Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2016-11-29

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

All times shown according to UTC.

Time Nick Message
00:01 Rodya_ joined #git
00:07 Rodya_ joined #git
00:09 d0nn1e joined #git
00:10 DolpheenDream joined #git
00:10 renlo joined #git
00:13 phishy_ joined #git
00:15 phishy_ left #git
00:16 jaguarmagenta joined #git
00:18 dreiss joined #git
00:18 Siegfried joined #git
00:19 jimi_ joined #git
00:20 Intee joined #git
00:20 venmx joined #git
00:21 Electrometro joined #git
00:21 nicksloan joined #git
00:22 barteks2x with git add -p, is it possible to split single remove/add many lines into smaller chunks?
00:23 JosephFerano joined #git
00:23 JosephFerano Can I discuss LFS issues here or is there a better channel for that?
00:24 fission6 joined #git
00:24 clemf joined #git
00:25 JosephFerano I'm getting some odd behavior with git lfs where I'll do a git lfs pull and it smudges every single lfs tracked file in the repo
00:25 JosephFerano what gives?
00:26 whitby joined #git
00:28 barteks2x already done by manualy editing hunk
00:28 renlo joined #git
00:29 JosephFerano ok, so it's not running a smudge, it's calling "git lfs clean" per file
00:30 dsdeiz joined #git
00:30 JosephFerano When doing a simple git status
00:31 JosephFerano http://hastebin.com/acocasodud.pas
00:32 mischat joined #git
00:33 madewokherd joined #git
00:34 mischat joined #git
00:36 dirtyroshi joined #git
00:37 nikivi if I remove .git file, I have essentually 'ungitted' the folder right?
00:37 nikivi .git from the root directory that is
00:37 weaksauce joined #git
00:38 sangy joined #git
00:38 nikivi I git cloned another repo and want to make it my own and start from no commits
00:38 Rodya_ joined #git
00:40 kadoban nikivi: Not particularly necessary or useful usually, but yes you can do that if you really want.
00:40 weaksauce left #git
00:42 zdexin[m] joined #git
00:45 zivester joined #git
00:48 rominronin joined #git
00:48 renlo joined #git
00:50 boz_v1 joined #git
00:50 Gsham joined #git
00:50 iveqy joined #git
00:50 larrybolt joined #git
00:51 j416 joined #git
00:52 styx_ joined #git
00:52 txdv joined #git
00:53 Zoxc joined #git
00:57 jnewt_ joined #git
00:57 b1tchcakes joined #git
00:57 jnewt__ joined #git
00:58 mischat joined #git
00:58 mischat joined #git
00:59 anuxivm joined #git
00:59 mischat joined #git
01:00 mischat joined #git
01:01 k6ekb joined #git
01:01 mischat joined #git
01:02 mischat joined #git
01:02 mischat joined #git
01:03 mischat joined #git
01:04 Derperperd joined #git
01:04 durham_ joined #git
01:04 mischat joined #git
01:05 mischat joined #git
01:06 treehug88 joined #git
01:07 wxg joined #git
01:08 buzzLightBeer joined #git
01:09 jstimm joined #git
01:09 ResidentBiscuit joined #git
01:09 netj joined #git
01:10 Peng_ joined #git
01:11 Doginal joined #git
01:15 dsantiago joined #git
01:15 zivester joined #git
01:15 Nilesh_ joined #git
01:18 rwb joined #git
01:19 drbean joined #git
01:20 jnewt joined #git
01:21 renlo joined #git
01:21 Ferdroid joined #git
01:22 ResidentBiscuit joined #git
01:26 ResidentBiscuit joined #git
01:29 jnewt_ joined #git
01:32 przemoc joined #git
01:35 frivol joined #git
01:36 GavinMagnus joined #git
01:41 jacob123 joined #git
01:44 askb joined #git
01:45 GT4066 joined #git
01:46 durham joined #git
01:47 tnecniv joined #git
01:49 jstimm joined #git
01:49 atesc joined #git
01:49 Jonno_FTW joined #git
01:51 jstimm joined #git
01:52 wxg joined #git
01:56 jimi_ joined #git
01:57 buffal0 joined #git
01:57 venmx joined #git
01:58 CubiK joined #git
01:59 GavinMagnus left #git
02:00 am2on[m]1 joined #git
02:00 renlo joined #git
02:00 renlo left #git
02:01 MTecknology joined #git
02:01 CubiK joined #git
02:01 Remramm kadoban: it's probably a good idea actually, more efficient than going through gc when you just want everything gone
02:04 b1tchcak_ joined #git
02:04 kadoban Remramm: I was speaking of "wipe out everything and start fresh" as not necessary or that useful usually.
02:05 kadoban If you actuall need to, I don't really have a strong opinion of which is better, orphan branch or nuke git.
02:06 mischat joined #git
02:06 t-mart joined #git
02:06 ejb joined #git
02:07 MTecknology Dumb question... I made a commit. Now I can't find it. I have a master and a detached head. How can I find that commit?
02:07 kadoban MTecknology: Check 'git reflog' is probably the easiest way
02:07 a3Dman joined #git
02:08 MTecknology interesting.. and helpful; tahnks!
02:08 MTecknology thanks! *
02:08 t-mart joined #git
02:09 kadoban Anytime
02:09 Gsham joined #git
02:09 kadoban Also, don't be in detatched HEAD much, definitely try to avoid making commits in that state
02:10 finalbeta joined #git
02:11 CEnnis91 joined #git
02:12 giraffe joined #git
02:13 dsdeiz joined #git
02:13 dsdeiz joined #git
02:14 d^sh joined #git
02:14 am2on[m]1 left #git
02:15 MTecknology I get scared and confused when my head is detached. I try to avoid it.
02:15 kadoban xD
02:15 fat_ joined #git
02:16 jaguarmagenta joined #git
02:18 anonymuse joined #git
02:18 jnewt_ joined #git
02:20 BackEndCoder joined #git
02:20 rominronin joined #git
02:21 Peng_ joined #git
02:22 pks joined #git
02:23 Qria joined #git
02:27 Gsham joined #git
02:29 jaguarmagenta joined #git
02:30 Didac joined #git
02:31 Vampire0_ joined #git
02:33 rchavik joined #git
02:34 Rodya_ joined #git
02:34 anuxivm left #git
02:37 jnewt__ joined #git
02:38 UserOO7 joined #git
02:40 nowhere_man joined #git
02:46 davidkrauser joined #git
02:48 ilbot3 joined #git
02:48 Topic for #git is now Welcome to #git, the place for git help and serious business | Public logs at http://jk.gs/glog | First visit? Read: http://jk.gs/git | Current stable version: 2.10.2 | Getting "cannot send to channel"? /msg gitinfo .voice | The git-jokes project has been suspended for lack of contribution. Help revive it! https://madeitwor.se/git-jokes
02:50 dodobrain joined #git
02:51 Phrohdoh I did it with git filter-branch -f --env-filter " GIT_AUTHOR_NAME='Taryn Hill' GIT_AUTHOR_EMAIL='taryn@phrohdoh.com' GIT_COMMITTER_NAME='Taryn Hill' GIT_COMMITTER_EMAIL='taryn@phrohdoh.com' " HEAD
02:51 Phrohdoh left #git
02:53 refried_ joined #git
02:54 am2on joined #git
02:55 raspado joined #git
02:55 thekrynn joined #git
02:56 DolpheenDream joined #git
02:56 clemf joined #git
02:59 jnewt_ joined #git
03:01 aavrug joined #git
03:01 Maggie joined #git
03:02 MTecknology really? I lost my stuff again.
03:02 kaldoran joined #git
03:02 MTecknology I think I mighta really screwed up this time, though...
03:02 Maggie joined #git
03:03 anonymuse joined #git
03:05 MTecknology git reset --hard origin/master .. if I had extra commits on master, did I blow those away?
03:06 jaguarmagenta joined #git
03:06 mischat joined #git
03:07 chachasmooth joined #git
03:08 jnewt__ joined #git
03:08 Peng_ joined #git
03:09 aidalgol joined #git
03:11 jaguarmagenta joined #git
03:12 UserOO7 joined #git
03:12 jimi_ joined #git
03:14 sunri5e joined #git
03:15 dreiss joined #git
03:16 Peng_ joined #git
03:16 hexagoxel joined #git
03:16 serialoverflow joined #git
03:17 davidkrauser joined #git
03:17 thiago joined #git
03:21 Goplat joined #git
03:24 anonymuse joined #git
03:28 venmx joined #git
03:28 Gsham joined #git
03:30 inflames joined #git
03:30 syg_ joined #git
03:30 inflames joined #git
03:30 buffal0 joined #git
03:33 fahadash joined #git
03:35 jnewt joined #git
03:35 kbs joined #git
03:39 jnewt_ joined #git
03:40 jnewt__ joined #git
03:43 jimi_ joined #git
03:44 lb1c joined #git
03:46 fission6 joined #git
03:47 railssmith joined #git
03:50 Tobarja joined #git
03:51 rominronin joined #git
03:53 dreiss joined #git
03:56 buffffal00 joined #git
04:01 peepsalot joined #git
04:01 realz joined #git
04:02 a_thakur joined #git
04:03 ColdKeyboard joined #git
04:03 chachasmooth joined #git
04:04 jost_ joined #git
04:07 mischat joined #git
04:10 navidr joined #git
04:10 einnjo joined #git
04:15 Vortex34 joined #git
04:15 xall joined #git
04:16 Andrew_K joined #git
04:22 Darren_ joined #git
04:27 jnewt joined #git
04:28 jnewt_ joined #git
04:29 venmx joined #git
04:29 jnewt__ joined #git
04:33 TinFury joined #git
04:33 TinFury Hey i need help. I setup a git server on my local network
04:34 TinFury and on my local computer I setup that server as the origin
04:34 TinFury Then I did my initial push
04:34 TinFury And it said success but it seems to have sent the html folder and none of the contents
04:35 jnewt joined #git
04:35 howdoi joined #git
04:36 jnewt_ joined #git
04:36 Gallardo joined #git
04:37 jnewt__ joined #git
04:38 UserOO7 joined #git
04:38 fstd_ joined #git
04:39 ochorocho__ joined #git
04:42 Puffball joined #git
04:43 jaguarmagenta joined #git
04:54 subhojit777 joined #git
04:54 kyle_ joined #git
05:01 pks joined #git
05:02 robotroll joined #git
05:04 dsantiag_ joined #git
05:07 Cabanossi joined #git
05:07 donught joined #git
05:08 mischat joined #git
05:12 cqi joined #git
05:14 lindenle joined #git
05:14 durham joined #git
05:15 cerkauskas joined #git
05:16 Shentino joined #git
05:16 durham joined #git
05:17 inflames joined #git
05:17 hhee joined #git
05:19 ShekharReddy joined #git
05:19 pbrewczynski joined #git
05:21 SimpleName joined #git
05:23 rominronin joined #git
05:23 SimpleName Now I am working on a branch named “hello”, my friend workon branch named “world”, now “hello” need the function of world, what should I do
05:24 kadoban SimpleName: Probably merge world into hello
05:25 kadoban You can also consider rebasing hello onto world, or maybe a cherry-pick sometimes.
05:25 SimpleName in fact “world” just modify one function, so can he modify master directly, and then I rebase?? kadoban
05:25 SimpleName 豆瓣
05:25 SimpleName ..
05:26 sssilver joined #git
05:26 kadoban That's an option, sure, either my merging his thing into master, or ... ya whatever you want
05:27 Peng_ joined #git
05:28 kadoban 角番 xD
05:28 anry joined #git
05:28 eehackr joined #git
05:28 SimpleName :D 逗
05:29 kadoban ;)
05:29 arnsa joined #git
05:29 arnsa hello
05:29 kadoban Hello
05:29 SimpleName hello arnsa
05:29 arnsa I've aciddently been working on the wrong branch, and I need to transfer 2 last commits to master branch. How can I do that?
05:30 venmx joined #git
05:30 kadoban arnsa: Have you pushed the branch anywhere yet, with the 2 new commits?
05:30 arnsa kadoban, nop
05:31 kadoban arnsa: Easiest way is probably to cherry-pick them onto master then, and then edit the branch to remove them (' git reset --hard HEAD~2 ')  make sure you have no uncommited changes and such before you start
05:32 arnsa how can I "cherry-pick them onto master"?
05:33 kadoban git checkout master && git cherry-pick thebranch~1 && git cherry-pick thebranch    or something
05:33 kadoban You can do it as one, but 1/2 the time I fuck up the order when specifying. I don't cherry-pick much
05:34 eehackr joined #git
05:35 arnsa kadoban, shouldn't it be thebranch~2 instead of thebranch~1?
05:35 kadoban Naw, unless I'm simply making a mistake. Seems right though.  thebranch is the latest commit, thebranch~1 is the one before that. So those are the two most recent commits
05:35 _ikke_ arnsa: thebranch is the last commit, thebranch~1 is the 2nd commit, thebranch~2 is the third commit
05:36 durham joined #git
05:36 arnsa ah I see
05:36 _ikke_ For rebasing, you actually need to specify the parent of the last commit you want to rebase
05:37 arnsa kadoban, error: Commit abb4d159d4b8b6e707f15fe79ee26fe2ef4d3caf is a merge but no -m option was given.
05:37 arnsa fatal: cherry-pick failed
05:37 SimpleName I work on branch “hello”, and want to go to master change a file named A_FILE, now git status: there have three files modify, but just want to commit A_FILE, what should I do
05:37 _ikke_ cherry-picking merges is not going to copy the actual merge
05:37 arnsa ugh, I suppose I'll just transfer my changes manually
05:37 arnsa this is too much work :D
05:37 _ikke_ SimpleName: git add A_FILE && git commit
05:38 kadoban arnsa: Hmm, should have asked if they were merges ... you might be better off just redoing the merge commit, unless it was a terribly compilcated conflict resolution.
05:38 subhojit777 joined #git
05:40 ochorocho__ joined #git
05:40 arnsa kadoban, _ikke_ thanks :)
05:41 inflames joined #git
05:43 SimpleName _ikke_: I am working on “hello”, modify three files, now I want to checkout to master, but git hint me: can’t do that, except I commit or stash
05:46 pks joined #git
05:47 a_thakur joined #git
05:51 SimpleName ok, I solve it
05:51 SimpleName use git stash
05:51 SimpleName and then git stash apply
05:51 _ikke_ right
05:52 bocaneri joined #git
05:54 ToBeCloud joined #git
05:54 armyriad joined #git
05:56 svm_invictvs joined #git
05:56 tnecniv joined #git
06:01 jfr joined #git
06:03 Macaveli joined #git
06:03 Peng_ joined #git
06:04 Cabanossi joined #git
06:09 overlord_tm joined #git
06:09 mischat joined #git
06:10 Mowgli joined #git
06:12 InfoTest joined #git
06:12 rscata joined #git
06:14 Vangelis joined #git
06:16 lordjancso joined #git
06:16 Timvde joined #git
06:17 tnecniv joined #git
06:23 function joined #git
06:24 function is there a nicer way to do "git branch --contains && git tag --contains"
06:26 ngperfex joined #git
06:26 clemf joined #git
06:27 function also, is there a way to get "git branch --contains" to not show the leading * ?
06:27 clemf joined #git
06:30 venmx joined #git
06:31 dermoth_ joined #git
06:35 suy joined #git
06:41 nowhereman joined #git
06:42 tnecniv joined #git
06:45 inflames joined #git
06:45 Macaveli joined #git
06:46 a3Dman joined #git
06:47 LucaTM joined #git
06:47 pks joined #git
06:50 ThomasLocke joined #git
06:50 ThomasLocke joined #git
06:50 suy joined #git
06:52 qhp_ joined #git
06:54 _ng joined #git
06:54 rominronin joined #git
06:58 suy joined #git
06:58 qt-x joined #git
06:59 qhp joined #git
07:01 venmx joined #git
07:02 dsdeiz joined #git
07:02 dsdeiz joined #git
07:02 Raging_Hog joined #git
07:02 dermoth joined #git
07:04 overlord_tm joined #git
07:06 dreiss joined #git
07:06 robattila256 joined #git
07:08 overlord_tm joined #git
07:09 cdown joined #git
07:09 overlord_tm joined #git
07:10 mischat joined #git
07:10 suy joined #git
07:12 overlord_tm joined #git
07:13 Andrew_K joined #git
07:15 lindenle joined #git
07:16 derk0pf joined #git
07:17 Anja_ joined #git
07:20 roelmonnens joined #git
07:21 freimatz joined #git
07:25 seishun joined #git
07:25 King_Hual joined #git
07:25 kerrick joined #git
07:28 cqi joined #git
07:29 JeroenT joined #git
07:30 nidr0x joined #git
07:30 cqi joined #git
07:30 mischat joined #git
07:31 UserOO7 joined #git
07:33 Gallardo joined #git
07:33 Alienpruts joined #git
07:34 xall joined #git
07:34 cqi joined #git
07:36 rominronin joined #git
07:36 overlord_tm joined #git
07:36 jaguarmagenta joined #git
07:37 jnavila joined #git
07:37 UserOO7 joined #git
07:38 jknetl joined #git
07:41 Silox| joined #git
07:43 dunpeal joined #git
07:44 MrMojito joined #git
07:44 MrMojito joined #git
07:44 brillenfux joined #git
07:45 Sceorem joined #git
07:47 nowhereman joined #git
07:48 dunpeal_ joined #git
07:49 PCatinean joined #git
07:51 derk0pf joined #git
07:53 dunpeal joined #git
07:53 chele joined #git
07:53 e01 joined #git
07:53 elect_ joined #git
07:55 buzzLightBeer joined #git
07:57 Vangelis joined #git
07:59 inflames joined #git
08:00 spacelord_ joined #git
08:00 Siegfried joined #git
08:02 venmx joined #git
08:04 cdown joined #git
08:04 jnhy left #git
08:05 adac joined #git
08:06 _ikke_ function: git branch is a so called porcelain tool, not built for scripting
08:06 function _ikke_ sure, and I want this for human viewing, not scripting
08:07 function its just that "git tag --contains" has a slightly different format
08:07 function http://sprunge.us/OdIc
08:07 function see this for example
08:08 function I'm also happy to script something like for-each-ref and reconstruct "git branch --contains" but that would be annoyin
08:10 rominronin joined #git
08:14 subhojit777 joined #git
08:15 ayogi joined #git
08:15 Darcidride joined #git
08:16 cdown joined #git
08:16 lordjancso joined #git
08:17 jfr joined #git
08:18 mischat joined #git
08:19 troulouliou_div2 joined #git
08:21 LucaTM joined #git
08:21 krillr joined #git
08:23 fees joined #git
08:25 cdown_ joined #git
08:27 israelzuniga joined #git
08:30 Guest88455 joined #git
08:31 howdoi is there a easy way to compare two PRs in github (I understand it's a github specific question)
08:31 lb howdoi if i recall correctly, there are refs pointing to the PRs ...
08:32 howdoi git diff PRcommit1...PRcommit2 ?
08:32 lb also possible :)
08:32 JanC_ joined #git
08:32 howdoi fatal: Invalid revision range 🙈
08:33 lb howdoi man gitrevisions
08:33 gitinfo the gitrevisions manpage is available at http://jk.gs/gitrevisions.html
08:34 howdoi oki
08:39 leeN joined #git
08:40 ispn joined #git
08:41 encod3 joined #git
08:42 Balliad joined #git
08:44 cdown joined #git
08:46 leeN joined #git
08:46 cdown_ joined #git
08:48 JeroenT joined #git
08:48 dodobrain joined #git
08:52 TomyWork joined #git
08:53 ahmedelgabri joined #git
08:54 Balliad joined #git
08:55 acetakwas joined #git
08:56 cdown joined #git
08:57 rossome joined #git
08:58 systemovich joined #git
09:02 roelmonn_ joined #git
09:03 venmx joined #git
09:03 azerus joined #git
09:03 rominronin joined #git
09:07 bongjovi joined #git
09:07 systemovich joined #git
09:09 marcogmonteiro joined #git
09:09 kurkale6ka joined #git
09:10 thekrynn joined #git
09:12 willlondon joined #git
09:16 mikecmpbll joined #git
09:16 venmx joined #git
09:18 Alienpruts joined #git
09:21 Shentino joined #git
09:21 adac joined #git
09:25 ferr1 joined #git
09:29 manuel joined #git
09:31 cqi joined #git
09:31 zeroed joined #git
09:31 Snugglebash joined #git
09:32 mischat joined #git
09:32 Puffball joined #git
09:32 finalbeta joined #git
09:32 mischat joined #git
09:32 Serpent7776 joined #git
09:34 brk_ joined #git
09:34 sweatsuit joined #git
09:35 MalteTimmermann[ left #git
09:36 willlondon joined #git
09:36 willlondon left #git
09:37 willlondon joined #git
09:37 clmsy joined #git
09:37 jaguarmagenta joined #git
09:38 Charliechin joined #git
09:38 rominronin joined #git
09:38 willlondon joined #git
09:39 WhereIsMySpoon left #git
09:40 rorro joined #git
09:41 OOPMan joined #git
09:41 Gallardo joined #git
09:46 cdown__ joined #git
09:47 subhojit777 joined #git
09:51 donught joined #git
09:53 achadwick joined #git
09:54 Puffball joined #git
10:03 aelgabri joined #git
10:04 Macaveli joined #git
10:07 mablae joined #git
10:07 chll_ joined #git
10:07 styx_ joined #git
10:09 rominronin joined #git
10:15 Peng_ joined #git
10:16 Darcidride joined #git
10:16 ochorocho__ joined #git
10:16 aelgabri joined #git
10:18 OOPMan joined #git
10:18 garethdaine joined #git
10:20 SwiftMatt joined #git
10:25 rnsanchez joined #git
10:27 netj joined #git
10:28 hhee joined #git
10:29 abec0_ joined #git
10:32 rominronin joined #git
10:34 cdown_ joined #git
10:36 roelmonnens joined #git
10:37 cdown__ joined #git
10:37 NwS joined #git
10:41 subhojit777 joined #git
10:42 pks joined #git
10:46 arnsa joined #git
10:46 donught joined #git
10:48 reznord left #git
10:51 jfr joined #git
10:54 alhariel joined #git
10:54 fefa2k joined #git
10:54 fefa2k hello
10:55 xall joined #git
10:56 fefa2k just one quick question, there is a repository that is not mine, I want to create a branch and do a pull request, the normal behaviour would be -> clone the repository -> make branch -> commit -> ask for pull request ?
10:56 howdoi can we boil down: git pull upstream develop && git reset --hard upstream/develop; to a single line?
10:56 _ikke_ fefa2k: Is this repository hosted somewhere?
10:56 fefa2k I've seen examples but all are assuming the repository is mine
10:56 fefa2k github
10:56 selckin howdoi: replace pull with fetch, looks like a single line to me
10:56 _ikke_ You normally fork the project, and in your fork you can push your branch too. Then on github you can make a pull request
10:57 fefa2k ohh ok
10:57 fefa2k I already did a fork some time ago and now it's outdated, how could I update a fork?
11:00 iliv joined #git
11:00 iliv hi
11:01 fefa2k nevermind I found it, thanks _ikke_
11:02 iliv i was wondering if there is a way to sync (with a PR perhaps) all branches of one repo into another repo which is upstream for the first one.
11:02 iliv essentially, I have a fork of upstream and I created a bunch of branches that I'd like to sync with upstream
11:02 iliv I'm using github by the way
11:04 tobiasvl iliv: !sync
11:04 gitinfo iliv: [!pull_upstream] To incorporate changes from an "upstream" or forked-from repository: git remote add upstream $URL; git fetch upstream; git merge upstream/$BRANCH. https://help.github.com/articles/syncing-a-fork/
11:04 sdothum joined #git
11:06 Gallardo joined #git
11:06 a_thakur joined #git
11:06 stamina joined #git
11:07 cdown joined #git
11:10 byte512 joined #git
11:11 Silox| joined #git
11:12 bdeferme joined #git
11:13 iliv tobiasvl, well, suppose I don't own upstream but I want to submit a new branch(es) for upstream to review and add if they're content with proposed changes
11:13 Sceorem_ joined #git
11:13 tobiasvl oh. so a pull request
11:15 Sceorem__ joined #git
11:15 iliv well, github seems to require that upstream already has a branch and if I don't own upstream repo and cannot create a branch.. a contributor would not be able to make a PR unless it is against some existing branch of upstream.
11:15 redfly joined #git
11:15 iliv that's what happens on the web interface
11:15 tobiasvl what
11:15 iliv I wonder if there's a way to do this on command line
11:16 tobiasvl I don't understand what you mean but surely upstream has *a* branch
11:16 tobiasvl it doesn't need to have your branch
11:16 iliv yes, by default, typically, it is master
11:16 Siegfried joined #git
11:16 tobiasvl yes
11:16 iliv in this case it does need to incorporate other branches
11:16 iliv and that is the crux of the matter
11:17 tobiasvl you will need to create one PR per branch, if that's what you mean
11:17 lindenle joined #git
11:17 tobiasvl you can do it with the command line tool http://hub.github.com
11:19 iliv let me explain it differently. i have a fork of repo A that I call B. in my B fork I created several new branches. repo A, which is upstream, doesn't have these branches. I would like to get my branches added to the upstream repo A. Again, it doesn't have them and I want to let other people to be able to do something similar to a normal PR that will create a new branch with whatever commits it contains IF I/upstream aprove such a PR.
11:20 iliv I don't know if it makes any sense.. that's just what I'd like to happen but I don't know if it's even possible?
11:20 tobiasvl that last sentence doesn't make sense to me. what other people? what do you want them to do?
11:20 tobiasvl perhaps #github is a better place for this
11:20 iliv make a PR that will create a new branch in upstream (my repo)
11:21 tobiasvl upstream is your repo?
11:21 tobiasvl I thought upstream was repo A
11:21 iliv I have a fork of repo A that I call B
11:21 iliv :)
11:22 tobiasvl yes. and then you said "repo A, which is upstream" but "new branch in upstream (my repo)"
11:23 iliv it has to read without (my repo)
11:23 iliv like technically I have access to both but other people will not
11:23 tobiasvl aha, so you have forked your own repo
11:24 iliv only in a sense of ownership. those are two distinctly different repos
11:24 tobiasvl but do you want other people (what people?) to create pull requests from your repo B to repo A? why? I don't think that's possible, but people can clone repo B and then make PRs from that to A
11:25 tobiasvl but head over to #github, maybe they understand the issue better
11:25 Doginal joined #git
11:26 iliv in my workflow I have this one "central" repository that multiple contributors will work and will submit not only changes to existing branches but also submit new branches that previously did not exist in upstream repository. again, i'm not sure if this is technically possible but what I'd like to happen is to be able to review a PR that creates a new branch (+commits that it may have) and merge it or decline.
11:26 iliv will fork *
11:27 tobiasvl no, a PR can't create a new branch. and in your scenario there would be no merging involved anyway
11:27 tobiasvl a pull request is a request to pull a branch, and a pull is a fetch + merge into another branch
11:27 iliv yeah, as I said I'm not even sure this is possible. I'm just entertaining this idea and trying to understand how to do this properly and in an effectvie way.
11:28 selckin do you currently have a system where this happens? because planning to have that many branches sounds like a nightmere
11:29 tobiasvl to get a branch to github, that branch needs to be pushed up (or it could be fetched, but you can't fetch from inside github)
11:29 tobiasvl I'd just add them as contributors to repo A, but protect the master branch
11:29 selckin its perfectly possibel with git, just not github UI
11:29 tobiasvl or, I guess "restrict" it in GitHub terms
11:29 tobiasvl https://help.github.com/articl​es/about-branch-restrictions/
11:31 ismithers joined #git
11:32 redfly howdy folks; i have a question that's been bothering me for years
11:32 redfly how does one fetch only a single tag from a remote?
11:32 stamina joined #git
11:32 redfly according to the docs, using `git fetch <repo> tag <tag>` will only fetch everything up to that tag
11:33 iliv selckin, shouldn't be too many branches and only under 5 would be actively worked on all the time. what I'm setting up here is automatic snap package builds on Launchpad. It requires that snapcraft recipe file is in a root of a branch. The only way to have it build multiple versions of the same package is to have multiple branches. That or multiple repos which is even worse :)
11:33 redfly however, whenever i do that, it fetches tags that aren't in the history of the tag i'm requesting
11:33 ronskisika joined #git
11:34 Siegfried joined #git
11:34 redfly i've also tried the alternate syntax of `git fetch <repo> refs/tags/<tag>:refs/tags/<tag>` with the same result
11:34 iliv selckin, I'm okay with doing it on command line. Just not quite sure how to do it. So, any tips?
11:35 tobiasvl redfly: add --single-branch perhaps?
11:37 nemesit|znc joined #git
11:37 redfly tobiasvl: that isn't documented in either https://git-scm.com/docs/git-fetch or the git fetch manpage; i'll nonetheless give it a try and get back to you
11:37 redfly nope
11:37 redfly `error: unknown option `single-branch'`
11:38 jaguarmagenta joined #git
11:39 tobiasvl oh yeah, that's just for cloning, sorry
11:41 redfly perhaps it isn't possible; over the years i've simply updated a script that prunes all the unwanted tags that get fetched every time i fetch a specific tag
11:41 nemesit|znc joined #git
11:48 dsdeiz joined #git
11:48 glebihan_ joined #git
11:52 xall joined #git
11:53 govg joined #git
11:53 buzzLightBeer joined #git
11:55 Macaveli joined #git
11:56 rorro joined #git
11:58 brk_ joined #git
11:59 Starky joined #git
11:59 fmcgeough joined #git
12:02 paladinn joined #git
12:04 jeffreylevesque joined #git
12:05 ToBeCloud joined #git
12:05 netj joined #git
12:06 fling joined #git
12:12 nikivi joined #git
12:13 iliv joined #git
12:13 JeroenT joined #git
12:16 _ikke_ apparently tig can also stage single lines (as alternative to git gui)
12:16 zeroed joined #git
12:16 ochorocho__ joined #git
12:16 cdg joined #git
12:17 byte512 joined #git
12:21 m___1 joined #git
12:26 djb-irc joined #git
12:26 Darcidride joined #git
12:27 zeroed joined #git
12:27 fission6 joined #git
12:31 muahah joined #git
12:32 ok91 joined #git
12:33 cqi joined #git
12:34 isysd joined #git
12:36 Shentino joined #git
12:36 Peng_ joined #git
12:40 m___ joined #git
12:41 astrofog joined #git
12:42 dminuoso I've been wondering a while about non-functional commits, I'm just not sure how to deal with them in my work flow.
12:43 dminuoso I mean, do you just squash everything together for release/master branches? Do you just rely on tagging to denote stable commits?
12:45 iliv joined #git
12:45 _ikke_ What do you mean exactly with non-functional commits?
12:45 _ikke_ I never squash everything into one commit
12:46 Rhonda joined #git
12:46 _ikke_ I prefer merging in branches
12:46 Rhonda Tagging for the releases regardless of everything else.  :)  If you want to squash them or not depends on your personal working habits.  I don't squash because of the fear of losing some intermediate step that followed a certain thought process, but that might be just me and I know others squash a lot.
12:49 dminuoso _ikke_: Well I have commits which leave my software in a non-working state.
12:50 dminuoso _ikke_: Okay so you basically have a feature branch and toss tons of commits in there, and then you merge this into a development branch once a feature is done?
12:52 mostlybadfly joined #git
12:52 m___1 joined #git
12:52 osse dminuoso: when working, squash/amend commits as you go along
12:53 _ikke_ dminuoso: I (try to) make my commits so that they are always in a working state
12:54 _ikke_ So I do use fixup commits to squash small edits into other commits
12:54 osse git commit --amend / --squash is your friend
12:54 _ikke_ git commit --fixup!
12:54 m___ joined #git
12:55 synthroid joined #git
12:55 _ikke_ Right
12:55 davimore joined #git
12:55 dminuoso osse: Well honestly I've started quite chaotic - but I don't think it's a reason to not start git'ing properly.
12:58 DolpheenDream joined #git
13:00 iasco7 joined #git
13:00 ok91 joined #git
13:01 p_d joined #git
13:01 johnmilton joined #git
13:01 p_d a
13:01 _ikke_ b
13:01 p_d left #git
13:02 jephsf joined #git
13:03 masuberu joined #git
13:03 elect joined #git
13:04 zeroed joined #git
13:04 zeroed joined #git
13:07 dviola joined #git
13:07 howdoi selckin: git pull upstream develop && git reset --hard upstream/develop; fetch in this will become a single liner?
13:08 selckin howdoi: fits on my screen
13:09 howdoi selckin: I mean can this be compressed to a single command
13:09 tobiasvl no
13:09 _ikke_ The last fetch is redundant
13:09 selckin man git-alias
13:09 tobiasvl why is the first one a pull and not a fetch
13:09 _ikke_ It can be done
13:10 _ikke_ git fetch upstream +refs/heads/develop/refs/heads/develop
13:10 canton7 ^ missing a colon
13:10 _ikke_ git fetch upstream +refs/heads/develop:/refs/heads/develop
13:10 _ikke_ thanks
13:10 tobiasvl and -f ?
13:10 canton7 the '+' means the same as -f
13:10 _ikke_ tobiasvl: That's what the + is for
13:10 tobiasvl ooooh yeah right
13:10 canton7 I don't think you have the / after the : though?
13:10 fefa2k left #git
13:10 canton7 or does it get ignored?
13:11 osse dminuoso: what do you mean?
13:11 canton7 .... but, bear in mind, you still need a checkout to make your working copy line up
13:11 _ikke_ better to leave it out
13:11 cqi joined #git
13:11 canton7 (if you have develop checked out to start with)
13:12 paladinn joined #git
13:14 ToxicFrog joined #git
13:14 dminuoso osse: I've been slacking a lot, committing very sporadically with undescribtive commit messages. Mostly because I'm the only developer on this project.
13:14 iliv joined #git
13:14 dminuoso Sometimes commiting partial work.
13:15 InfoTest joined #git
13:15 p_d joined #git
13:15 p_d a
13:15 zivester joined #git
13:15 dminuoso And at times it feels like "there's no point to start adhering to some git flow and rules if it's started bad already"
13:16 mischat joined #git
13:16 re1 joined #git
13:17 osse dminuoso: you can still save your soul, maybe. use rebase -i to combine related commits into actually meaningful commits and give them useful messages
13:17 _ikke_ p_d: ...
13:17 CheckDavid joined #git
13:18 p_d _ikke_: sorry actually i'm testing if m IRC is working properly and my nickname is registered!! I recently migrated to ubuntu
13:19 a_thakur joined #git
13:19 osse dminuoso: and fi you do that regularly then there is no problem. the flags i mentioned helps you with this. as long as the finals commits make sense then who cares how they got there
13:19 osse dminuoso: !sausage
13:19 gitinfo dminuoso: [!sausage_making] Some developers like to "hide the sausage making", transforming their commits before presenting them to the outside world. See http://sethrobertson.github.c​om/GitBestPractices/#sausage and !perfect
13:19 dminuoso osse: I see. git update-ref -d refs/heads/master && git commit -s -m "Initial import"
13:19 dminuoso :o)
13:19 dminuoso And then git push -f
13:20 rbern joined #git
13:20 venmx joined #git
13:23 derpsauce__ joined #git
13:24 p_d left #git
13:24 dminuoso osse: Mmm, it's just too chaotic to clean up the history honestly. I'm unsure whether I should keep the history as is, or just squash it together and then start with better practices.
13:24 osse that is up to you
13:25 osse how many commits are there?
13:25 osse the start of a project is always bit bumpy imho
13:25 dminuoso osse: 604
13:26 LionsMane joined #git
13:27 osse dminuoso: if it's all a mess then the only useful thing it can be used for is bisection
13:27 osse ie. helping tracking down which particular change introduced a bug
13:27 osse if that's not relevant then I'd just nuke the entire history from orbit
13:27 osse it's the only way to be sure
13:28 breezy_ joined #git
13:28 guampa joined #git
13:29 dminuoso Right, so after this I basically need to stop working on development branch, and work in small feature branches that I can then squash together and then merge back into developnebt,
13:29 dminuoso Except for small hotfixes perhaps.
13:30 Peng_ joined #git
13:33 osse Imo feature branches can very well contain many commits
13:33 osse The main point is that each commit should make sense
13:33 osse !atomic
13:33 gitinfo Use this to atomically switch deployments: http://rcrowley.org/2010/01/06/th​ings-unix-can-do-atomically.html
13:33 osse Ehh disregard that
13:34 dminuoso osse: For a split second you got me very interested what that was about, then for 2 seconds you triggered a real "wtf" expression on my face. :-)
13:34 dminuoso But I get the idea I think.
13:35 cdown joined #git
13:37 osse dminuoso: Eg if I'm implementing a logging facility then I might have a commit or two that implements it, then maybe some commits more that leverages it in existing code. Along the way I make spelling errors in function names and write +1 when I meant -1 etc. Those latter changes I commit with a --fixup referencing the commit that introduced the mistake.
13:37 jephsf joined #git
13:37 osse After that rebase -i magically makes everything better
13:37 osse Inshallah
13:37 ugjka joined #git
13:37 ThiefMaster how can i get the diff for a renamed file when specifying the filename explicitly? i'm currently using this in a pre-commit hook to run a tool on only the modified areas of my staged files but it chokes on renames: for FILE in `git diff --staged --name-status HEAD | awk '$1 != "D" {print $2}'` ; do RESULT=`git diff -U0 --cached $FILE | $PEP8_CMD --diff $PEP8_OPTIONS`
13:38 ThiefMaster if i just do `git diff --staged` i get the proper diff but i don't see how i'd specify the two filenames for the rename there
13:38 ThiefMaster nvm, got it, just needed `--` before the two arguments
13:38 oaao joined #git
13:39 dminuoso osse: How do you rebase? Do you just squash fixups into their referenced commits?
13:39 jaguarmagenta joined #git
13:39 kyan joined #git
13:39 osse dminuoso: git rebase -i --autosquash HEAD~10, for example.
13:40 osse The autosquash can be made default with a config option
13:40 osse dminuoso: I'd you use --fixup then rebase can automatically figure out the reordering of commits
13:41 osse very automatica
13:41 osse such useful
13:41 osse wow
13:43 osse Read the descriptions of those flags in the man pages and you'll get it pretty quickly.
13:43 KevinDF joined #git
13:43 bongjovi joined #git
13:44 dminuoso osse: Yeah. Already reading the manual - I guess the "-i" was just an example, it could also be a non-interactive rebase I suppose.
13:44 osse dminuoso: Actually no.
13:44 osse For some reason
13:45 dminuoso This option is only valid when the --interactive option is used.
13:45 dminuoso Mmm.
13:45 osse Sometimes when it's trivial and I can't bother with the editor I do GIT_EDITOR=cat git...  :P
13:45 fission6 joined #git
13:46 _ikke_ dminuoso: Yeah, autosquash without interactive does not make sense
13:49 f3r70rr35f joined #git
13:49 re1 joined #git
13:49 dreiss joined #git
13:50 osse but it's auto! I don't want to interact with it
13:50 _ikke_ You should at least verify if it's doing the right thing ;-)
13:50 Shentino joined #git
13:50 osse over my dead code base
13:51 qqx Currently interactive and non-interactive forms of rebase have implementations that are almost completely independent of each other.
13:51 mkoskar joined #git
13:51 osse yes. it's due to the implemention; not whether it makes "sense"
13:51 qqx Hopefully at some point that will be changed and autosquash can be supported by both.
13:51 osse "sense" is in the eye of the bee-holder
13:53 mischat joined #git
13:56 wwspider joined #git
13:56 inflames joined #git
13:56 jeffreylevesque joined #git
13:57 dan2k3k4 joined #git
13:58 rnsanchez_ joined #git
14:01 cdg joined #git
14:02 cdown joined #git
14:02 elect joined #git
14:03 osse ThiefMaster: you can add --diff-filter=R to find renames
14:04 osse then you can prolly add --name-only
14:04 ThiefMaster osse: i'm trying to fix this script to work with renamed  files too: https://gist.github.com/ThiefMaster/7ca3c48ef936​04be0478e30852d0c444#file-pre-commit-sh-L25-L27
14:05 ThiefMaster my current approach is to get $FILE either as 'filename' for additions/changes or 'oldfile newfile' for renames so it works either way when passing it to git diff
14:05 ThiefMaster so i need both renames and other things
14:05 ThiefMaster oh nvm, you replied to my question in here - thought it was related to my related question in #bash
14:06 osse hmm
14:06 osse can't you just  git diff | pep8 --diff ?
14:06 osse or isn't pep8 that smart
14:07 ThiefMaster i don't want it to take context into account
14:07 ThiefMaster ah, you mean just `git diff -U0 --staged | pep8 --diff`
14:07 ThiefMaster hmm that could work
14:07 * ThiefMaster checks
14:08 anonymuse joined #git
14:08 osse it says that it only checks modified lines
14:10 dfaught joined #git
14:10 _Ching joined #git
14:12 _Ching Hey I did 'rebase master' to point feature branch to head of master. After which I made some more commits(while others pushed and merged to master). So I rebased again. Now it shows that local and remote have diverged. Should I force push?
14:13 tobiasvl _Ching: that's up to you, if you want to !rewrite remote history
14:13 gitinfo _Ching: 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
14:13 tobiasvl but yes, if you want the remote to reflect you rebased local branch, you will have to force push
14:13 _Ching Move on? What do you mean? I have to push right?
14:14 xall joined #git
14:14 _Ching But my feature branch was never merged to master.
14:15 tobiasvl "Move on" = do a regular merge and live with messy history
14:15 _Ching I create a feature branch -> Added commits -> Rebased to head of master ->PR-> Rejected -> Commits on feature branch -> Rebase again on local -> Push to remote fails
14:15 tobiasvl -> force push
14:16 _Ching Yeah. In this case does it rewrite any significant remote history as no one has branched off my feature branch
14:16 tobiasvl my warning about rewriting history applies to remote branches that other people might have pulled. is this github? are you working on your own, private fork? if so, there's nothing wrong with force pushing
14:17 _Ching Github. Working as a team. Remote branch that only I work on.
14:17 osse it is probably expected that your force-push,t hen
14:17 tobiasvl only you can know if you can force push, but it sounds like it'd be unproblematic
14:18 _Ching If I'm to work as a team on the feature branch, should I refrain from rebasing to head of master and just live with the messy history instead of hoping to maintain linear history?
14:19 mindriot101 joined #git
14:20 tobiasvl depends on what your team has decided. I'd say yes, don't rebase/force-push branches other people work on. for them to manually merge in their changes every time the remote branch changes its history is way harder to deal with and nerfs some of the most useful features of git
14:20 jimi_ joined #git
14:21 noecc joined #git
14:21 _Ching So the advice to rebase to maintain linear history doesn't really cover a lot of cases right?
14:21 _Ching Thanks. I think I'll follow that.
14:21 andlabs joined #git
14:22 CEnnis91 joined #git
14:22 osse linear history is for losers anyway. #keepgitloggraphcool
14:22 [gmi] joined #git
14:23 JeroenT_ joined #git
14:23 shiriru joined #git
14:23 tobiasvl rebasing to maintain linear history applies when touching up your local work before pushing it (not after), and when pulling in upstream changes that have been done since you branched off, to make the future merge clean
14:24 tobiasvl IMO
14:24 osse put you have to push to get feedback on your PR
14:24 osse *but
14:24 selckin having a merge just because of the time someone decide to git pull has no value, pull --rebase or gtfo
14:24 Stag joined #git
14:24 Dougie187 joined #git
14:25 ShekharReddy joined #git
14:25 tobiasvl yeah exactly, that's what I tried to say
14:25 kbs joined #git
14:25 tobiasvl osse: sure, pull requests are a special case. force-pushing to those should be fine in most scenarios
14:26 tobiasvl except _Ching's scenario where they work as a team on a feature branch
14:26 Stag Hi folks, say I have a repo with branches develop and update-develop, what is the best way to push update-develop into develop without preserving files in develop ?
14:27 blackwind_123 joined #git
14:27 _Ching osse: You describe my situation precisely. Team working on feature branch requiring feedback on PR
14:27 osse coordinate with your team.
14:28 JeroenT_ left #git
14:28 JeroenT joined #git
14:28 osse what we once did was that we decided to push fixups and stuff as regular commits. then we "pair-programmed" when we did a rebase. worked pretty well
14:28 ExoUNX joined #git
14:29 venmx_ joined #git
14:29 rivarun joined #git
14:30 JeroenT joined #git
14:31 howitdo joined #git
14:31 tvw joined #git
14:33 _Ching osse: Pair programmed? Like basically alerted the other guy that you're rebasing?
14:33 osse _Ching: he pushed his work. he sat next to me
14:34 osse might not work if the other guy is on the far side of the earth though
14:34 xberg joined #git
14:35 ash_workz joined #git
14:35 _Ching osse: Sounds like a bad idea. Even if he's not on the far side, he's not in my line of sight most of the time. I think I'll just stick with no rebasing.
14:36 qqx Even if you don't want to do pair programming for rebasing, you could still coordinate a time when rebasing is expected by everybody working on a branch.
14:37 _Ching Ultimately doesn't that screw up history anyway?
14:37 qqx That would likely be desirable when you're about to merge the branch or open a PR, when there shouldn't be much if any outstanding work on the branch anyway.
14:37 bongo joined #git
14:38 qqx Rebasing mainly causes problems when it changes commits that are a parent of other commits. If you pick a time well, there will be few (ideally no) child commits.
14:40 _Ching qqx: That's again a very small case right? I mean I expect people to put up a couple of commits before I rebase again.
14:41 qqx It also depends on how the feature branch is developed. The history that gets pushed might already be quite clean, and therefore not benefit much from rebasing.
14:42 qqx _Ching: If you do the rebase while all the outstanding work is already on the branch that you're rebasing it shouldn't cause any problems.
14:42 osse _Ching: if you tell him to let you know when he's pushed all his stuff. then let him that you'll rebase at that point, it will work fine.
14:42 qqx Then once the rebase has been done other people can fetch the rebased version and base new work on that rather than the old version.
14:43 _Ching I don't follow. Do you mean like without much master branch activity? Our development involves simultaneous development of feature branches that span a long interval of the entire project instead of quick and easy feature branches.
14:43 osse also, if he has unpush stuff:  gti fetch; git rebase --onto=origin/foo origin/foo@{1}   # Works like a charm!
14:43 InfoTest joined #git
14:44 ejb joined #git
14:44 qqx _Ching: It really shouldn't matter what's being altered in the rebase. Just as long as you're not changing commits that are the parent of other things.
14:45 nickabbey joined #git
14:45 bongo I have a question regarding filter-branch.  I converted a repository for a web site but the file permissions are want is desired.  so I wanted to run something like: """git filter-branch -f --tree-filter 'git ls-files -z | xargs -r -0 -n 15 -P 100 chmod 755' --tag-name-filter cat -- --all """.  But this also doe not appear to have the desired result.
14:45 qqx Although if the change caused by the rebase is large it does make it more difficult to review. But being hard to review would likely apply to the merge commit that you'll get otherwise.
14:45 bongo in the end I plan to push changes to new remote.
14:45 bongo with all branches having files with the desired permissions.
14:47 _Ching qqx: My point was that my feature branch would probably span the a considerable amount of total project duration. Does that make rebasing moot since I almost expect the master branch to significantly change.
14:48 shinnya joined #git
14:49 osse just merge in master and screw linear history .who cares anyway
14:49 qqx _Ching: It depends why you're wanting to do a rebase. If it's just to pull in changes from a parent branch I personally wouldn't do it.
14:49 bongo Any ideas how to change permissions on all files for all history of all branches on the remote??
14:49 qqx I was more covering *how* to handle it than if it should be done.
14:49 _ikke_ bongo: First do it local
14:50 _Ching osse: Ah. Okay. qqx: I see. Thanks
14:50 _ikke_ bongo: man git filter-branch, look for tree-filter
14:50 gitinfo bongo: the git-filter-branch manpage is available at http://jk.gs/git-filter-branch.html
14:50 chigang joined #git
14:50 bongo I am.  but must be missing something.
14:50 qqx bongo: Are you on a system that uses file permissions (e.g. are you doing this on Windows)?
14:51 bongo yes
14:51 venmx joined #git
14:51 bongo readhat
14:51 _ikke_ Not that git only tracks the executable bit
14:51 WayToDoor joined #git
14:51 bongo redhat
14:51 _ikke_ so it only knows 755 and 644
14:51 bongo right
14:52 b4x joined #git
14:52 Hello71 technically it does 777 and 666
14:52 bongo I'm ok with that but want all files to be u+x
14:52 Hello71 but with a default umask of 022 then you get those results
14:52 _ikke_ Hello71: Correct
14:52 cqi joined #git
14:52 _ikke_ But you often see those modes in the tree
14:53 _ikke_ 100644 and 100755
14:53 zeroed joined #git
14:53 zeroed joined #git
14:54 nicksloan joined #git
14:55 DaveTaboola joined #git
14:56 d0nn1e joined #git
14:56 e14 joined #git
14:56 bongo I've also tried: """git filter-branch -f --tree-filter 'fit -ls-files -z| xargs -r -0 git update-index --chmod=+x ' --tag-filter cat -- --all"""
14:56 bongo which is much fasted but had same problem.
14:57 bongo *faster
14:59 whitby joined #git
14:59 Hello71 huh. it says 100644 but actually sets 666
14:59 * Hello71 sends patch
15:00 qqx bongo: If you have a recent enough version of git you could try using an index filter with: git ls-files -z | git update-index --chmod=+x -z --stdin
15:01 boombatower joined #git
15:01 dunpeal joined #git
15:02 bongo thanks qqx will try that now.  running git v2.10
15:02 al-damiri joined #git
15:03 portalesman joined #git
15:04 TinFury joined #git
15:05 AaronMT joined #git
15:08 _ikke_ qqx: are you by chance qsx? or is that someone different?
15:08 qqx _ikke_: Somebody different
15:08 _ikke_ ok
15:09 dunpeal joined #git
15:09 DeltaHeavy joined #git
15:10 DeltaHeavy Can I ask GitHub questions in here? I have a GitHub account I forgot about using my work e-mail and I now want a particular repo something I'm using my main account to send e-mails to my work account. Is this possible?
15:10 DaveTaboola joined #git
15:11 crose joined #git
15:12 DeltaHeavy Got it figured out I think
15:12 DeltaHeavy Always right after asking :/
15:14 _ikke_ Rubber Duck effect
15:15 bongo I ran on a subset output looked promising but no joy when I search for files that are not u+x.
15:18 bongo I see tags are updated, checked out a tag and still find files that are not u+x
15:19 bongo does --all select remote and local branches?
15:20 LionsMane When I do an git svn clone {url}, I am getting the trunk folder as part of the fetched files.  I was hoping to have the *contents* of the trunk folder as part of the clone instead.
15:20 whlai When I push, git is pushing way more than is on my stage (according to git ls-files --stage)
15:20 whlai any idea why?
15:20 LionsMane bongo:  For git branch, yes.
15:20 ash_workz is there a more github specific channel?
15:20 _ikke_ #github
15:21 ash_workz whaddya know
15:21 _ikke_ whlai: How do you know, is pushing more? git always pushes complete commits. It cannot just push individual files. So if something is pushed, it's because it's part of the history
15:21 shiriru joined #git
15:22 aarobc joined #git
15:22 bongo Your right. It uses rev-list so it is the same for filter-branch..
15:22 whlai _ikke_ The size is too big for what I added to the index using git add
15:22 Kodax left #git
15:23 _ikke_ How do you determine this?
15:23 whlai Is there any way it will try to push anything tha I am not seeing when I  git ls-files --stage?
15:23 _ikke_ whlai: git ls-files is not a good tool to determine what's getting pushed
15:24 DaveTaboola joined #git
15:24 _ikke_ That only deals with the index, while git pushes things that are comitted
15:24 whlai I determine because the directory I'm trying to push is smaller than what is trying to be pushed by git. I can see the size giot is trying to write
15:24 whlai (it's failing)
15:24 whlai ohhhh
15:25 whlai _ikke_ How can I see the commits? and then how can I clear old commits?
15:25 _ikke_ git log --all shows you all commits
15:26 wrksx joined #git
15:26 wrksx I got a weird issue
15:26 roger_padactor joined #git
15:26 ResidentBiscuit joined #git
15:27 selckin full of anticipation
15:27 xberg joined #git
15:27 wrksx lol
15:27 Jellyg00se joined #git
15:27 ResidentBiscuit joined #git
15:27 wrksx nope okay, I got a simple issue, cannot understand what's goin wrong
15:27 wrksx ssh gitolite3@host works well, while git pull doesn't work
15:27 roger_padactor hey, is there a way to make a commit rule that if certain text is in the file it doesn't make the commit when merging into a specified branch?
15:28 xberg joined #git
15:28 wrksx I mean using ssh command the ssh key works correctly, while git pull doesn't use the key
15:28 wrksx so I tried GIT_SSH_COMMAND="ssh -vvv" git pull
15:28 _ikke_ roger_padactor: pre-commit hook is the only thing that comes to mind, but that does not come into play when git does a fast-forward merge
15:28 selckin wrksx: git remote -v, check if same user/hsot
15:29 sssilver joined #git
15:29 roger_padactor _ikke_, thank you.
15:29 wrksx selckin: origin  gitolite3@Neptune:gitolite-admin (fetch)
15:29 wrksx and another line for (push)
15:30 wrksx selckin: but I don't understand what u meant by same 'user/host'
15:31 selckin wrksx: 'ssh gitolite3@Neptune' works?
15:31 andlabs joined #git
15:31 wrksx selckin: Fuckin hell u nailed it
15:31 wrksx selckin: the capital N
15:31 wrksx Damn me
15:31 GodGinrai joined #git
15:31 wrksx selckin: ty so much
15:31 selckin np
15:32 qqx roger_padactor: Like I can't think of a way to do exactly that either. But you may be able to use a pre-push hook to prevent that going to a remote.
15:33 ELFrederich I'm looking at using Git (via Gitolite) as a backend for storing some application's data.  This application will collaborate using Git as a backend.  The issue now is that I want clients to be able to add new repos.  To do this, the gitolite-admin project needs to be modified.  How can I protect this project to make sure clients are only adding repos and users, and not modifying existing repos
15:33 ELFrederich permissions or modifying existing users keys.  Possible with some kind of hook?
15:34 GodGinrai joined #git
15:35 mikecmpbll joined #git
15:36 roger_padactor qqy could do it in a pull request maybe?
15:37 cheater joined #git
15:37 roger_padactor qqx, ^
15:37 cheater hi
15:37 anuxivm joined #git
15:37 cheater i have ~10 git repos under one directory and constantly add more. is there a way to make them all use the same committer info?
15:37 cheater name/email
15:38 jimi_sanchez joined #git
15:38 _ikke_ Only on a per user global level
15:38 _ikke_ Not per directory or something
15:38 cheater e.g. if i could just put one file in that parent dir that would tell all git repos to use config inside it that would be great
15:39 cheater there's something like this for mercurial
15:39 _ikke_ That doesn't exist
15:39 tnecniv joined #git
15:39 cheater i think it might be a plugin or something
15:39 cheater ok _ikke_... that's too bad
15:40 jaguarmagenta joined #git
15:40 jimi_sanchez joined #git
15:41 Darcidride joined #git
15:41 GodGinrai cheater: you could write a simple script that does this
15:41 GodGinrai cheater: and then even set it up as a git alias or something
15:41 Starky joined #git
15:42 cheater yeah, but... meh
15:42 ash_workz is the 'graph thingie' maintainer in here?
15:42 _ikke_ jast?
15:42 andlabs joined #git
15:43 ash_workz ah, jast
15:43 _ikke_ He's here once in a while
15:43 ash_workz better channel to talk to him about his graph thingie?
15:43 re1 joined #git
15:43 _ikke_ No idea
15:43 _ikke_ https://jk.gs/
15:43 qqx roger_padactor: I don't understand that question.
15:43 _ikke_ It lists his e-mail address
15:43 gchristensen left #git
15:44 Noldorin joined #git
15:45 Siegfried joined #git
15:45 ash_workz yeah
15:45 qqx cheater: You could use something like direnv to set the EMAIL environment variable when you're in some directory.
15:45 ash_workz thanks _ikke_
15:46 crayon joined #git
15:46 ok91 joined #git
15:48 subhojit777 joined #git
15:48 re1 joined #git
15:50 kbs joined #git
15:50 zivester joined #git
15:51 LBV_User joined #git
15:52 anddam left #git
15:52 User458764 joined #git
15:53 ELFrederich qqx: never heard of direnv before.  Can't decide if that is a great idea or a terrible one
15:53 User458764 joined #git
15:53 ELFrederich qqx: would make debugging issues incredibly conplex.  Put it this way, I wouldn't want to my users to use it ;-)
15:53 __marco joined #git
15:54 salamanderrake joined #git
15:56 tnecniv joined #git
15:56 [0xAA] joined #git
15:58 GodGinrai ELFrederich: yea, I'm not so sure I'd like to debug issues with phantom directory-based env vars
15:58 tnecniv joined #git
15:58 muahah left #git
15:58 eivarv joined #git
16:00 a_thakur joined #git
16:02 __marco Hello. I have a file that I use for my tests and that I need to include in a git repository. This file, a sqlite database, changes every time the tests are run and I *don't* what to commit every time I run the tests nor having it in the list of changed files
16:02 __marco How can I do?
16:02 GodGinrai __marco: why does it need to be included in the repo, then?
16:03 qqx __marco: You can't. git doesn't support ignoring changes to files that are tracked.
16:03 diget joined #git
16:03 __marco GodGinrai: because it contains some initial data
16:04 GodGinrai __marco: why not make a "initial data" file that gets copied to the filename that the tests use during the testing process
16:04 tnecniv joined #git
16:04 qqx You could possibly track a copy with a different name. Ignore the file name that is actually used, copy that over when you want to commit it.
16:04 krs93 joined #git
16:04 Random832 __marco, have the version with only initial data in some other directory, with instructions to copy it after they clone the repository
16:04 GodGinrai __marco: that way, the file never changes, and you can add the file the tests use to the .gitignore
16:04 Pseudocrat joined #git
16:04 mda1 joined #git
16:04 ELFrederich GodGinrai: could get crazy when scripts call other scripts from other directories, etc, etc.  We need Linus to write a rant on it
16:05 __marco GodGinrai: Yes, I was thinking about that
16:05 tnecniv joined #git
16:05 GodGinrai ELFrederich: indeed
16:06 ManiacTwister joined #git
16:07 dunpeal joined #git
16:07 __marco GodGinrai: I will. Thanks
16:07 ok91 joined #git
16:07 robotroll joined #git
16:07 __marco qqx: thanks, I'll do so
16:08 cheater joined #git
16:08 bongo qqx: so look like the filter-branch is having no impact.
16:08 djb-irc joined #git
16:09 bongo I see tag updates like tag-1 -> tag-1 ( sha1_a -> sha1_a )
16:09 bongo to me that looks like a noop.
16:09 ManiacTwister joined #git
16:10 e14 joined #git
16:10 bongo should the branch and workspace on filter-branch exit reflect the change?
16:11 maxxe joined #git
16:11 bongo Also in my new clone I noticed only the default branch checked out after clone was processed.  but reported """is unchanged"""
16:11 cqi joined #git
16:11 ojdo joined #git
16:11 bongo something is not working as expected.  this last run I tried with previous git install 2.0.
16:11 Rodya_ joined #git
16:11 ManiacTwister joined #git
16:11 maxxe is it ok to add a binary to a repo if it won't ever change, for example a image of type png?
16:12 _ikke_ Sire
16:12 _ikke_ sure
16:12 endiruna joined #git
16:12 maxxe ok
16:12 _ikke_ Git can handle binaries. Problem is when they're large binaries and change often
16:12 maxxe yeah
16:13 mikecmpbll joined #git
16:13 clemf joined #git
16:13 maxxe I have been fetching the binary stuff during build else. but to fetch a small png seems a bit overkill. :)
16:14 Rodya_ joined #git
16:15 osse pngs, icons, small resources like that. no problem.
16:15 bongo I've saved log and inspection shows even when I run for --all the there is """WARNING: Ref 'refs/remote/origin/*some branch*' is unchanged"""  for all the branches.  perhaps my tree-filter is not running as expected.
16:15 osse maxxe: To be perfectly honest, we have DLLs > 100 MB and it's still not a problem.
16:15 osse they never change, though
16:15 maxxe osse, yeah it's a icon
16:15 osse we have some 15 MB ones that change once a month or so :P
16:16 maxxe *an
16:16 maxxe yeah
16:16 ShekharReddy joined #git
16:16 whlai I have a bunch of commits in my history (according to git log) that I want to delete. I don't want to touch my working tree. How can I do this?
16:16 mischat joined #git
16:16 bodhi joined #git
16:17 Derperperd joined #git
16:17 jimi_sanchez joined #git
16:18 aarobc joined #git
16:18 _ikke_ whlai: You can only do that by rewriting history
16:18 GodGinrai _ikke_: I think he's aware of that since he mentioned he wanted to delete them
16:19 GodGinrai _ikke_: he needs to know *how* to commit the horrible act :P
16:19 whlai Let me give more details. I had done come commits and pushed. I now have deleted and recreated a, now empty, remote repo/
16:19 whlai I still have commits that are in my history according to git log
16:19 whlai I want to remove those and only push a much smaller amount of files
16:19 todd_dsm joined #git
16:20 whlai I don't want to leave all of my working files in tact
16:20 whlai How can I do this?
16:20 GodGinrai whlai: you want to start a new repo with the content of your current repo?
16:20 Rodya_ joined #git
16:20 GodGinrai whlai: with no history?
16:20 whlai I want to start a new remote repo, but I want to clear out the commits from my local git history
16:20 whlai the remote has no history
16:20 whlai the local does
16:21 GodGinrai whlai: just delete your .git folder, and then git init again
16:21 whlai I basically want to rest everything and start from scratch but not touch my local files
16:21 whlai that's it?
16:21 whlai kk
16:21 whlai cool
16:21 whlai thanks
16:21 GodGinrai np
16:21 NeverTired joined #git
16:22 _UniFreak_ joined #git
16:23 nfk joined #git
16:23 _UniFreak_ joined #git
16:23 derpsauce__ left #git
16:25 lindenle joined #git
16:26 maxxe osse, it should be cleaner to fetch those dlls during build of the project. but that's only my opinion.
16:26 kpease joined #git
16:26 osse maxxe: i agree. this was a quick temporary solution that became permanent
16:27 osse the big question being: fetch them from where?
16:28 maxxe upload them to a site/mirror. during 'make' (if you use make for building) fetch them at the right place
16:28 jokajak joined #git
16:29 osse true
16:29 GodGinrai osse: And if you wanted something more complex, there are always dependency managers, like apache ivy
16:29 osse these are proprietary in-house thingies. maybe we can just put then on gitlab as a "release" somehow
16:30 maxxe I fetch LibreSSL in a native .cab file when building my irc client for Windows
16:30 madewokherd joined #git
16:30 maxxe within the Makefile
16:30 GodGinrai maxxe: you're making an IRC client for Windows?
16:31 osse GodGinrai: this is a C/C++ project and we use CMake.  Hunter works fairly well. Ideally we'd create a separate repo with the DLL stuff in it
16:31 maxxe not only. but it runs on windows
16:31 maxxe https://dataswamp.org/~markus/swirc/
16:31 maxxe still a lot of work to do
16:32 osse Aaand there are some proto files. Goddammit, the whole thing is a mess
16:33 flavius joined #git
16:33 eivarv joined #git
16:34 GodGinrai maxxe: neat.
16:35 maxxe thank you. have been working hard on it. finally now 2016 I got a release ready. but it took a long time. :)
16:35 maxxe started around 2012
16:35 ok91 joined #git
16:36 sangy joined #git
16:38 osse maxxe: now i know your full name and what you look like. how weirded out are you now on a scale from 4 to 17
16:38 akanksha__ joined #git
16:39 maxxe 17 :D
16:40 akanksha__ left #git
16:41 osse How much of that is because I know this, and how much is because I keep asking about it like a creepy fucker
16:43 oskarkv joined #git
16:43 svm_invictvs joined #git
16:43 garethdaine joined #git
16:43 maxxe osse, I don't see a problem with it. you probably know the name of the creator of weechat and irssi as well.
16:44 osseswirc joined #git
16:44 ManiacTwister joined #git
16:44 maxxe but anything is better than to be part of facebook
16:44 iLembus joined #git
16:44 GodGinrai you don't like Facebook telling all your friends you died? :P
16:44 zacts joined #git
16:45 maxxe https://stallman.org/facebook.html
16:46 osse osseswirc: hi
16:46 maxxe I trust Stallman. however I don't agree with him that there's no God
16:47 osse hmm, how do I switch between buffers :P
16:47 howdoi maxxe: how do you know there is one?
16:47 osseswirc found it!
16:47 howdoi osseswirc vs osse
16:47 ManiacTwister joined #git
16:48 osse relevant nick is relevant
16:48 GodGinrai I think osswirc would have been better
16:48 govg joined #git
16:49 howdoi :D
16:49 tyreld joined #git
16:49 rorro joined #git
16:50 maxxe howdoi, I don't know for sure. But happenings. I can never prove either
16:50 ManiacTwister joined #git
16:51 howdoi maxxe: neither can it be proved that there is none; It's something that needs to be experienced and can't be explained; Just like air or hunger ;)
16:51 UniFreak joined #git
16:51 osse I am horribly offended that ctrl-g clears the input and not ctrl-u. other than that i like it
16:51 maxxe howdoi, yeah :)
16:51 GodGinrai ctrl-g?  now that's just weird :P
16:52 howdoi maxxe: if there is a creation, there must be a creator; This universe was created, it did not happening out of nothing...so...(:
16:52 grawity osse: it kiiiinda fits emacs' "C-g is cancel"
16:52 * howdoi I like offtopic philosophy :D
16:53 osseswirc but the creater doesn't have to be sentient or anything
16:53 maxxe howdoi, yes :)
16:53 mindriot_ joined #git
16:53 osse osseswirc: surely you mean creator
16:53 osseswirc osse: yes
16:53 kbs I wonder if numbers are a creation
16:53 howdoi okies I have great respect towards RMS as well!
16:54 howdoi creator, yup
16:54 ManiacTwister joined #git
16:54 jimi_sanchez joined #git
16:55 howdoi gtg, later
16:55 e14 joined #git
16:57 rbern joined #git
16:57 ManiacTwister joined #git
16:59 Darren_ joined #git
17:00 bongo I've saved output of the filter-branch execution when I run for --all the there is """WARNING: Ref 'refs/remote/origin/*some branch*' is unchanged"""  for all the branches.  perhaps my tree-filter is not running as expected.  is there a good way to confirm this?
17:00 ManiacTwister joined #git
17:01 m___ joined #git
17:02 emg joined #git
17:04 ManiacTwister joined #git
17:04 robattila256 joined #git
17:04 einnjo joined #git
17:07 ManiacTwister joined #git
17:07 jccn joined #git
17:07 rominronin joined #git
17:08 LucaTM joined #git
17:09 thiago joined #git
17:10 axisys joined #git
17:11 ManiacTwister joined #git
17:12 railssmith joined #git
17:14 TinFury Having a problem... git pull says it's up to date but it's not
17:14 ManiacTwister joined #git
17:14 TinFury I have a local copy that pushes to my git server..... and a have a production server that pulls from git server
17:14 TinFury I'm the only one that uses this.... just set git server up for the first time
17:14 re1_ joined #git
17:14 TinFury can anyone help
17:15 qqx TinFury: What protocol are you using to pull?
17:16 TinFury qqx, ssh
17:16 jstein_ joined #git
17:17 ManiacTwister joined #git
17:18 TinFury Take a look at this output
17:18 TinFury http://laravel.io/bin/DeLwj
17:19 aelgabri joined #git
17:20 nowhereman joined #git
17:20 JosephFerano Can I get some help with git lfs? I duplicated a project (just simply did cmd + d on OSX) and I've tried working with the repo, but for some reason, I checked out another branch, and now it tries to run the clean filter over every single lfs file
17:21 ManiacTwister joined #git
17:21 UniFreak joined #git
17:21 donmichelangelo joined #git
17:21 JosephFerano When I run git lfs status, "Git LFS objects not staged for commit:", it shows all my lfs tracked files
17:22 JosephFerano all I did was just switch branches
17:22 aarobc joined #git
17:22 JosephFerano Any help will be appreicated
17:22 rxo joined #git
17:22 synbit joined #git
17:24 ManiacTwister joined #git
17:24 durham joined #git
17:25 Abbq joined #git
17:25 HardlySeen joined #git
17:27 ManiacTwister joined #git
17:28 ochorocho__ joined #git
17:30 Siegfried joined #git
17:30 ManiacTwister joined #git
17:30 nickabbey joined #git
17:31 tnecniv joined #git
17:32 rts-sander Hi I type git push -ff to mean "fast forward"
17:32 seishun joined #git
17:32 rts-sander but it does forced update, where did it go wrong? :(
17:32 bremner oh, is that a legit option?
17:32 bremner oh, yeah, -f means force
17:32 bremner you shouldn't assume ;)
17:33 dreiss joined #git
17:34 ManiacTwister joined #git
17:36 qqx rts-sander: There's no such thing as a fast-forward push. And if it did exist you generally need to use double dashes to get multiple letters treated as a single option.
17:36 encod3 joined #git
17:36 aarobc joined #git
17:36 qqx With just a single dash -ff would be seen as the same as -f -f, so being even more forceful.
17:37 ManiacTwister joined #git
17:38 rts-sander I see, acronyms are evil
17:38 jimi_sanchez joined #git
17:39 ShekharReddy joined #git
17:40 aviraldg joined #git
17:40 aviraldg joined #git
17:40 jaguarmagenta joined #git
17:40 ManiacTwister joined #git
17:41 Rodya_ joined #git
17:43 manuel joined #git
17:43 invisbl_ joined #git
17:44 ManiacTwister joined #git
17:45 dviola joined #git
17:46 manuel joined #git
17:46 vuoto joined #git
17:46 dviola joined #git
17:46 irqq joined #git
17:47 ManiacTwister joined #git
17:48 svm_invictvs joined #git
17:51 ManiacTwister joined #git
17:51 synbit joined #git
17:52 bremner ITYM AAE
17:53 [0xAA] joined #git
17:54 manuel joined #git
17:54 ManiacTwister joined #git
17:54 frem_ joined #git
17:55 xall joined #git
17:55 Hink joined #git
17:57 hahuang65 joined #git
17:57 ManiacTwister joined #git
17:58 e14 joined #git
17:58 nickabbey joined #git
17:59 PioneerAxon_ joined #git
18:01 troppie joined #git
18:02 ochorocho__ joined #git
18:02 Macaveli joined #git
18:03 manuels joined #git
18:04 Oatmeal joined #git
18:04 DeltaHeavy left #git
18:05 ash_workz joined #git
18:05 invisbl joined #git
18:05 invisbl joined #git
18:06 clemf_ joined #git
18:07 Rodya_ joined #git
18:08 publio joined #git
18:13 cheater joined #git
18:17 Rodya_ joined #git
18:18 andlabs left #git
18:20 bamb joined #git
18:21 Gsham joined #git
18:22 chardan joined #git
18:22 Drone4four joined #git
18:24 cheater joined #git
18:25 dp_ joined #git
18:26 Drone4four i have two repositories. i'd like to hide one inside the other. is that possible? can I achieve this by putting a directory in the .gitignore file of the 'parent' repo, and then put the second git repository within that  directory referred to in .gitignore?
18:26 thiago Drone4four: yes
18:26 thiago note that the directory in the outer repo must not exist yet
18:26 thiago not as a file, not as a directory with stuff inside
18:26 cjohnson not be added to the repo, anyway
18:26 cjohnson it can exist
18:27 jimi_sanchez joined #git
18:29 qqx There can be content already in a directory that is to be ignored. Git will still continue to track changes to those files, it just won't suggest new files as ones you may want to track.
18:29 jimi_sanchez joined #git
18:30 clemf joined #git
18:32 roger_padactor_w joined #git
18:32 aswen joined #git
18:33 dreamseeker_ joined #git
18:33 Drone4four i've been working on these two repos separately and now i'd like to do as i suggest.  but are you ppl saying that i can't copy the second repo, keeping its commit history.
18:33 Drone4four thiago, cjohnson, qqx: are you three saying that i can't keep my commit history for the second repo (the one i'll be hiding)?  I would like to retain the tracked changes
18:34 Drone4four btw cjohnson i really like the throw back to San Andreas. i had alotta fun with that game =D
18:34 oskarkv_ joined #git
18:34 thiago Drone4four: if they're independent repos, you can leave them anywhere
18:34 thiago mv them just fine
18:35 thiago if you put one inside the other, then the outer will see the inner one as a submodule
18:35 railssmith joined #git
18:35 thiago so either add it or ignore it
18:35 cjohnson are you trying to merge their commit histories in some way?
18:35 thiago if you replace an existing dir with it, then you'll have to do something
18:35 viktor joined #git
18:35 cjohnson or you just physically want to place one repo inside another
18:35 thiago also note that there's a "private" gitignore file: .git/info/excludes
18:35 thiago you can add stuff there that is only for you
18:36 Drone4four cjohnson, i would prefer to keep the commit histories independent
18:36 Drone4four i do not want to merge the history
18:37 cjohnson cool then yep do as said. make a dir called otherrepo, add otherrepo to .gitignore, move your entire repo including its own .git folder into otherrepo
18:37 cjohnson then if you cd into otherrepo you can issue git commands on that repo
18:37 cjohnson and if you cd to the parent you can issue git commands on the parent
18:37 Drone4four ok perfect. thanks for the clarification, cjohnson
18:38 Drone4four and thanks thiago and qqx
18:39 rominronin joined #git
18:43 clemf joined #git
18:44 willlondon joined #git
18:45 mikecmpbll joined #git
18:46 nickabbey joined #git
18:46 SwiftMatt joined #git
18:46 UserOO7 joined #git
18:47 UserOO7 joined #git
18:47 oskarkv joined #git
18:48 rbern joined #git
18:50 finalbeta joined #git
18:50 clemf joined #git
18:51 ok91 joined #git
18:52 Sasazuka joined #git
18:56 nickabbey joined #git
18:58 clemf joined #git
18:58 PioneerAxon joined #git
18:59 dsantiago joined #git
18:59 dan2k3k4 joined #git
19:01 rorro joined #git
19:02 cheater hi
19:02 Ryanar joined #git
19:03 cheater someone has been doing git push -f and i checked out his branch (origin/foo) by doing git checkout foo, and now when i did git fetch -a i have an unmerged file. i would like to throw that away and just take what he did. what do i do?
19:04 sssilver joined #git
19:04 PioneerAxon joined #git
19:05 anonymuse joined #git
19:06 GodGinrai cheater: -f?
19:06 dreiss joined #git
19:07 kadoban cheater: What's the actual message? Unmerged file? I have no idea what that'd mean in this case.
19:08 aelgabri joined #git
19:08 cyan__ joined #git
19:08 nicksloan joined #git
19:09 cagedwisdom joined #git
19:10 spacelord_ joined #git
19:11 cheater yeah, so what i did was to do git reset <hash of the head of origin/foo> and then git checkout src/conflict-file
19:11 cheater and now it's all fine
19:12 nTQ5QarKE9 joined #git
19:14 UserOO7 joined #git
19:15 synthroid joined #git
19:16 nicksloan joined #git
19:16 kyan joined #git
19:18 netj joined #git
19:19 Flowinho joined #git
19:20 kulelu88 joined #git
19:20 einnjo joined #git
19:20 Phylock joined #git
19:20 UserOO7 joined #git
19:21 mostlybadfly joined #git
19:23 TomyLobo joined #git
19:25 mseyne joined #git
19:25 endiruna joined #git
19:26 nidr0x joined #git
19:28 chachasmooth joined #git
19:29 shashin joined #git
19:32 paul424 joined #git
19:33 cdown joined #git
19:35 fiddlerwoaroof joined #git
19:36 perrier-jouet joined #git
19:36 lungaro someone checked in a bunch of binary crap into git. Can I delete it somehow?
19:36 lungaro it would be nice if I could zero out the specific object to reclaim the space
19:37 [0xAA] revert or reset it probably
19:37 GodGinrai there's some way to deep-clean
19:37 GodGinrai I forget how to get the bot give the url for it
19:37 nwkj86 joined #git
19:39 hhee joined #git
19:39 nicksloan joined #git
19:40 sarbs_ joined #git
19:41 jaguarmagenta joined #git
19:41 roger_padactor_w left #git
19:42 cheater repack?
19:45 Flowinho_ joined #git
19:45 dan2k3k4 joined #git
19:46 ejb joined #git
19:47 pks joined #git
19:48 mischat joined #git
19:49 nicksloan joined #git
19:49 manuels joined #git
19:50 dsantiag_ joined #git
19:50 multi_io joined #git
19:51 encod3 joined #git
19:57 msonntag joined #git
19:57 overlord_tm joined #git
19:58 monokrome joined #git
19:59 ELFrederich when I run "git clone git@some-host:gitolite-admin" what process does Git go through on some-host to determine where gitolite-admin is?
19:59 nicksloan joined #git
20:00 ELFrederich ... or is there something special about the git user's ~/repositories directory that Git has hard coded to look in there?  I'm curious, because Git knows nothing about Gitolite correct?
20:01 bremner ELFrederich: that's something built in to gitolite, not to git
20:01 _ikke_ git does not know anything about gitolite
20:01 ELFrederich bremner: I'm not running gitolite.... I run "git clone git@some-host:some-repo" ... that's just running Git
20:02 vlt !bfg | lungaro
20:02 gitinfo A tool designed to remove large files, or passwords from history: https://rtyley.github.io/bfg-repo-cleaner/ (!rewrite applies)
20:02 bremner ELFrederich: you are ssh'ing into the remote host and running gitolite there
20:02 ELFrederich it just so happens that the "git" user on "some-host" has installed gitolite.  But Git doesn't know that.  I'm wondering how it is hooked up or "installed"
20:03 bremner ELFrederich: it's explained in the gitolite docs
20:03 GavinMagnus joined #git
20:03 ochorocho__ joined #git
20:03 ELFrederich bremner: how?  "git clone git@some-host:some-repo" how does that invoke gitolite?  I know Git is installed on both the client and server, but somehow git on the server winds up invoking gitolite... right?
20:04 Rodya_ joined #git
20:05 ELFrederich bremner: I added a new user called "git" and have been watching changes in it's home directory but I don't see any that would intercept the git stuff... nothing has been modified in ~/.profile or ~/.bashrc or anything
20:06 Salander27 joined #git
20:06 bremner http://gitolite.com/gitolite/glssh.html
20:07 seishun joined #git
20:08 mindriot101 joined #git
20:08 ELFrederich bremner: THANKS  if I read the docs that would probably be the last page I would have read ;-)
20:09 Guest80353 joined #git
20:09 irqq joined #git
20:10 jieguyan[m] joined #git
20:10 NeverDie joined #git
20:10 nickabbey joined #git
20:10 rominronin joined #git
20:10 sr___ joined #git
20:11 jieguyan[m] left #git
20:12 aidalgol joined #git
20:13 NeverDie joined #git
20:13 nickabbey joined #git
20:13 lungaro thanks vlt. i'll check it out
20:14 sharpy joined #git
20:14 Darren_ joined #git
20:17 muahah joined #git
20:29 bmahe joined #git
20:30 daynaskully joined #git
20:31 TikityTik joined #git
20:31 TikityTik my .gitignore file in my repo is not working.
20:31 GavinMagnus joined #git
20:31 GodGinrai TikityTik: I doubt that.  Care to elaborate?
20:31 GavinMagnus left #git
20:32 mdw joined #git
20:32 TikityTik I merely added .gitignore in the root of the repo, and the file only has: __pycache__/*
20:32 TikityTik and going git status still shows unstaged from that folder.
20:33 TikityTik wait, hold on
20:34 TikityTik nevermind, sorry.
20:34 TikityTik My problem was that someone added that directory to the repo anyhow, or it was stuck in my cache or something, so I had to git -rf it.
20:34 aelgabri joined #git
20:35 overlord_tm joined #git
20:41 guampa joined #git
20:42 sdlnv joined #git
20:44 re1_ joined #git
20:47 axl_ joined #git
20:48 axl_ left #git
20:50 mischat joined #git
20:50 axl_ joined #git
20:50 Jackneill joined #git
20:54 Jackneill joined #git
20:54 TikityTik What does MASTER~4 mean?
20:56 epist3me joined #git
20:56 GodGinrai 4 commits before master, using the first parent each time
20:56 TikityTik I couldn't find anything on google for ~4, where is a good tutorial on this?
20:57 GodGinrai TikityTik: https://git-scm.com/book/en/v2​/Git-Tools-Revision-Selection
20:57 cheater try symbolhound.
20:58 Rodya_ joined #git
21:00 TikityTik I'm trying to erase a bunch of commits I made to a readme because it's too much spam to history
21:00 TikityTik And they're already pushed. I'm doing: git rebase -i origin/master~6 master. Do I just change all the picks to d? Because I'm getting: error: could not apply 3b58ee0... Update README.md
21:01 masuberu joined #git
21:01 TikityTik it's reporting an error for that where I leave the pick
21:02 aarobc joined #git
21:02 GodGinrai TikityTik: you don't rebase anything that's public
21:03 dan2k3k4 joined #git
21:03 TikityTik GodGinrai: Well I have a bunch of commits I want to get rid of.
21:03 GodGinrai TikityTik: to the point that you are going to cause problems for everyone who has pulled from the repo?
21:04 TikityTik is there no other way to cleanup the commit history?
21:04 GodGinrai No.
21:05 GodGinrai Well, there are other ways.  But they also would cause problems for other users once they have pulled a public copy
21:05 GodGinrai That's why the Pro Git book (and even the git manpages, IIRC) stress that you shouldn't rebase anything that's already been made public
21:06 TikityTik well it's only one other individual who hasn't pulled yet
21:06 TikityTik but i'll just leave it I suppose
21:07 ahmedelgabri joined #git
21:13 felixjet joined #git
21:13 cdg joined #git
21:15 cdown_ joined #git
21:17 e14 joined #git
21:17 qws-user-1228 joined #git
21:18 dunpeal joined #git
21:19 alex1a joined #git
21:20 nanga joined #git
21:20 dsdeiz joined #git
21:21 wikkie joined #git
21:23 anonymuse joined #git
21:23 m___ joined #git
21:25 Samual joined #git
21:26 marcus_ joined #git
21:26 User458764 joined #git
21:26 m0viefreak joined #git
21:30 spacelord_ joined #git
21:31 Virox joined #git
21:31 pks joined #git
21:31 jimi_ joined #git
21:32 Spec-Chum joined #git
21:33 jimi_ joined #git
21:34 seishun joined #git
21:35 jimi_sanchez joined #git
21:39 jimi_sanchez joined #git
21:39 senaps joined #git
21:41 senaps hi... i have a config file with secret key and database connection string, it's on deployment branch and i have the same config file witch is empty and is on development branch. i make changes in development, but is there a way that i  merge development into deployment, and config file doesn't change?
21:42 rominronin joined #git
21:42 jaguarmagenta joined #git
21:42 Seveas senaps: yes, don't put configs in the same repo as the code.
21:42 chadhinton joined #git
21:42 Seveas configs go in your config management system
21:43 GodGinrai ^ this
21:43 GodGinrai Anything that should be different based on environment (configs, etc.) shouldn't be in the codebase
21:44 GodGinrai or should be accounted for in all branches of the codebase
21:44 tang^ joined #git
21:44 iliv joined #git
21:44 senaps so how do i push it alongside other files to the server, say heroku for example?
21:45 Nicmavr joined #git
21:45 Nicmavr joined #git
21:49 Guest80353 left #git
21:51 jstimm joined #git
21:51 thiago joined #git
21:51 mischat joined #git
21:53 rwb joined #git
21:54 Seveas Not only is git not a configuration management tool, it is also not a deployment tool.
21:54 GodGinrai dingdingding
21:54 GodGinrai we've got a winner here!
21:55 aarobc joined #git
21:55 Seveas senaps: heroku is really annoying in this respect as it insists on using git as a deployment tool.
21:57 askb joined #git
21:57 Seveas what I'd do in this case is have a special clone only for deployments. It's master branch has the config. Whenever a deployment needs to be done, you pull from the main repo, keeping the config. And you *never* push from this repo to the master repo, so no other branches in any clones of the master repo have the config
21:58 Seveas Or what I'd really do is find a better way of hosting my app than herokupoop, but I realize that that's not a helpful answer right now :)
21:59 senaps it's just for testing!! i can't buy the VPS yet... i have to make sure that my app works fine and then give it a shot on a good vps...
21:59 nikivi joined #git
22:00 Vangelis joined #git
22:00 Seveas vps'es come as cheap as less than $5/month
22:00 qcube joined #git
22:01 senaps so i should basically have multiple clones of the same thing, with one being the one that is being online and never pull data from it, just push my new files on it and i'm okay....
22:02 senaps Seveas,  will, here in iran, that 5$ translates to a large amount of money! :)))
22:02 cheater does git now have a ncurses based gui for staging parts of a file (single lines) rather than the whole file at a time?
22:03 cheater like hg crecord
22:03 solars joined #git
22:03 aarobc joined #git
22:04 zacts joined #git
22:04 kadoban cheater: git add -p ?
22:04 Electrometro joined #git
22:05 dunpeal joined #git
22:05 solars hi, quick question: if I have a feature branch that takes a bit longer, and I also push it, then have to rebase to dev or master, I basically have to force push it if I want the changes to be on the remove repository. I have to be careful to not pull because otherwise I will have it duplicated, also I cannot pull changes someone else might add there after my force pushes - how do I handle this right in general? :)
22:06 kadoban Either be careful or ... don't do the rebases. You can just merge stuff in instead, and optionally clean up history before you merge it into master/dev.
22:07 [0xAA] joined #git
22:07 cheater no kadoban that's specifically what i don't want
22:07 howitdo joined #git
22:07 cheater look at hg crecord, this is 100% what i want
22:07 kadoban cheater: *shrug*
22:07 cheater git gui does that too but that's a gui tool and.. yuck
22:08 chardan joined #git
22:08 kadoban If you want a fancy interface, ya I'd look for some external tool.
22:08 cheater *sniff*
22:10 GodGinrai solars: never rebase something already public
22:10 btree joined #git
22:10 solars so in this case I should only merge?
22:11 btree left #git
22:11 cheater yes
22:15 Cavallari joined #git
22:16 solars thanks guys
22:19 cdown joined #git
22:21 tang^_ joined #git
22:21 Sasazuka_ joined #git
22:23 mdw joined #git
22:23 thiago joined #git
22:25 NeverDie joined #git
22:25 chardan joined #git
22:27 grapheti joined #git
22:27 cdown joined #git
22:28 mkilivan joined #git
22:29 grapheti left #git
22:29 grapheti joined #git
22:33 cheater joined #git
22:36 mkilivan joined #git
22:38 highlander747 joined #git
22:38 highlander747 left #git
22:40 mkilivan joined #git
22:40 highlander747 joined #git
22:45 ToBeCloud joined #git
22:45 WaReZ joined #git
22:46 [0xAA] joined #git
22:47 aidalgol joined #git
22:47 malber joined #git
22:51 mdw_ joined #git
22:51 [0xAA] joined #git
22:58 Seveas .version
22:58 Topic for #git is now Welcome to #git, the place for git help and serious business | Public logs at http://jk.gs/glog | First visit? Read: http://jk.gs/git | Current stable version: 2.11.0 | Getting "cannot send to channel"? /msg gitinfo .voice | The git-jokes project has been suspended for lack of contribution. Help revive it! https://madeitwor.se/git-jokes
22:59 mkilivan joined #git
22:59 venmx joined #git
23:00 sangy joined #git
23:01 fmeerkoetter joined #git
23:01 malberDev joined #git
23:03 Rodya_ joined #git
23:06 finalbeta joined #git
23:06 synchronous joined #git
23:07 Rodya_ joined #git
23:08 chachasmooth joined #git
23:11 qsx already?
23:12 qsx even i remember "whoa 1.9 is gonna break the whole numbering scheme"
23:12 wizardedit hi there. Is there anyway to have git svn report the revision number and if it's modified? e.g., if I edit a repo but don't commit, svnversion shows 1234M, but git svn info | grep ^Revision shows 1234
23:12 zifnab was kicked by Eugene: Every time a git release happens with no new jokes I will terminate a hostage
23:12 wizardedit and git svnversion doesn't exist
23:13 zifnab joined #git
23:13 Eugene wizardedit - man git-svn; I believe find-rev is what you want
23:13 gitinfo wizardedit: the git-svn manpage is available at http://jk.gs/git-svn.html
23:14 wizardedit Eugene, I can get the revision, but it doesn't change at all if there's a diff versus not
23:14 rominronin joined #git
23:14 wizardedit git svn find-rev HEAD shows 1234, not 1234M (but git diff shows a diff)
23:14 dviola joined #git
23:15 Eugene HEAD is the last commit that you created. `git diff` shows new work that has not yet been commited, so SVN doesn't know anything about it
23:15 boombatower_ joined #git
23:15 Eugene I guess I may not be understanding the question?
23:15 malber joined #git
23:15 Seveas Eugene: maybe git *is* the joke...
23:15 bluezone joined #git
23:15 qsx HEAD IS ... oh, svn.
23:16 dermoth joined #git
23:16 Seveas was kicked by Eugene: That's not a punchline
23:16 Eugene Yeah, I don't use git-svn very often myself
23:18 wizardedit Eugene, yeah, git svn is a bit of a mess
23:18 wizardedit used as a last resort
23:18 wizardedit anyway, issue is that I added support for valgrind to report it's svn revision in --version, kinda like git describe
23:18 wizardedit with svn, it will report 1234, or 1234 if you're running svn and you have a diff
23:19 wizardedit or 1234M*
23:19 wizardedit but with git svn, I only get 1234, not any indication that the source has been edited
23:20 Vampire0_ joined #git
23:20 einnjo joined #git
23:21 buffal0 joined #git
23:21 dsdeiz joined #git
23:23 Eugene If you can get the revision IDs into git tags(this might be helpful: http://stackoverflow.com/questions​/23760674/svn-revision-to-git-tag) you can use `git describe` directly
23:25 YottaByte joined #git
23:26 YottaByte I'm in branch A with some local, uncommitted changes. I want to move to branch B to commit them. does git checkout branch_b keep my local uncommited changes and then it's as simple as git add ., and commit -m "my msg", git push?
23:26 YottaByte or should I do a git pull after I checkout the other branch first?
23:26 Eugene YottaByte - yup; !float
23:26 gitinfo YottaByte: If you have made a change in your working directory and have NOT YET COMMITTED, you may "float" that change over to another (`git checkout oldbranch`) or new (`git checkout -b newbranch`) branch and commit it there.  If the files you changed differ between branches, the checkout will fail.  In that case, `git stash` then checkout, and `git stash apply` and go through normal conflict resolution.
23:27 Eugene pull would do a fetch+merge from whatever B's upstream is
23:29 YottaByte Eugene: ostensibly I'd have to do a pull first, right? because I might not be able to push?
23:30 Eugene YottaByte - not really. Commit what you have, then merge(or rebase) as you need to if there's a push conflict
23:31 Eugene If you are in an uncommited state doing a merge is a terrible idea
23:31 YottaByte why is that?
23:31 Eugene Well, for one pull will refuse because you have uncommited work
23:31 chachasmooth joined #git
23:31 axl_ joined #git
23:32 Eugene For two, if you do manage to force a merge commit, there's no good way to tell what changes are from where
23:32 Eugene Which is known as an Evil Merge
23:33 YottaByte oh so you should always commit before you pull? I often pull in an uncommitted stage and it'll auto-merge I think
23:34 Eugene If it can fast-forward cleanly, sure
23:38 Marbug_ joined #git
23:39 nowhereman joined #git
23:40 felixjet joined #git
23:43 felixjet joined #git
23:43 tnecniv joined #git
23:43 mellernoia joined #git
23:44 jeffreylevesque joined #git
23:45 dunpeal_ joined #git
23:45 cbegin joined #git
23:49 mkilivan joined #git
23:50 dunpeal joined #git
23:52 buffal0 joined #git
23:53 cagedwisdom joined #git
23:54 mischat joined #git
23:56 [0xAA] joined #git

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