsham_, if you need more, there are many options, including a private instance of GitHub or GitLab installed at your companies server
23:20 HedgeMage zealsham_: You said you were afraid of using the command line, and that is the most basic of skills needed to do this sort of thing.
23:21 Vampire0 zealsham_, another option is !gitolite
23:21 gitinfo zealsham_: Gitolite is a tool to host git repos on a server. It features fine-grained access control, custom hooks, and can be installed without root. Download: https://github.com/sitaramc/gitolite Docs: http://gitolite.com/gitolite/ Quick example: http://gitolite.com/gitolite/overview/#basic-use-case
23:21 HedgeMage I <3 gitolite, but I'm a bit nervous about pointing someone who's never been outside the gui at it.
23:22 Vampire0 zealsham_, or !gogs as the others pointed out
23:22 gitinfo zealsham_: Gogs is a self-hosted platform similar to !gitlab, written in Go, aiming to be easier to deploy. http://gogs.io/
23:23 Vampire0 ah, now I found the right trigger I had in mind zealsham_, it is !hosts
23:23 gitinfo [!host_gui] There are several options for self-hosting git repositories with a web GUI: !gitblit !gitlab !rhode_code !gerrit and !gogs. Keep in mind that these all take far more work than a (non-GUI) !gitolite install. An overview of nearly all GUIs is available at: https://git.wiki.kernel.org/index.php/Interfaces,_frontends,_and_tools
23:23 zealsham_ yes
23:24 zealsham_ i use the commandline but not at an exceptional level
23:24 pietv joined #git
23:24 peepsalot joined #git
23:25 Vampire0 zealsham_, heavily depends on what features you expect / need
23:25 TbobbyZ_ joined #git
23:25 zealsham_ Vampire0: all i need is just a server to push code to
23:28 Bhoi joined #git
23:32 mattcen joined #git
23:35 roelmonnens joined #git
23:35 gaucheph joined #git
23:40 kexmex joined #git
23:41 Vampire0 zealsham_, as I said, then just use a network share
23:41 Gurkenglas_ joined #git
23:41 sondr3_ joined #git
23:42 zealsham_ okay, thanks
23:42 arecaceae joined #git
23:43 misosoup joined #git
23:43 Jellyg00se joined #git
23:44 secb joined #git
23:45 nidr0x joined #git
23:47 Starkythefox joined #git
23:50 Achylles joined #git
23:50 rominronin joined #git
23:51 gopar joined #git
23:52 Noodlewitt joined #git
23:52 kerwincsc joined #git
23:52 kadoban joined #git
23:53 bilal joined #git
23:53 rkazak joined #git
23:53 mel00010 joined #git
23:54 mquin joined #git
23:56 johnflux joined #git
23:56 chipotle joined #git

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